NAD PP3 via USB - Loud noise in Win7 but not in XP

[corboid]

Hi,
I have similar problems as it has been discussed earlier, but I have a feeling there is no solution yet.

I can make pure digital noise free recordings via USB using my laptop Lenovo T60p running Win XP, but when I use my mini PC HP 210-1023 running Win7 I get an awful hiss making the recording totally useless. I means that my PP3 should NOT be deffective but the problem is connected to Win7 and/or VinylStudio. The noise is somewhere 20-40 dB below the normal classical music levels. If I wash it out digitally it affects the music too much.

Listening to the characteristics of the noise it gets me to belive that windows7 makes the sound analogue and then digitalises it or adds some analogue noise to the digital signal from USB (Yes, I checked that nothing more than USB/NAD PP3 is selected for recording)
Even if PP3 does not have any level control I can access the recording level in Win7 (but not in XP!) and the noise is scaled with the level.

I use USB-headset so I should get pure digital path NAD PP3 >> Win7 + VS >> file >> USB-headet. But it sounds very "analogish".
My mini-PC has only  a single analog sound socket (4-pin combined in/out) so the by pass AlpineSoft uses to propose by recording the analog signal is not applicable. Anyhow I purchased the NAD to get the signal digitally and not mess with poor A/D-converters inside the PC.

I hope you have a solution for this problem.
Best regards,
Corboid

(PS: I used the "set gain to 1" setting in VS level control, for win7)

[Paul Sanders (AlpineSoft)]

Can you send me a sound sample please?  Please record about 1 minute to MP3 format (to keep the file size down) and send it to:

p DOT sanders AT alpiinesoft DOT co DOT uk

Thanks.

You might also like to try a different USB cable and / or port.  I'm sure this problem is hardware related.  If the PP-3 didn't work with Windows 7 (or with VinylStudio on Windows 7) we would certainly know about it and I have tested this combination explicitly.  Also, ensure that the phono connections are tight and that the turntable is earthed to the lug on the PP-3.

Setting the digital gain to 1 is the right thing to do.  Any other setting cases Windows 7 to scale the digitised signal and that will result in clipping or loss of dynamic range, depending which side of the line you are on.  The lack of a recording slider in XP is a blessing, not a curse.

I think the analog route is closed off because, according to the specifications for your machine, your socket is for a Microphone and hence mono only.

[corboid]

Hi Paul,
sure, I'll send you samples from win7 and xp tonight.

First I expected the problem to be some electric interference so I tried three different USB-cables with and w/o ferrit cores.
I also tried to isolate the PP3 from the computer by a powered hub w/o any difference. Same result when running the PC on batteries or 230V/AC.

[Paul Sanders (AlpineSoft)]

OK, thanks.  It's very odd, I must admit.  Anyway, I'll take a listen and see what I make of it.

[corboid]

A follow-up note for other people looking for a solution:

I discovered that the USB-socket on the left side of the HP mini-PC produces much higher noise that the two USB's on the right side.
And both sockets have higher noise levels than I get on my Lenovo.

Paul Sanders believes that my PP3 (purchased in september 2008) can be the problem and needs to be replaced.

[Paul Sanders (AlpineSoft)]

*Possibly* needs to be replaced.  I need to follow this up with NAD.

[mgd]

I just want to add my 5 cents.

I purchased a NAD PP3 which was very noisy both on PCs and on my Mac. The noise was not clear white noise but rather typical computer noise with clearly audible tonal components around 1000Hz. A friend told me that the USB clock is 1ms so the 1Khz noise is most probably USB clock noise that reached the ADC in the NAD PP3.

After a bit discussion, Hifi klubben replaced my NAD PP3 as they told me this was a known problem for units with a serial number below or at H88PP3G02501.

My new unit is much better, BUT there is still clearly audible noise which is about 15dB louder than the noise from my Macs built-in sound card (using analog in). Also the noise is still not white noise but contains audible tonal components (peaks) around 4kHz and 17.5kHz (to high pitched to be heard by many) as well as a lot of harmonics.

My Mac's built-in sound card has only white noise (no peaks at any frequencies) at a very low level.

I am going to return my NAD PP3 and I cannot recommend this unit.

If I am going to digitize my old vinyl records and spend a lot of hours doing it, I want to be sure the recordings is not masked in audible noise. (And remember, we are not talking white noise hiss, but annoying computer noise.)

Instead, I would recommend finding a decent RIAA preamp (maybe NAD PP2) and use another external sound card (or maybe your built-in sound card if it is without noticeable noise).

/Martin

P.S.

quote author=corboid link=topic=183.msg547#msg547 date=1281601769]
I discovered that the USB-socket on the left side of the HP mini-PC produces much higher noise that the two USB's on the right side.
And both sockets have higher noise levels than I get on my Lenovo.
[/quote]

I believe the main problem is that the PP3 picks up the USB noise which it shouldn't. Different PCs and Macs have different levels of noise from the USB ports. That could explain the differences you experience. In any case, this noise shouldn't reach the ADC in the PP3.

[Paul Sanders (AlpineSoft)]

I'm, sorry to hear that.  The problem does seem to be dependent on the PC as well as the PP-3 - most people are OK.

In corboid's case (after some off-line discussion and talking to NAD), he can get acceptable recordings from all but one particular USB port on one particular PC, which demonstrates that the PC is injecting the noise in his particular case.  He tells me he will be getting a new PC soon so it will be interesting to hear how that pans out.  We have scotched the theory that it was anything to do with Windows 7 per-se.

[mgd]

In corboid's case (after some off-line discussion and talking to NAD), he can get acceptable recordings from all but one particular USB port on one particular PC, which demonstrates that the PC is injecting the noise in his particular case.

As far as I understand (from what I have been told), USB will always be more or less noise (transmission of clock signal, multiplexing of packages sent/received) and the USB device must be able to handle this. In case of a USB connected A/D converter it must be able to suppress the noise from the USB port.

So you are right that in corboid's case one of his PCs is injecting more noise than the other into the PP3, but the PP3 ought to be able to handle that.

If you have a look at the attached file: The red graph is a spectrum analysis of the signal I record from the PP3 when the line-in signal to the PP3 is silent (a stopped Logitech Squeezebox). The green graph shows the result of recording exactly the same signal (silence) at the same level using my Mac's built-in sound card (analog line-in).

As you can see, the noise introduced by the PP3 is around 13dB higher, but more important it has a lot of audible peaks. Especially, the one just below 4kHz is annoying. This is my second PP3. The first one was much worse (you have heard it yourself as I send the recording to you earlier this summer.

Regards,
Martin

[Paul Sanders (AlpineSoft)]

I do feel a certain conflict of interest here but NAD use a TI PCM2904 converter chip which includes the A/D converter stage and it is this chip which is fed from the power supply in the PC.  The 2904 is widely used in the industry for other devices in the PP-3's price range and I suspect that these will also suffer from the same problem when the PC power supply is noisy, e.g.

http://forum.audacityteam.org/viewtopic.php?f=27&t=33018&start=0

(the ART uses the same chip).

None of which helps you of course if you are afflicted by this problem.

Thanks for posting the sound plots.  They make interesting viewing. 

[corboid]

Hi,
I still did not have time to analyse if the noise is sent to the PP3 via the VCC/5V-connector (pin 1), ground (pin 4) or via the data connectors (pin 2/3). Need to get an oscilloscope first.

A well designed A/D should assure having a clean supply voltage and not let the ripple march into the analog signal and fool the A/D conversion.

I have some ideas to isolate the PP3 from the computer. Unfortunately the ready made opto-isolators for USB are many times more expensive than the PP3 (and can only handle USB1.1 and USB2.0/low speed, ie up to a little more than 1 MBaud), so it needs some hands-on experiments.

It seems NAD are good on analogue design but they have a lot to learn about digital design. PP3 is excellent as an analogue amp. But the digital signal is just "fair" but not as good as it should/could.

//Corboid

[Paul Sanders (AlpineSoft)]

Yes, I know the USB chip gets its power from the PC but I'm not sure whether it's a requirement that it does so.  Anyway, I will be interested in the results of your findings.  It's possible that the noise pickup happens in the phono stage, rather than in the A/D convertor itself, of course, and attaching a scope lead is bound to affect the results you get...

[corboid]

No, the analogue phono signal is totally unaffected, I would hear it via my amp (I have only high-end equipment all the way so I would hear it immediately).
I.e. the ripple is caught in the digital part of the PP3 and not in the phono-amp.

I contacted my NAD reseller, they were well aware about the problem and promised to replace my box (S/N H7ZPP3Gxxxxx ==> manufactured december 2007!).
I will see how much better it will be before I make further improvements (if needed).

[Paul Sanders (AlpineSoft)]

That's interesting.  We had another case where the customer _did_ hear it via his amp.  I guess it's a more complex problem than at first appears.

Anyway, good luck.  Noise levels are definitely lower on the later models - NAD redesigned the PCB layout in the light of what they learned from the first model.

[corboid]

I found an interesting gadget providing a clean USB power voltage
http://www.aqvox.de/usb-power_en.html

If it works as it claims it could be a solution for computer noise picked up by the external A/D-converters:s.
I'm a bit worried that they only inject a clean VCC, and let the data signal pass through. If the disturbance sneak in that way, it will not help.

However the product costs a bit, EUR 87.00.