You said:Davey said:
You said "in this case" - that is, the case of the unbalanced receiver. It would make a huge difference if the source used a transformer output - no connection to -ve output means no signal from +ve output. Only later did you say that a short on an active output might create problems.Davey said:
I am. Are you reading them too?
You were the one that brought up a transformer output.
Obviously, in the case of a xformer output, pin 3 needs to be taken to ground.....otherwise it wouldn't work at all.
I brought up the (possible) issue with active outputs...and I specifically said "active circuitry" to eliminate passive outputs (like transformers) from that instance. You were the one that invoked the straw man discussion regarding a transformer-coupled output.
This is a basic topic. I'm surprised there is so much misunderstanding and semantics at work here.
My goodness.
Dave.
Last edited:
We were talking about balanced outputs. You, it now appears, were talking about the subset of balanced outputs which don't use a transformer - yet at first you did not make this clear. You only mentioned active circuitry in your second paragraph, having already in your first paragraph said that no -ve connection was needed - but it is needed for a transformer.
Anyway, I think we can agree that a grounded output may be required, prohibited or tolerated depending on how the source works.
Anyway, I think we can agree that a grounded output may be required, prohibited or tolerated depending on how the source works.
"But this is to short the XLR 3 (-) to ground! would this burn the output stage / +/- converter of the USB DAC?"
The OP's original query makes it relatively clear he's referring to a non-transformer coupled output. (A little extrapolation on my part...but not incorrectly I don't think.)
I can't think of any USB DAC's (off the top of my head,) but I'm sure there probably are some, that have transformer-coupled outputs and would absolutely require this pin 3 ground connection to operate.
Apologies for not identifying EVERY possible output scenario immediately.
So, to recap:
If it is a transformer-coupled balanced output.....pin 3 ground is appropriate and required.
If it is just a resistor to ground internally on the (-) output.....a pin 3 ground is irrelevant.
If it is an active stage driving both +/- outputs.....pin 3 ground is not appropriate and MAY cause issues with the output circuitry.....in some cases......possibly.....maybe.
You fellas can tell me all you want this is not a potential problem, but I've seen it. (This is the reason I responded to this thread initially.)
I repaired a piece of gear a few years back where the (-) balanced output was damaged by the use of an XLR/RCA adaptor plugged externally. When I repaired it, I added a build-out resistor to both (+) and (-) outputs to prevent the issue from recurring.
Goodness.
Dave.
The OP's original query makes it relatively clear he's referring to a non-transformer coupled output. (A little extrapolation on my part...but not incorrectly I don't think.)
I can't think of any USB DAC's (off the top of my head,) but I'm sure there probably are some, that have transformer-coupled outputs and would absolutely require this pin 3 ground connection to operate.
Apologies for not identifying EVERY possible output scenario immediately.
So, to recap:
If it is a transformer-coupled balanced output.....pin 3 ground is appropriate and required.
If it is just a resistor to ground internally on the (-) output.....a pin 3 ground is irrelevant.
If it is an active stage driving both +/- outputs.....pin 3 ground is not appropriate and MAY cause issues with the output circuitry.....in some cases......possibly.....maybe.
You fellas can tell me all you want this is not a potential problem, but I've seen it.
(This is the reason I responded to this thread initially.) I repaired a piece of gear a few years back where the (-) balanced output was damaged by the use of an XLR/RCA adaptor plugged externally. When I repaired it, I added a build-out resistor to both (+) and (-) outputs to prevent the issue from recurring.
Goodness.
Dave.
Last edited:
Why not upgrade to a THAT1246 used for preamp?
Or, upgrade the existing preamp with a THAT1200?
THAT IC Selection Guide
Or, you could diy any of that with the OP275. It is very small! But, it is very useful too.
Looks like some buffers might be needed as well--Keantoken has that extraordinarily transparent 4 transistor buffer.
Well, I'm just saying that there's no need to waste a balanced drive source. Using the balanced source correctly--balanced all the way, will reduce intermodulation distortion by approximately 4x, or at least sound a whole lot clearer. So, it is probably worth re-fitting the preamp to support high quality XLR sources.
Or, upgrade the existing preamp with a THAT1200?
THAT IC Selection Guide
Or, you could diy any of that with the OP275. It is very small! But, it is very useful too.
Looks like some buffers might be needed as well--Keantoken has that extraordinarily transparent 4 transistor buffer.
Well, I'm just saying that there's no need to waste a balanced drive source. Using the balanced source correctly--balanced all the way, will reduce intermodulation distortion by approximately 4x, or at least sound a whole lot clearer. So, it is probably worth re-fitting the preamp to support high quality XLR sources.
Whether or not you can ground pin 3 (or 2 for that matter) and still get good performance from the differential output depends on the output circuit. Some circuits just invert the signal from pin 2 and feed the inverted signal to pin 3. With those circuits, you'd usually just experience a slight degradation in performance by shorting pin 3 to ground, but you may get no output at all if you ground pin 2. Grounding pin 3 (or 2) will result in only half the output voltage, though. So you go from 2 V differential to 1 V single-ended.
The THAT1646 is interesting in that it allows either the inverted or the non-inverted output of the differential pair to be grounded with no performance degradation.
A better solution is to use a differential input circuit. The THAT1200 is a good example and it performs very well. THAT also included some circuit tricks to increase the common-mode impedance of the input which improves performance as well.
Should you decide you need a differential-to-single ended converter (aka a differential receiver), I happen to have one available. I released the THAT Receiver circuits over the weekend. I have both a stereo and a mono version available.
THAT Receiver – Rev. 1.0
They're based on the THAT1200.
Tom
The THAT1646 is interesting in that it allows either the inverted or the non-inverted output of the differential pair to be grounded with no performance degradation.
A better solution is to use a differential input circuit. The THAT1200 is a good example and it performs very well. THAT also included some circuit tricks to increase the common-mode impedance of the input which improves performance as well.
Should you decide you need a differential-to-single ended converter (aka a differential receiver), I happen to have one available. I released the THAT Receiver circuits over the weekend. I have both a stereo and a mono version available.
THAT Receiver – Rev. 1.0
They're based on the THAT1200.
Tom
Last edited:
To me your OP was pretty clear. You were concerned about the fact that various XLR->RCA connection methods sometimes ended up shorting one of the active outputs of many source components.
That makes all posts involving transformers and those compare and contrast various active and passive approaches off topic in my book.
The correct answer is IME and IMO Fuhgeddaboutit. This has been going on for decades and every designer worth his salt adapted to it decades ago.
The OP said
It is not beyond the realms of possibility that a genuinely 'professional' DAC would have a transformer feeding the XLR output. In fact, one could argue that if it lacks a transformer then it is not truly 'professional'. Therefore I did not assume that it lacked a transformer.HiFiNutNut said:
Now THAT is good advice! The board can upgrade that fun mode preamp to support the performance capacity of high end XLR sources.
Why is everyone getting so excited about a perfectly reasonable assumption (that a "professional" balanced XLR output might use a transformer) just because they happened to make a different more restrictive assumption (that a "professional" balanced XLR output definitely will not use a transformer)?Pano said:
As far as I know, the OP still has not told us. Even his references to active outputs could be assumptions.
Are the THAT chips as good as or better than some good transformers?
I hesitate to consider BAL interconnections because of this: if all of my equipment are of true BAL type then there is definitely a value to do it. But in reality, if all active components in the audio chain are UNBAL, to use BAL interconnections in each of the component (except the source) input we convert BAL to UNBAL and at each component output (except power amp) we convert UNBAL to BAL. That is a lot of conversions. The result is merely to reduce noise pick up on the interconnections and reduce ground noise.
On the contrary, if we use UNBAL interconnections and make the interconnect cables having very heavy shield (low resistance and good RFI rejection), and keep the interconnects short - no longer than 1-2 metres, it might be possible that the distortion is less than the distortions happening in those UNBAL-BAL, BAL-UNBAL conversions. I am not sure.
I hesitate to consider BAL interconnections because of this: if all of my equipment are of true BAL type then there is definitely a value to do it. But in reality, if all active components in the audio chain are UNBAL, to use BAL interconnections in each of the component (except the source) input we convert BAL to UNBAL and at each component output (except power amp) we convert UNBAL to BAL. That is a lot of conversions. The result is merely to reduce noise pick up on the interconnections and reduce ground noise.
On the contrary, if we use UNBAL interconnections and make the interconnect cables having very heavy shield (low resistance and good RFI rejection), and keep the interconnects short - no longer than 1-2 metres, it might be possible that the distortion is less than the distortions happening in those UNBAL-BAL, BAL-UNBAL conversions. I am not sure.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.