[FallenAngel]

Uhm... I think I see a few little problems.

First off, you're mixing MJE253 and BD139, I don't think that's a good idea, even those they are PNP / NPN accordingly, I don't think they work together well, they're opposites - one is much lower output current than the other, you're asking for trouble mixing them.

Reading 45mV across the output resistors, even with 2R2 resistors is only 20mA of current, which is pretty low and you're likely to want to raise it to decrease the chance of oscillation. Even using 2R2 isn't advised because it might be unstable, I upped mine to 4R7 to stabilize the buffer.

The AD843 is also a little cranky in the PPAv2 for some reason, I read Tangent had problems with this chip, especially in ground (as well as on Pimeta) not sure why, but I didn't chance it using this chip.

 

[d-cee]

Thanks.

I didn't actually build the amp and I'm not too up with the technical things. I don't have any other transistors to swap in, so for the mean time I'll try to tweak the bias.

I somehow mistook mV for mA and I read on tangen't site that 30-40mA was the max before no more improvements kicked in. I read 45mV and thought that must be ok and considering the amp stayed fairly cool I thought it ok.

I've now adjusted it so that I'm reading ~85mV across the 2R2 resistors (these resistors were easy to read

)

One thing I'm a little confused about is if it's lack of output current why it's happening to the DT880. It's quite high impedance so shouldn't something like my 62ohm K701s heard the problem first?

 

[FallenAngel]

Quote:

Originally Posted by d-cee /img/forum/go_quote.gif Thanks. I didn't actually build the amp and I'm not too up with the technical things. I don't have any other transistors to swap in, so for the mean time I'll try to tweak the bias. I somehow mistook mV for mA and I read on tangen't site that 30-40mA was the max before no more improvements kicked in. I read 45mV and thought that must be ok and considering the amp stayed fairly cool I thought it ok. I've now adjusted it so that I'm reading ~85mV across the 2R2 resistors (these resistors were easy to read ) One thing I'm a little confused about is if it's lack of output current why it's happening to the DT880. It's quite high impedance so shouldn't something like my 62ohm K701s heard the problem first?

Wait! 85mV is WAY over what's safe! That's almost 50mA, you're about to burn your output stage! Get it to about 66mV! That'll be 30mA. I had mine running at 35mA and the top of the case gets pretty warm within an hour, after a few it gets VERY warm.

 

[d-cee]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif Wait! 85mV is WAY over what's safe! That's almost 50mA, you're about to burn your output stage! Get it to about 66mV! That'll be 30mA. I had mine running at 35mA and the top of the case gets pretty warm within an hour, after a few it gets VERY warm.

i see. i thought under 40mA (85 / 2.2 = ~38) was ok... thank goodness it's only been on for about 6 minutes (1 test song)

i touched the top of the output transistors and they seemed warm but not too hot, i will drop it as per your recommendation

i have been listening to it just now with 2 x OPA637 and 1 x OPA627 and it's working fine no more funky crazy distortion/white noise.

 

[FallenAngel]

Quote:

Originally Posted by d-cee /img/forum/go_quote.gif i see. i thought under 40mA (85 / 2.2 = ~38) was ok... thank goodness it's only been on for about 6 minutes (1 test song) i touched the top of the output transistors and they seemed warm but not too hot, i will drop it as per your recommendation i have been listening to it just now with 2 x OPA637 and 1 x OPA627 and it's working fine no more funky crazy distortion/white noise.

Nice, I had some trouble with those opamps with 2R2 and a low bias as well, upped the bias and got stability everywhere except IEMs, then switched to 4R7 and everything worked perfectly at 25-30mA.

Edit: My mistake on the bias setting, it's very late here, I'm going to sleep

 

[d-cee]

ok i'm measuring about 65mV across the output resistors which works out to be about 29.5mA working stably with the 2 x OPA637 + OPA627

still no luck with the AD843 though. oh well.

so what's the general concensus about opamp biasing? i've got a 1k resistor there and reading 1v so that's 1mA what are the limits?

thanks =)

edit: one more thing, when measuring DC offset at the jack between L + G and R + G, i get <10mV up until about 80% on the pot, at which point is very rapidly increases until it reaches about 45mV @ 100% pot.

in fear of deafness I never see myself going past 50% on the pot but is this kind of behaviour normal?

 

[FallenAngel]

Quote:

Originally Posted by d-cee /img/forum/go_quote.gif ok i'm measuring about 65mV across the output resistors which works out to be about 29.5mA working stably with the 2 x OPA637 + OPA627 still no luck with the AD843 though. oh well. so what's the general concensus about opamp biasing? i've got a 1k resistor there and reading 1v so that's 1mA what are the limits? thanks =) edit: one more thing, when measuring DC offset at the jack between L + G and R + G, i get <10mV up until about 80% on the pot, at which point is very rapidly increases until it reaches about 45mV @ 100% pot. in fear of deafness I never see myself going past 50% on the pot but is this kind of behaviour normal?

I bias my opamps into Class-A to about 3mA, no harm in it, but probably minimal performance increase as well, but for an extra 2mA of current draw, why not

This amp generally should not see ANY DC offset (as long as the transistors are matched - but in your case with MJE253 + BD139, they are definitely not). <10mV is OK, 45mV is pretty bad.

Personally, I would recommend replacing the BD139 with MJE243 which should definitely help stabilize your amp, MJE253 and MJE243 are complimentary pairs and I think you're only supposed to use complimentary pairs in this situation.

 

[d-cee]

Well, maybe it was just waiting to happen and this sped it up but as I increased the opamp bias one of my AD744 kicked the bucket.

As I did my routine checks after each change, I noticed that the DC offset in one of my channels was 2.5v ! yikes, so I swapped the AD744 between left and right and sure enough the offset moved to the other channel.

Thankfully I got a spare AD744, popped it in and it's all good.

I have it biased to 2mA at the moment and nothing else has happened, is this a safe value or should I drop it further?

I'm using the comp pin of my AD744 if that makes a difference.

 

[n_maher]

Quote:

Originally Posted by d-cee /img/forum/go_quote.gif Thanks. I didn't actually build the amp and I'm not too up with the technical things. I don't have any other transistors to swap in, so for the mean time I'll try to tweak the bias.

I'll check my parts box and see if I have what you need. I agree that mixed output transistors at least sounds like a bad idea to me, not sure if it would actually cause any problems.

Quote:

One thing I'm a little confused about is if it's lack of output current why it's happening to the DT880. It's quite high impedance so shouldn't something like my 62ohm K701s heard the problem first?

Impedance doesn't tell the whole story, you should look at it in concert with the phones relative sensitivity. Those two put together are a much better measure (from what I understand) of how hard they are to drive.

DT880 = 96 dB /mW
K701 = 105 dB / mW

 

[d-cee]

thanks nate

i'm currently researching my options re: output transistors

if anyone has an opinion i'd love to hear it

 

[tsaavik]

Hello all, have a ppaV2 ad8610 running 30ma bias with 1ma class A. Power is 24v from steps running around .16ma

dc offset is (gnd probe to input gnd):
-0.4mv right
0.1mv gnd
-1.0mv left

I'm seeing large (1.5v) dc output between input ground and output upon poweroff. My meter goes to 0, then jumps up to 1.5v or so, then slowly drains down to 0. I also have turn on thumb, but i only see about .5v during poweron.

Is this normal/okay? Should i consider adding a ε12 type circuit? I'm double checking this amp because i had a mysterious headphone failure that I'm hoping not to repeat.

 

[FallenAngel]

Quote:

Originally Posted by tsaavik /img/forum/go_quote.gif Hello all, have a ppaV2 ad8610 running 30ma bias with 1ma class A. Power is 24v from steps running around .16ma dc offset is (gnd probe to input gnd): -0.4mv right 0.1mv gnd -1.0mv left I'm seeing large (1.5v) dc output between input ground and output upon poweroff. My meter goes to 0, then jumps up to 1.5v or so, then slowly drains down to 0. I also have turn on thumb, but i only see about .5v during poweron. Is this normal/okay? Should i consider adding a ε12 type circuit? I'm double checking this amp because i had a mysterious headphone failure that I'm hoping not to repeat.

DC offset is measured from output ground

 

[tsaavik]

Whoops, here is the numbers measuring output ground to left/right channel.

Code:

[left]ACTION LEFT RIGHT ------------+-----------+------------ Min Vol | -1.1mv | -1.2mv Max Vol | -1.1mv | -1.3mv MinVol+bass | -4.1mv | -3.9mv MaxVol+bass | -4.2mv | -3.8mv Turn off | +1.2v peak| +1.2v peak Turn on | +.367 peak| +.5v peak[/left]

I have an ancient fluke that doesn't have a min/max button so the peaks are kinda guesses. I turned it on and off a bunch and just grabbed the highest number i saw. All tests were run with a really crappy pair of Walkman style headphones plugged in.

 

[FallenAngel]

Quote:

Originally Posted by tsaavik /img/forum/go_quote.gif Whoops, here is the numbers measuring output ground to left/right channel. Code: [left]ACTION LEFT RIGHT ------------+-----------+------------ Min Vol | -1.1mv | -1.2mv Max Vol | -1.1mv | -1.3mv MinVol+bass | -4.1mv | -3.9mv MaxVol+bass | -4.2mv | -3.8mv Turn off | +1.2v peak| +1.2v peak Turn on | +.367 peak| +5v peak[/left] I have an ancient fluke that doesn't have a min/max button so the peaks are kinda guesses. I turned it on and off a bunch and just grabbed the highest number i saw. All tests were run with a really crappy pair of Walkman style headphones plugged in.

That looks fine, the turn on/off thump isn't great, but I think I get something similar and haven't had any problems.

 

[tangent]

Quote:

Originally Posted by tsaavik /img/forum/go_quote.gif I'm seeing large (1.5v) dc output between input ground and output upon poweroff

Unless you're using super-sensitive headphones, this will just be a loud click.

How much rail capacitance do you have?