The Objective2 (O2) Headphone Amp DIY Project

[Splatem]

I am building a desktop case for my o2, where all sockets and buttons will be mounted off the PCB. I was wondering if twisted pair cat5 cable is the best option for wiring, or if shielded audio cable would do a better job at isolating noise. Twisted pair is effective at isolating hum for balanced operation, but I am unsure how it helps minimize noise in this unbalanced application?

[Atilla]

Originally Posted by agdr

For the input signal to make it through to the output without power, given the common ground, it would pretty much have to either punch through the signal path, or out to the power supply rails via the internal circuitry of the op amp(s).

No change from the volume control should rule out the signal path. The power supply bypass capacitors C8 and C9 should also do an effective job at bypassing any signal current on the supply rails, leaving just whatever voltage results from signal current x capacitor series resistance.

Well that is an interesting one! I don't have any ideas.

Well. I've looked at it, twisted and turned it, inspected all areas for potential mistakes and I can't say I can find anything that might give me a clue about how could I possibly get mysterious output with power off. What a funny, strange thing to debug.

If anyone accidentally hears output, when the amp is off, let me know

[Atilla]

Hm, actually. There is 1 thing. I forget what I gain resistors I set exactly (yeah, I know) but they were 2x and probably 6x. The output from my sound card is high enough to drive the 6x into clipping immediately. I haven't really measured it, since it's just fine on 2x with my headphones.

If I lower the volume from the PC a bit, then I don't get this coupled in.

Still ... what the hell is going on exactly?

[agdr]

Originally Posted by Atilla

Hm, actually. There is 1 thing. I forget what I gain resistors I set exactly (yeah, I know) but they were 2x and probably 6x. The output from my sound card is high enough to drive the 6x into clipping immediately. I haven't really measured it, since it's just fine on 2x with my headphones.

If I lower the volume from the PC a bit, then I don't get this coupled in.

Still ... what the hell is going on exactly?

Here is something else to try testing. If you have a DMM measure the DC voltage between pins 4 and 8 of U3 when it happens (with the amp still off). If the incoming signal is large enough that the amp would be clipping if it were on, maybe the input voltage is making it through the U1 op amp internals to the power supply rails and actually charging the rails to a volt or two. Looking back into the mosfets with the amp off would be an open circuit, cutting off the LED as a DC load.

If that is what is happening the solution may be adding more power supply load after the mosfets. A 4.7k resistor between the positive end of c8 and the negative end of c9 under the PCB would pull about 5mA at +/-12.0Vdc, or 400uA at +/-1.0Vdc.

Something else to try would be back-to-back 2.2V zener diodes across R14 and R20

http://www.mouser.com/ProductDetail/...PmeqxzyI0f8%3d

That would hard-limit the incoming signal to around +/-2.8V peak. Back to back red LEDs may work too.

[Atilla]

Back-to-back LEDs would be fun as they'd be a warning indicator. Well, if the source can feed them enough current to light up.

I'll open and measure it when I bring it to my tools. Sounds quite plausible.

[Twinster]

Got the O2 working like a charm now. It was a bad ground. I sand down the inodize paint properly this time and mount the RCA jack like you recommend and it fixed my hollow sound.

Thanks for the help!

Originally Posted by Twinster

So you also used a ground wire to each RCA jack.

My wires are about 2" long and I have also twisted them around the signal wire. Maybe I need to sand down more inodize paint and I'll redo the nut order like your.

Thanks again.

[sgrossklass]

Originally Posted by Splatem

I am building a desktop case for my o2, where all sockets and buttons will be mounted off the PCB. I was wondering if twisted pair cat5 cable is the best option for wiring, or if shielded audio cable would do a better job at isolating noise. Twisted pair is effective at isolating hum for balanced operation, but I am unsure how it helps minimize noise in this unbalanced application?

Definitely go for shielded here. Make sure you don't add any ground loops when wiring up the volume pot (or anywhere else for that matter), or just extend that one mechanically.

[billyk]

Originally Posted by Splatem

I am building a desktop case for my o2, where all sockets and buttons will be mounted off the PCB. I was wondering if twisted pair cat5 cable is the best option for wiring, or if shielded audio cable would do a better job at isolating noise. Twisted pair is effective at isolating hum for balanced operation, but I am unsure how it helps minimize noise in this unbalanced application?

Shielded is good, I get all my wire from this guy items in John's Silver Teflon Wire Shop store on eBay! It is just so nice to work with. I have been buying this wire from him for 10 years!

[jfrana]

I do know that a click and thump is expected on turn off and on (as per documentation), and while it is not startlingly loud, it is a solid click and thump. Voltage measurements are close to expected (if a tad low throughout, especially U2-p7 at 8.5v). Hopefully it is nothing to worry about. Voltage at output is under 4mv, with ac, no batteries, and no input. Measuring voltage at on and off, the most extreme value I have seen was -7 volts at turn off which quickly comes below a volt and decreases to < 10mv within a minute.

All good? Would you risk $1000 headphones without further testing?

With headphones plugged in it is an entirely different story. The highest I saw was -170mv and dropping to 0 within a second of turning off. The turn on max was less than -100mv and dropping to 3.4mv within a second.

Yes the -7v was a surge for a moment. On turning off, with probe on P2 and headphones unplugged, the DMM has -5v (or -6.5 or 2.1) immediately on turn off and is below 200mv within a couple seconds. On turn on, P2 comes to 3.4mv within a second and holds steady, never to rise above 4mv until the surge at turn off.

U3, pins 4 and 8 on turn on nearly immediately rise to -11.7 and 11.7. At turn off, both drop below 1v within a few seconds. (Same observation from U4, pins 4 and 8).

Thanks agdr for you attention and shared knowledge. This project has been fun and educational.

After resoldering and many tests, measurements remain the same. Is that on/off transient spike of up to -7v (with no headphones) likely to damage headphones? This might be a question for RocketScientist, I have completed the tests indicated in the "details" page.

"Check the U2 pin 1 and pin 7 voltages" p1= -11.72, p7=8.5
"battery removal test" U4 pin 4 and pin 8 are below 1v at .68
with the amp on the stable output voltage is below 4mv
But still up to -7v surge at turn off from P2 square to each channel of P2.

Any concerns? (Are there tests to determine if Q1, Q2, C1, C16, and/or C21 are damaged? Do my U2 readings indicate that mine likely are?)

[agdr]

jfrana - yep that is definitely one for RocketScientist or someone else who knows a lot more about transient voltage effects on headphones than me.