Re: Transmitter gain calibration

Mel Farrer, K6KBE <farrermesa@...>

Hello David,

As a reference on the junction of the RF transistor and the heat sink, I speak from personal manufacturing experience.  There are two issues.

First, the flatness of the mating heat sink surface.  Flatness can be measured with a micrometer or by using a straightedge and lay down a slick of heat sink compound and look for areas of white, none should be visible.

Second, is the flatness of the transistor base.  If one wishes to know if the transistor is seated properly with maximum contact with the heat sink, before soldering it down, one must put a .005 mil slick of heat grease compound on the two hole mounted transistor base plate and finger tighten the screws down, then take the screws out, remove the transistor and look at the morue pattern.  If the pattern is spectacled, that is, has a uniform pattern across the surface with lots of metal showing through, the surface tension was adequate for the test.  If areas that show white grease that is not compressed, the flatness of the base plate is a problem.  Address that problem now.  Then when final assembly is made, only go another 1/4 turn----no more.  On Stud mounted transistors, the ring around the stud must be uniform in display. 

There is a manufacturing MIL spec on this, but memory fails.

I hope this helps a little.

Mel, K6KBE

On Mon, Jun 8, 2020 at 11:19 AM David Cutter via <> wrote:

It's a lot easier to achieve correct and even pressure using a spring clip that presses on the body of the device, rather than one end.  Several shapes available.  I can understand the desire to use a screw that you tighten to the limit of the thread.  It's the same syndrome as turning off a tap with all your strength: it's simply unnecessary and can lead to premature failure. Sorry for the drift.

David G3UNA/G6CP

On 08 June 2020 at 15:12 Ray Maxfield <wa6vab@...> wrote:

around 0,8Nm for T0-220 …….   Let me see What range my torque wrench is Calibrated in.

I will check the Flatness of the Devices, and if needed.. File it Flat( or Lap the surface) and use Thermal compound.

Non Flat devices have been around for Years.   They were Found in Motorola Power devices

In the 70’s, I was repairing RF Amplifiers Then. ……

Ray  WA6VAB. 


Sent: Monday, June 8, 2020 2:11 AM
To:; Ray Maxfield
Subject: Re: [Elecraft-K3] Transmitter gain calibration


I would caution against the Hefty Fist technique.  This can distort the tab and/or the panel it is attached to so that part of the tab is no longer in contact.  I recall a torque setting for optimum contact pressure and it's not very much.  Even for very large devices (as big as your hand) it's only 5Nm from memory. 

Have a read through this document, it recommends around 0,8Nm for T0-220


David G3UNA/G6CP



On 08 June 2020 at 05:57 Ray Maxfield <wa6vab@...> wrote:

Hi Robert………… Thanks for the Feed Back. 

This seems to be a Somewhat of a failure in the K3.

The Sticky Pad, is suppose to help transfer Heat from the Fet to the Bottom Cover.

They Work, but are NOT as Good a the Old Original Thermal Compound.

I went to the Computer Parts store and purchase a small syringe of Thermal compound

And applied some to Each Tab of the Fets, and re-assemblied the bottom cover.

Making Sure to get the Star Lockwashers on each…. Tighten with a Hefty Fist.

I would order 2 Each, One to keep as a Spare! 


Oh Yes, we use to call the thermal compound, FBI Finger print ink !

It will get on Everything if your not Careful.


Great Job…. So far.

When you get the Fet installed, Run the Gain Caliberation Test,

So the radio will Reset to the Correct Power output…..

Use dummy Load if at all possible.   ……….. VAB 




Sent from Mail for Windows 10


From: Robert W5AJ
Sent: Sunday, June 7, 2020 8:19 PM
Subject: Re: [Elecraft-K3] Transmitter gain calibration


Ray,   thanks what a GREAT test method – a testing tool that is universal and easy to understand!


Checked the K3 schematics and figured that the single driver in 10W stage was good or nothing would happen.

Of the two finals in 10W section, appear to be in push/pull of some type, one is COLD!


Should have part / repair done by FD, so this is all good.


Noted the FETs had sticky pad between them and bottom cover

wonder if that needs to be replaced every five years or so type thing….

I plan to use same grease used on CPU to fan on the repair.


73 W5AJ



Midland, Texas


From: [] On Behalf Of Ray Maxfield
Sent: Friday, June 05, 2020 11:06 PM
Subject: Re: [Elecraft-K3] Transmitter gain calibration


Hi Robert…….. I would look for a bad Driver Fet Transistor

There are 3 Fets that are Heat sinked to the bottom of the radio.

Check your manual for the Correct location……

They are connected to the bottom plate for use as a Heat sink

Using Phillips head screws……..

Place the radio so it is raised up using the Front Bail.

Place your fingers on the 3 Screw that are holding the 3 Fets.

Key the PTT so the XT is on.   All 3 should get Warm, Not HOT

Or Cold.   Hot= Shorted, Cold Open,  Warm Operation normal.

I found one of my Fets was Cold……  Give it a try……… Ray WA6VAB  




From: Robert W5AJ
Sent: Friday, June 5, 2020 6:36 PM
Subject: [Elecraft-K3] Transmitter gain calibration


K3/100 -> noted that K3 was indicating 50watts output when the RF power was set at 100watts.

opened the K3 utility ( and started the Calibration test

5 watt test, failed first band - 160 - gets to about 4 watts out and fails.

50 watt test, everything going fine until six meters, fails - couldn't get 50 watts out....

anything that I can tweak or check local or is the party over....

73 W5AJ










Join to automatically receive all group messages.