My PPA: measures well, sounds bad. HELP!
Jun 5, 2004 at 9:30 AM Thread Starter Post #1 of 17

JMS

100+ Head-Fier
Joined
Sep 16, 2002
Posts
106
Likes
19
I built my PPA about 3 months ago, using basically the standard recommended configuration/parts. I always thought it sounded a little thinner than my META42 (also built to standard configuration/parts), but figured it was due to the improved stereo separation. I hadn't done a serious subjective comparison between the META42 and the PPA, but when I tried them recently as a preamp for my C370 to power speakers, the difference became more apparent. The PPA sounded shrill, and somewhat lean on bass. I have a setup where I can switch the PPA/META42 (set to unity gain) in and out of the preamp loop, and the META42 sounded much closer to no-amp than the PPA. I don't think this difference is only true of the amps when serving as preamps -- the sound was consistent with what I heard through headphones (Senn HD580), only it was more obvious in this usage which sound was better and more accurate.

Now, I also measured the 2 in RMAA using the loopback method through an Audigy 2 NX, and both amps seem to measure as on Tangent's web site, where the only major difference is that the PPA has better stereo separation.

What's going on here? Perhaps my PPA is built wrong. The only non-standard thing I did was to use a low gain (3x), with R3-R6 of:

R3L/R: 1k
R3G: none
R4L/R: 2k
R4G: 4.7K
R5L/R: 3.32K
R6L/R: 1M

Could that be the culprit? Other values are:

R1L/R: 4.7K
R1G: 4.7K
R2L/R: 1M
R7L/R: 5K
R8: 100
R9: 1K
R11: 1K
R12: 1K
R13: 1K
R14: 1K

I hope it's something that can be fixed. Please help!
 
Jun 5, 2004 at 6:29 PM Post #2 of 17
R1 = R3 + R5

You're off a bit. That can cause a source impedence mismatch that can increase distortion. R1 should be 4.32k, as should all the other 4.7 you have listed.

Which op-amps are you using? The R8 of 100 kinda makes me think you didn't measure the current to bias the op-amp, but I could be wrong and you just got a really well matched set of transistors.
 
Jun 5, 2004 at 7:06 PM Post #3 of 17
I think I used 4.7k because I couldn't find 4.32k. I'm using:

4-4-4 Intersil
3x AD8610
50k Alps Blue
Q1: 3x Digikey 2N5484
Q2: 3x Digikey 2N5486
Q3: 12x Digikey 2N5486
You're right, I didn't measure the current to bias the opamp. Should I? Thought I was OK with the default config.
 
Jun 5, 2004 at 10:39 PM Post #4 of 17
Have you ever heard what a bad ground loop condition sounds like?
Well, you have now.
smily_headphones1.gif


As far as your PPA sounding thinner then the META goes:
What is the gain set at on the META?
wink.gif
 
Jun 6, 2004 at 12:04 AM Post #5 of 17
The gain on my META42 is 10x:

R1: 1k
R2: 500K
R3: 475
R4: 4.75K
R5: 2.21K
R6: 221K
R7: 200
R8: jumper
R9: jumper
R10: 100
RL: 10K
Pot: 50K stepped attenuator
AD8620, 2x EL2002
 
Jun 7, 2004 at 12:32 AM Post #7 of 17
Some people just don't like the AD8610 in the PPA. Try something else.
 
Jun 7, 2004 at 6:47 AM Post #8 of 17
or simply throw $30 on Diamond Buffer kit and things will start to shaking up
smily_headphones1.gif
 
Jun 7, 2004 at 10:07 PM Post #9 of 17
I would suggest swapping the op-amps for something else too. If you like smooth and warm sound with good bass, you could try e.g. OPA637/OPA627, LT1360 or LT1363. I personally don't like AD8610 in PPA much.
 
Jun 9, 2004 at 5:41 AM Post #11 of 17
Yes, I have in C6G: 10pF 500V silver mica, Digikey 338-1061.

Just to check if it's the op-amp, I just tried using some cheap LM741CN's in place of the AD8610's, and the shrillness remained. Perhaps less so, but still there. I dunno, is this just the PPA's sound?
 
Jun 10, 2004 at 7:09 AM Post #13 of 17
Yes, I followed the instructions as on Tangent's site, and have things wired properly. In S2, I had wires running to a switch. I just replaced the wires with a jumper, but no help. I have some OPA227's on order, will see if that helps.
 
Jun 10, 2004 at 8:05 AM Post #15 of 17
What source are you using with the amp?
 

Users who are viewing this thread

Back
Top