One of the greatest transceivers ever produced.     

This page provides:

· Some information not available anywhere else which should be useful to any TS-940 owner,

· Information at single web site, easy to follow, [no other site provides this about the 940]

· Information remaining available so more TS-940s will be repaired + functional, (and probably improved),


Version 2: 4 April 2005,   Version 3: 25 April 2005,  Version 4: 27 May 2005,  Version 5: 31 May 2005,   Version 6: 10 June 2005:  Version 7:

16 June 2005:  Version 8: 25 July 2005, Version 9: 30 July 2005. Version 10: 4 August 2005, Version 11: 13 Sep 2005, Version 12: 18 October 2005,

Version 13: 23 October 2005, Version 14:  22 March 2006,   Version 15: 8 April 2006,   Version 16: 27 May 2006, Version 17: 12 July 2006

Version 18: 22 Aug 2006, Version 19: 23 Sep 2006, Version 20:  5 October 2006, Version 21: 7 Jan 2007, Version 22: 17 Jan 2007, Version 23: 5 Oct 2007,

Version 24: 19 April 2008, Version 25: 27 July 08,  Version 26: 20 Sept 08, Version 27: 26 Sep 09, Version 28: 13 Feb 10,  Version 29: 25 Jan 2012, Version 30: 25 Feb 2012 ,

Version 31: 19 Jan 2013



The intention is to acknowledge the person who discovered the information

When information is already well documented and reliably maintained on another site then a hyperlink is made to that site to avoid yet another slightly different version.

I will publish all email feedback of new information at the end of the page, so that whatever is discovered by others can be shared by all. Please email to

Yours sincerely

Jeff King ZL4AI  /  DU7


Disclaimer: No liability or responsibility whatsoever is taken for any of the information on this site. You assume total liability for any modification you undertake.
Copyright to Author: Everything on this page 
See also TS-950sdx page which hold valuable information:



















R1. KENWOOD PRODUCED 3 SERVICES BULLETINS which do considerably improve the receiver. 39




Mike KC8ZNW on 25/4/05 describes this same behaviour to the 54

Executive Summary of AGC Mod. 55

Independent Feedback on how Receiver Improves. 55








4.1: Background on how Pin Diodes were discovered to improve radios. 67

4.2 RadCom Technical Topics explains what Pin Diodes were supposed to achieve. 69

Intermodulation properties of switching diodes, by Dr. Ing. Jochen Jirmann, DB1NV.. 81






To Access tHE SECTIONS Below click here ------>:




4.4 Experience from Persons who modified the TS-940. 2

4.5 : Pin Diode Modification for TS-440. 13

4.6: So summary of Pin Diode Modification. 13



6.1 Observation of the Phase Noise problem by Thomas Hohlfeld DF5KF in 2005. 23

6.2 Extensive History of the Phase Noise Issues upto about June 1987. 23

6.3 Data Sheet for Problem BA718 OpAmp. 24

6.4 What is Phase Noise and how to measure it by John Grebenkemper, KI6WX.. 24

6.5  Phase Noise measurements on Elecraft K2 by John Grebenkemper, KI6WX.. 46

6.6 Significant Phase noise improvements developed by Thomas Hohlfeld DF5KF.. 47

6.7 Information on Ultra Low Noise OpAmps available in 2008. 53

6.8 Further assessment and Phase noise improvement evaluations planned by Jeff King ZL4AI 57

6.9 First plots of phase noise comparisons below show some very exciting results: 57

6.10  Real Experience of the Ultra Low Noise LT1028: 58

7.0 SM-220 mods to remove Ghost Signals generated by the 940. 63



PLL BOARD 0: Reseat Connectors.. 66

PLL BOARD 1: Remove the Black Foam from Behind the Board.. 66

PLL BOARD 2: Remove the Wax from the VCOs.. 67

PLL BOARD 3: Identify which PLL is not locked.. 68

PLL Board 4:  PLL Board and RF Board and PLL out of lock.. 70

PLL Board 5:  PLL Board and setting voltages:  Comprehensively updated in 2012.. 71






Power Supply HEAT SINK RUNS TOO HOT.. 81





VK5KYO includes Larger Computer Fan and relocated Rectifiers. 87

28 Volt Crowbar safety circuit for the TS-940. 87

TS-940 Power Supply Improvement in 2012: Read this page first.. 87




RF BOARD 1: Board Runs very Hot.. 2


TS-930 (and TS-940…) Power amplifier repair.. 4

Alternative replacement for MRF485. 6



Repairing old fluorescent displays (to brighten up dim digits) 12





Service Manual & Serial Numbers.. 15

Identifying When Radio Manufactured.. 17

BULBS.. 20

Sub display bulb: 20

S Meter Display Bulbs. 24







PARTS.. 37










































From: "Vaso Nastasic" <>

To: <>

Sent: Thursday, January 04, 2007 9:09 PM

Subject: [Kenwood] TS940S CW heterodyne tone


Hello from this part of world and HNY 2007!  I have problem with ……

Anyone can help?

73s de Vaso YT1XX / YT6XX.


Hi Vaso,


did you find the home page of Jeff King ZL4AI,

Regards, Nermin S58DX



On Behalf Of Lynn Baustian

Sent: Friday, 5 January 2007 2:02 p.m.


Subject: [Kenwood] TS-940S


Does anyone happen to have all the info on Jeff Kings sight saved to a .pdf

It could be invaluable should his sight go down for any reason.

73.....Lynn  WA7ADY




From: jaking []
Sent: Saturday, 6 January 2007 10:48 a.m.
To: ''
Cc: 'Lynn Baustian'
Subject: RE: [Kenwood] TS-940S Page and updates


PDF of TS-940 page, or TS-950sdx page. If you really need this, print to a printer or a pdf writer and just print off your own version, Or just save the site as an html. Ihug is very reliable and this site is unlikely to go down. If that happened I would shift it to another ISP.


Taking a copy in any form is not recommended. because I update the TS-940 page regularly. I record my notes of working on the 940 on the site page. After that it takes about 5 minutes to post a new page. That way the site progressively becomes better.


The reason copies are not so good is they will not include improved / revised information. It is better to have just one authenticated source of information. For example I used to get lots of questions about how to follow circuit diagrams. Since I posted my ideas on how to read, all those questions seem to have stopped. If you have an old copy of the site you will not have that information. Ironically the reason I first published the site was I had collected the information and published it because I could see it would be helpful to others. Now I have to defend the advantages of keeping the information in just one place.




In the next years or longer as time permits I intend doing on both a late and an early 940


1.         phase noise tests,


2.         receiver tests, such as MDS [minimum discernable signal], IMD3 etc, because no source of this important information exists for modified or later 940s,  [ZL4AI measured on his 20 mil unit MDS CW = -144 dBm, MDS SSB -139 dBm]


3.         band filter adjustments / alignment


4.         FET 2SK125 reversal with before and after laboratory MDS tests, to establish what the change in receive signal level is. Plus I intend telephoning Kenwood USA Amateur radio technical support on 001-310-639-4200, option 5, option 2, option 1, to find out what the real truth of Kenwoods official position on FET reversal is? I find it difficult to understand why reversing a single FET should overload or make the receiver unstable? I anticipate the measured change in MDS due to reversal will be only a couple of dBs.


5.         possibly replacing 2SK125s with J310s to discover if later (year 2005 made) (of higher quality manufacturing standard) pre-amp FETs reduce the noise,


6.         try installing better and additional shielding around the entire pre-amp sections to block out and reduce noise,


7.         replacing band filter diode switches with relays to see if MDS and other receiver qualities improve,


8.         Figure out the switching in and out of filters in SSB mode, so I can develop a rewiring of jumpers which allow switching in the empty CW filter bays while in SSB mode. Then try INRAD, and / or Kenwood AM 6K wide or other filters as additional or alternative switch ins, to enable filtering capability similar to the TS-950 and TS-850. [Measure receiver tests, for various combinations.]


9.         publish audio spectrum graphical outputs of the 940 performance. [Possibly try replacing audio components such as capacitors to give a more hi-fi higher quality audio].


10.       publish files of 940 performance with and without DSP filtering. Experience so far is 940 plus DSP is an exceptional improvement]


11.       -evaluate and publish information on how to link 2 940s together so as to have and simultaneously use, a second 940 as a dual receiver,


12.       Information about replacing bulbs with LEDs [published version 23]


13.       Details on 28 volt crowbar circuit to protect the radio against power supply failure, [base data was published in Version 20]


14.       An adjustable resistor for activating temperature of the power supply fan,


15.       More experience on removing the very hot power diodes off the AVR board onto the heatsink


16.       publish details of all the changes Kenwood made to the 940 between 1985 and November 1986. There are many more changes to the 940 than those on the published services bulletins.


17.       publish info about filter input of SM-220 to remove ghost signals from 1 MHZ generated by the 940. [Published version 24]


All that information, should be of real interest to any 940 owner. For that reasons I suggest you keep reading the site.

[If anyone has done or undertakes any of the above earlier please send your results for publication. There are definitely more mods that can be undertaken to improve the 940.]


Jeff King ZL4AI











Remember the only reason most 940 owners ever sell, is their radio has now developed faults, and they cannot bear the large cost of getting another person to repair the radio. So sadly they replace it with a radio that is working, for economic reasons. Most say if it had not developed faults I will still be using it because its performance is as good as all the latest models.


INRAD has released the 940 roofing filter #903 which give close in signal rejection equivalent to 2012 latest receivers.


Almost all radios after 1991 have been made using surface mount technology. The components are very small and difficult to work on.

For many amateurs servicing these radios is difficult, and requires soldering equipment with good temperature control, very fine tips and electrostatic safe.


The TS-940, (and others of the very early 1990s like the Icom 765, and Yaesu FT-1000D) can still be serviced, and most parts are still readily available at local stores.


For the 940 almost all parts can be replaced by substitutes of readily available common equivalent parts, easily obtained, which are not expensive, and often of higher quality than original part.  In most instances for the 940 there is no reason to abandon the radio because a part is not available from Kenwood. You can use standard off the shelf equivalent capacitors, resistors, ICs and often equivalent replacement transistors.


The only parts hard to get are ICs specially programmed by Kenwood [Only a a couple of these in the 940]. Even then you can get the code and have replacements burned for your radio.

You can buy second hand TS-940 boards on EBay and elsewhere at very reasonable prices. With second hand boards and parts you can still obtain every component. Don’t sell your 940… Repair it.


If you like kit sets, you could say the 940 is like one large kit set, preassembled.





this section is first because it is key to working on the Ts-940. It is not explained in the service manual.

Once you understand this you can easily understand the TS-940.

First step is download the manual from, or other sites.


On page 107 find the BLOCK DIGRAM. This is probably the most useful diagram in the book because it illustrates conceptually how the radio works.

It should have been at the very front of the manual.  From this you can easily follow through where major signals and functions travel.


Then go to page 12 Receiver and page 17 transmitter and read through how the receiver / transmitter functions while following through the block diagram. This will give very good idea of how the radio works.




Then on page 103 find the overall schematic of the 940.  This lays outs out all the major boards. There is a coding system, to trace between boards.




Now zoom into part of that schematic and look at example connectors:



Best illustrated by worked example.

On The PLL board X50-2020-00 find Connector Plug 2.

It has five wires:


Two example traces are shown below;


Wire 1-DA0       shown with a destination tag  CAR-1-DA0”,                   whose destination is Carrier Board (X54-1840-00)(A/2) Connector Plug 1 –DA0 Wire 8,  

Wire 2-DA1

Wire 3-DA2

Wire 4-DA3

Wire 5-UL1        shown with a destination tag  CAR-1-UL1”,                   whose destination is Carrier Board (X54-1840-00)(A/2) Connector Plug 1 –UL1 Wire 4,  


As you can see if you trace along all grouped wires through on the diagram, the grouped wire line leads from one board to another.


Using this system you can trace from any part of the 940 to any other part.


On each schematic or board layout diagram all you have to do is find the Connector Plug Number. Then you can easily trace through the board.


Page 83 to 89 of the service manual shows every Connector Plug on every board is listed out with description of its function. This helps identify what the wire function is doing.


Page 80 and 81 shows the physical location of most boards inside the radio.






Using the above decoding method solved the following problems.


From: W5EJ

Sent: Friday, 18 August 2006 7:45 p.m.


Subject: RE: Need help on 940 Wires


Jeff, after seeing your documentation on the 940 I thought I'd send you a note and see if you could help me out.


I picked up a used 940 and appears whoever had it before me had replaced several connectors inside the unit with direct solder connections. (yes a mess).


Question. On the Speaker/internal switch unit (left front top with cover top off) there are 3 connectors that plug into the right side of the unit. One has three wires, red orange and black. All three of mine are disconnected and I cannot tell which connector terminal on the PCB to re-solder them too. (your wires would be in a connector)


I don’t have another unit to reference and cannot find any pictures. do you have any pictures or 940's with the tops off could you let me know in what order front to rear the three wires are in your connector? (Of the three connectors on the right side facing the unit this is the connector towards the rear of the unit and only connector on that board with just three wires) Crazy question I know. …….



Any thoughts or help or pictures would be greatly appreciated.


thanks 73 - John (W5EJ)



From: jaking []
Sent: Monday, 21 August 2006 8:59 a.m.
To: ''
Subject: RE: Need help on 940 Wires




I don’t have a TS-940 apart right now.  But it is easy enough to figure out. Down load a service manual from


Then on page 103 overall schematic, find those connections.

You may have to look at board layout middle of p 102 to identify the connector number and wire letter. Looks like connectors 48 29 and 30. You will see the letters for each number there.


Then find the same places on the schematic p 103.


Follow the wires through on the schematic to see which board and connector they end on another board.


Find that other connector inside the radio, and get the colour of the wire from where it connects to the other board.


Then use a multi meter to verify you have the same wire at both ends. You would be best to un-plug the other end connector, before measuring continuity, because that way you are 100% you are only measuring on that wire only.


Then you can reconnect.


There is no color coding in the manual so this is the only way to decode the wire colours.


For example ON PAGE 103 YOU WILL FIND 48 LG appears to connect to Switch L-49-LG


Or 28 CV1 connects to IF board 14-CV1


So find IF board 14-CV1, on the IF board and you will get the colour of that wire there. Un plug IF board 14-CV1, and use the meter to verify you have the same wire.


This decoding systems works right through all 940 connections.


Let me know how you get on.


Jeff ZL4AI




From: W5EJ

Sent: Sunday, 20 August 2006 6:41 p.m.


Subject: Re: RE: Need help on 940 Wires


Took your advise and downloaded the manual, had the unit fixed in 30 minutes.  Thanks for the help - John








I was initially lost about what and how to look for faults. Developed the following methodology. (All suggestions about repair methodology are welcome)


Step 1 is to download a service manual


Step 2: Read how radio works.

Read though the receiver and transmitter written explanations, and then follow those paths through on the circuit diagrams.

Easiest way to follow through is photocopy circuit pages and get a highlighter and highlight those paths on the circuits.

Only then you get feeling for where you look to find a particular fault.



Step 3:

Find the region of the radio responsible for the function causing your fault,

And locate that on the main wiring diagrams and detailed wiring diagrams.

Then progressively check every connection and component.

That is slow and tedious.

Write down every check you made.

By a process of logical and elimination you will find the fault.

It may take days or weeks, but if you are methodical you will find the fault.



Step 4:

In suspect region:

- Do go over and tighten all connectors. These white plugs have metal female clamps which get stretched open after many insertions. For each pin you need to take each connector out of the plastic sleeve and clamp it more closed with small pliers. [Female outers are only designed for so many insertions. After that they just become loose][Lot of discussion on using deoxit. I have never used it and suggest it is not an oxidation problem, but a problem of connectors not have strong spring devices


I would guess tightening plugs would fix 80% of ts-940 faults, because the connectors are the weak point of 940s.


-       Re-solder regions of boards where suspect components are located. [On my three 940s (2 ex USA, 1 ex Australia) I have never found soldering defects that others go on about at length. I wonder if the soldering issue is highly over verbalised].



Step 6:

Using a digital volt meter continuity checker, verify all suspect connections are connected both inside and across boards.

(You can make your own RF volt meter for a couple of dollars by following articles  in the ARRL Hand book.)

[Please note that I have had the phase lock loop dots appear because of connectors being loose on the PLL and Digital A Boards. Yet reading other persons posy the first step would be to work on PLL board wax. My advice is first verify the connectors on each and every PLL is connected]



Step 7:

If you can locate replacement board for the suspect region and try that that will get fault absolutely defined to that board.

The fastest way and possibly cheapest way to fix is probably by elimination


All boards on the 940 are plug and play. When you put in replacement boards there are service adjustments to make afterwards , before seriously using the radio. But if the board is defective you can replace it and the radio should spring back to life.


For example your can buy a PLL and / or Digital A Board on EBay for between $50 and $100 each.


If you have problem in these regions I would suggest you buy, plug and play to eliminate if it is any of these boards.


You can always sell the boards you don’t need again afterwards.


You can also probably sell your defective board as well for the parts on it. [Those parts are valuable to others]


Max cost is one board, plus postage plus eBay fees.

A lot cheaper than sending the radio away, and waiting and waiting.



Step 8:

To find the fault, with radio running

- Mechanical connection fault find by tapping components with plastic non conducting object.


-       Fault that occurs as radio heats up. Use Freeze spray to cool suspect regions. If fault reappears when cooled you have found the region.

-       Use a hair dryer to heat up regions to see if you can make a fault appear or dis-appear,

-       Use shields make of cardboard around suspect commonest to try and isolate the cooling or heating to just one of two suspect components,

-       Use digital volt meter to verify correct voltages shown on circuit diagrams,

-       -Use oscilloscope to verify correct patterns as shown on circuit diagrams



Step 9:

Progressively replace components in suspect regions. Except for programmed ICs, You can use alternative parts,



Find parts description on parts list in service manual,

Or more detail at



If you are methodical,

make written notes of everything you do,

re-think about what functions you have verified are working, 

re-read the service manual, re-checking your understanding,

you will fix the 940.


This maybe safer than the risk of damage during shipping.









R1. KENWOOD PRODUCED 3 SERVICES BULLETINS which do considerably improve the receiver.

 AGC circuit improvement

TS-940S Signal To Noise Ratio Improvement With Noise Blanker

TS-940S VCO/Carrier To Noise Ratio Improvements


In early March I [Rich Maher] talked to someone at International Radio regarding the reciprocal mixing noise problem with the Kenwood TS-940S.  I had been in the process of installing the fix described in your newsletter (late 1986 issue) and found that it had already been installed on my TS-940S (S/N 7100269).  The factory installation had one problem, the resistors used for R120/R129 were color coded for 900 ohm (close enough to the 1K in the newsletter), but in actuality measured 465 ohms.  Apparently, Kenwood had gotten a bad batch of resistors from some supplier and had not discovered the problem.

At the time you indicated that was the first report you had received of the resistor value problem and recommended that I contact Kenwood. I called them and was told that they had not heard of the problem before.  They also stated that a new fix for the reciprocal mixing noise problem had been developed and was described in a Service Bulletin dated March 2, l987. I requested a copy of the bulletin and have attached a copy of it to this letter for your information.  (See Issue No. 76, Pg. 30 and 31 for Kenwood Service Bulletin No. 917 and schematics pertaining to this subject.)

Since receiving the bulletin from Kenwood, I have installed it on my TS-940S and found it to make a very significant improvement in weak signal handling in the presence of nearby strong signals.  I would recommend highly that anyone experiencing reciprocal mixing problems install the new fix. It should be noted that some of the newer TS-940S have the fix installed.  I was preparing to install the fix on a friends's TS-940 which had a serial number 100 lower than my own and found that the fix had been factory installed. Apparently, more than one manufacturing site is used and serial numbers are given to each in blocks.  Consequently, it is possible for higher serial numbers to be produced at one location without the fix, while another site any have cut in the fix but is using numbers from a lower block.

The quickest way to verify whether the fix has been installed is to check R120 and R129 on the PLL Unit (X50-2020-00).  If these two resistors are 3.3 ohms in value [Editor correction Service Bulleting 917 says 3.3 Kilo-Ohms], the fix is already installed.  Do not depend on the on the absence of C176, C180 or C181 as an indication, as earlier attempts (factory or field) to correct the mixing noise problem may have removed these same capacitors.  The instructions in the bulletin state that when making the modifications to the RF Unit (X44- 1660-00), it is easiest to move C132/C133 to the foil side of the board.  As the component side of the section of the RF Unit containing these two capacitors has been filled with wax, it is definitely not easier.  The factory installation of the fix left C132/C133 on the component side and installed the R154/C193 and R155/C194 series RC networks on the foil side.  This is definitely easier.  As a side note, the installation of the fix took me about 2 hours.  Both the PLL Unit and RF Unit modifications must be completed before the transceiver is usable.  If you install just the PLL Unit modifications and then try the receiver, CW signals will should like raw AC.  Also, to make life simple, do not remove each of the boards above the PLL Unit individually.  The easy way to gain access to the PLL Unit is to remove the top two screws (one on each side) holding the front panel and loosen the bottom tow screws.  This allows the front panel to be tilted forward.  The speaker assembly and all the boards above the PLL unit may then be removed as a unit by removing only 4 screws and tilting this unit towards the front of the TS-940S.  No cables need be removed from the boards above the PLL Unit.

I hope the above information is helpful to you in dealing

with the reciprocal mixing noise problem.  (Thanks, Rich

Maher, WZ4Z, 1117 NW 7th St., Boynton Beach, FL 33435)



In September 2004 PY1NR announced he had discovered:

-on RF board Preamp Q10 and

-on the IF board 2nd balanced mixer Q4,

had been drawn on the circuit boards and mounted in the reverse orientation to that shown in the Kenwood Circuit Diagram.


See PY1NR web site



Initially ZL4AI  found it hard to understand this website and actually what PY1NR had discovered. Starting with the circuit board layouts I tried to draw out the circuit: What I found was that apparently the FETS were mounted with the drain where the source was supposed to be and vice-versa.

As FETs normally allow current flow until the gate has a potential, I wonder if this really makes that much difference.

PY1NR suggest that reversing these transistors will provide 10 dB of gain. But this claim does not appear be based on before and after measurement. It would be useful to have some feedback on whether others have had much improvement by reversing the FETS.
Garey Barrell provides Kenwood’s advice 

========================================================= =============================
From: [] On Behalf Of Garey Barrell
Sent: Wednesday, 9 March 20055:53 a.m.
Subject: Re: [Kenwood] RE:TS-940 What is the correct FET direction?

Jeff -

OK.....  Just in from Kenwood...


Dear Kenwood Customer:

This information pertains to the TS-940S component location.

The circuit designer said the installation of Q10 in the actual TS-940S transceiver is correct. 

The PCB view in the Service Manual is correct too. The schematic is the only section that is in error. The schematic indicates the drain of one FET connected to the source of the second FET. The correct installation is to have the source of one FET connected to the source of the second FET. 

In addition, testing at Kenwood Communications in Long BeachCA showed poor results. Sensitivity can become unstable. The most important point about the Q10 pair is that both FET's must be replaced at the same time (like a matched pair). Replacing only one FET at a time can affect sensitivity.

If you need further assistance, please e-mail us again.


Kenwood Amateur Radio Customer Support 


73, Garey - K4OAH


From: Garey Barrell []

Sent: Friday, 11 March 2005 7:26 a.m.


Subject: Re: [Kenwood] RE:TS-940 What is the correct FET direction?

Jeff -

OK.   I just had a discussion via phone with the Amateur service department at Kenwood.

The Q4 situation is not quite as clear.   The schematic appears to be correct, (sources tied together or push-pull,) and the board layout

drawing appears to be incorrect.   According to a tech in Japan, the FET's in the actual unit are correct.  They have not found any instance  where they were reversed in the actual radio or any 'in-house' docs that could have resulted in such an error.  I guess someone is going to have to open one up and look at the traces!   Looking at the board traces in the component layout, it certainly appears that one FET has the Source and Drain connections reversed if the FETs are installed in the orientation shown.   Perhaps the board traces were changed?  [ZL4AI editor comment: Boards made exactly as shown in the Service Manual]

The guys at Kenwood, both in LA and Japan, are pretty frustrated with the whole mess!   They tried to duplicate the Q10 situation, and found that performance was degraded considerably when the PY1 "correction" was  made.  They also mentioned that replacing one of the pair was not  recommended.  The original circuit used a matched pair and they  recommended replacing them only with a matched pair.  They were unable  to describe the "matching" process, but we surmised they selected for  Idss, and possibly transconductance.

The big question is, these transceivers have been working and meeting  specs for 15+ years, so who cares!?  :-)

73, Garey - K4OAH





-----Original Message-----
From: Traian Belinas []
Sent: Tuesday, 5 April 2005 9:34 p.m.
Subject: Re: Ts-940 All problems SOLVED .. Possibly for you too!


Hi Jeff,


The website is and will be great.

Will look carefully at.


I have two things to say.

First is that the second mixer Q4 JFET is indeed wrong mounted.

Here are attached pictures, you can use them on the website.

The PCB traces are symmetrical, the mixer should be balanced, and as the two

FETs are identical type, the way that they are mounted is obviously wrong.

As I said, I have reversed the Q4 and the improvement exist, but it is not

so great as other had reported (the sensitivity goes improved by 2 to 3 dB)














After reading Kenwood’s Garey’s and Traian’s advice, I turned around only Q4 on the IF board.

The result was a quieter receiver. I do not believe that there was any significant gain increase in the receiver.

I would appreciate (and will post on this page) emails describing others experience regarding this change.

From on 25/4/05

 Hi Dale

I also became interested in the RX mod you mention. Before opening my 940, I decided to first check whether drain and source of the 2SK125 are symmetrical or not. This was easy for me because I own a "dead" 940 RF board as a source of parts for future repair of my rig.

I collected one of the 2SK125s from this board and built a source-grounded test configuration with a 5K resistor connecting drain to +8V. Then, I fed a sawtooth test signal (about -6 to -1V) via 10K into the gate. U(drain) was recorded against U(gate) on a DSO (Tek 468). Thereafter, I repeated the measurement with drain and source exchanged.

I obtained the characteristical FET response curves and these were exactly (!) identical in both configurations. This did not change when the test frequency was increased to 10 MHz. It seems, therefore, that the 2SK125 is symmetrical.

As a consequence, i decided not to correct the layout error in my 940.

Like others, I also believe that there is not much to improve. My 940 has an RX sensitivity of about 0.15 µV (10 dB S/N) on all bands (well, I must say it was worse until I re-aligned the entire RX). The IP3 is +18 dBm (I once replaced the band switching diodes by PIN diodes).

Like others, I often had connector problems after working in the 940 - another reason only to go into this rig when necessary.

Best 73,
Thomas (DF5KF)




-----Original Message-----

From: Traian Belinas []

Sent: Friday, 29 April 2005 2:05 a.m.


Subject: Re: FW: ts-940


Regarding the TS940 2SK125 preamp, yes the FETs have this interesting

feature: for low signal/low freq and/or low DC, they are symmetric. This is why they are used as passive variable low resistance/attenuator/switching  for low signal with rather good results. The things are changing at HF/VHF amplifiers where the interelectrode capacitances became important (do you remember about neutralising a FET preamp?), and these are not quite symmetrical, as the devices are manufactured so that the drain to gate capacitance to be as low as possible for obtaining lower out to in feedback when used for common source applications... So, even if symmetrical, why to use it as for having the greatest unwanted out ot in capacitance/feedback? The gain obtained by inverting the D/S for the TS940 Q10 may be still not high (I don't intend to do it because of the reason explained before), but the engineering feel tell us that something is not ok there... And regarding the second mixer, there it is obvious that it is not ok, even if it works... An counterexample is also the TS950 (both SD and SDX) which use the same Rx preamp as the TS940 with 2SK125 and 2SK520 (they are all FET cascade preamps) but for the 950 it is actually build as shown in the diagram, no drain/source inverting there (maybe the same for their second mixer), so which of them is the best regarding this, the 940 or the 950?!


Please let me know if any other new info about the 940/950.






PY1NR provides feedback and re-endorses previous statements on turning the FETs around




From: [] On Behalf Of John Rotondi

Sent: Friday, 17 March 2006 10:51 p.m.


Subject: [ts-940] FET Reversal Fix Notes


Dear Fellow TS-940 users-


Just a quick post to let others know this information, which you can use as you see fit:


I have now fixed 2 TS-940SATs according to the findings of PY1NR who first detailed the reversal of 2 FETs in the TS-940, based on factory mistakes in the PCB silk-screening. After doing my own radio, I absolutely found a significant increase in received signal levels, with no audible increase in noise floor. I wondered why other users were not rushing to do the fix- and then saw several posts denying the validity of the fix. However, since I did not effectively document this in a scientific manner, I could not effectively offer valid 'proof' of the results.


When I mentioned this to one of my RACES group leaders- who also owns a TS-940- he decided that we would to do the 'fix' to his unit- but this time, we would document the results using a repeatable local test signal. The documented results: after each FET was reversed, we found a 1 S-unit improvement in received signal level using our local test signal in the 20 meter band, for a total of 2 S-units receive gain improvement.


Now, there is much conjecture regarding the dB value of S-units, and other TS-940 users may know what these 2 S-units on the TS-940 meter mean in terms of dB. Generally, from my research, each S-unit may represent 5 or 6 dB of signal, which means the fix has increased receive gain 10 to 12 db. Certainly nothing to sneeze at: being able to give one of the finest receivers made the full scope of RF gain that it was originally intended to have

- at no cost, and without negative repercussions? As the bands wane on the downside of the sunspot cycle, and running only a vertical 10 feet off the ground, I am finding I can use all the noise-free gain available to hear DX!


At any rate, this was my experience, which I humbly offer to the TS-940 user community.


Wishing you all good DX!


John, WA2OOB

Ventura, CA




On Mar 19, 2006, at 1:53 AM, Jeff King wrote:




found your report very very interesting.

Despite all the controversy, some of which I have reported on


I would appreciate if you could you please confirm you turned one FET around, ran signal test, identified improvement 1 S unit and then Turned other FET around and ran signal test, identified improvement 1 S unit?


You know it would be helpful if Kenwood would actually confirm their view of whether the FET in correct position results in too much gain.


hope to work you one day! and


Yours sincerely

Jeff King zl4ai




*From:* John Rotondi, WA2OOB []

*Sent:* Sunday, 19 March 2006 11:56 p.m.


*Subject:* Re: [ts-940] FET Reversal Fix Notes


Hello Jeff!


Very nice to hear from you! Thank you for your interest in my posting  on this topic.


I have seen your excellent website- thank you for providing such  valuable information to the user community. I am still reading through

all the information regarding PIN diodes, and may mod my radio in that area as well.


Just a bit on my background: I am a professional sound engineer, and  have been designing/building/maintaining/operating professional music

recording and TV/Film post production facilities for many years. When  I first did the FET fix to my TS-940, the results were obvious to my

ears. In doing the second radio with my friend, we systematically  followed these steps to document the results relative to an external  repeatable test signal, independent of band conditions, QSB, etc.:


1) Set up the signal source: my MFJ-259 antenna analyzer with whip  antenna, to generate a signal near 14.200 MHz.

2) Set up the TS-940 with a small whip antenna on the work table,  about 4 feet from the test source.

Note that the MFJ-259 RF test signal is fixed in level, so this would  not be a variable in these tests.

3) Tuned the TS-940 to this test signal, peaking the carrier reception  in USB mode, and recording the maximum S-meter reading.

Note that I moved the radio around a bit to ensure that the reading  was stable and repeatable, and not sensitive to relative position.

4) Shut off the test source so as not to deplete the battery while  working on the radio.

5) Reversed the first of the FETs, reinstalled it's PC board,  installed the whip antenna, and positioned the radio as for the original measurement.

6) Powered up the test source, and tuned the TS-940 to it as before.

There was a full 1 S-unit increase in received signal level.

7) Shut off the test source.

8) Reversed the second of the FETs, reinstalled it's PC board, installed the whip antenna, and positioned the radio as for the original measurement.

9) Powered up the test source, and tuned the TS-940 to it as before.

There was now another full 1 S-unit increase in received signal level  over the previous measurement, giving 2 full S-units total over the original base reading.


 While this is probably not as sophisticated as if we would have used a  Communications Monitor (IFR, Marconi, etc.) or other test system

directly coupled to the receiver, with stepped calibrated attenuators, and RF voltmeters coupled to the IF of the TS-940, we felt that it

would be a fast way to have valid empirical data to verify that we had created an improvement, rather than a disability, for the TS-940. BTW,  post fix listening on air clearly showed the significant gain improvement.


In listening today on 10 meters on my own TS-940, I know that this additional gain has brought signals to the readable level that would

otherwise have not been readable. I have also done extensive listening tests with extremely strong local broadcast signals to determine if

this fix has compromised rejection of extraordinarily strong out-of-band signals, or has resulted in compromised receive RF or

audio intermod or other non-linearities resulting from component saturation, imbalance, or interstage distortion- but have heard no

such issues. I will mention that my recently purchased IC-706 MK II  (for mobile use), of more recent design and with some DSP, totally

folds up from same broadcast interference that has no effect on the 940!

The 940 receive audio quality remains exemplary. I have been pleased with the results of the fix, and feel it was worth the effort

to realize the full potential of the original design intent.


I can only think that some amateurs did not have the same results  because perhaps the FETs were not closely enough matched to begin

with, or they had other problems, such as bad solder joints as often  found in these units?


I hope this information is helpful to you! And yes- it would be nice  if Kenwood would enlighten us on these issues- but as the radio is not

a current product, and did quite well even with this 'defect', they  have little motivation to do so.


I will look forward to a QSO with you on HF!


John , WA2OOB






Editors Note:

John has undertaken some very useful measurements and it is very useful to have some measurements.

Measurement outcomes could be more factual if a change in signal to noise ratio was measured by laboratory methods described by the ARRL. For example MDS.

If someone could do an MDS  noise floor test before and after the FET swap, it would be more complete evidence of the assumed improvement.


Garey Barrell sensibly advises:

Even a good test, i.e., s+n / n measurements before and after, or _accurate_  noise figure measurements really wouldn't impress me that much, since a receiver meeting the Kenwood specs would be limited by external noise regardless!


I suspect Garey is correct about the noise floor: This is a less than 0.2 microvolt receiver: Maybe turning the FETS around produces more noise, [which of course lifts the S meter] but does it produce any more signal or better signal to noise ratio?  

If first before an FET swap the S meter was calibrated against a signal generator, then signal strength against independent signal source measured, then an MDS measured, then after the FET swap the s meter was again re-calibrated, then a reading of the independent signal sourceand separately MDS again would show that it was just not an increase in noise.

I wish Kenwood would behave like a responsible manufacturer and explain the technical reasons they do not recommend turning the FETs around.


Have a look at the following links which show how measuring receiver improvement is a difficult undertaking. Even definition of what you are measuring requires some considerable reading and comprehension.






From: Bruce Bennett []
Sent: Monday, 21 September 2009 12:49 p.m.
Subject: Kenwood TS-940S questions


Hi Bruce

Answers are in text below.


Hello Mr. King -

 I was hoping to get some advice from you concerning the 940. I am about to buy a 940SAT from a ham here in Indiana.

…… I am really looking forward to this as I have read for years about the great receiver and superb audio quality for both ham use and SWLing, which will be a big part of my use of the rig.

 I am aware of the FET problem and I wanted to ask you about your opinions about my rig. The serial number is 8,2xx,xxx and I was wondering what you think about whether I should make an effort to change the installation of the two FETs.



I do not think not will make much difference. Cliff of Aavid the well know Kenwood repair expert only found 2 db. He is the only person I know who has used calibrated spectrum analyser to make the measurement. If you look at the function of an FET it is and on off valve like a water tap. It should function much the same no matter which way round it is.

I will soon do some tests now that I have a calibrated signal generator and spectrum analyser to prove what the real difference is.


I am confused about the potential effectiveness of this procedure. I saw something you wrote several years ago in which it appeared that there was little difference in gain after the leads were changed. However, there was also a test engineer who achieved about 5-6 db increase in gain with each FET making an impressive 2 S unit increase in gain.


6dB = 1 S Unit


In this time of poor propagation I imagine that could make a significant difference in being able to copy weak signals, DX or otherwise.


Yes but if you buy or make an outboard before radio pre-amp you could add 20 dB difference much more easily. Good pre-amps are not that expensive.



 I would very much appreciate your comments about this. I should tell you that I was close to buying an Icom 746Pro for about the same amount that I paid for the 940. However, I really like the pre-DSP design of the Kenwood and the way that it operates, the smooth tuning and that wonderful audio. One owner said that the way it sounded for AM shortwave listening was wonderful.


You can add DSP at the outlet end and it helps alot.


That decided it for me. Frankly, I also like the "bigness" of it. It is very enjoyable to sit down to a large rig that looks like it is ready for serious operating.



Bruce Bennett   K6RQR/9

Bloomington, Indiana


p.s. If there is anything else that you think I should be aware of concerning the operation of my 940 I would be glad to hear it.


If I find anything else I will publish it when found.


I did a calibrated minimum discernable signal test on my 20 Million 940 last year.

ON 20 meters

SSB it measured 134 dB

CW with AF tune, peaked, measured 142 dB


My 20 mil 940 has no extra CW filters.


ARRL test on a 5 mill with CW filters engaged gave only 138 dB


Much is made of ARRL tests, but unfortunately they do not specify which filters or mode was engaged for the test. This makes a huge difference to the result.

The ARRL TS-950 and TS-950SDX tests were done different filter combinations. Hence they are incomparable. Yet many Ham buyers (like I once) did place far too much weight on that one test result when making their decision.


If you look at Sherwood Engineering’s site you will see variability between various similar models. I suggest you do a MDS to verify your radio is operating to spec.


Then if there is a problem you can work your way through the receiver flow path to identify where performance drops off. All you need to the test is a signal generator, a step attenuator and an AC volt meter cable of resolving at 2 KHz or less. Be aware that most cheap volt meters will only measure AC to 400 HZ and hance cannot be used.



The suspicion without any measured evidence I hold is that those capacitors are 20 years old now and some may have deteriorated, and replacement may on some radios improve receiver performance by returning it to specification.



Hope this helps

Yours sincerely

Jeff King








Short (Simple) Version of What you need to verify on your TS-940 is the next 5 lines


Leave R 149 and R 150 in their original positions.


With the radio upside down and the front facing you,

R149 is on the right of the pair and should be 150K or 68K.

R150 is on the left of the pair and should be 2.2 meg.


If you want to upgrade (the way Kenwood changed the Service Manual), change R149 to 150K.


Thanks to Dennis WB8WTU who suggested a short Version of what to verify.



If you are interested in the history of this Verification read the rest of this Section 3,

If you are NOT interested in the history of this Verification go to Section 4.




This R149, R150 issue was first discovered in about 1986, and is mentioned in International Radios Bulletins


STOP: This modification was suggested following Kenwood Japan’s advice, that

“The I.F  circuit diagram was correct and the I.F. board was labelled incorrectly.”



Kenwood Japan have now changed their mind and confirmed

“The I.F  circuit diagram is incorrect.

 Communication 3 with Kenwood Japan


Swapping R149 and R150 probably increases sensitivity to similar degree as achieved by just turning the AGC off


Please review KI4NR’s email below advising the (Kenwood intended) correct construction was electrical layout of the AGC identical to the TS-930.


KI4NR advises the rising S meter caused is leaking in C128 and C130. On the Editors radio C128 has been replaced and does not fix the rising S meter.

When time permits C130 [and / or other AGC capacitors] will be replaced and when replacement has been shown to remove the rising S meter this web page will be updated to confirm that. At that time this section of the web page will be restructured to separate communications about IF circuit diagram from the rising S meter problem.





Kenwood appears to have done the following:  Please note there are 2 mistakes.


1. First incorrectly labelled the schematic: (with resistor values around the wrong way)



2.  Then incorrectly labelled the PC board [to correct the mistakes on the schematic] so correct resister values put in circuit.

(For example the position of R150 was labelled as R149 on the PC Board, which resulted in a 150K resistor being put at the R150 position.)





Areas in grey below should be disregarded.




Significantly improves the AGC timing function: After modification:

- You hear weak signals a lot better.

- S meter with AGC SLOW ON becomes quite responsive and lively in the region of S1 to S4 signals. 

--Before S meter did not move much in S1 to S4 region. 

--Before it would take a strong signal to lift the meter suddenly to S4. 

I always wondered why the TS-940 behaved differently to other transceivers [TS-930S, TS950SDX] which react much faster over S1 to S4.




Mike KC8ZNW on 25/4/05 describes this same behaviour to the

Hello everyone I have a question about the movement of my 940's meter. It seems that it
barely moves on some signals which are perfectly readable, other sigs give me 8 or 9 and I
have even heard an occasional 10DB+ movement. My TS830S will give me a 2 or 3 s-unit
increase when I switch the antenna to it for the same signal.
Is this an effect of the sensitivity of the receive section? Or do I have a malfunction? In
addition my VFO exhibits the occasional hiccup on the last 2 digits on small movements of
the knob. I understand this may be caused by solder joints.



Executive Summary of AGC Mod

Its easy to modify a TS-940S to hear better (or as well as) a TS-950SDX.

When fixed, TS-940 really pulls out those very weak signals.

Simply swapping 2 resistors around, will enable this rig to hear as Kenwood designed and intended in Kenwood’s original circuit diagram.


The error is on the IF board:

Kenwood printed labels for R149 and R150 around the wrong way!!!

As assembled by the factory, (the outcome is) in the main signal path, a 2,200 Kilo-Ohm resistor ends up where a 150 Kilo-Ohm Resistor should be.

Being 14 times larger the 2,200 Kilo-Ohm resistor (incorrectly) significantly degrades the signal.

Swap the resistors around and the receiver hearing improves significantly!!!


Kenwood have confirmed the resistors are in the wrong place. Their emails are below:

Probably “these resistors in the wrong place” occurs in every TS-940S produced.




Independent Feedback on how Receiver Improves



From: Ed []
Sent: Monday, 23 May 2005 7:18 a.m.
Subject: TS-940S


Also, your  AGC Timing Correction was applied on my rig (SN 806XXXX) and

worked great! Sure enough, resistors R149  (68K on my equip) and R150 2.2Meg

had been incorrectly installed by the Mfr.  The board markings for those

resistors were wrong.



Ed Alves KD6EU



Full email at:  FeedbacK_3







From: []
Sent: Thursday, 23 June 2005 4:46 p.m.
Subject: agc modification

Hi Jeff,

I was looking through the 940 page and found my feedback to you(regarding the AGC modification with resistors 149 and 150) under the alc setting portion. Im sure I mislabeled my original email to you on this(think I wrote alc). I am having some luck with changing out the 2.2 meg for a 1 meg resistor. Im thinking maybe a little lower value might be worthwhile to test also, like a 6-800k ohm value.

 I know I received another email from you on this but I just wanted to let you know it looked like my feedback was in the wrong spot on your page.



-----Original Message-----
From: []
Sent: Sunday, 12 June 2005 9:43 p.m.
Subject: Re: alc mod


 I thought that mod might be a little better than it was for the alc. It made my radio appear as if it was in fast agc mode all the time. There wasnt a lot of smoothness in the ssb signal that Im used to. Like I said, maybe something like a 1.1 meg is worth considering in there. There isnt much room to solder at all in there. Geez, its tight.





[Editors Note: ZL4AI questions the validity of these observations but has included them to keep feedback information unbiased. Varying the resistors from Kenwoods values was never recommended or intended. With resistors changed around on the Editors 940 AGC slow is still very much slower than AGC fast.]







-----Original Message-----

From: Michael Feryok II []

Sent: Saturday, 9 July 2005 9:57 a.m.


Subject: AGC Mod


Hey Jeff,

Thanks so much for your TS940 page it helped a co-worker and I today to swap the R149-150 resistors for the AGC mod. Very apparent improvement in noise level and gain. I can hear stations that are buried into the noise floor now. Mike, KC8ZNW


From: [] On Behalf Of mikeferyok
Sent: Saturday, 9 July 2005 9:53 a.m.
Subject: [ts-940] AGC mod works great!!

My friend and I  did the R149-R150 swap and it improved the gain and
noise level. Adjusted the VR3 for a proper zero on the meter and
worked  LZ1YE and YV5YMA right after on 17 meters!
Very low noise compared to before the swap. I highly recommend it.
Thanks to everyone here, and Jeff ZL4AI,  Mike KC8ZNW
I'm still debating the transistor gain swap.....????






From: [] On Behalf Of Dale
Sent: Tuesday, 12 July 2005 5:37 a.m.
Subject: [ts-940] Re: AGC mod works great!!

Hello Mike, I'm having both mods done to my 940 now and I hope the out
come is like yours. I'll post after I get my 940 back and let everyone
know how it goes. I have a very late model serial number which is
20700050 and it still had both mistakes in it, so I hope this will
improve on the already great recieve on the 940. 73 and enjoy your
improved TS-940S. Dale, KD5UVV

--- In, "mikeferyok" <mikeferyok@y...> wrote:
> My friend and I  did the R149-R150 swap and it improved the gain and
> noise level. Adjusted the VR3 for a proper zero on the meter and
> worked  LZ1YE and YV5YMA right after on 17 meters!
> Very low noise compared to before the swap. I highly recommend it.
> Thanks to everyone here, and Jeff ZL4AI,  Mike KC8ZNW
>  I'm still debating the transistor gain swap.....????





-----Original Message-----

From: []

Sent: Sunday, 24 July 2005 3:52 p.m.


Subject: [Articles] Improve TS-940 Receiver for Weak Signals


Posted By KB9IV

Well I finally got around to the AGC mod.  What a fantastic also improves CW to my ears.  In addition the AGC mod also seems to improve useable weak sensitivity and decreases distortion.

Forget the "FET reverse" project.  NO difference here, it's not worth the risk and time.

Best 73,

Bill  KB9IV



-----Original Message-----

From: Bill & Becky []

Sent: Sunday, 24 July 2005 4:00 p.m.


Subject: 940 AGC Change


Hello Jeff,


Thank you for the info on the "AGC" correction.  What a fantastic difference here!!

Makes a good 940 a great 940........I can now hear much better not.  I found the FET reversal change useless...........not worth the bother.


Have a Great Day!!



Bill  KB9IV .......Minnesota











-----Original Message-----
From: John []
Sent: Friday, 29 July 2005 4:03 a.m.
Subject: Your 940 observations




I appreciate your efforts on the 940. I have to say the AGC deal is not quite right. I have work on more 940 that I can remember. I have known for years the silk screening of the numbers on the circuit board is wrong. but the resistor placement on the board is correct. also the service manual is wrong on the schematic. The 2.2 Meg ohm resistor is in parallel with C-127 .....the 68K or 150K resistor is in series with C-126 which give you the base line time constant when AGC switch is in the fast postion. This is the CORRECT arrangement. Also if you look at the TS -930 that has the identical AGC circuit this is how it is on that radio too. The reason why you get the AGC rise when the radio has been sitting is the Capacitors are leaky and by swapping the resistors around helps correct that problem. I have had 940's have the rising S meter problem and changing and the caps C128, C130 in the AGC fixed it. This circuit is a Hi impedance type with FET very sensitive and crazy things happen. I have check many, many 940 I have repaired new and old serial numbers and have not found one yet that had the resistors in wrong. Look at the TS-930 schematic to see what I am taking about.


73 John KI4NR


Editors note:

On the TS-930 signal board the equivalent AGC resistors to R150 and R149 are:

R730  2.2M 


R710 68K




-----Original Message-----
From: LPC Wireless, KI4NR []
Sent: Friday, 29 July 2005 5:39 a.m.
Subject: More Info ... Your 940 observations




I forget to add something. When you swap the resistors around. you are putting the 2.2 Meg ohm in series with C-126. this effectively removes the Base line time constant to all AGC positions on the switch including AM even thou the switch does not function there. That why people say the meter is more jumpy. plus the 150 K or 68 K bias the gate of Q23 more heavy and allows the receiver to stay more sensitive to low level signals. if you look at the TS-930 schematic this is the correct circuit in every way and the way Kenwood intended it to work and how the 940 is


One other thing .....on all the older 940   4, 5 and early 6 mil serial number ...the IF board is different. The gain distribution in not the same. All the 940 ... late 6 and newer  had better IF boards. They have more gain TX & RX  the radio are hotter sensitivity wise, better AGC compression. I use a 5 mil TS 940 with a later model 8 Mil IF board in it  .......much , much better !!


Also Kenwood put an  S meter slam mod in those boards. all the older 940 when you shut the radio off, pin the S meter over. The newer boards are fix for that.







From: Jeff King []

Sent: Monday, 1 August 2005 8:01 a.m.

To: ''

Cc: ''

Subject: RE RE: Is your advice Correct about TS-940 R149 and R 150: being in wrong places???


Dear Mr T.Soranaka


Thank you so much for your 2 emails sent in March 2005 [attached as below].


From your advice I understood:


“The I.F  circuit diagram is correct about positions of R150 and R 149 and the I.F. board is labelled incorrectly.”


Because your advice was valuable I recorded this to a small web page:

This has been seen by some TS-940 enthusiasts. It enables one to adjust a TS-940 to operate as (you advised) Kenwood designers really intended.


A very experienced Kenwood repair expert from the USA very strongly suggests your advice may not be correct.  The reasons he states sound correct and are very convincing: Those reasons are summarised below.


With the greatest of respect to Kenwood Corporation and yourself I ask please:

Could you please review your advice and advise again if R150 and R149 on the IF Board should be swapped around to make the TS-940 to operate as Kenwood designers really intended?




30 July 2005:

Abbreviated summary of key points in Emails from KI4NR Kenwood Repair Expert in USA


When R149 and R150 are swapped around the AGC does not function as Kenwood intended.


- The service manual is wrong on the schematic.

- The silk screening of numbers on the circuit board are reversed to the schematic and wrong in relation to the schematic (only).

- But the resulting resistor placement on the board is correct.


I believe the silk screening on the 940 IF board is correct and the IF schematic is wrong.



The 2.2 Meg ohm resistor is in parallel with C-127 .....the 68K or 150K resistor is in series with C-126 which gives the base line time constant when AGC switch is in the fast position. This is the CORRECT arrangement.


The 2.2 meg ohm resistors in both the TS-930 and TS-940 sets up the bias to the FET from the 3.2 volt AGC reference voltage. The 68k or 150k in series with the Cap set up the base time constant. The other FET switch in for slow AGC on SSB and Fixed AGC on AM.


Also if you look at the TS-930 (both schematic and signal board) that has the almost identical AGC circuit.

(R730 2.2M and R710 68K, are the equivalent resistors on the TS-930.) The TS-930 is the correct circuit in every way and the way Kenwood intended "the AGC of the TS-940" to work.



When you swap R149 and R150 around. you are putting the 2.2 Meg ohm in series with C-126.

This effectively removes the Base line time constant to all AGC positions on the switch including AM even thou the switch does not function there. That is why people say the meter is more jumpy. Plus the 150K or 68K bias the gate of Q23 more heavy and allows the receiver to stay more sensitive to low level signals.



73 John KI4NR


LPC Wireless

Phone: 386-774-9921




Mr T.Soranaka I look forward to receiving your advice.


Yours sincerely

Jeff King






-----Original Message-----
From: Customer Service Section []
Sent: Tuesday, 2 August 2005 6:01 p.m.
Subject: Re: RE RE: Is your advice Correct about TS-940 R149 and R 150: being in wrong places???



Dear Mr.King,


Please accept my apologies for having supplied incorrect information.

A very experienced Kenwood repair expert from the USA is right.

The service manual is wrong on the schematic.


Yours sincerely,




Customer Support Center

Kenwood Corporation












From: John Brush []

Sent: Monday, 12 September 2005 2:29 p.m.


Subject: TS-940S R149/R150 More Info




I absolutely agree with the comments made by John KI4NR. A rising S-meter reading is due to a leaking capacitor, and not the incorrect placement of R149/R150. In my case, I did the resistor swap and noticed that the S-meter’s response was the same for both the AGC’s Fast and Slow positions – not good. After undoing the resistor swap, I now had the rising S-meter problem (a problem I didn’t have before the modification). In my case, the problem was resolved by replacing C126, the capacitor that is in series with R150 (2.2M) as shown on the schematic. Per John’s advice, I also plan on replacing C128 and C130.


I must have one of those old IF boards, because my S-meter pegs when I turn the radio off.


73, John (WA3CAS)










Below is Page 92 of the Revised Service Manual

Observe that:

-R 149 and R 150 are mounted between almost the same connections. I.e. between the junction of C127- R148 - C128 - C130 - R156   to->  C126 - Q21 - Q22 

-The difference being that “additional C 126” is between R150 and C126 - Q21 - Q22









Below is page 93 of the Revised Service Manual

You will notice that R 149 is connected between C126 and the junction of C126 - Q21 - Q22. That is R149 has been mounted where R 150 should be.







Does putting the 2.2M ohm resistor where the 150 ohm Resistor should be make a difference. Yes! You bet. Change the 150 ohm back to the direct circuit and the AGC responds very quickly. [similar to the AGC in a TS-930]. AGC could not respond quickly before because it had to wait until C126 charged up.

This is in the heart of the AGC timing section. 

Probably all TS-940s have R149 and R150 in the wrong place.






Change around is easy.

You will need to take the IF Board out.


The difficult part is removing and putting all the connectors off / on the board.

Before starting, draw a diagram of the board showing each connector and position and colour of its wires.

That makes it certain you put the right connectors back in the right  places.

If you don’t draw a diagram you will not know where all the connectors go back. Some two pin connectors could easily go in more than one place. That’s could be disastrous

These colours are not shown in the service manual.


I suggest you put in new resistors, because with longer leads they will slightly easy to hold in place while soldering.







ZL4AI discovered this while searching for the a fault described below 


Only happens in SSB: 

If TS-940 left not running for a couple of days, when you turn it on, 

with the AGC turned off or set in fast position, then the meter needle 

goes to up 25db + 9 (approx). The signal is diminished like RF gai

turned up. Over the next 25 minutes the meter needle slowly moves it way 

back to S0. 

SSB in normal position, and TS-940 turned on this does not happen. 

Needle is initially at s0. 

During the first 25 minutes if you switch between off - fast - normal 

then the needle goes back to zero in less time ... say 20 minutes. 

If TS-940 left for a couple of months, and then turned on same behaviour 

but worse. 

Meter needle goes full scale right in all positions (off - fast - 


It takes longer say 40 minutes for the needle to move to the s0. then 

ts-940 functions as described above. 







After R149 and R150 changed back to positions Kenwood intended in the circuit diagram, the result was: 

-The fault of the rising S meter when cold disappeared. 

- S meter dropped back to S1 on both AGC OFF and AGC SLOW, with no antenna signal.  Needed to adjust VR3 to bring the S Meter to S0. 









T.Soranaka Kenwood Japan was most helpful. You will see in the emails below Kenwood have readily confirmed that these components are around the wrong way. Then in a third communication (above) confirmed they are correctly installed.



Traian Belinas

who diagnosed the problem and really understands these circuits. Traian appears to have amazing skill and after reading the symptoms pointed me to look at R149. From there it became obvious the circuit was not assembled according to the circuit diagram.



Garey Barrell


Who provide some very useful advice on functions of components and explanations how to read the circuit diagrams. 










Not all IF boards are identical.

I installed another IF board installed as per factory spec with R149 and R 150 in their other components position in my TS-940. It did not have the rising S meter problem.  But it was not sensitive to weak signals







-----Original Message-----

From: Customer Service Section [

Sent: Tuesday, 15 March 20057:11 p.m.


Subject: Re: Question about TS-940 R149 and R 150: Appear to be in wrong places!

Dear Customer,

Thank you for your reply. I suppose that currently R149 and R150 are mounted correctly as the screen printing lettering R149 and R150 are reversed. Please confirm actual resistors comparing the circuit diagram. The circuit diagram is correct.

Yours sincerely,




Kenwood Corporation






----- Original Message -----

From: Jeff King

To: 'Customer Service Section'

Cc: ; ; Bill Bailey ; Ken McVie

Sent: Tuesday, March 15, 20051:53 PM

Subject: RE: Question about TS-940 R149 and R 150: Appear to be in wrong places!

Dear T.Soranaka

Thank you for your advice.  

Could you please advise if it would be advisable to swap  R149 with R 150 and vice versa, so the TS-940 functions in accordance with the circuit diagram?

Yours sincerely

Jeff King

-----Original Message-----

From: Customer Service Section []

Sent: Monday, 14 March 200510:29 p.m.



Subject: Re: Question about TS-940 R149 and R 150: Appear to be in wrong places!

Dear Customer,

We are sorry for inconvnience. I have checked with our communication department as to R149 and R150.  Unfortunately reference number of R149 and R150 on the borad are reversed. R150 and R149 are 2.2M and 68K or 150K respectively as shown in the Service Manual.

Yours sincerely,




Kenwood Corporation






----- Original Message -----

From: Jeff King

To: ; ;

Cc: ; ; Bill Bailey ; Ken McVie

Sent: Saturday, March 12, 20056:41 AM

Subject: Question about TS-940 R149 and R 150: Appear to be in wrong places!

Dear Kenwood Customers Services,

I have found that when emailing Kenwood USA about a Kenwood USA product I got redirected to contact a Kenwood representative close to my home location. I am not sure who is best to send this to. So I am sending it onto to all Kenwood contacts.

Thank you for your recent replies.

While trying to find a fault in my TS-940 I have been going over the IF board. It appears to me when the board was made it was marked with the screen printing lettering of R149 being where R150 should be and vice versa. I have followed the board traces both in the Service Manual and on the back of a board, and these resistors both seem to be in the wrong place.

This means:

Specified in First Service Manual: 

R149 68K 

R150 2.2M

Specified in Revised Service Manual:

R149 150K

R150K 2.2M

Resistors as actually installed on my board if you follow the logic of the circuit diagram. 

R149 2.2M

R150 150K

I have two IF boards here and they both have the resistors installed as required by the screen printing and hence on both boards both resistors are reversed. Possibly this is the case for every TS-940 ever made.

I cannot understand how the circuits would function as the designer intended, as the installed resistors are very different to those shown on the schematic diagrams. Could you please advise if my observation is correct, and after later when Kenwood has investigated if it would be advisable to swap  R149 with R 150 and vice versa?

At this time could you please just confirm that the question will be investigated?

I look forward to your reply.

Yours sincerely

Jeff King ZL4AI














This improvement is not fully documented yet. Please send in information.


4.1: Background on how Pin Diodes were discovered to improve radios.


[TenTec] Pin Diodes / Paragon

Chester Alderman
Wed, 17 Sep 1997 17:14:45 -0400


TenTec builds a great amateur radio and obviously to give you a 'million dollar radio' that cost the user five bucks, economics really does have to enter the  picture. PIN diodes have been around for many years, however they were initially invented, designed, manufactured, and sold to be microwave switching devices, obviously because you can not use mechanical relays for internal switching within a microwave RF circuit. It has not been developed as a 'low frequency' device until within the last 20 years or so. 
Dont quote me on this because I've been out of microwave design for too long. A regular diode is a piece of silicon (or germanium) that has a junction. One side of the junction is doped, during mfg process, to have an excess of electrons (+P) and the other side of the diode is doped to have an excess of holes (-N), therefore the term PN junction. And of course a PN junction diode will pass current (a signal) in only one direction. A PIN diode has these (almost) same characteristics (but each side is doped differently than than a standard diode), and in addition, between the +P and the -N sections is a third section that is doped specifically to allow the diode to switch VERY rapidly, and this region is called the Intrensic junction, thus the name PIN. The speed that a PIN diode can switch is basically determined by the width of this Intrensic area, the narrower the I junction, the faster it will switch and therefore the higher the rf frequency that you can pass throught it. Unfortunately, to build a PIN diode that will switch at HF freqs (below 30 MHz), the I junction has to be made wider, or the diode will switch too fast to allow one cycle of an HF signal to pass through it.
What all of this means is that PIN diodes are relatively expensive. A regular PN junction diode (typically a 1N4148 for instance) may cost 5 cent apiece, 
however a 'cheap' true PIN diode will cost between one and three dollars apiece; and thats why you do not see them in very many amateur radios. 
ECONOMICS. (I'm not sure why it took that much verbage to explain, but it did.)
Corsair II's used a regular silicon switching diode, 1N4148 to switch the filters. The Omni 6 does use PIN diodes, but probably because of the above 
mentioned economics, TenTec uses diodes that 'will do the job' verses expensive PIN diodes.
I read Rhode's article on PIN diodes and decided I could improve the IM performance of my Omni 6 (it didn't need it!!), so I bought the expensive Hewlett-
Packard PIN diodes that Rhode stated were the best, and installed them in my Omni 6. Over the past five years using my Omni, chasing DX and participating 
in some serious DX contest, I have yet failed to see where these expensive HP PIN diodes made any substantial improvement.
TenTec runs about 10ma of current through their production PIN diodes, in order to gain the full IM advantage of the HP PIN diodes, you must run 80ma 
through the HP PIN diodes!
So the bottom line is if you replace the PIN or silicon diodes in a rig, you will see (hear) practically no improvement, UNLESS you redesign the circuitry to 
utilize the diodes operating at their optimum design specifications. Probably if you find the filter switching diodes in   your rig are running 'hot' to your touch, it 
probably means that someone has taken the time to change the current running through the switching diodes to really improve the IMD.
At 01:32 PM 9/17/97 -0400, you wrote:
>H. M. 'Puck' Motley W4PM wrote:
>> I have the feeling that the pin diodes in question are a modification
>> suggested in an article by Ulrich Rhode (not sure of the spelling of his
>> name) a few years back concerning 2nd order IMD in modern rigs. One of
>> the rigs mentioned was the Paragon.  The article stated that by replacing
>> the common switching diodes used to switch the receiver front end band
>> pass filters with a certain type of pin diode, 2nd order IMD could be
>> improved. Maybe some of our more technically oriented folks remember this
>> article and can comment in greater detail. This is all I remember so if
>> you have additional questions don't ask me!
>Thanks, Puck. I was certain it was something Rohde said, just wasn't
>quite sure when or what the exact reason was. I just spoke to Ten Tec
>about this, and they actually said they had tested different types of
>diodes to switch the Paragon's receiver filters, and settled on regular
>switching diodes because there wasn't much difference with other types.
>So, I guess replacing the receiver filter switching diodes with PIN or
>other (hot carrier, etc.) types is probably a mod that some users have
>done themselves. At least I know for sure it's not a factory
>Is there anyone out there who knows this for sure? Has anyone done the
>aforementioned mod? I know one fellow recently mentioned in a message
>that a rig he had for sale had the mod. Now I'll go search for the Rohde
>article. 8^)
>73, KE3KR







4.2 RadCom Technical Topics explains what Pin Diodes were supposed to achieve.






TT FEBRUARY 1993 REPORTED briefly an important article by Dr Ulrich Noble, KA2WEU/DJ2LR, which was published simultaneously in English and German QST and CQ‑DL November 1992) on "Recent advances in shortwave receiver design". He subsequently published a series of three articles (QST May, June and July 1994)on Key components of modem receiver design, and a recent follow‑up Key components of modern receiver design: a second look" (QST, December 1994). In these articles he stressed that for receivers intended to have a very wide dynamic range, the intermodulation distortion that arises from the use of unsuitable RF switching and tuning diodes imposes an important limitation. He has recommended the use (or substitution) of such special‑purpose RF diodes as the Hewlett‑Packard HP5082‑3081 PIN diodes.


Dr Rohde's articles encouraged Tom Thomson, WOIVJ, to investigate how bad in practice are the more distortion‑prone switching diodes and how good are those designed for low distortion ('Exploring intermodulation distortion in RF switching and tuning diodes', QST, December 1994). He carried out laboratory tests on four types of diodes: The IN4153 generic PN switching diode: the Motorola MPN 3700 PIN diode intended for RF switching; the BAT‑17 Siemens PIN switching diode; and the low‑cost 1N4007 which is a generic 1 kV‑PIV rectifier diode with a PIN structure but not intended for RF switching


He has tabulated results in terms of diode switch insertion loss (dB) at 1O MHz with 0, 5, 10 and 20mA bias currents; and similarly the second‑ and third‑order intercept points (IP2, 1P3 and dBm). He draws the following conclusions: "RF‑specified PIN diodes are the devices of choice for low‑distortion switching at HF and above, for band pass filter selection and C switching in a narrow‑band pre‑selector. Although the presence of a PIN structure in the 1 N4007 makes it seem attractive as a low‑cost alternative to RF‑specified PIN diodes, its insertion‑loss performance When unbiased and reverse‑biased ‑ and its IMD performance when unbiased ‑ is demonstratively interior to RF‑specified PIN diodes.


He adds: 'The manually switched and tuned front‑end filters of the 1960s and 1970s had much to offer in terms of second‑order IMD, but we need not retrogress to those techniques to achieve improved 1P2 and 1P3 performance today. More attention paid to front‑end filtering in general can produce the improvement we need."


Dr Rohde in commenting on W0IVJ's finding, notes that many amateurs had reported difficulty in obtaining HP5062‑3081 diodes. He recognises that even with the Motorola MPN3700 with a US price Of less than £11 replacing all 20‑plus filter‑switching diodes can be expensive. Nevertheless he recommends changing all the diodes between the antenna and the first mixer, which includes the diodes on both sides of the band pass filters of a transceiver but not the transmit/ receive switching diodes which typically are already high‑quality PIN types. He also adds some notes on Japanese switching diodes which might be used to replace the 'bad' diodes seen in the past".




























THE ITEM 'RF Switching Diodes Controversy 'TT, July 1995, included G4HRY's criticism of the advice given by Dr Ulrich Roinde, DJ2LR/KA2WIEU/4, in his excellent articles in OST that the second order IMD performance of several popular amateur HF transceivers could be improved by judicious substitution of PIN diodes, such as the Hewlett Packard HP5082‑3081 , specifically intended as RF switching diodes. I pointed out that the criticisms were based solely on RF losses and that G4HRY had not made any IMD measurements. I added that I was sure that Dr Rohde could provide a convincing reply. However, in view of his experiences with GKPT's Omnii VI 1 felt it would be right to include G4HRY's view that it was unwise to put unquestioning faith in published articles including even those in OST and RadCom.


The detailed measurements provided by G3SBI (TT, November) and those published by DB1 NV in VHF Communications showed clearly that there is a wide difference between different diodes used for RF switching both in insertion loss and in IMD performance and that IMD is significantly affected both by the forward current through the diode and by frequency.


As a result of an unfortunate delay, the November item was written before I received a fax sent by Dr Rohde on July 18th. This, in a slightly abridged form, reads: I feel really concerned and sorry about G4KPT and the results of his experiments. As a matter of record, I would like to point out that intentionally 

 1 had not changed any of the diodes myself. but had the authorised service departments of AES, Milwaukee replace the diodes in the Yaesu FT890; ICOM changed the diodes in two IC765s; and Kenwood made the same changes in a TS50. The itemised ICOM repair bill shows 0.12uV for 12dB SINAD, I also had the other companies involved validate that following the diode changes, the receivers were within specifications.


'This validates my statement that this was a repeatable effort and the changes were not done at the expense of performance in any respect. It is also a matter of record that the HP5082‑3081 diodes were used in the production of the Collins KWM380, one of which I still own and whose noise figure is on target with 0.3uV without a pre‑amplifier and whose 2nd order IMD is superior to other diode applications. This should remove any doubts as to the correctness of my OST article.


1 have had no experience in modifying an Omni VI nor did I do any measurements or modifications with it. The ARRL edited in the Omni VI because it is a popular US‑made transceiver and there had been some discussion as to whether or not the European version had different diodes or relays. Before fingers are pointed at specific diodes, I would like to examine the circuit diagram because there can be no need to change all 40 diodes. As an experiment, 1 may want to supply one set of more modem diodes.


"Everyone who has contacted me as the result of the OST articles had been advised not to use the HP 3081 (for reasons of cost and availability) but rather to use a Siemens BAR17 diode or M1204 diode, which is available through ICOM dealers/repair centres. Those diodes are much less expensive and more readily available.


"To the best of my knowledge, the companies who changed the diodes in the equipments involved did not change the diode bias. It is questionable why any one should wish to change diodes in the IF section; similarly diodes in the transmit / receive switches should not be touched.


"Finally, there is no question that relays provide the best of all worlds as far as IMD characteristics are concerned, but not necessarily the best solution in terms of space and costs. 1 have just tested a soon‑to‑be‑re leased transceiver which uses PIN‑type di odes and exhibits superb IMD characteristic while maintaining a good noise figure.


"As to multi‑tone functionality, once 2n and 3rd order IMD tests have been done, on can predict the higher‑order IMD effects, especially since they are based on diode characteristics and this type of test is a legitimate test to evaluate receivers.


"Hopefully, your readers will not deduct from this experiment that QST or other reputable magazines publish articles which are technically incorrect."


In a subsequent letter, dated September 19, 1995, Dr Rohde confirms that he has run into a lot of people who have modified their RF switching diodes and have been extremely happy with the results. Further, after refining his test set‑up he finds the improvement is now slightly more dramatic than outlined in his QST article.


In regard to Dr Rohde's endorsement of the technical accuracy of articles, I would enter a caveat. While most writers strive lo complete accuracy. the mechanics an Murphy's Law of publication make it difficult to avoid some errors, particularly in column produced to a tight deadline. Many years ago I stressed that I regard Technical Topics as forum for new ideas, not all of which are likely to prove repeatable or even strictly accurate No guarantees can be given on experimental ideas still under development!  I welcome comments from sceptical readers or those spotting printing errors etc. Fortunately, there is good evidence that the vast majority of 7T items do work as intended, and often provide useful additions to amateur lore!





Intermodulation properties of switching diodes, by Dr. Ing. Jochen Jirmann, DB1NV

ZL4AI was contacted by a neighbouring ham, (known for many years). Peter Johnson ZL4LV. peter.Johnson@Paradise.Net.Nz

Peter designed and developed from scratch in the early 1970s an HF transceiver. (Actually it is still under development and may soon have BA479s installed.) The local Branch of the New Zealand Amateur Radio Transmitters Association under Peter’s guidance sold this as a kitset. Peters design was the first use of diodes for band switching. Peter published this technique in English Radio magazines the early 1970s and thereafter the first commercial transceivers appeared with diodes switching bands. As the inventor of the concept Peter has collected articles on diode switches, and provided the following.
















 for tHE SUBSEQUENT SECTIONS click here:
















































ts940 battery

ts950 sd

ts940 cold solder

ts950sd fan

ts950 serial numbers

TS-940S HF Transceiver

Feb 1986               

kenwood serial number

LO Local (TS-940 : Remote control of SLOPE, TUNE, VBT)       

1.8khz ssb filter ts850s        

Resistors in wrong place in the Kenwood Ts-940             

950SDX + diode mod         

rtty key input ts570

buy mI204 pin diode

ten tec omni vi transverters  

difference between ts930 and ts940   

TS-830 drift          


ts-930s unlocking

GU74 hybrid


gu74 plate voltage               

TS50 data operation rtty interface schematics   

how to replace s-meter lamps in the Kenwood ts-940      

ts50 psk interface schematics

ts850 color schematics         

kenwood 940        

ts850 repair rx attenuation   

kenwood amateur ts-940 antenna tuner repairs 

ts870 number serial              

kenwood amplifier heats up and blows fuse      

ts930s no display no receive

kenwood ts 940    


kenwood ts 940s late serial number    


kenwood ts-930 display vfo

ts940 lithium battery            

kenwood ts-940 tuning upgrade         

TS940 VCO Loop

kenwood ts940     

ts940s problem solved         

ts870 number serial

how to replace s-meter lamps in the Kenwood ts-940