[Turing]

I recently built a CMoy amp using Tangent's guide and although it works pretty well (ie. plays music), there are still problems with it.

1. a pretty audible hiss when the source is hooked up, even when it's not on.
(I think this is more then the hiss that Tangent suggests is removable using the R5 resistors)
2. the sound starts to distort after turning up the source volume to above the midpoint.

After doing a whole bunch of testing, here is what I've found:

With power (9V battery) only i get the desired ~4.5/-4.5 split at V+ and V-.
But, with power and outputs hooked up (having the inputs hooked up doesn't seem to make a difference one way or another), it changes to ~8V at V+ and ~-0.5V at V-.

That shouldn't be happening should it?

reisistances at the opamp socket:
(in K)

1. 11
2. 1
3. 100
4. 4.7
5. 100
6. 1
7. 11
8. 4.7

All those look as expected to me.

I've cleaned the board several times and gone over it with my loupe, looking for solder bridges, etc. I've also tested for bridges using my meter.

I've run out of things to test, anyone have any suggestions on what might be causing the voltage problem (which I think is responsible for both issues mentioned above) ?


FYI:
I'm testing using the alligator clip method suggested by Tangent, with both grounds hooked up to the virtual ground (which I think is right). No pots or jacks are hooked in yet, and the R5 resistor spots are jumpered. All parts are the standard ones that Tangent suggests (R3s are 1Ks).

 

[tangent]

Quote:

a pretty audible hiss when the source is hooked up

Could be the source is just noisy. What is it?

Quote:

even when it's not on.

It may be in a standby mode, and not actually powered down. Have you removed its power source?

Quote:

the sound starts to distort after turning up the source volume to above the midpoint.

With only a 9V supply and the nasty voltage split you're getting, this is not surprising. The distortion you're hearing is clipping.

Quote:

anyone have any suggestions on what might be causing the voltage problem

Can you measure the current the amp draws? I'd be interested to know if it changes significantly when you plug headphones in.

 

[Turing]

The source is an portable CD player (i've also tried with my iPod).

You are right, the player was on standby. When the power is off the hiss disapears. There is also a hiss when the outputs are hooked up and inputs are not (power on).

How do I measure the current the amp draws?

Also:

When I have only right out and ground out hooked up I get:
~5.3V at V+ and ~-2.7V at V-.

When I have only left out and ground out hooked up I get:
~4.7V at V+ and ~-3.3V at V-.

Both gives me
~8V at V+ and ~-0.5V at V-.

 

[Pars]

Quote:

How do I measure the current the amp draws?

Pop the 9v connector off one side of the battery. With the meter set to mA mode, put one side onto the diconnected battery terminal on the connector, the other side to the disconnected battery terminal itself (i.e., wire the meter in SERIES with the battery).

Chris

 

[Turing]

Quote:

Originally Posted by tangent Can you measure the current the amp draws? I'd be interested to know if it changes significantly when you plug headphones in.

0.01A regardless of whether headphones are plugged in or not.


(thanks for the info Pars)

 

[Pars]

Hmmm, 10mA sounds reasonable I think for an OPA2132. Do you have an LED in yet? I would suspect not. Does your meter have a mA range?

Also, try measuring V+ and V- with the opamp out, but phones plugged in (check the output DC offset BEFORE you plug phones in... output L to ground, output R to ground... anything over 20mV or so DO NOT plug the phones in unless they are expendable!). If the DC offset is OK, do your V+ and V- behave the same without the opamp in? Did you use the stock PS component values of 4.7K resistors and 220uf caps? Increasing the cap size to 470uf may help some.

There is a good thread on headwize regarding the CMoy PSU architecture and improvements to it (sorry, I don't have a link to it right now). You may want to check over there and at least read it (it was titled something to do with updating or CMoy Rev. 2 or something like that... if you can't find it, let me know and I'll see if I can get a link for ya).

Chris

 

[tangent]

The noise you're hearing when the source is plugged in is most likely just source noise. The best way to reduce it is to reduce the amp's gain.

The noise you're hearing with nothing plugged in should be very low, more on the order of "I can just hear the difference when the amp is on" rather than "sounds like a flippin' waterfall!" Reducing this is the purpose of R5.

Regarding your voltage measurements, it sounds like something is shorting out in the output jack. Either you've hooked it up wrong, or you melted something while hooking it up. Try changing it out.

 

[jasonhanjk]

Did you have a virtual ground TLE 2426?

 

[alane]

Turing,
I know that you have checked all this but i had a similar problem because i had overlooked the jumpers from the two capacitors in the power section ( the ones that go to ground ) .. it would distort badly at anything aproaching a good volume and the sound was generally bad and scratchy.. my eyes were sore looking at the board till i noticed that...

Hope this helps

Alan
Dublin/Ireland

 

[Turing]

maybe these might help

 

[Turing]

Arrg!

Ok, so I cleaned the board a bunch more times, re-soldered a few connections, cleaned the board a few more times...still didn't work. (bad voltage split with headphones attached.)

So I did the only sane thing, pulled all the components off the board, tested each one individually, and then soldered them back on to the other (clean) half of my radioshack 150.

The voltage split is still there, same problem as before.

I then tried submerging the board in 99 iso for 30 minutes, then brushing it off with water. The board is super clean now, I can't see any flux at all, even with my hand lens.

But the bad voltage split still remains with the phones attached. Without the phones hooked up, the split is a perfect +4.4/-4.4

Anyone have ant ideas on what I can measure to try and nail this problem down to one section of the board?

 

[Pars]

I assume this is just mislabeling, but in your picture, you have the inputs and outputs reversed, i.e., the left-in should be the left-out and vice versa. The inputs should be going to one end of the film caps. The outputs should be coming off the three jumpered outer pads formed by R4 and R5 (or its jumper).

Chris

 

[Turing]

oops, yeah, that's mislabeling.

 

[Turing]

does anyone know if dropping in another battery (bringing it up to 18V) would help?

 

[tangent]

Have you checked the jacks? Perhaps even go to the extent of removing them and temporarily using alligator jumpers to attach cables to the amp.

Regarding the battery change, yes, that may help, but it's attacking the symptom not the cause.