TOPIC: Is SHD balanced internally?

Thinking about getting a SHD, I was wondering if it is an internally balanced device?

Specs seem to imply 1 DAC (4490EQ) per output pair, so I’m wondering if the signal is going through a phase splitter into the DAC then being balanced afterwards for output from the XLR connectors?

Or can it be confirmed that the whole internal signal chain is balanced?

My intended use-case is to use it as a Roon streamer ->DIRAC -> balanced into my amp.

Thanks

As far as I know it converts balanced to single ended before A/D and the opposit at the D/A.
So short answer: no

In the case of the using the streamer, I believe the digital would go straight into the D/A, but since it's just a single D/A, I'm not sure if the 4490 allows balanced input or if you need to split the signal into single ended. Then if using balanced, it would have to be converted again.

Can MiniDSP confirm that the digital signal chain isn't balanced from input to output?

If that's the case, you'd think that unless you're doing a very long run or a run through a noisy environment, that the single ended performance would be superior on the SHD?

Shame because I have a balanced amp and would have liked to have an end-to-end balanced signal.

If this device isn't fully balanced, my wishlist for the successor product would be to have dual DACs for a fully balanced path and to add a 12V trigger - so close to the perfect product otherwise.

I believe that the device measures well as-is, but this should lead to some improvement. May also explain the lower SNR for XLR on ASR.

Uh... the DAC chip has balanced outputs.

speakee wrote:
Are you sure? The SHD having balanced outputs doesn't mean that the signal is balanced from input to output within in the device.

It could be taking a split output from the DAC and converting to a balanced output, which isn't the same - but also what I'm trying to confirm / figure out.

eisenb11 wrote:
The datasheet is there -> www.akm.com/content/dam/documents/produc...0eq-en-datasheet.pdf

eisenb11 wrote:
I can imagine some performance gain from having a pair of A/D and D/A connented to the XLR connectors, it could be a 3dB better SINAD.
But having the digital (floating point) path also balanced would make less sense. The digital noise floor would be <140dB, so no need to improve on that...

speakee wrote:
Thanks, data sheet confirms that the chip has a differential output.

mvs0 wrote:
It seemed like that when researching products, many of the high end components used dual DAC chips.

One curios thing is that on ASR, the single ended test had a better sinad by about 3 dB over the balanced output?

To be honest, single-ended outputs are better in every way for audiophile usage. If you want balanced, just put a matched source impedance on the extra pin, but only actively drive one pin of the XLR. It will always be lower distortion. Also, noise pickup on the cable shouldn't be an issue for such short runs so the matched source impedance isn't even necessary.

If you insist on double-driven balanced outputs, you want the differential DAC output converted to single-ended before leaving the box - this way you don't send all that common-mode noise down the cable and into your next device. Also, it's nearly impossible to filter out that common-mode noise because it's nearly impossible to maintain the common-mode rejection (due to component tolerances), and then the front-end of your downstream device will be susceptible to extra distortion from the high frequency modulation. Btw, every DAC has common-mode noise on its output - it's the same kind of noise as a class D amp for those not familiar with the nuances. The class D noise isn't audible at a speaker because the motor is linear at those frequencies, but it's pretty hard for a nice pre-amplifier to deal with. Plus it's gonna radiate on the cables which you don't want either.

However, don't short an actively driven pin on an XLR because the large power supply currents will affect the driven channels (unless it's a transformer coupled output).


It turns out these are rather deep engineering topics, but the point is that we don't want balanced all the way thru the box. It's really important to have a single-ended stage after the DAC at some point in the signal chain.

mebentz2: Thanks for the taking the time to write that - that was super educational and informative!

eisenb11 wrote: You might want to ignore much of what he said.
The Class-D analogy is nonsense.

Also, it's (generally) not a problem to short one side of an active signal to ground via an appropriate XLR/RCA cable or adapter. Most pro equipment has significant build out resistance in place to prevent damage. In the case of miniDSP gear the output resistance is 600 ohms and you will not damage anything by doing that.

I'm not sure why you're hung up on an internally balanced configuration. Almost all audio electronic equipment is unbalanced internally......only the input/output stages might be configured for balanced interfacing. The SHD does have balanced interfacing and if you're connecting to other equipment that also has balanced interfacing there's no reason not to use those connections.

Dave.

Haha, I don't mind if you ignore me, but it's funny to hear that after all these years of electrical engineering experience on some of the highest performing products in the pro audio world

But seriously, don't take my word for it....go take a look at the measurements for yourself. I wouldn't trust me either.

Btw, I agree that shorting an active pin to ground won't damage anything. However, it dumps a bunch of Class B power supply currents into your power and ground planes. When you think Class B, think diode rectification...you will see a marked increase in second harmonic distortion from the increased power consumption. We are talking high fidelity here right? There's a huge gap between high-fidelity and not smoking equipment... Also, proper gear that does the common-mode rejection in the box will require a second series element for the actively driven balanced leg, which means cascaded distortion. Again, this is easy to measure. It's even worse with the cross-coupled servo outputs.

And just to clarify, my comments are about the design of the equipment. I've not torn apart an SHD (yet), but I hope to run it through its paces someday. Just because I think unbalanced makes more sense for the highest fidelity, doesn't mean I think you should drive the SHD into an unbalanced load.

I'm new to this forum, but does anyone have the service manual for these units, or at least a block diagram of the internals?

Who do you think you're talking to here? Some nub who has no familiarity with the miniDSP equipment?
I have a schematic for the SHD. In fact, the guy who designed the DAC/analog portions of the SHD was recommended to miniDSP by me.

I suggest to put away the smart-ass attitude.

Dave.

I think most of us here (myself included) are more interested in a technical discussion. Care to share the schematics with the group? Or at least just the output stage?