pshz - The config cable connector varies between installations. But if you cannot work this out, you will have no chance of doing anything with it afterwards, except possibly to totally mess up your autopilot. You need to find an electronics engineer who can read the manuals and find out what is wrong.
Hiflyer - I think you must be right in the case of a DG (not HSI) installation. There is no Course Pointer, so the autopilot must be getting its "heading" signal from somewhere, and the heading bug is the only option. Then it uses the deviation bar value to adjust the heading.
I have a KLN94 too, and the autopilot requires the HSI course pointer to be correctly set - equal to the current GPS track. The autopilot uses the CP as the "heading" to fly.
The KAP-140 Pilot's Guide shows several possible installations.
Actualy, I did not try with VOR. I did with KLN-94. And result was the same. KLN-94 does not require setting heading bug for sure.
Are you shure that you set heading bug and vor indicator on the same value when flying nav? In our c172r it is like that. The autopilot takes heading as course to fly and vor indicatior deflection to compensate the deviation.
If heading bug and vor indicator deviate from each other, the plane eventually flies a constant offset to the desired track.
Thank you, Peter. I found also this page, but I'm a little confused with A and B schemes. From other forums I found, that KAP140 on 172 is using US NATO and these connectors more looks like microphone or headsets jacks.
Anyway, I'm not a good technician, but I think, that it's not a problem to order such cable from any electric workshop. What should I know is exact connector type - how can I understand, wich one should I use in my airplane?
Page 57 in the IM.
yes, and I found information regarding terminal in it, but the cable...
Everything you need is in the Installation Manual - see link I posted.
I would be happy to try to fix it using laptop. But I have no cable required for connnection :( It was started after electrical trim failure. I have restarted avionics in flight (probably, it was my fault?) and on the next flight it happened.
I know only Konekrohonen in Finland, but they are no more in Malmi and they can't help unfourtunetely.
So, I'll try to contact Avionik Straubing and also will try to find someone to do tests.
There are "avionics shops" everywhere, but most are not electronics engineers and unable to troubleshoot from first principles with e.g. a scope. Most just follow wiring diagrams, and have specialisations according to their company capabilities.
In this case I would look for a good electronics engineer. The former USSR is full of excellent hardware and software engineers. They had (have) an excellent scientific/technical education system - unlike the West where things have been watered down so nobody sees themselves as a failure 
The kind of 1970s "technology" involved in autopilots will be trivial to one of these.
Once the problem item has been identified, you can send it off to Straubing, or better still to a specialised firm in the USA.
The hardest job will be to find the wires. Ideally you need the wiring diagram for the aircraft. But if you can get installation manuals for e.g. the DG then you can check signals right on the back of that. What DG (and what aircraft) is it? Can you post some pictures?
It is possible that the problem may be fixable purely through the terminal (laptop) connection. But then I have to ask how did it start in the first place? If it is a "digitally" set parameter which is out, how did it get that way? Did somebody change the autopilot computer "quietly" without recalibrating it?
It sounds like there is no significant ROL error. Is the HDG error totally constant? The heading bug output is trivial to check with a voltmeter.
