Haha! The 7912 was fine! I just didn't re-attach the lead after I had to cut the trace.
That's sort of funny!
Anyway, music started coming out of the system as of 12:23PM, Hell's Bells, if you must know! It's the first song in my library due to the AC/DC alphabetical boost.
So far, so good. I don't have very good speakers up here at work, so I don't know if it sounds "right" or not, but so far it is at least up to the limit of my MP3/bad speaker combo. Qualitatively, it seems to have good, solid bass, not too "boomy," with a nice, even response throughout the range of the speakers (maybe a tiny touch bright past ~8k, but that may just be my imagination). I am hearing some MP3 artifacts that were smeared with my sound card, so the "resolution" and "definition" are most assuredly a step up. No real news there! There isn't any noise at zero volume, and turn on/off thump is non-existent. Oh, and I can sit my hand directly on the heat sink, and it is barely warm after playing through half of Dark Side, so it seems as if I can tentatively call this one a winner! It definitely plays music well and is quite compact for an integrated solution, which was the goal.
Dark Side of the Moon is still playing now, and I'll put on some Herbie Hancock after this to get a better feel for it, but first impressions are definitely positive. It definitely won't play "loud" but it is more than capable of driving my office speakers "too loud" for work. Still, I may bump up the gain of the 2134 to see how it likes that, or I may just sit back and enjoy it for a while!
PS I bypassed the attenuation pot since the leads were picking up some goofiness and introducing a pretty substantial offset into the 1876. I'll have to track that down, but the pot isn't really needed since I'll primarily be controlling volume via the computer. Also, the headphone amp works perfectly!
That's sort of funny!
Anyway, music started coming out of the system as of 12:23PM, Hell's Bells, if you must know! It's the first song in my library due to the AC/DC alphabetical boost.
So far, so good. I don't have very good speakers up here at work, so I don't know if it sounds "right" or not, but so far it is at least up to the limit of my MP3/bad speaker combo. Qualitatively, it seems to have good, solid bass, not too "boomy," with a nice, even response throughout the range of the speakers (maybe a tiny touch bright past ~8k, but that may just be my imagination). I am hearing some MP3 artifacts that were smeared with my sound card, so the "resolution" and "definition" are most assuredly a step up. No real news there! There isn't any noise at zero volume, and turn on/off thump is non-existent. Oh, and I can sit my hand directly on the heat sink, and it is barely warm after playing through half of Dark Side, so it seems as if I can tentatively call this one a winner! It definitely plays music well and is quite compact for an integrated solution, which was the goal.
Dark Side of the Moon is still playing now, and I'll put on some Herbie Hancock after this to get a better feel for it, but first impressions are definitely positive. It definitely won't play "loud" but it is more than capable of driving my office speakers "too loud" for work. Still, I may bump up the gain of the 2134 to see how it likes that, or I may just sit back and enjoy it for a while!
PS I bypassed the attenuation pot since the leads were picking up some goofiness and introducing a pretty substantial offset into the 1876. I'll have to track that down, but the pot isn't really needed since I'll primarily be controlling volume via the computer. Also, the headphone amp works perfectly!
Oh, and cutting the part out is great in theory, but with the board fully populated, I can't really get in there that easily with cutters! Still, I emphatically agree that the knife blade and solder suckers are quite handy in situations like this, and I greatly appreciate the "tip" (haha, glad I noticed that unintentional little punBWRX said:
).
Well, I knew it was too good to be true! The USB DAC is pretty noisy as of right now. I am not 100% sure what is causing it, but as soon as I turned it on when I got home I noticed it straight away (I guess my office was too noisy to notice!). If the amp is on and the DAC is powered off, there is no hiss. As soon as the DAC is online, hiss like a champ.
Time to trouble shoot! (Isn't that part of the fun??)
Time to trouble shoot! (Isn't that part of the fun??
)
I remember I got a nasty hissy noise from my PCM2707 when I built it. I then discovered i'd forgotten to add a capacitor! I'd forgotten to solder in the capacitor from VCOM to ground and that really sounded bad.
To be honest I think all of the decoupling capacitors need to be as close to the chip as you can get them. The chip has an internal 5V->3.3v DC-DC switch mode supply so the decoupling is critical. This is why I suggested the 3.3V supply as you can generate that in a linear regulator and eliminate a noisy switcher.
Here's the schematic and board I designed. Notice how all of the decoupling is right near the chip. It took me many hours to get this layout on only single sided board! This worked well but sadly I negated to put it in a box, and after a year of being a bare board on my desk plugged into my headphones, the PCM2707 bit the dust My own stupid fault!
To be honest I think all of the decoupling capacitors need to be as close to the chip as you can get them. The chip has an internal 5V->3.3v DC-DC switch mode supply so the decoupling is critical. This is why I suggested the 3.3V supply as you can generate that in a linear regulator and eliminate a noisy switcher.
Here's the schematic and board I designed. Notice how all of the decoupling is right near the chip. It took me many hours to get this layout on only single sided board!
This worked well but sadly I negated to put it in a box, and after a year of being a bare board on my desk plugged into my headphones, the PCM2707 bit the dust My own stupid fault!
After playing around with it a little, I think the hiss is actually acceptable for this device, but of course not ideal for "Hi-Fi". The hiss is only noticeable when the pot is set to max gain, and the software volume is all the way down. If you get closer too moderate amp gain, and moderate software volumes, the hiss disappears.
Unfortunately, if I use the potentiometer to control volume, I seem to be inducing an unacceptable DC offset into the amp! I cannot for the life of me figure this one out!
I attached a .jpg of the pertinent area that is introducing offset in case someone sees something that I did and was being stupid at the time.
For more info, when the pot is set to 50K, there is very little offset, and when the pot is set to 0 ohms, there is very little DC offset (below 50mV). When set to moderate levels (~20K-30K), there seems to be a good deal (~1-2V) of DC offset induced! I'm going to marinade on this one for a while since it doesn't really make much sense to me!
Anybody seen this before?
David
Unfortunately, if I use the potentiometer to control volume, I seem to be inducing an unacceptable DC offset into the amp! I cannot for the life of me figure this one out!
I attached a .jpg of the pertinent area that is introducing offset in case someone sees something that I did and was being stupid at the time.
For more info, when the pot is set to 50K, there is very little offset, and when the pot is set to 0 ohms, there is very little DC offset (below 50mV). When set to moderate levels (~20K-30K), there seems to be a good deal (~1-2V) of DC offset induced! I'm going to marinade on this one for a while since it doesn't really make much sense to me!
Anybody seen this before?
David
Attachments
Thanks for the suggestion,jaycee said:
I definitely tried to get as many of the decoupling caps close to the chip as possible, but the 47uF elect.'s were a little difficult.
As I said in the post above, though, I think the noise is a little better than I originally thought since I seem to need pretty aggressive attenuation to play at moderate volumes if the computer outputs a higher volume level. Still, thanks for the alternate board suggestion. When I build my separate USB DAC I'll definitely consider some of your layout since I'll most likely be etching the boards myself that time, and single sided is a heck of a lot easier than x2 sided!
Brian, thanks as always for the suggestions!
There's no offset on the OPA2134 (one of the first things I checked), and I normally would agree with you that there would be no way R2 could cause a problem, but right now I am flummoxed.
Fortunately, I do have a more direct "symptom" as to what is causing this offset: when I leave the input pins of the LM1876 floating, they sit at ~+5V. I have no idea why the inputs are sitting at this voltage, but it does mean that if I tie them to ground via a 500 ohm resistor, it gets rid of the DC offset. It's almost like these pins are somehow picking up a potential from the USB PSU or the standby/mute voltage. I can't figure that one out!
The solution that is working as of now is to eliminate R1 and R2 and to change the value of R3 to 500 ohms. I am listening to this configuration right now with no offset and no hum, but I am concerned as to WHY there is a potential building up on the input of the 1876. Time to do some homework, I guess, but this doesn't make any sense as to why this is happening.
There is still a little bit of DC offset that varies as I change the pot, but it is always under 40mV, so that's not the end of the world as far as I am concerned. Fortunately, the 500ohm load should be pretty easy for the opa2134 to drive, so this could be the long term solution if need be.
Now the REALLY good news! This thing sounds great!
Getting some attenuation of the PCM2707 via the pot dropped the hiss WELL below that of the other stuff associated with my computer, even if I am using headphones that block most of the environmental noise. There is some transformer hum, but with the case on and the computer going, this is not even slightly noticeable.
Though the bookshelf speakers I am using are definitely NOT a set of reference monitors, they kick the living daylights out of the old set of computer speakers I did have hooked up to my computer.
I don't think this gizmo is the ultimate high-fi solution, but it definitely sounds very good for what I am asking it to do! It is nice and compact and can drive a set of bookshelf speakers louder than I would want to listen to them with good sound quality. I can't really ask for much more out of this project!
There's no offset on the OPA2134 (one of the first things I checked), and I normally would agree with you that there would be no way R2 could cause a problem, but right now I am flummoxed.
Fortunately, I do have a more direct "symptom" as to what is causing this offset: when I leave the input pins of the LM1876 floating, they sit at ~+5V. I have no idea why the inputs are sitting at this voltage, but it does mean that if I tie them to ground via a 500 ohm resistor, it gets rid of the DC offset. It's almost like these pins are somehow picking up a potential from the USB PSU or the standby/mute voltage. I can't figure that one out!
The solution that is working as of now is to eliminate R1 and R2 and to change the value of R3 to 500 ohms. I am listening to this configuration right now with no offset and no hum, but I am concerned as to WHY there is a potential building up on the input of the 1876. Time to do some homework, I guess, but this doesn't make any sense as to why this is happening.
There is still a little bit of DC offset that varies as I change the pot, but it is always under 40mV, so that's not the end of the world as far as I am concerned. Fortunately, the 500ohm load should be pretty easy for the opa2134 to drive, so this could be the long term solution if need be.
Now the REALLY good news! This thing sounds great!
Getting some attenuation of the PCM2707 via the pot dropped the hiss WELL below that of the other stuff associated with my computer, even if I am using headphones that block most of the environmental noise. There is some transformer hum, but with the case on and the computer going, this is not even slightly noticeable.
Though the bookshelf speakers I am using are definitely NOT a set of reference monitors, they kick the living daylights out of the old set of computer speakers I did have hooked up to my computer.
I don't think this gizmo is the ultimate high-fi solution, but it definitely sounds very good for what I am asking it to do! It is nice and compact and can drive a set of bookshelf speakers louder than I would want to listen to them with good sound quality. I can't really ask for much more out of this project!
Many potentiometers do tie both ground leads together and onto the mounting point as well.
There's an optional "ground loop buster" for input circuits.
In the attached photo of my "standard five" board here, you can see it as the resistor (center left) that's connecting the input star ground. It can be from 4R to 10R.
The picture shown isn't using a preamp, but the "ground loop buster" could probably work well for a preamp too. It controls the problem for whenever a grounded potentiometer contacts a grounded metal enclosure.
EDIT: The problem is most frequently seen with shielded input cable, because the thick wire can become a "more attractive" grounding point for various power circuits. Lifting by 4R to 10R insures that the input circuit isn't the path of least resistance.
There's an optional "ground loop buster" for input circuits.
In the attached photo of my "standard five" board here, you can see it as the resistor (center left) that's connecting the input star ground. It can be from 4R to 10R.
The picture shown isn't using a preamp, but the "ground loop buster" could probably work well for a preamp too. It controls the problem for whenever a grounded potentiometer contacts a grounded metal enclosure.
EDIT: The problem is most frequently seen with shielded input cable, because the thick wire can become a "more attractive" grounding point for various power circuits. Lifting by 4R to 10R insures that the input circuit isn't the path of least resistance.
Attachments
Nope. No ground loop. Seriously. During testing the pot was hanging out in free air above the amp. And the pot I am using doesn't have the signal ground tied to the chassis mount, so even with it bolted down there isn't a chance for a ground loop.
PLUS, a ground loop wouldn't account for the +5V on the input pin when it is floating, so even if there WAS a ground loop it would only account for some hum and not the main culprit of voltage on my input pins!
PLUS, a ground loop wouldn't account for the +5V on the input pin when it is floating, so even if there WAS a ground loop it would only account for some hum and not the main culprit of voltage on my input pins!
Thanks Daniel. I guess I got spoiled from my last LM3886 amp that had less than 1mV DC offset! I'm sure I'll figure the cause of the offset some day, but for now the case lid is on, so it will have to wait!
Now that this project is done, I've decided that I would do it slightly differently if I had to do it over, probably using a single ended output from the PCM2707 to drive an OPA2134 "preamp" that would double as a headphone amp. I think the TPA6120 is a little superfluous in hindsight.
Either way, here are a few pictures of the final product sans knob since I still don't know what I want to use!
First up is just a neat picture of the PCM2707 and surrounding circuit. I know, I can't keep parts straight and my soldering leaves something to be desired, but I still think it looks pretty cool!
I'm sure I'll figure the cause of the offset some day, but for now the case lid is on, so it will have to wait! Now that this project is done, I've decided that I would do it slightly differently if I had to do it over, probably using a single ended output from the PCM2707 to drive an OPA2134 "preamp" that would double as a headphone amp. I think the TPA6120 is a little superfluous in hindsight.
Either way, here are a few pictures of the final product sans knob since I still don't know what I want to use!
First up is just a neat picture of the PCM2707 and surrounding circuit. I know, I can't keep parts straight and my soldering leaves something to be desired, but I still think it looks pretty cool!
Attachments
And a little more context: Here it is where it will live, right under my monitor with the crappy speakers it is driving in the background. I am sure I will build a pair of nice speakers for my computer at some point, but that is a project for a different day!
Though the color scheme seems to match my computer stuff, the more I look at the case, the more I wish I had gone for the black anodized version. They had it at Mouser, but I couldn't bring myself to pay the extra $$$ for the shipping since I was ordering everything else from Digikey.
Oh yea, and at one point somebody asked if I would be offering boards. Honestly, at this point I know the boards work, but I don't think this design is ready for prime time so to speak, but if people are really interested, I'll be happy to fix up the board and offer the files to anyone interested. I seriously doubt there is enough interest to get a group buy together or anything, but if you want the Gerbers and bill of materials to put together on your own, I'm happy to oblige. Just drop me an e-mail or post here!
Though the color scheme seems to match my computer stuff, the more I look at the case, the more I wish I had gone for the black anodized version. They had it at Mouser, but I couldn't bring myself to pay the extra $$$ for the shipping since I was ordering everything else from Digikey.
Oh yea, and at one point somebody asked if I would be offering boards. Honestly, at this point I know the boards work, but I don't think this design is ready for prime time so to speak, but if people are really interested, I'll be happy to fix up the board and offer the files to anyone interested. I seriously doubt there is enough interest to get a group buy together or anything, but if you want the Gerbers and bill of materials to put together on your own, I'm happy to oblige. Just drop me an e-mail or post here!
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.