Alien DAC v1.1 Construction Thread - Page 55

Now that I am playing around with my Alien DAC more, I am experiencing some issues that I ignored before but is probably real. Most of the time, if I hot plug the USB cable, it works fine. But sometimes doing so would result in crackly/scratchy playback when the volume level is low. And other times it would only produce a loud random digital noise during playback. In each of these cases the OS recognizes the Alien DAC PCM2702 just fine and there is no indication that anything is wrong other than the sound. The only way to clear up the problem condition seems to be to unplug the USB cable and replug it again.

I tried a couple of different USB cables and on different computers so I could rule those out as the problem.

My Alien DAC is USB-powered. I wonder if there is an internal power-sequencing issue during hot plug? Anyone else experience this?

Edit: Oh, I should mention that this is on Windows XP SP2 with ASIO4ALL 1.8 installed. On Linux (Kernel 2.6.13) I also experience these problems, but the crackling noise problem can sometimes be "fixed" just by changing songs (depending on the player). The digital noise problem seems to always require a re-plug, though.

Amb,

Except for your description of a "loud random digital noise," all of this seems fairly typical from what I've noticed. As you state, uplug it and then plug it back it in. That usually works. In my case, I typically have to readjust the PC volume level, or in a worst case, select the Alien DAC as a source again (USB Speakers). It will also "clog up" Foobar, and that may need to be killed and re-started as well.

I would note that I've experienced these same idiosyncracies with the M-Audio Transit. It seems nothing more than a minor irritation and considering the high-level, complicated OS decisions that must take place with switching drivers, communication interface, and PC sound selection, I don't believe it rises to the level of a fault, IMHO.

If the Alien DAC is connected on boot up, it seems to work flawlessly without exception.

AMB,
I've been playing a bit with my Alien DAC as well. I have used it on two desk tops and my laptop.....it is recognized as USB Speakers on all three and they behave similarly. I did have to adjust the computer volume up for all three, as it defaulted to 50% upon recognition. My gain 11 Pint was okay at that level, but my gain 8 Pint did not have suitable volume with adjusting the computer.

I have had no problems with scratchy sound or random digital noise so far, either when hot-plugging or pre-plugging. I'm running XP on all of the computers, and playing back from the Alien DAC through one of my Minified Pints. It acts the same with USB power or external power. I do get a thump powering on the DAC and thump with a diminishing whizzz when powering off, but that is easily combatted by turning down/off the amp vol prior to switching the DAC on/off.

I'm posting a couple of pics on the Post Pics thread. I copied your switch/power jack layout front and rear, but did not shorten the cases.....you must work with small layouts better than I

Thanks rjkdivin. I too think that something is not quite right with my Alien DAC. Unlike tomb's experience, my M-Audio Transit almost never exhibit these issues when hot-plugged (or for that matter, the M-Audio Firewire Audiophile), so I'm inclined to think the problem is specific to my Alien DAC. I am going to try and build a second one to see if it behaves any differently.

Btw, I re-cased my Alien DAC into a black Hammond 1455C802, not shortened, and with custom FPE panels to match my Mini³. I also eliminated the switch and the external power jack, as my previous tests showed no benefit with external power.

Do I order this at your site?

as mentioned above, i'm having some trouble with (predictably) soldering the PCM2702 down. i have a couple of questions regarding this process. first, some background:

i started out by soldering one of the corner pins down to get the chip in place, then continued to try to do the pins one at a time. this worked for a few pins, but i was frequently bridging pins and having a hard time clearing them (i tried the trick of placing my iron horizontally across the pins and pulling away - it just doesn't seem to work as well on a chip with such a fine pitch). i struggled with it for quite a while yesterday, then eventually got frustrated and decided i'd try the "flood-and-suck" method. unfortunately, this isn't working at all for me - i guess i just can't get my desolder pump in there quickly enough to catch the solder while it's still molten. this leads to my questions:

1. what's the best way of clearing a BUNCH of pins that are bridged? any tips on making the "suck" part of "flood and suck" actually work? i've been removing small amounts of solder using the above horizontal iron techique, but it's terribly slow going, and seems to stop working JUST before i've removed enough solder to clear a given bridge.

2. how much heat can a PCM2702 take? in the process of attempting the flood-and-suck method, the chip got too hot to touch several times. with all the pins bridged, as they are with this method, the chip is basically acting as a heatsink for the giant gob of solder on the pins. should i assume that i've fried it and start anew, or do they tend to be pretty resilient?

3. relating to the above, are there any straightforward tests that can be done while the chip is soldered down to determine whether or not it's still in working condition?

4. does thin-gauge solder make a difference when soldering difficult chips such as this? i'm using 0.031", and am wondering whether it's worth picking up some 0.015" for my next attempt (or if i ever manage to salvage this one).

Your experience with flood and suck sounds like mine. The key to the horizontal method is to apply the solder to the soldering iron - not to the chip/pins. Then what you're doing is actually wiping the pins with a "wet" iron. Use as little as you possibly can get away with to make the pins "shiny."

0.031" is OK if you apply it to the iron and wipe, but it may be more difficult to control the volume of solder. I've heard that 0.015" is almost impossible to work with, breaking at the slightest provocation. 0.020" or 0.025" or something like that might be a better compromise - from what I've heard. I have one or the other of those, but can't remember which.

As for the chip, the less heat the better. I use the Hakko 936 and make sure the temp is not set above the maximum recommended soldering temp. I don't know if that helps, but if you have a regular iron, it will continue to increase in temperature, depending on how it's used. You can refer to TI's data sheet on the PCM 2702 for recommended soldering limits:
http://www.ti.com/lit/gpn/pcm2702
Looks like it says 260 deg. C. as the maximum soldering temp on Pg. 3. That's probably conservative and it also says 5 secs, which when you think about it, is an awfully long time to be in contact with the chip. You should solder in a "drive-by" fashion on the pins - hitting them with quick whacks - at least that's how I did it.

'Course, I've been extraordinarily lucky to have done it twice now.

Wow, 260C is pretty low. I just assembled mine this weekend and used about 650F. It was my first time doing an SMT project and I lost C1 in it's packaging entirely (waiting for Jeff to get more) and had a learning experience with tweezers...spring loaded == bad. I also have to thank Jeff for helping me out and responding to my e-mails of (oops, I just lost C21, and there goes C11 too before I switched to manual tweezers).

I did a test run and all was OK.

For soldering PCM2702 I found it mostly easy. Bear in mind this is this is the fourth SMT item I've ever soldered, so take it as some notes from a n00b for other n00bs.

The items that bothered me the most were actually the smaller SMT items like the resistors and caps. Trying to center mount them and then finding enough room to get solder onto the pads was difficult. In my earlier attempts I would move the resistor all the way over so I have more room to apply to the pad, but I got the knack of it towards the end and those are laid down a little more evenly, relative to the pads.

Anyhow, for the PCM. I used 0.031" Radio Shack solder. I did the wipe method, but not fully. I partially flooded 1 pad, then used that to draw across the others and applied more solder to a pad and drew again. Note, it's the pad, not the pins. I never applied the solder directly to the pad or pins. Always to the soldering iron tip first then to the object. When applying solder to a pad, I put solder on the tip of my iron and touched the pad, letting fluid dynamics take it's course.

Sometimes it got messy from a momentary spasm or too much solder on the tip, but using a wick to clean up the excess and wiping again to reflow and clean. If some solder got behind the pins, I cleaned my tip good, then wiped the upper bends that elevate from the pads. That melts the solder and it flows down. I used paste flux liberally too, at one point, to clean I put paste all over the pads and pins and quickly wiped and then flooded with alcohol and some brushing to remove all the flux for inspection. My pad and pins are quite clean, except for the 2 pins that are bridged to ground. I may have gone overboard with the brushing as some lettering off the ICs are gone.

I'll try to post pics some time, for evaluation.

I hope that's helpful to you, xmokshax.

the above advice has definitely been helpful. unfortunately, i should have probably asked before flooding the pads entirely.... i think i now have quite a job ahead of me if i'm going to get this sucker cleaned off and all of the pins un-bridged. i'll probably take another crack at it tonight, and if that fails, i'll be sending Jeff an e-mail for a new chip and board... :P

as far as temperature goes, i started out at ~500F, then bumped it up to 600F when joints were taking a long time to melt. as i said, the chip sometimes got too hot to touch, so i'm wondering whether it's still OK, especially since i was soldering a bit above the recommended max temperature.

600 should be fine as I did ~600-650 and everything works, at least from the short test I did. What you have to watch out for are the hold times. Keep it short and quick. For that, I like a hot iron. I find lower temps makes me hold longer to reflow, which doesn't work as well.

Use a wick and clean up as much as you can on the pads, ignore the legs for the time being. Then clean the tip and wipe along the legs that raise upwards. That will help some. Keep repeating, but take breaks in between to let it cool. It's like welding a bead on a casing with parts inside, like an axle. Do it all at once and the internal components will seize.

It wouldn't hurt to smear flux all over the legs, I believe. I didn't use liquid flux at all. I used a small screwdriver and wiped paste flux on all the pads I was working on.

Also, if you really are having issues with heat, you may be able to put a small heatsink on the chip, and hold it in place with your finger, offset so you can work on the pins/pads on one side. Hopefully that works, I've never done it. That will help dissipate some of the heat from the chip casing.

i did consider heat sinking the chip, but i don't have any small sinks lying around. i'm a bit worried that i've held too long on this chip, and i wish that there were an easy way to test whether or not i've fried it, before i put a lot of time into fixing it and solder the rest of my kit's components onto this board.

Anything metal can be used as a makeshift heat sink for this purpose. I often place a pair of needlenose pliers on a component or directly on the lead if possible to dissipate some heat. It makes quite a difference.

anyone having luck locating the REG101UA chips lately? Both RS-components and Farnell list 'em as "no longer available"... drats, I have a PCB laying here waiting to be built and wanted to order a couple more as well from Jrossel...

BOM check-up before ordering from Digikey

Last quick check of the BOM before order from digikey. I think I cleaned everything that needed to be cleaned

You don't need IC4 unless you plan to power the amp via virtual ground amp (most likely you're not).

How will you power this thing? I personally never saw a reason to use a fixed regulator because you'll have to use external power, and if you're doing that already, might as well run at 5.4V instead (hense adjustable).

If you're powering it using USB, you're still probably better off using adjustable, just regulate it down to 4.7-4.8V. You can also leave it completely unregulated, but I personally don't know how good the performance is, I think it's pretty similar.

Just remember, if you're going for an adjustable regulated, you'll need 2 more resistors - check out NOTES on this page.