RMI-FC100, a single stage audio power amplifier

[Rudi_Ratlos]

Andrew: the BAT46 - Schottky diode has a negative temperature-coefficient as well and I recommend to fasten it on the heatsink.

When I built my prototype (having problems with the Vf of the internal TT-diodes), I did the following test:

I could not adjust the quiescent current / BIAS with 5 internal TT-diodes being active.
I could do the quiescent / BIAS adjustment with only 4 diodes being active.
I played music very loud for some time and then stopped playing. I could see that it takes a second or so, until the voltage that dropped
across 2 emitter resistors settled to the initial value that I adjusted during quiescent / BIAS adjustment.

I did the same test with the Schottky diode, bypassing the "5th" internal TT-diode.
After playing music for some time and then stopping it, the voltage that I measured across the 2 emitter diodes,
dropped to the initial value within the "blink of an eye".

This was the proof for me that the Schottky diode enhances the temperature compensation and is very near to the
originally designed "5 TT-diodes - solution".

In my current build, I use 5 internal TT-diodes and everything works as designed.

Best regards - Rudi

[DRZ1]

After some tests it works well Andy.(and i don't have many other simple solution).

In the spec sheet of the Schottky diode i use (ONSemi MBRS family), the temp curve shows a more sensitive variation than the thermaltrack ones.

[Telstar]

There are any pcb (still) available?

[Keith Taylor]

I have finally sorted out my problem mentioned in post 1244 (more than 2mA through the diode string). I think I must be getting too old and dopey for electronic projects as I failed to notice that LED D8 was not glowing! The reason was that it was installed back to front resulting in Q17 sourcing 5.9 mA instead of 1 mA.

For anyone experiencing biasing problems it would be worth checking the voltage drop across R17 in the current mirror. It should be around 0.33V when 2 mA is passing through the TT diodes.

Keith

[AndriyOL]

Can anybody advise me, why my fronted PSU with 39V zener installed have on outputs 39 positive and 39.5 negative VDC? I have matched mje devices to 18% between each other.
I've tried different transformers with 34VAC and 38VAC on input of the PSU. According to published bom, zener has to be 36V type for 39V PSU (I want to have 42V PSU) or I understood something wrong?

[AndrewT]

measure the voltage across the zener diodes.

Then add on the two measurements you get for the series connected semis.
This total should be your output voltage.

I don't have the sch in front of me, Is this correct?

[Rudi_Ratlos]

AndriyOL: "standard" ZENER diodes usually have a variation of 10%.
No reason to get worried.

Best regards - Rudi_Ratlos

[AndriyOL]

Originally Posted by AndrewT

measure the voltage across the zener diodes.

Then add on the two measurements you get for the series connected semis.
This total should be your output voltage.

I don't have the sch in front of me, Is this correct?

37 and 37.5 VDC on zeners and 39-39.5VDC on outputs.
Yes Andrew, this is correct, according to designer's ammendment, zener should be 36V type. Don't know that zeners shall be matched likewise.

Originally Posted by Rudi_Ratlos

AndriyOL: "standard" ZENER diodes usually have a variation of 10%.
No reason to get worried.

I understand, thanks! Can they be matched?

[AndrewT]

These Zeners are operating at low current. Well under the 5% to 10% of current required for maximum power dissipation.
This results in lower than expected voltage across the Zeners. and also results in a bigger variation in voltage.

Your choice is take what you first pick up or select what you want/need from your stock.

400mW and 500mW Zeners cost about a cent/penny each when you buy 100.
Then you can select just about any value you want from the 100 on the band.

I strap a resistor to one the DMM probes.
Then apply +ve to one probe and -ve to the other probe.
Now go down the strip and read off the voltage at the current the resistor allows to flow. Write it on the paper strip.

[roender]

Originally Posted by AndrewT

These Zeners are operating at low current. Well under the 5% to 10% of current required for maximum power dissipation.
This results in lower than expected voltage across the Zeners. and also results in a bigger variation in voltage.

Any Zeners diode pakaged in less than 1W case and with Vr > 5v will be fully on reverse conduction for Ir >= 0.2...0.5 mA. There is no need for higher current trough a reverse polarized PN (or NP) silicone junction in order to have a true avalanche conduction and far away from the noisy corner of the reverse I/V curve