[Aguirre]

Hello

I built an amp with an AlienDAC running to a PIMETA. The PIMETA is powered from an external dual voltage power supply. The AlienDAC is powered directly from USB.

Originally I had electrolytic output caps in the AlienDAC and no input caps in the PIMETA. There was no problem with DC offset that I noticed. The sound was better than the integrated audio from my computer, but still not so great.

I then removed the output caps from the AlienDAC and jumpered across their postions on the PCB. I cut the traces on the C1L/C1R positions on the PIMETA board and then installed 4uf 200v Audience Auricaps. The sound is easily ten times better.

Ok enough with the lead in, here comes my problem. When I measure DC voltage with music playing it is about 5-10 millivolts. BUT when I turn the volume pot DC jumps up into the volt range. Like 3 or 4 volts worth. And then takes maybe about 20 seconds to come back down to several millivolts. I am using the ALPS blue velvet which installs right into the PIMETA board.

I wanted to know if this is normal. I don't seem to recall this happening when I had the electrolytic caps but I don't know if I actually tested for it before. Also if this is unusal, would it make any difference if I moved the caps. Right now signal comes to the pot first and then into the caps.

Altogether I am very happy with my amp even if I can't use the volume control.

thanks in advance.

 

[Pars]

Is the DC offset you are measuring with the volume turned up still with music playing? Normally, DC offset is only measured with no source playing, and sometimes with the inputs shorted. If the DC offset measures OK with no source playing, thru the range of volume control, then I wouldn't worry about it. If it still goes up to an unreasonable level (above 15-20mV), then there is a problem somewhere.

It shouldn't be an issue of where you have the caps. Is the Alien DAC meant to be used without output caps? In your configuration, with the input caps on the Pimeta and no chance of connecting the DAC to a DC coupled amp, its probably irrelevant.

 

[Aguirre]

Measured DC voltage between ground and either left/right at the headphone jack is 5-10 mV for any position of the volume control (max, min, half, etc.). It is only when I adjust the volume control and for about 20 seconds after that the DC voltage spikes and slowly comes back down. This is with or without music playing as long as the AlienDAC is powered and connected. With no source connected to the PIMETA output DC is ~ 4mV and doesnt change as I adjust the pot.

 

[picklgreen]

Thank you for noticing the difference proper film caps make in the output stage of the ALIEN DAC. The orginal design is defintly flawed in using electrolytes in the CL/CR positions! He should have specified some WIMA box caps that are small enough to fit in that position perfectly. For something as low voltage as the ALIEN DAC, there was no excuse for putting electrolytes directly in the signal path!

 

[Aguirre]

It really does sound a LOT better with the auricaps. To Alf's credit I think his webpage does mention you can use film caps or botique electrolytic. I went with the Elna Silmic II since you can get them from digikey, but they dont sound so great.

 

[Pars]

The majority of CD players on the market today use electrolytics as the output coupling caps. Of course they also use opamps for I/V. Not to say that any commercial CD players are designed correctly

 

[Aguirre]

I removed the auricaps for the input positions on the PIMETA amp (C1R/C1L) and then jumpered across their positions. I then moved them back as output caps on the AlienDAC. The DC voltage now remains locked at 5mV when I turn the pot.

I think this is because the pot is a variable voltage divider and when I turn the pot it presents a changing signal which the caps pass.

I had wanted to use the caps as input caps on the PIMETA in case I wanted to use a source other than the AlienDAC. I guess I might cut more of the traces in that input section of the PIMETA and rewire it.

My problem is fixed. Just a heads up to anyone else who might be thinking of using the C1R/C1L on the PIMETA.

 

[Pars]

That is strange... I would have thought exactly the opposite. The input caps on the Pimeta are between the pot and the amp. I have a Pimeta that I am playing around with a sort of mini3fied config for a bipolar opamp. In this, you pretty much have to dispense with the input caps in order to get the feedback resistors low enough in value to eliminate hiss (with the caps, you have to have resistor values of ~680K or so in the feedback loop with the 100K input resistor in order to balance the + and - inputs to the opamp). With the high resistor value and input caps, the offset is low and stays low, irregardless of the pot setting. In the mini3fied config without the caps, the offest starts very low in mine, but increases to ~50mV at full volume.