Dear Ricardo,
the two amps are totally different. The dual driver in the output stage is to deliver the needed base current for the 8 pairs of power BJTs.
The 3rd change:
You are using lateral MOSFET in your output stage. Better you have a look at the development of these projects:
You can find links in the first post. The project E2 was also tested to run with +/-62V supply with this results:
Have fun, Toni
the two amps are totally different. The dual driver in the output stage is to deliver the needed base current for the 8 pairs of power BJTs.
The 3rd change:
- the lower 10R gate stopper betters the overal gain/phase margin
- the 4.7R emitter are for better current sharing and for thermal stability
- mostly a result from many automated simulations to have optimum THD vs stability vs clipping.
You are using lateral MOSFET in your output stage. Better you have a look at the development of these projects:
E1) SA2016 - L-MOSFET (70W@8R, 140W@4R) - 3 matched pairs Exicon ECX10N20 and ECX10P20
E2) SA2016 - L-MOSFET (100W@8R, 200W@4R) - 3 matched pairs double die Exicon ECW20N20 and ECW20P20
You can find links in the first post. The project E2 was also tested to run with +/-62V supply with this results:
SA2016: using lateral mosfets ECW20N20 ECW20P20
Test conditions: +/-62V regulated; RgateN 270R; RgateP 180R; overcurrent protection in situ
THD20k+N@200W@8R: 0.0052% 🙂
This measurement was taken with sagging rail voltages from power supply (~ +/-58V) just before power supply current limit:
THD20k+N@320W@4R: 0.008% 🙂
Have fun, Toni
Thank you for your prompt reply.
I will look to SA2016 ... does it uses two pairs of drivers ?
Best
Ricardo
I will look to SA2016 ... does it uses two pairs of drivers ?
Best
Ricardo
Found the schematic....
You are using 2SC4793 and 2SA1837 with a 68ohm biasing resistor.
As those Toshiba are now almost impossible to obtain, would you please suggest an alternative ?
I need fast drivers (Ft> 100MHz) able to run at 40mA into 125V without SOA issues. (That is why I chose to double the 2sc3503 2sa1381)
You are using 2SC4793 and 2SA1837 with a 68ohm biasing resistor.
As those Toshiba are now almost impossible to obtain, would you please suggest an alternative ?
I need fast drivers (Ft> 100MHz) able to run at 40mA into 125V without SOA issues. (That is why I chose to double the 2sc3503 2sa1381)
I have tons of 2SC4793 and 2SA1837 on stock for my projects and fans of SA20XX series. If you only need a few of them for this project I can send you some.
For a new commercial project this is of course no option ...
For a new commercial project this is of course no option ...
NPN: 4027 in stock at Mouser, 5000 more expected in september
PNP: 2548 in stock at Mouser
🤔
You could also look for KEC KTC4370/KTA1659.
2SC4793,A1837 manufactured by Unisonic are available in profusion https://www.profusionplc.com/parts/2sc4793lb-tf3t
Almost ready, it is a crowded pcb I am a bit nervous.
Dear chat72,
looks very good!
Note: the bias LED3 must be a 3mm RED led and need to be soldered on the heatsink side as well as the bias BC550C which should have good contact to the heatsink using some thermal paste.
BR, Toni
looks very good!
Note: the bias LED3 must be a 3mm RED led and need to be soldered on the heatsink side as well as the bias BC550C which should have good contact to the heatsink using some thermal paste.
BR, Toni
Not sure if this will affact the operation point, V value of all zener diode are smaller than spec for about 5%. V for green leds are less than 2v, around 1.95-1.96v.
I suppose a general procedure is to set a trim pot to maximum value, full counter-clockwise, then increasing it by turning clockwise.
I suppose a general procedure is to set a trim pot to maximum value, full counter-clockwise, then increasing it by turning clockwise.
The thermal curves mV/degree celsius of green and red led's are different . To avoid thermal runaway this is a must to use red LEDs. See here the simulations for the bias generator tempco: Bias tempco sim
To safely test the amp I have a different procedure. I start with a dual power supply at +/-15V set to a maximum current of 100mA. I will post the procedure the next days...
To safely test the amp I have a different procedure. I start with a dual power supply at +/-15V set to a maximum current of 100mA. I will post the procedure the next days...
Here you can see the 2 small holes in the heatspreader where the BC550C and the red led have thermal contact (using little bit of thermal grease).
Judging by the size of the coil, are you sure that it's 0.8uH? I try to do it with 18awg wire, it's bigger and longer than that.