Roender's FC-100 prototype and builder's thread

[AndriyOL]

Are you offering to connect the tester to FC100 only? 60-100W bulb will be fine?

[AndriyOL]

I tried with 60W bulb to FC100 only, it's still shorting but leds on the apm were shined

[Rudi_Ratlos]

I have sent you a PM.

Best regards - Rudi_Ratlos

[AndrewT]

Originally Posted by AndriyOL

Hi,
Can anyone help, what could be a reason of fc100 amp shorting after powering it up with 40W light bulb tester?
Used it with 30VDC main PSU and 37VDC fronted PSU. They are work fine alone.

by "shorting" do you mean the bulb lights up?

One cannot check a high bias amplifier using the bulb tester method.
All ClassAB amplifiers can be checked prior to setting the bias current. Once all the wiring has been checked and the voltages at the amplifier have measured as "normal" then the bulb is removed and the bias is then set with the amplifier direct on line.

The FC is no different.
You set the bias to "OFF" by shorting each of the diodes that set the bias voltage. That turns the FC into a an unbiased ClassAB amplifier. NOW everyone can test the wiring and the voltages for "normal".

Now remove one diode short at a time and see the effect on the bias voltage.

[Rudi_Ratlos]

Andrew,

Andriy owns a very "individual" version of Mihai's FC-100, which he calls "FC-100 MKII", and which he has layouted and etched by his own.

From the pictures of the PCB that he sent me, I can see that he has done several modifications to the frontend, he is using Toshiba 2SC5200
and 2SA1943 (?) as OP-transistors and "simulates" the TT-diodes of the NJLs by the base-emitter junctions of some TO126-case transistors
that he screwed on top each of the Toshibas.

The problem he has, is, that the backend rail-voltages drop to very low voltages (about +/-2VDC) after powering-on the backend.

I find it very hard to assist with his build, though it is interesting, because his version differs in many aspects from Mihai's design and evaluations.
I suspect a problem in the TEMPCO / BIAS adjustment.

Maybe Andriy tells in detail, what he changed compared to the original FC-100 design.

Best regards - Rudi_Ratlos

[AndriyOL]

Originally Posted by AndrewT

by "shorting" do you mean the bulb lights up?

One cannot check a high bias amplifier using the bulb tester method.
All ClassAB amplifiers can be checked prior to setting the bias current. Once all the wiring has been checked and the voltages at the amplifier have measured as "normal" then the bulb is removed and the bias is then set with the amplifier direct on line.

The FC is no different.
You set the bias to "OFF" by shorting each of the diodes that set the bias voltage. That turns the FC into a an unbiased ClassAB amplifier. NOW everyone can test the wiring and the voltages for "normal".

Now remove one diode short at a time and see the effect on the bias voltage.

Yes, lights up.
I shorted one of 5 junctions and noticed a little higher voltage on FC with tester (2.3VDC, with 5 transistor junctions was 1.5VDC) and some hum from speaker. Shall I try to short them all and measure voltages?

Originally Posted by Rudi_Ratlos

Andriy owns a very "individual" version of Mihai's FC-100, which he calls "FC-100 MKII", and which he has layouted and etched by his own.
Maybe Andriy tells in detail, what he changed compared to the original FC-100 design.
Best regards - Rudi_Ratlos

I used Toshiba TTC5200\TTA1943 devices. It's Mihai's design also, but he didn't published the layout for this one. RMI-FC100 MKII, single ended folded cascode VAS

[AndriyOL]

When only 3 "TT diodes" left everything seems working fine. The bulb glows once during capacitors charge. Output dc offset is not stable and rising starting from 3mV and higher. Perhaps this is bacause jfets are not maintaining the same temperature.
So, bias for this amplifier should be set not like on a schematic or I've used not fully suitable base to emitter transistor junctions?

[atupi]

I give up in matching Q1,Q3 (15034/15035) because it seems impossible to find some overlaping hfe values.

I've asked few more fellow DIY-ers to measure their transistors in order to match with mine and the results are discouraging.

Did anyone of you managed to found mathed pairs?
What will the results of using unmatched pairs ?

Thanks,
Adrian

[roender]

Originally Posted by atupi

I give up in matching Q1,Q3 (15034/15035) because it seems impossible to find some overlaping hfe values.

I've asked few more fellow DIY-ers to measure their transistors in order to match with mine and the results are discouraging.

Did anyone of you managed to found mathed pairs?
What will the results of using unmatched pairs ?

Thanks,
Adrian

Hi Adrian,

The unmatched q1 and q3 will increase even order distortions.
Go ahead with what you got and be happy. Life is short, listen to the music and don't worry much about not perfect parts.

Regards,
Mihai

[atupi]

Multumesc pentru incurajari Mihai !

Thanks for your support Mihai !