Got a strange problem with a power amp I just picked up

[Derringer]

it's a Hafler TA-1600

schem here:
http://www.hafler.com/pdf/archive/TA1600-schematic.pdf

The RCA ins work ok, but I'd rather use the balance ins.
When I do use the balanced inputs, I get volume when both channels are turned to zero
Then, as I turn them up, the tone gets darker, "nulls out" about halfway up, and then gets real bright.

I need to test a different source with balanced out jacks to see if it's something with the interface I am currently driving the amp with ... but I did a little internet searching and found someone else with the same issue and no conclusion.

When I look at the schem though, I don't see where the "-" portion of the balanced signal is flipped and mixed with the "+" sign. It looks to me like mid-way through the volume rotation, out of phase signals are hitting each other and canceling stuff out ... just like I'm experiencing. Plenty of people here know more than me and I appreciate your advice and insight.

Thanks for looking folks.

[Rob Strand]

The RCA ins work ok, but I'd rather use the balance ins.

I believe the inputs are stereo not balanced.   

Normally you put in a stereo signal and it mixes it to mono.   The two stereo signals are mixed (ie added)  in-phase.   

When you apply a balanced input the two "halves" of balanced signal get added by the amp and so they cancel each other out.   The balance control simply makes them cancel less.

[DIY Bass]

The spec sheet says it is balanced TRS or unbalanced RCA inputs

[Derringer]

Spec sheet says balanced TRS input and unbalanced RCA
http://www.hafler.com/pdf/archive/TA1100_TA1600_datasheet.pdf

Here's the manual as well
https://www.manualslib.com/products/Hafler-Ta1600-227406.html

[Rob Strand]

Spec sheet says balanced TRS input and unbalanced RCA

So it does! I've screw up.  I saw the non-inverting opamps and "STEREO" on the schematic and didn't look any deeper.

The key is further-up in the circuit.  The non inverting opamps eventually feed the non-inverting and inverting inputs of the power amplifier.  Normally power amps only use the non-inverting input but here the inversion for the balanced input is achieved at the power amp.   The volume control is a little weird because of that.  U2A is the non-inverting path and U2B is the inverting path.

There's a trim pot there R8 which I presume is to trim the common-mode rejection.

I'd probably drive one side of the balance input then the other side separately to see if the level on each half is the same.  If it's not the same then it won't work correctly with the balanced input.  *However* be very careful about driving the lower part of the circuit U1B separately.  The amplifier will be at full volume regardless of the volume setting, so you will need to set the input low using the external source. 

If that seems OK then drive the same signal into both inputs to check you get no output.  Then if necessary adjust R8 to null the output when the volume control is set to maximum.

If you have an oscilloscope you can trace the signals through from inputs to the opamp outputs to see if it all looks good.

[Derringer]

Thank you, Rob!

Yes, I do have a signal gen and scope.
First chance I get to pull it apart, I will follow your advice.


That sensitivity/volume control .... am I understanding it correctly?
When fully clockwise, the resistance between the "+" and "-" signals is equal to the parallel resistance of the dual-gang pot ... which is enough to keep them separate.

When it's CCW, the signals are supposed to mix with each other and null out?

[Rob Strand]

That sensitivity/volume control .... am I understanding it correctly?
When fully clockwise, the resistance between the "+" and "-" signals is equal to the parallel resistance of the dual-gang pot ... which is enough to keep them separate.

When it's CCW, the signals are supposed to mix with each other and null out?

Not quite.  It's a bit weird.

The pot is acting like a voltage divider (actually two in cascade).

When the pot is on minimum the "power amp +ve input"  (U2A) and the "power amp -ve input" (U2B) are at equal voltage,  which is the voltage at the output of U1B.    Since the power amp +ve input and -ve input are at equal voltages the differential action cancels the two and the amp produces no output.   In the balanced case the voltage at U1B isn't zero but the subtraction process still produces zero output.

When the pot is on full is lets the "amp +ve connector input" connect through to the "power amp +ve input" and
"amp -ve connector input" connect through to the "power amp +ve input".  In that case it just amplifies the difference, which is full output.

If you go through the cases of single-end and  balanced input the scheme works in both cases.

I don't think I've ever seen a set-up like that on a power amp.  It's pretty weird passing more difficult task of canceling signal to the power amp - the least likely part to succeed in the process.

[Derringer]

scoped it, both channels, tip and ring separately

Its doing pretty much what I was hearing. The tip takes signal and boosts it continuously throughout the entire volume sweep on both channels.

The ring, at channel level zero puts the output at about what the tip does with dial at 6. It stays there till turned to about 4 at which the signal dips, then comes back up around 6, but does not come anywhere near the output of the tip signal no matter what i turn it to.

It's the exact same behavior on both channels.

I fiddled with the trim, but nothing changed.

Any ideas?

[Derringer]

also, there is a bias adjustment in the power section, r136... but the schematic doesn't say what it should be set to.

[Rob Strand]

also, there is a bias adjustment in the power section, r136... but the schematic doesn't say what it should be set to.

If I had to pull a number out of the air I'd say about 100mA.   Set it, leave the amp running for at least 1hr then check and re-adjust.   I generally check the current a few times before the hour is up in case the current skyrockets.   

There's a bit of set at you own risk going on when you don't set it up like the manufacturer recommends.  There's probably some instructions that go along with the schematic.

[Derringer]

Thanks again.

I think I'm going to leave well enough alone. It's 80% SMD components in there anyway and I am not up to tackling that if there is something to be replaced/re-worked. The unbalanced ins work fine and looked clean all the way up on the scope. I'll just adjust my EQ to compensate for the lack of perceived bass when using them.