Collins KWM2-A AGC slow by simple modification?

[kostas_sv3ora]

I have a Collins KWM2-A HF transceiver for service (early model, schematic attached) and I wonder if the AGC can be made to have slower decay by a simple modification.
Collins has released some modifications and included them in the later models. However all of these require serious work to be done.
I am not exactly familiar of how the AGC is set, but I suspect I can add/alter a capacitor/resistor combination somewhere and make the AGC having slower response, especially in the decay.

Here is the schematic
http://www.ccae.tm6cca.com/DOWN/KWM-...0schematic.pdf

[G3PIJpeter]

Originally Posted by kostas_sv3ora

... I suspect I can add/alter a capacitor/resistor combination somewhere and make the AGC having slower response, especially in the decay.

What should the rise / decay times be (according to the handbook)? If the AGC action is now too short, it may be that the voltage on the AVC line is decaying too quickly due to a leaky shunt capacitor.

Peter G3PIJ

[kostas_sv3ora]

Hm...
According to these mods https://collinsradio.org/archives/se...M2-2A_SB_8.pdf that were included in the latest production radios (but my early one mine does not have), the problem seems to be known.
So I do not think there is a leakage, based only on this information though.
I think I won't have a good chance to just replace a capacitor and see good results, I think I have to do at least mods A and C, to achieve slow agc without agc overshoot....

[G3PIJpeter]

Looking at the KWM 1970 Service Bulletin, I think that the Part A modification most closely matches your wish for slower AGC delay - copy attached. The Part A modification is to "... provide dual time-constant action", which seems to be your aim. I am not sure that also carrying out B and C will add much to A - although overall performance will depend on your actual QTH circumstances.

I have spent some time looking around the AGC line on the main circuit diagram and cannot see where adding extra capacitance will have the effect you desire. I think that you have to carry out the instructions contained in the Service Bulletin to sort out an early design fault (shortcoming) in these early models.

Good luck!

Peter

[kostas_sv3ora]

Peter,
I did both modifications A and C as per the document you described.
It did fix the AGC overshoot and provided somehow more slow AGC decay.
The radio is now more proper to listen to, but I would like an even slower AGC decay. The AGC now if like the "medium" AGC setting in other transceivers I have used.

Any idea of what can I do about it?

[G3PIJpeter]

You have carried out modifications A and C. The Service Bulletin SB 8 (from https://collinsradio.org/archives/se...M2-2A_SB_8.pdf - post No. 3 above) shows the revised circuit for Modification A on page 2 and Modification C on page 5.

The handbook circuit diagram shows that the AVC line is connected to each of the control grids of the 1st IF amplifier V17 and the 2nd IF amplifier V38 via a 1 M-ohm resistor (R74, R78). The AVC line is also connected to the RF amplifier V7 via a 100 k-ohm resistor (R34) decoupled to earth by a 1000 pF capacitor (C41).

The muting terminals J20 and J21 are shorted together when the receiver is operative, which connects C276 0.05 mFd (added in Part C mod) to earth. The unmodified AVC rectifier feeds the AVC line via R180 (150 k) shunted by C265 (0.001 mFd) and R82 (4.7k) - modification A shunts the AVC line to earth via C92 (0.01 mFd).

Therefore, the total capacitance shunting the AVC line to earth is C41 + C276 + C92 i.e. 1000 pF + 0.05 mFd + 0.01 mFd = 0.061 mFd. If you want to make the AVC decay more slowly, then the above suggests that you could simply add extra capacitance from the AVC line to earth - increase the value of either C276 or C92 or add an extra capacitor at some convenient point. Add a further 0.05 mFd at the start and then increase or decrease until you achieve the desired effect.

The AVC line in many receivers has several series resistances, each shunted to earth by a capacitor. The KWM AVC is relatively simple, so just adding more C to the AVC line is the way to go.

Peter
G3PIJ