The Objective2 (O2) Headphone Amp DIY Project

Unfortunately solder bridges wouldn't show up well in a photo. Best thing to do is use a magnifying glass and take a look at them directly. Especially the solder joints on U6 - it is very hard to see a bridge between those pins without the magnifier.

You should get -24.9V or so to ground (metal shell of the gain switch) at the non-banded side of D4, going into the regulator.

I'm a bit suspicious of the PCB at the moment. I have a spare one here. I thought I'd check out the traces.

It seems that the topmost pin (relative to the jack) of the U6 pinholes is shorted to both the top of pin of C7 and the ground of the jack. I first noticed that the top pin of U6 shorts to the middle pins of the battery area (which are ground yes?)

Looking at the PDF of the circuit, that doesn't seem right to me.

The center pin for U6 does NOT short to the ground where the batteries are. But it does for U5. Fishy?

EDIT: Tried to edit and failed. I realise now that the central pin of U6 is not ground (thanks to the trace diagram). I have ~106 Ohms from the bottom pin of U6 (furthest from power plug) to ground. That must be going over R1/R2 in parallel? (1/220+1/220)^(-1) ~ 100
 
Last edited:
I've found a short from the non-banded side of D6 (and R2) to ground. Now I need to work out how it got there.

Good work! :) Yes, with the batteries out and the power switch off that D6-R2 junction shouldn't have a (shorted) path to ground. Keep in mind that one leg of the power switch also connects to that junction. Maybe you have a solder bridge to ground on one of the power switch terminals. Would be easy to do - those are very small solder joints.

You are right, the top pin of U6 is ground, the middle is the voltage input and the bottom is the voltage output. The pinout of U6 is different from U5. I just did a resistance measurement (not the diode check function) from from the ground pin to the output on a LM7915 and I get about 18K either way. With a 1N5818 diode I get about 2K in one direction and 250K in the other. So 106 ohms to ground on the output pin of U6 is way to low.

If it were me I would start unsoldering things. Unsolder the unbanded end of D6 and see if the short goes away. If not do the same for the end of R2 that goes to the same junction. If not, then the unbanded end of D5, etc. until you find out what part or trace has the short.
 
I'm a bit suspicious of the PCB at the moment... The center pin for U6 does NOT short to the ground where the batteries are. But it does for U5. Fishy?

Why would you find that fishy? Have you looked at data sheets or pin-outs for the 2 IC's? Why would you suspect the PCB? While anything is possible, the PCB's are identical and work perfectly for virtually all builders. If it isn't working for you then in all likelihood you have made an error. Perhaps you inserted a part incorrectly, perhaps you made a bad solder joint.

The symptoms you describe (0V on voltage regulator output, hot regulator) indicate exactly one thing: a short to ground from the regulator output. It is not the circuit board or a mysterious defective component, it is a construction error. Find it and fix it, that's what DIY means. All you need is a multi-meter and patience.

I am not trying to be a jerk, but I get frustrated when people keep trying to blame everything but their workmanship despite all contrary evidence. The PCB is fine, the circuit works as advertised, you made a short somewhere.
 
How hot is too hot to touch? It is a voltage regulator: there's going to be some heat generated. What power supply are you using?

Well, yes and no. With no batteries and no load, those regulators pass a few tens of milliamps. They do not get hot, they barely get warm (a couple degrees above ambient). A newborn could hold that regulator all day without discomfort.

If the regulator is actually hot, or even noticeably warm, something is wrong, either a dead short or a downstream component shorted or damaged.
 

Antoinel

Member
Paid Member
2012-01-01 10:23 pm
Pennsylvania
The Objective (O2) Headphone Amp

This volume-controlled and high fidelity headphone amp (O2) can be used to exemplify a specific facet of the management of Gain Structure per the teaching of the article by DIYer Pano which is entitled: What is Gain Structure? Given that this headphone amp is driven by a pristine line level signal from a high quality CD player, then its output can be fed directly to the input of a power amp via a 1/4 inch headphone to RCA stereo conversion plug (sold at Radio Shack). The input impedance of common power amps is normally high; for example greater than 20 Kilo Ohm. Thus, the operational status of this headphone amp is an ideal quiescence. It is doing minimal work in this high impedance load. It is conceptually akin to that of STASIS (R) of Threshold Corporation; inventor Nelson Pass. A fall back to (O2) maybe the volume-controlled headphone amp which may happen to be internal to the CD player at hand. This amp of unknown fidelity relative to (O2) also operates in an ideal quiescent state driving the same power amp of high input impedance.
 
This volume-controlled and high fidelity headphone amp (O2) can be used to exemplify a specific facet of the management of Gain Structure per the teaching of the article by DIYer Pano which is entitled: What is Gain Structure? Given that this headphone amp is driven by a pristine line level signal from a high quality CD player, then its output can be fed directly to the input of a power amp via a 1/4 inch headphone to RCA stereo conversion plug (sold at Radio Shack). The input impedance of common power amps is normally high; for example greater than 20 Kilo Ohm. Thus, the operational status of this headphone amp is an ideal quiescence. It is doing minimal work in this high impedance load. It is conceptually akin to that of STASIS (R) of Threshold Corporation; inventor Nelson Pass. A fall back to (O2) maybe the volume-controlled headphone amp which may happen to be internal to the CD player at hand. This amp of unknown fidelity relative to (O2) also operates in an ideal quiescent state driving the same power amp of high input impedance.

Um... OK?
 
Found it!!! I desoldered R2, D6, D1, D5, and the gain switch but couldn't find anything wrong. All the joints looked visibly good. (Not any more...)

I was following the trace around looking for anything, when I noticed that a joint for the big cap on the side of board nearest D4 was just touching the trace running to the gain switch. The solder mask there is a little bit scratched. I removed some solder from the joint and the short is now gone.

Listening at the moment. The amp has passed all the voltage tests on NwAvGuy's blog. Only defect I can notice at the moment is channel imbalance at the very bottom of the volume pot.

Do you think it's most likely that I scratched the PCB, or that I killed the solder mask with heat? How hot is too hot? Do I need to keep my iron below 350 C? Is there anything that can repair these little trace scratches?

@nezblue: I understand your frustration at us beginners. I did edit my post almost immediately after posting though, as soon as I saw that I was wrong. I always expected it to be my fault but was baffled momentarily by the different pin layout on U5, U6. I know that people come and ask silly questions, but I'm willing to put in the work to learn and know better, eventually.

@agdr: Many thanks for your help - I was going to have a hard time finding this error. Hopefully I didn't make more in there!

I am not trying to be a jerk, but I get frustrated when people keep trying to blame everything but their workmanship despite all contrary evidence. The PCB is fine, the circuit works as advertised, you made a short somewhere.
 
Listening at the moment. The amp has passed all the voltage tests on NwAvGuy's blog. Only defect I can notice at the moment is channel imbalance at the very bottom of the volume pot.

Hey congratulations on the working amp! You probably meant "power switch" instead of "gain switch". Yep that trace is the one going from that shorted node to the power switch and that joint with C3 is ground. The capacitor lead gets grounded on the top of the board via a through-hole. :)

Imbalance at the very low end of the pot is actually normal. Analog pots are only accurate to 20%, even the good ones like this Alps pot. That innacuracy really shows up on the low end. To do better you would have to go to a stepped attenuator which is pricey and the size of a golf ball.

The scratches on your traces and soldermask won't hurt anything and won't affect the sound. The biggest problem there, like you say, is allowing solder to stick to traces in places where it shouldn't.

Hey one more thing. Your D5 might be shorted. With batteries out and power adaptor unplugged try testing it with the DMM diode test function in both directions and compare to tests on D1. Thinking about your previous measurements that 106 ohms you got on the output of U6 was probably measuring what was left of D5. :) When the anode of D5 got shorted that would have pulled short circuit current through U6 and D5. U6 is short circuit protected, but not so much for D5.
 
Last edited:
Continued tech support, great service. :D

I see voltage drops of ~166 mV for both D1 and D5. (More like ~145 mV for D6 and D2). Safe ranges? Looks like I may have dodge a bullet?

RE: the volume. Since I'm running from a Xonar D1 in 24 bit mode, I'm probably better off lowering the volume a little at the PC side than using the pot below 9 o'clock or so, I suppose. Better to lose something in the digital regime where I have some headroom. I'm using Fostex T50RPs and can't handle full volume from the PC + O2 with the pot near 12 o'clock.

Hey one more thing. Your D5 might be shorted. With batteries out and power adaptor unplugged try testing it with the DMM diode test function in both directions and compare to tests on D1. Thinking about your previous measurements that 106 ohms you got on the output of U6 was probably measuring what was left of D5. :) When the anode of D5 got shorted that would have pulled short circuit current through U6 and D5. U6 is short circuit protected, but not so much for D5.
 
Continued tech support, great service. :D

I see voltage drops of ~166 mV for both D1 and D5. (More like ~145 mV for D6 and D2). Safe ranges? Looks like I may have dodge a bullet?

Yep those look good, given that they are Schottky diodes.

I agree, you would probably have better luck with channel equality controlling the volume on the digital side. If you made your O2 with the standard 2.5x and 6x gains you might want to consider changing the resistors and backing that down to 1x and 2.5x. That is what I'm running my O2s at. The higher gain pair apparently is most useful for some portable players with 0.5Vrms outputs.
 
Last word that I can find from NWAVGuy in response to " Could you give a rough estimate of when we might see the ODA/ODAC for sale? Thanks in advance!":

"...the plan is to release the design before the end of March. That's only a goal, no promises. It will take at least a few weeks beyond the documentation release before there are completed boards/amps for sale. But it should be well before summer."
 
Found it!!! ...
Listening at the moment. The amp has passed all the voltage tests on NwAvGuy's blog.

I am genuinely glad that you found the problem and fixed it, and that you have a working amp. I am also sorry that I was a bit grumpy the other night. We are all here to help, and I dare say most of us here have benefited from help freely offered by others on this forum. I know I have. If it helps I really just wanted to get you looking in the right direction and not distracted by red herrings.

Good catch, by the way, finding that solder bridge, those can be very difficult to locate.