Thanks Salas..Tweaking with the op-amp installed might not get you the result you expect because you will be fighting with the DC servo. I second Salas' suggestion to wait until everything is packaged, then remove the op-amp and trim for zero offset, and finally replace the op-amp, of course. This, in theory, will give you lowest distortion performance.
It's also about hitting the intrinsic op-amp offset limit that varies from chip to chip in production and/or between its channels. AD823 has good DC performance but its not a precision spec op-amp. There are bit better and bit worse samples. In any case the final offset range reported by builders until now is low mV. Never caused any practical concern.
Salas,
Thank you.. I'll hold off on any further tweaking until I get the chassis and the paint starts to dry. I noticed last night after I installed the opamp that the left channel was more stable than the right.. Possible that it was just tuned more closely prior to installing the opamp. I'll get everything in place, reset the offset without opamp in place and see where things pan out.
Quick quesiton.. after opamp is installed should I have left the input pins shorted while measuring the offset? I left them shorted..
Thank you.. I'll hold off on any further tweaking until I get the chassis and the paint starts to dry. I noticed last night after I installed the opamp that the left channel was more stable than the right.. Possible that it was just tuned more closely prior to installing the opamp. I'll get everything in place, reset the offset without opamp in place and see where things pan out.
Quick quesiton.. after opamp is installed should I have left the input pins shorted while measuring the offset? I left them shorted..
Hello,
I have a problem with my DCG3/DCSTB combo. It was working fine (for a week) then this morning upon switching on, the main fuse popped. Replacing the fuse and turning them on again I noticed that the the right channel Dref2 led bar is brighter than the others. Powering off, the same LED bar stays on a bit compared to the other 3.
Voltage measurements at output (V+/V-):
Left Channel: 17.28
Right Channel with the glowing LED DREF 2 bar: 0
I removed the DCG3 preamp connected.
Thanks for your help.
I have a problem with my DCG3/DCSTB combo. It was working fine (for a week) then this morning upon switching on, the main fuse popped. Replacing the fuse and turning them on again I noticed that the the right channel Dref2 led bar is brighter than the others. Powering off, the same LED bar stays on a bit compared to the other 3.
Voltage measurements at output (V+/V-):
Left Channel: 17.28
Right Channel with the glowing LED DREF 2 bar: 0
I removed the DCG3 preamp connected.
Thanks for your help.
Hi
I suggest you first check the BC5xx transistor on the problematic side of the DCSTB. Then the MJE big pass transistor. For normal Vbe when still on board (0.6-0.7V).
Lastly measure the voltage drop across that side's 33Ω resistor pins (R1 or R2). Voltage/33 should result around five mA if the associated PF5102 is still ok.
I suggest you first check the BC5xx transistor on the problematic side of the DCSTB. Then the MJE big pass transistor. For normal Vbe when still on board (0.6-0.7V).
Lastly measure the voltage drop across that side's 33Ω resistor pins (R1 or R2). Voltage/33 should result around five mA if the associated PF5102 is still ok.
You miss a plus polarity voltage rail, right? It's Q2 that has failed it seems.The only one not showing a Vbe reading. Try to replace it.
There can very well be random failures in small transistors.
On the other hand to actively kill some BCxx in a DCSTB it usually takes a rail output short. Even momentary.
There can very well be random failures in small transistors.
On the other hand to actively kill some BCxx in a DCSTB it usually takes a rail output short. Even momentary.
Check mV drop pin to pin on each polarity's 33 Ohm resistors R1 R2. mV/33=ImA. If the not as bright bar side has less mV across its 33Ω resistor then it also has less mA available to light up same Vf in Leds due to a new non matched J2 (J2 weaker than J1). You can equalize mA by smaller value R2. ImA=mV/R. DCG3 is working fine already, for aesthetic reasons you could try equalize led bias and brightness leading to bit more Vout also, as I described.
No it hasn't sonic consequence as long as it is not widely different. Talking for this particular output stage. It's not a push pull. It's a single ended Mosfet cascode class A with constant current bias.