Peavey USB-P DAC Teardown

[amirm]

This is a teardown of the recently reviewed Peavey USB-P transformer coupled DAC. The internals are simple enough:

Data and power are provided via the USB bus. The core of the system is a on-chip USB DAC, the TI/BurBrown PCM2705. This is a low performance DAC designed for quick implementations. Here are the TI measurements for it:

I have converted the THD+N percentages to SINAD (in red). As you see, there are two implementations: one with USB power with worse results (left) and the other with stand-alone power supply (right). Neither is much to write home about. Ours is the one on the left which best case has a SINAD of 78 dB. In the review I got around 74 dB. The reduction is due to use of transformers on the output to provide isolated/balanced output (the main reason for this device's existence).

Transformers can generate a lot of distortions especially in lower frequencies. Let's go ahead and measure that using USB-P:

At 20 Hz and full digital level input, distortion shoots up to nearly 2% for a SINAD of around 35 dB! Most of our measurements are at 1 kHz which gives these a ton of benefit of doubt.

Lowering the output signal level improves transformer distortion but then signal to noise ratio worsens so there is not a lot of gain to be had as you see in the green and blue lines above.

I suspect this level of distortion is very much audible. So only use this device if you have to and performance is not critical.

------------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

 

[trl]

The reduction is due to use of transformers on the output to provide isolated/balanced output (the main reason for this device's existence).

Glad to see what's inside the box. I guess that on the stage or even on studios ground lifters or other sort of isolators, like this Peavey DAC, will help a lot, despite their poor measurements.

As per https://peavey.com/usb-p-usb-playback-device/p/03001370: "The dual transformers in the USB-P provide balanced outputs and ground isolation required when interfacing a computer to a sound system".

I wonder how the final measurements will look like with a different DAC chip, but same output transformers.

 

[amirm]

I don't think a better DAC would make much of a difference.

 

[Bruce Morgen]

Transformer are apparently An Equal Opportunity Destroyer when it comes to DAC performance, whether it's this relatively cheap Peavey or an absurdly overpriced "high-end" unit from PS Audio.

 

[wwenze]

Question for the floor:

Is this level of transformer performance at this price cheap or expensive

 

[restorer-john]

Transformer are apparently An Equal Opportunity Destroyer when it comes to DAC performance, whether it's this relatively cheap Peavey or an absurdly overpriced "high-end" unit from PS Audio.

Not all transformers are created equal and good transformers cost real money. This is a US$50 item.

Consider all balanced gear from the distant past used 600 ohm transformers. I have CD players and DACs with transformers for the balanced outputs and the performance is very close to the RCA line outs.

 

[restorer-john]

Question for the floor:

Is this level of transformer performance at this price cheap or expensive

It's very cheap and built for a purpose which it achieves. No purist is going to go near it, but that's OK.

 

[IVX]

probably this is the old design, maybe 90th.

 

[Bruce Morgen]

Transformer are apparently An Equal Opportunity Destroyer when it comes to DAC performance, whether it's this relatively cheap Peavey or an absurdly overpriced "high-end" unit from PS Audio.

Not all transformers are created equal and good transformers cost real money. This is a US$50 item.

Consider all balanced gear from the distant past used 600 ohm transformers. I have CD players and DACs with transformers for the balanced outputs and the performance is very close to the RCA line outs.

The transformers in that pricey PS Audio DAC that Amir found extremely wanting are (supposedly) top-notch "audiophile" components -- whether they were designed-in properly is an open question, of course. What is clear in that particular case is that they were functionally unnecessary unless they were there to intentionally introduce "vintage" coloration -- AKA supposedly ear-pleasing distortion that some "high-end" luminaries are all about nowadays -- into the signal chain.

 

[restorer-john]

What is clear in that particular case is that they were functionally unnecessary unless they were there to intentionally introduce "vintage" coloration -- AKA supposedly ear-pleasing distortion that some "high-end" luminaries are all about nowadays -- into the signal chain.

I don't think you understand the benefits of a transformer coupled stage. That's fine, some of us do.

Bob Carver supposedly also thought it was OK to rectify mains AC and use it for his prototype power amplifiers in the 1970s and subwoofers in the 1990s, thus saving money on transformers. How did that work out? Hint: it was dangerous.

 

[Francis Vaughan]

One thing that could have improved this design would have been a buffer between the transformer and DAC. The DAC can only source quite low current. Adding a driver with a good low impedance output can significantly improve the low frequency performance of a transformer, and might pull 10dB of distortion out of it. I really expected to see some such driver. Disappointed it isn't there. I guess the feeble output swing was a clue.
The 2705 is a vernerable part. Treated nicely it was a competent DAC. Nothing compared to modern devices, but TI were rightfully proud of it when it first came out. However that was nearly 2 decades ago. It is now listed as not recommended for new designs. One suspects this Peavey DAC has been around for a while. Which might not be a surprise.
BOM is going to be roughly $5 for the DAC and $5 each for the transformers in quantity. Add a nice case, connectors and PCB. Really this is a fair price. Upgrading it significantly would well over double the price. More like quadruple. The next step up in transformers gets you to $20 parts, better DAC, proper power management, driver stage. Think well north of $200. Good transformers start at $100 each.
Galvanic isolation, affordable, audio quality: pick any two.

 

[Bruce Morgen]

I don't think you understand the benefits of a transformer coupled stage. That's fine, some of us do.

Bob Carver supposedly also thought it was OK to rectify mains AC and use it for his prototype power amplifiers in the 1970s and subwoofers in the 1990s, thus saving money on transformers. How did that work out? Hint: it was dangerous.

I understand fully that transformers have real benefits in some applications when designed in properly, but there's no earthly reason to put them in the signal path of a home stereo DAC when there are clearly other -- and measurably better, not just cheaper -- solutions available. You also might want to consider putting your know-it-all arrogance on a time-out, John -- bringing up irrelevant apples vs. oranges comparisons like Carver's ill-considered power supply adventure just makes you come off as silly.

 

[restorer-john]

You also might want to consider putting your know-it-all arrogance on a time-out, John -- bringing up irrelevant apples vs. oranges comparisons like Carver's ill-considered power supply adventure just makes you come off as silly.

Hardly, sunshine. Transformers in this reviewed product are there for one very valid reason, just as the lack of similar devices in the Carver implementation were implemented for another, perfectly valid, but equally misguided reason.

Just because silicon derived line drivers have become reasonably silent and offer great on-paper specs, doesn't mean that in certain situations, an isolated, transformer driven solution, cannot provide a better overall performance.

As they say, there are many ways to skin a cat.

 

[anmpr1]

I understand completely the criticisms of this device (and others like it) from an absolute technical standpoint. Yet it kind of reminds me of changing a tire on your car. The hydraulic lift and air gun at the garage is the best and easiest way to do it, but if you are stranded on the side of a dirt road in the middle of nowhere, the hand jack and tire iron in your trunk will do the job.

Similarly, if you have all your gear thrown in the back of your station wagon, driving to the gig in order to set up your Peavey Audio Performer Pack (on sale now for $500!) , you might wish that you could bring your better equipped home studio with you, but this will do the job.

Thanks for the tear-down.

 

[restorer-john]

but if you are stranded on the side of a dirt road in the middle of nowhere, the hand jack and tire iron in your trunk will do the job.

Exactly, and for USD$49 on sale, it's a bargain solution. Why hold it to standards it was never designed to aspire to?

 

[kopczas]

Output transformer could perform better with implementation of Walt Jung`s Mixed Feedback circuit as an input impedance matching buffer to transformer. Maybe it`s not as perfect as true high performance output stage for DAC but still much better then direct connection from dac to transformer.

 

[IVX]

kopczas, taobao $10 permalloy 600:600 E26mm trafo has 0.000*% of THD@1kHz and 1Vrms 20Hz 0.1%. BTW, a good idea to test the frequency response with different loads.

 

[kopczas]

taobao $10 permalloy 600:600 E26mm trafo has 0.000*%

That`s nice. The mixed feedback I`ve mentioned work great with Lundahl transformers as Jung`s circuit suggestion in Opamp Application Handbook. Whole idea was to reduce distortion at low frequencies. Lundahl 1517 is an old design. Maybe modern transformers are better in specs.
Anyways, this mixed feedback buffer works for Lundahls. I`ve checked it.

 

[IVX]

kopczas, did you see Audio Precision trafo-feedback patent? They still use that as an output for their analog generator.

 

[kopczas]

did you see Audio Precision trafo-feedback patent?

No, I didn`t.