Doodlebug - USB Isolator

[mcandmar]

 

 
For those following this thread here is the difference i measured using a Doodlebug. I think the noise floor is a limitation of the sound card i was using to measure the results.

 

[tomb]

Very cool!  That's pretty similar to what I got awhile back - mine was maybe worse to start with, but that's because I've been using a POS netbook with an M-Audio Transit.  M-Audio quit making the Transit, I suspect because they couldn't keep up with the driver changes in follow-on Windows OS systems.  The little netbook is the only thing I have that still runs Windows XP and the Transit works fine under that.  The USB power system in the netbook, however, is another thing entirely.

 

[tomb]

I've finally been doing some work (pdf downloads available by clicking):
 


 
Hammond has been contacted and they're working on a quote for case work (1455C801), so we should be able to get cases going soon.  Next up is the website and a production proto.  I'll continue to report on the progress - hopefully, it won't be too many more weeks before we have kits available.

 

[Avro_Arrow]

 I'll continue to report on the progress - hopefully, it won't be too many more weeks before we have kits available.

Great job Tom, keep up the good work!

 

[SilverEars]

Hi, why does pupDAC pick up so much noise compared to skeletonDAC?

 

[tomb]

  Hi, why does pupDAC pick up so much noise compared to skeletonDAC?

 

• It depends on how much noise already exists on the particular PC USB buss.
• It's a much more sensitive, detailed DAC.  Some of the noise is masked by the lower-grade chip in the SkeletonDAC
• The output opamp in the PupDAC is one of the finest low-voltage opamps on the market.  Strictly speaking, it's not there for amplification, but for combining the balanced output of the DAC chip into a single-ended output.  Nevertheless, it's going to amplify every detail.  If that detail is noise, it's going to get amplified, too  There's probably more going on with the balanced-to-single-ended conversion, too, but I'm not knowledgeable enough about that to explain it.
• Another way to look at it, is that the potential of the PupDAC is much, much higher.  It just needs a reasonably good environment to ensure the best performance. 

 
In my case, the power supply of my testing environment is probably as bad as it could get and still provide operation (Acer Netbook running Windows XP).  I have wondered for years why my RMAA tests were noisier, sometimes going as far as depending on other people to test Beezar products (dsavitsk, cetoole, cobaltmute, Linux-works of Arduino fame).  As mentioned previously, AMB once suggested that my cables were picking up extraneous noise in my environment, etc.  The entire exercise in testing the DoodleBug has proven what's really happening.
 
Anyway, in an environment other than the Acer Netbook, you will see a much better base performance for the PupDAC without the DoodleBug. I would still recommend the DoodleBug for any USB-powered audio device, though. 

 

[firev1]

Nice to see progress coming along! I posted on the ODAC thread about the possible benefits of USB isolation so I believe in their usage. As for testing, single laptops in my experience are really not idea for RMAA testing of these devices. What you really want ideally separate laptops for testing or in some others case, separating the interfaces(soundcard PCI and generator/DAC on USB). This is to avoid ground loop issues. I usually have 2 laptops on battery for testing. 

 

[SilverEars]

   

• It depends on how much noise already exists on the particular PC USB buss.
• It's a much more sensitive, detailed DAC.  Some of the noise is masked by the lower-grade chip in the SkeletonDAC
• The output opamp in the PupDAC is one of the finest low-voltage opamps on the market.  Strictly speaking, it's not there for amplification, but for combining the balanced output of the DAC chip into a single-ended output.  Nevertheless, it's going to amplify every detail.  If that detail is noise, it's going to get amplified, too  There's probably more going on with the balanced-to-single-ended conversion, too, but I'm not knowledgeable enough about that to explain it.
• Another way to look at it, is that the potential of the PupDAC is much, much higher.  It just needs a reasonably good environment to ensure the best performance. 

 
In my case, the power supply of my testing environment is probably as bad as it could get and still provide operation (Acer Netbook running Windows XP).  I have wondered for years why my RMAA tests were noisier, sometimes going as far as depending on other people to test Beezar products (dsavitsk, cetoole, cobaltmute, Linux-works of Arduino fame).  As mentioned previously, AMB once suggested that my cables were picking up extraneous noise in my environment, etc.  The entire exercise in testing the DoodleBug has proven what's really happening.
 
Anyway, in an environment other than the Acer Netbook, you will see a much better base performance for the PupDAC without the DoodleBug. I would still recommend the DoodleBug for any USB-powered audio device, though. 

Wouldn't you subtract out the noise with balanced?  If you have balanced outputs than you can subtract the noise before single ended output(like a differential signal line).  Since you have two set of signal lines, noise will be the same for each set of signal lines, and you can phase shift 180 and take the sum of the sets of signal lines, and take out the noise. All amps amplify noise, but some have better SNR, no?
 
So, I don't really understand why the single ended will have noise at the output if the balanced amps are picking up that much noise if differential mode is implemented.
 
Well, the doodlebug is suppose to take out PS noise.  How is the noise from amp related to PS noise?

 

[tomb]

 

   

• It depends on how much noise already exists on the particular PC USB buss.
• It's a much more sensitive, detailed DAC.  Some of the noise is masked by the lower-grade chip in the SkeletonDAC
• The output opamp in the PupDAC is one of the finest low-voltage opamps on the market.  Strictly speaking, it's not there for amplification, but for combining the balanced output of the DAC chip into a single-ended output.  Nevertheless, it's going to amplify every detail.  If that detail is noise, it's going to get amplified, too  There's probably more going on with the balanced-to-single-ended conversion, too, but I'm not knowledgeable enough about that to explain it.
• Another way to look at it, is that the potential of the PupDAC is much, much higher.  It just needs a reasonably good environment to ensure the best performance. 

 
In my case, the power supply of my testing environment is probably as bad as it could get and still provide operation (Acer Netbook running Windows XP).  I have wondered for years why my RMAA tests were noisier, sometimes going as far as depending on other people to test Beezar products (dsavitsk, cetoole, cobaltmute, Linux-works of Arduino fame).  As mentioned previously, AMB once suggested that my cables were picking up extraneous noise in my environment, etc.  The entire exercise in testing the DoodleBug has proven what's really happening.
 
Anyway, in an environment other than the Acer Netbook, you will see a much better base performance for the PupDAC without the DoodleBug. I would still recommend the DoodleBug for any USB-powered audio device, though. 

Wouldn't you subtract out the noise with balanced?  If you have balanced outputs than you can subtract the noise before single ended output(like a differential signal line).  Since you have two set of signal lines, noise will be the same for each set of signal lines, and you can phase shift 180 and take the sum of the sets of signal lines, and take out the noise. All amps amplify noise, but some have better SNR, no?
 
So, I don't really understand why the single ended will have noise at the output if the balanced amps are picking up that much noise if differential mode is implemented.
 
Well, the doodlebug is suppose to take out PS noise.  How is the noise from amp related to PS noise?

 
The cancelling effect with balanced circuits is a very theoretical abstract.  Yes, all things being equal - balanced drive inherently cancels even-ordered harmonic distortion - but not odd-ordered distortion, IM or power supply noise.  You can have power supplies interjecting signals at many given frequencies - 60 Hz, 120 Hz, 180 Hz, etc. - even higher with switching supplies, etc.  If the amplification section operates with very low distortion, it doesn't mean that distortion is created from those signals or any others (but most likely it will), but they certainly will add to the noise.
 
You need to re-phrase that last question or maybe my comments above answered it.
 
P.S. I'm not sure we can discuss this further in any meaningful way when the entire premise is based on conjecture - my guess for why the SkeletonDAC shows little improvement with the DoodleBug.  Let's review what's really known:

• The PupDAC is vastly superior in performance to the SkeletonDAC.
• The PupDAC shows remarkable improvement using the DoodleBug, depending on the original quality of the USB power buss.  Most tests indicate that the worse the USB power buss quality, the greater the improvement.
• The SkeletonDAC does not show as much improvement with the DoodleBug and performance is worse overall than the PupDAC.

 
I'm not sure we can make absolute, definitive statements beyond that.

 

 

[SilverEars]

Thank you tomb.

 
I was wondering if this is a good good as any power supplies that usually come with high-end DACs that have those fancy PS section with toroids and etc..
I'm guessing you will have built ones for sale soon?

 

[bada bing]

Any updates ?     

 

 

[Avro_Arrow]

I think Tom is still waiting on the prototype cases from Hammond.
There is also the laser etching to prototype and approve.
Should be soon.
Boards and most kit parts are ready. I think there were a couple
items that were back ordered.

 

[tomb]

  I think Tom is still waiting on the prototype cases from Hammond.
There is also the laser etching to prototype and approve.
Should be soon.
Boards and most kit parts are ready. I think there were a couple
items that were back ordered.

Yep - exactly.  I am awaiting Hammond's prototype case (1455C801) that's been machined for the DoodleBug.  Once I have that - and build one using the production board - we'll be ready to go.  Hopefully, I'll have the website done by then, too.  So ... not too much longer - I hope.

 

[kugino]

 
Yep - exactly.  I am awaiting Hammond's prototype case (1455C801) that's been machined for the DoodleBug.  Once I have that - and build one using the production board - we'll be ready to go.  Hopefully, I'll have the website done by then, too.  So ... not too much longer - I hope.

cool...looking forward to it!

 

[kugino]

any update, tomb?