[SSSSSmokey]

I just finished building my first PPA. My source is a car head unit. I am using a STEPS at 24v to power the PPA. I am using basically the default parts list on Tangents website, except my l/r opamps are OPA637s and my ground opamp is a 627.

I get audio, but it is very low quality. It is extremely quiet and destorted up until about halfway on the volume. At about halfway, it starts crackling and popping and gets very loud quickly.

I am also getting around 20 mV of DC offset (this is from measuring ground to output right?) at 0 volume and up to several hundred with the volume maxed.

I've checked over all my solder joints and don't see any shorts or any problems with the soldering. It just makes me sad to spend so much time on this and now it doesn't even work properly.

What should my first plan of action be? I don't know enough to really troubleshoot this by myself. :-/

 

[Megaptera]

Tangent has a troubleshooting guide that can walk you through a lot of problems.

A couple of questions, too:
Which version of the PPA did you build?
What is the DC offset of your source? (You're right, this is measured from a signal channel to its related ground.)
What gain did you set, 11?
Are you getting the same amount of offset in both left and right channels? If so, that suggests the problem may be in the ground channel.
Are any chips getting overly warm to the touch?

 

[SSSSSmokey]

PPA v2.0A
My meter really is a POS, but I am guessing from the randomly jumping numbers that it is around 2.2 mV. My DC offset on all 3 channels is 22.8.

I used the default resistor values, so if a gain of 11 is default then that is what I have. Otherwise please tell me which resistors set the gain and I'll tell you.

No chips are putting out more than barely perceptible heat, if any at all.

I have it in the case with the ends and top off now, if that makes a difference. Am I supposed to ground the board to the case in any way (such as soldering a wire from the ground plane to the case)?

EDIT: Another thing. Using my laptop as a source (Powerbook G4), the sound is EXTREMELY quiet even at the max volume and still distorted. My head unit I was using previously (Eclipse 8053) has 8v preamp outs, so that may have been a reason it was loud enough to hear.

EDIT Again: Also, what is B+ on the power supply connection thingy?

 

[Jazper]

first of all, don't plug in headphones with >10mV offset!

second, pictures of the board (top and bottom) would be useful for us to be able to help you.

 

[SSSSSmokey]

I know a picture would really help, but I don't have a camera right now.

Here are a couple of things I changed, if they could be responsible for the bad sound. Some of the R21, R22, R31, R32 (all 220 ohms) are different. For one of them I ran a 100 and a 120 ohm resistor in series cuz I ran out, and 6 of them are 1/8th watt instead of 1/4 watt. However the left channel is normal and still has the noise/distortion issues.

BTW, these are crappy earbud phones I don't care about. I am just using them for testing. Also, my DC offset jumps to 1000 mV+ when I have music playing. This seems really really bad.

Man something seems really screwed up. I wish I had a camera to take a picture. Anyway, considering all the info you have, what should be my best course of action?

EDIT: I dunno if this makes a difference, but I get 0 offset on the ground channel. I'll keep adding more stuff as I find it out. Tell me what to measure and I'll do it. Any help whatsoever is appreciated.

Thanks!

 

[Jazper]

Quote:

Originally Posted by SSSSSmokey BTW, these are crappy earbud phones I don't care about. I am just using them for testing. Also, my DC offset jumps to 1000 mV+ when I have music playing. This seems really really bad.

DO NOT PUT GOOD HEADPHONES ON THIS TILL THE OFFSET IS DOWN to <10mV

now.. what opamps are you using in the ppa?

(read your post)

check you're not shorting input to output, and that any diodes are the correct way around.

 

[SSSSSmokey]

Quote:

Originally Posted by Jazper DO NOT PUT GOOD HEADPHONES ON THIS TILL THE OFFSET IS DOWN to <10mV

I know.

Quote:

Originally Posted by Jazper now.. what opamps are you using in the ppa?

OPA637 for the left and right, OPA627 for the ground channel.

Quote:

Originally Posted by Jazper check you're not shorting input to output, and that any diodes are the correct way around.

There is only one diode and I am 100% it is the correct way. I had it reversed before and I got no power to anything.

I put all the polarized caps in the right way, they are clearly labeled.

Here is something that may be of interest: I am not getting ANY voltage across the output resistors (the big 2.2 ohm 1 watt ones). This this really has be perplexed.

Thanks for all your help, keep the suggestions coming!

 

[Jazper]

Quote:

Originally Posted by SSSSSmokey I know. OPA637 for the left and right, OPA627 for the ground channel. There is only one diode and I am 100% it is the correct way. I had it reversed before and I got no power to anything. I put all the polarized caps in the right way, they are clearly labeled. Here is something that may be of interest: I am not getting ANY voltage across the output resistors (the big 2.2 ohm 1 watt ones). This this really has be perplexed. Thanks for all your help, keep the suggestions coming!

you should be getting a small amount of voltage across them (I think) .. I'm going to go through the ppa design to help you in a sec.. (I have another project on the go now)

 

[Jazper]

have you adjusted the bias trimmers? step 7 here:

http://www.tangentsoft.net/audio/ppa/amp2/steps.html

 

[SSSSSmokey]

Quote:

Originally Posted by Jazper have you adjusted the bias trimmers?

I adjusted them a bit, but its hard since AFAIK you need to measure the current across the output resistors to do it properly.

I do get a measureable voltage across the R23/33 (the ones right "below" the output resistors if you are looking at it with the C1s running along the top). I adjusted the bias so those are 30 mV each. Not sure if thats correct but I can always change it later.

I'm at the point now where I'm going to start removing everything I don't need, like the biasing into class A, ect.

Also, I used 9 of the 470 uF Cerafines Tangent sells for the C1s. I don't know if too much capacitance would cause these issues, but if they could I'll take some off right away.


EDIT: Possibly major breakthrough. Looking at the pic on Tangent's site, I may have some of the output transistors (is that what they are?) backwards. If so, I am going to smack myself in the face before jumping for joy. I just really really want to chill with some good music and my Grados right about now.

 

[Jazper]

Quote:

Originally Posted by SSSSSmokey I adjusted them a bit, but its hard since AFAIK you need to measure the current across the output resistors to do it properly. I do get a measureable voltage across the R23/33 (the ones right "below" the output resistors if you are looking at it with the C1s running along the top). I adjusted the bias so those are 30 mV each. Not sure if thats correct but I can always change it later. I'm at the point now where I'm going to start removing everything I don't need, like the biasing into class A, ect. Also, I used 9 of the 470 uF Cerafines Tangent sells for the C1s. I don't know if too much capacitance would cause these issues, but if they could I'll take some off right away. EDIT: Possibly major breakthrough. Looking at the pic on Tangent's site, I may have some of the output transistors (is that what they are?) backwards. If so, I am going to smack myself in the face before jumping for joy. I just really really want to chill with some good music and my Grados right about now.

C4, C1 and C7 are polarized (so the orientation matters) the others shouldn't make any difference. I don't think high C1 would make that much difference.

As for measuring bias, note step seven says mV not mA. That's voltage not current. That's across R24 and R34.

 

[SSSSSmokey]

Thanks for all your help. I had the Q34s reversed.

Wheres the Grado smilie?

 

[Jazper]

enjoy! (you're welcome)

 

[SSSSSmokey]

I know this is going to sound insane, but...for some reason, my amp isn't quite right. I swear, the vocals are only playing from the left channel of my headphones. It isn't absolute, if I concentrate I can hear the vocals faintly on the right channel, but they are nice and clear on the left. It adds a really odd effect to the sound and not one I like...It makes my right ear sound deaf.

I'm assuming there is some kind of strange filter effect going on inside my amp causing this frequency range to be blocked and yet letting others to pass. I have tried multiple CDs on my CD player and now I'm using my laptop and it still has the same effect.

Anyone have any clues?

EDIT: OK, something is definitely effed up. I just disconnected the left channel to see if I possibly had it wired up wrong. However...both channels are still playing without the odd "vocal exclusion" effect. It does sound worse...maybe muddier? No need to respond to this...I'm just posting out my thoughts. I'll probably find it out and fix it before anyone posts a response anyway.

 

[Jazper]

Quote:

Originally Posted by SSSSSmokey EDIT: OK, something is definitely effed up. I just disconnected the left channel to see if I possibly had it wired up wrong. However...both channels are still playing without the odd "vocal exclusion" effect. It does sound worse...maybe muddier? No need to respond to this...I'm just posting out my thoughts. I'll probably find it out and fix it before anyone posts a response anyway.

try different opamps