PPA v2: clipping vs. voltage

[headfone]

(Haven't topically searched this forum thoroughly, so apologies in advance for any redundancy. Anyway...)

Parameters:
Input voltage: 18v (diy PSU), and 24v (ELPAC wal wart) (see below)
Op amps: OPA637 (L/R), OPA627 ground; have tried AD797 and AD825
Buffer bias: Tangent's default, 20 mA @ 44mV.
Op-amp class-A bias: Tangent's default, 1mA

When using large 'phones (Sennheiser HD650, beyer DT880, etc.), the audio output clips at high-volume setting during which the music signal is concurrently high-power (e.g. bass drum).

When I use the higher-voltage (24v) ELPAC wall wart, no clipping. At this input voltage level, however, PS pin voltage for the op-amps are +/- 19.5v -- a bit over the max rated (18.0v) for all op-amps I’m using. Not sure how “safe” higher voltages are but I don’t want to destroy these op-amps.

What’s the best solution here?

P.S.:
The PPA2 sounds great otherwise -- much better than any other amp I've heard to date.

 

[FallenAngel]

I don't think you're getting +/- 19.5V on those opamps, you're just getting 19.5V. You can safely raise it to 32V. Still, you shouldn't have the headphones clipping at 19.5V.

 

[headfone]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif I don't think you're getting +/- 19.5V on those opamps, you're just getting 19.5V.

Here's the way I measure:

For measuring postitive op-amp PS pin voltage (+Vs):
Pos. DMM probe: Op-amp pin 7 which +Vs
Neg. DMM Probe: "V-" (negative or ground of main PS input)

For measuring negative op-amp PS pin voltage (-Vs):
Pos. DMM probe: Op-amp pin 4 which -Vs
Neg. DMM Probe: "V+" (positive of main PS input)

Is this the correct way to measure + / - op-amp PS pin voltage?

Quote:

Still, you shouldn't have the headphones clipping at 19.5V.

You misunderstood (maybe my fault in not providing enough detail)...
The PPA2 clips with the 18v PS (with this PS, the opamp voltages are +/- 13.5v). The 24v PS allows opamp voltages around +/- 19.5v.

Please note that I'm not implying the op-amps are casuing the clipping -- I've always had enough voltage for them in any case. Rather, the output transistors or buffers may not have enough voltage, hence they clip.

Current may the issue. The ELPAC 24v adapter can output 0.33A, while my 18v PS (the one that causes PPA to clip) is limited to the max of this transformer:
Amveco Model #62045, Sec-Full Load V = 2x22, Current = 227mA. I'm only using one side of this trans, so 0.5 x 227mA, or only 0.1135A

 

[JamesL]

Quote:

Originally Posted by headfone /img/forum/go_quote.gif Here's the way I measure: For measuring postitive op-amp PS pin voltage (+Vs): Pos. DMM probe: Op-amp pin 7 which +Vs Neg. DMM Probe: "V-" (negative or ground of main PS input) For measuring negative op-amp PS pin voltage (-Vs): Pos. DMM probe: Op-amp pin 4 which -Vs Neg. DMM Probe: "V+" (positive of main PS input) Is this the correct way to measure + / - op-amp PS pin voltage?

You're measuring the power supply voltage.
The TLE's will split it into +/- 9.75v.

 

[FallenAngel]

I think you're misunderstanding the V+/V-. The opamp datasheet says they are rated for 36V operation (or +/- 18V when relative to ground). That's the key part, relative to ground.

The PPAv2 amp runs off a single supply, but creates a virtual ground. When you measure V+ relative to V-, you're measuring the full rail voltage. If you measure V+ (pin 7) or V- (pin 4) to ground (input signal ground), you will see it's half that much. You're running the opamps at +/- 9V, and that's plenty to avoid clipping.

Dropping 5V in from the PSU to opamps is A LOT, is the PSU not regulated? It should be.

At 13.5V feeding OPA637, there could be some clipping at high volumes on bass notes.

330mA is PLENTY for a PPAv2, it shouldn't draw more than 120mA.

 

[headfone]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif Dropping 5V in from the PSU to opamps is A LOT, is the PSU not regulated? It should be.

Yes, I think Tangent noted this, too, in an earlier msg. But where is it dropping?! All PSUs I use are regulated -- i'm an audiophool (call it a weakness)!
Quote:

At 13.5V feeding OPA637, there could be some clipping at high volumes on bass notes.

Do you mean +/- 9V, based on the method of measurement you noted: +Vs or -Vs WRT Ig (input ground). Quote:

330mA is PLENTY for a PPAv2, it shouldn't draw more than 120mA.

See my notes in the prev. msg (I edited it after your reply posted).

 

[FallenAngel]

How is your 62045 transformer wired? Details.

Measure V+/V- to ground with the 18V supply, what do you get?

 

[headfone]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif How is your 62045 transformer wired? Details.

Transf. wiring: For 115VAC, using the secondaries green and red (brown and blue are used to power a non-PPA application).
Schematic from AMVECO
Quote:

Measure V+/V- to ground with the 18V supply, what do you get?

+5.6v and -5.6v.

 

[FallenAngel]

Quote:

Originally Posted by headfone /img/forum/go_quote.gif Transf. wiring: For 115VAC, using the secondaries green and red (brown and blue are used to power a non-PPA application). AMVECO, INC: Miniature Low Profile Transformers +5.6v and -5.6v.

Is anything getting warm/hot (including the transformer)?

OPA637 will likely clip at 11.2V so lets just find out what's wrong with the PSU.

How are the secondaries wired?

What is the AC output of the transformer before going to your DIY PSU?

What is the output of your DIY PSU (not connected to PPAv2 and while connected to PPAv2)?

 

[headfone]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif Is anything getting warm/hot (including the transformer)? OPA637 will likely clip at 11.2V so lets just find out what's wrong with the PSU. How are the secondaries wired? What is the AC output of the transformer before going to your DIY PSU? What is the output of your DIY PSU (not connected to PPAv2 and while connected to PPAv2)?

Transf. runs cool. The LM317s run warm, which is normal, I think, considering the current draw -- I use several of these PSUs for other apps (CDPs, DACs, etc.) and they all run warm.
Secondaries wiring is noted in my last post: I use the red and green wires in this schematic. The secondaries are not wired series or parallel.

AC output of transf.: 20VAC (I use two cascaded LM317s, and each drops 1.25V)

Output of DIY PSU -- not connected to PPAv2 and while connected to PPAv2: 17.5V and 17.5V

 

[JamesL]

I'm a bit behind..
With 20vac from your transformer, wouldn't you have close to 28v or so after the rectifier. You'd be dropping a lot more than just 2.5v from the lm317's to regulate down to 17.5v.

Either way, that's a 6v drop from the supply to the opamp pins.

I have to sleep, but for reference, the opamps in my PPAv2 see a effective +/-13.5v from a 30v supply(29.5v after the diode).

Hope you can work things out.

 

[FallenAngel]

Well, if you're using only one of the secondaries (Red and Green), you're only able to draw getting 113mA. Connect the secondaries in parallel (connec Green+Brown and Red+Blue). Then you'll get the 227mA.

Try that and measure again.

Something very funky with the PSU. 62045 is 2x22, you should be getting MUCH higher voltage.

 

[error401]

Are you sure the JFETs aren't dropping the remainder of the voltage?

 

[00940]

let's see...

You have a DC source of 17.5VDC.

On the PCB of the PPA, you have a protection diode. After it, we're at 16.5VDC roughly. It's the voltage you should measure at the buffers.

Then you have the jfet isolating the opamps. 16.5 - 11.2 = 5.3VDC is what it seems they're dropping. Now you have to check if it's really there that voltage is lost.

Anyway, you are definitely on the low side for an amp with opa637. 24V is way better than 17.5V for high impedance cans.


edit: error401 beats me

 

[TzeYang]

Okay seriously,

Are you sure it's not your source?

I mean the PPA V2 has plentiful of voltage to drive the HD650 to INSANE levels. If you actually heard clipping, I sure am surprised you arent deaf yet.