[error401]

Parts arrived today (I love DigiKey!) and I had some time to work on this project.

DAC is completed and working fine, as expected. I've also wired up the CMoy on some annoying pad-per-hole board, but I can't test it until I get a working power supply.

The DCP02 chip works easily enough, no external parts needed other than a couple filter caps. I didn't forsee, however, that its lack of regulation would give me over 50V at the rails unloaded. I did some quick tests and it looks like even with a 4mA load it drops to about 17V on the rails. This should be reasonably safe for the opamp, but I'm a bit worried about startup. Voltage delta between the plus and minus rails is about half a volt, this performance is I guess OK, but could be better. My meter measures zero ripple, but I don't trust it at low ranges, especially in AC, it's a cheap Chinese knock-off. If I did this again, I'd probably get some suitable regulators and drop this down to +/-12V regulated. The voltage seems to fluctuate quite a bit, probably +/-200mV, but I haven't left it running for long enough to warm up and stabilize so it may be better after some time.

This weekend I'll try to test out the whole system and see how it sounds; it's basically all wired up right now except for the power supply and interconnects.

Anyone got hints on how to work with a Hammond enclosure with Al endplates? I've worked with the plastic ones before and those are bad enough...cutting a square hole in the Al is daunting me.

Oh, and why are standoffs so darn expensive at DigiKey? 0.25" #4-40 M-F standoffs are like 60c each, this seems excessive for such a basic fastener...

 

[MisterX]

Quote:

Anyone got hints on how to work with a Hammond enclosure with Al endplates? I've worked with the plastic ones before and those are bad enough...cutting a square hole in the Al is daunting me.

That is a major pain in the butt.

Do you have a Dremel?
A Dremel + A small high speed cutter bit makes it a lot easier.

 

[Yen]

Pls post pics and more info, especially schematics around dcp02 chip, when you are ready. I am also intrested usb driven amps.

Quote:

Originally Posted by error401 /img/forum/go_quote.gif The DCP02 chip works easily enough, no external parts needed other than a couple filter caps. I didn't forsee, however, that its lack of regulation would give me over 50V at the rails unloaded. I did some quick tests and it looks like even with a 4mA load it drops to about 17V on the rails. This should be reasonably safe for the opamp, but I'm a bit worried about startup. Voltage delta between the plus and minus rails is about half a volt, this performance is I guess OK, but could be better. My meter measures zero ripple, but I don't trust it at low ranges, especially in AC, it's a cheap Chinese knock-off. If I did this again, I'd probably get some suitable regulators and drop this down to +/-12V regulated. The voltage seems to fluctuate quite a bit, probably +/-200mV, but I haven't left it running for long enough to warm up and stabilize so it may be better after some time.

 

[Emon]

Quote:

Originally Posted by error401 /img/forum/go_quote.gif Anyone got hints on how to work with a Hammond enclosure with Al endplates? I've worked with the plastic ones before and those are bad enough...cutting a square hole in the Al is daunting me.

Aluminum is often easier because it's harder, but not too hard. With plastic, it's very easy to file the hole too large. The endplates should be thin enough that you can use a nibbling tool. Perfect for square holes. They cost about $15 for a cheap one, $30 for a good one. Try "nibbler" in Froogle. Dress the edges with a very fine, very small file, and take it slowly.

 

[error401]

Spent some time today and got everything wired up, but at the moment its still not in a usable state.

It seems that I might have to abandon the DCP02 chip and just try to get it to work from 5V. With 300R headphones it seems to be fairly stable, and the sound quality is decent, though fairly noisy. However, with 32R phones that I plan to use with it, the performance is much worse. Adjusting the volume control generally results in scary sounding clicks and pops, and occasionally the audio will fade to a hiss and then come back again. Sometimes it's possible to completely destabilize the circuit and a loud hiss results that doesn't go away when the volume is fiddled with. I think it's oscillating, but it doesn't seem to be getting overly warm. My meter has a freq/duty cycle setting and measuring the output gets a pretty solid result around 300kHz, so at the very least I think it's amplifying a component I don't want, and it seems to be fairly strong.

The voltages are also all over the place, adjusting the volume control can generate about a 5V swing on both rails, and the rails are matched even less well than before.

I've added additional bypass caps at the opamp and this doesn't seem to help. Does anyone have any other ideas? I'm not even too sure what's happening here, and unfortunately I don't have a scope. I've tried placing bypass caps around in various places in the circuit, as well as adding additional ground paths and this doesn't seem to help - much.

Using a battery and rail splitter, the amp performance is as I would expect. But I can't do this with the 5V input due to grounding issues. It does work without the grounds connected, but it's very noisy.

 

[splaz]

Well for my end plate I drilled a pilot hole then a hole same diameter as the holes dimensions, then filed out the corners with a triangle file, then used a flat file to get each side, well nice and flat. Then I rounded the corners out ever so slightly with a round file.

About 10-15 minutes work.

 

[error401]

It lives!

I played around a bit more tonight and tried a couple more things, and managed to get it sounding pretty nice. There's a bit of noise in the audio band, it seems to be heavily affected by stray capacitance (moving your hand near it changes the pitch a lot), so I think it's definitely noise from the DCP02 - the datasheet mentions that stray capacitance at the SYNC pin will change the switching frequency. It's not *too* noticable with music playing, but it could definitely be annoying during quiet passages. The amplitude seems less than the junky Kenwood receiver I'm using as my main headphone amp right now, but it's definitely not 'white' and thus is more annoying. I to reduce the gain in the CMoy - it's much too high for my headphones and that should improve the SNR. I'm also going to try bypassing the big electrolytics in the amp with some fast film caps and see if that helps with the noise, but I'm really not an analog guy - any suggestions?

To correct the issue I was having before I connected the DCP02 power supply directly to USBV+ rather than the post-regulator 5V rail on the DAC. I was hoping the regulator would kick out some of the USB noise, but it doesn't seem to be able to source enough current to keep the DCP02 running; it seems that it may have been going into a protection mode and then going near to saturation and back again in a loop due to the inability to source enough current. Maybe I should've gone with REG102s... I also connected all the grounds in a star configuration, with the DAC's OG port the star point (I'll probably move this to the chassis when I case it up). I'm now getting a pretty even +/-15.9V, the rails are maybe 50mV off each other.

Yen - here's a pic of how it's wired on my 'bench' right now. The power supply schematic is straight from the datasheet and is extremely simple. 2.2μF ceramic across the input pins, 1.0μF across each of the two outputs. There's no additional circuitry required for this chip. I might play around with capacitance a bit and see if I can quieten the noise, but we'll see. Note that I, in my infinite absent-minded wisdom, changed my wiring colour codes halfway through the project. At the start, green was ground, by the end I'd swapped it for blue.

 

[achina]

Hi error401,

I just finished my USB powered Pimeta/Alien DAC this past weekend. I used a TDK DC converter (CC1R5-0512DF-E). It is a +-12 which I bridged for a 24v output. I then ran that into a tread and dropped it to 20v. Everything combined draws 380ma off the USB +5. The TDK is rated for 60ma output and I am drawing 50ma. This is with 32ohm headphone attached and listening at much louder than normal (for me) levels. If I crank the volume up to max it approaches 420ma on the USB line and 60ma on the DC converter output. On extremely loud bass heavy music at maximum volume I have seen it draw 540ma on the usb line and 70ma on the output of the converter. When I first plug it in there is a surge current over 500ma for a split second. Not any worse than my USB powered harddrive.

The Pimeta is running OPA627 in L/R and OPA134 in ground. Biased into Class A. 1K bandwidth resistors on the BUF634. It's close to the maximum current draw right now, with a tiny safety margin, for normal/loud listening levels.

Your project is definitly doable with the amount of current you have available. I was thinking of adding digital volume to the Pimeta but there isn't enough power left to add anything. You could do it with your setup. Looks like you could fit the whole thing in the tiniest hammond case.

You could try running your lines through ferrite beads. They may help reduce some of the RF from the converter.

I think if I did it again I would get the CC1R5-0512SF-E which is a single 12v (125ma) output. Regulate it down to 10v, and feed that to the Pimeta. Then use the +5 off the rail splitter to power the DAC. Probably have to change the op-amp to something else though. +-5 is perfect for the PGA2311.

I look forward to seeing your final configuration.

Andrew

 

[error401]

achina:

I basically ended up just boxing the setup pictured. It's very susceptible to EMI, the noise changes significantly just moving it around on my desk. I think maybe I need to ground the case better; it's hard to solder to the Al and I'm not convinced it's conducting. I should try some ferrites as well, but I don't have any on hand so it'll have to wait until I make an order or go to my local shop and actually remember to buy some.

It's fairly noisy, but as long as I find a decent spot on the desk for it, it sounds pretty good. It may still be oscillating though, I get weird distortion if I change the volume control too quickly, but it seems to stabilize and go clean again within a second or so. I wish I had a scope...

 

[achina]

I wonder if the noise could be coming through the USB cable? I have seen some with ferrites built into them. My cheap USB powered speakers (uses USB for power only, analog in) pick up all kinds of RF.

Andrew

 

[error401]

Quote:

Originally Posted by achina /img/forum/go_quote.gif I wonder if the noise could be coming through the USB cable? I have seen some with ferrites built into them. My cheap USB powered speakers (uses USB for power only, analog in) pick up all kinds of RF. Andrew

It's possible. I'm pulling power from after the ferrites and small filter cap on the AlienDAC though; I was hoping most of the nasty USB noise would be filtered by this. I will try to source a clamp-on ferrite ring to place around the USB cable and see if that makes a difference. Some ferrite rings on the opamp power supply lines might help too if I can get some at my local shop.

It's not too bad though, the noise doesn't bother me at all.