I agree that the transformer approach is superior to simple adapters, especially when going from balanced output to unbalanced input. It allows you to use the whole signal (rather than half of it) and is less likely to cause hum issues. Whether the improvement will be significant is situation dependent, but it's the best approach, IMO. Adapters from unbal to bal sometimes ground one of the active pins (2 or 3 of XLR), but that is not advisable going from bal to unbal, as you would be grounding one output of your source component, which could damage it.
@mike_in_nc I'm unsure regarding your comment below. Can you elaborate?
I was very concerned about this very issue and a number of folks helped me out. Al ( @almarg ) was a primary advisor on this. Perhaps he can comment and add his usual dose of clarity to the discussion?
I also reached out to the designer and manufacturer of both the Preamp and Amp. Both verified that the approach @almarg and others recommended was safe. I also verified with Richard Gray who is local to me, before moving forward.
Adapters from unbal to bal sometimes ground one of the active pins (2 or 3 of XLR), but that is not advisable going from bal to unbal, as you would be grounding one output of your source component, which could damage it.
For a definitive answer, I’d go with the recommendation of the manufacturer of the balanced-output source component. (Some manufacturers even provide wiring diagrams for their recommended method of going from balanced to unbalanced.)
It makes me uneasy to short any output to ground, but I’m not an EE nor a circuit designer, so there’s a lot I don’t know.
Thanks for the mention, David. Most (but not all) XLR-to-RCA adapters ground the signal on XLR pin 3, i.e., they connect pin 3 to pin 1, which is the ground pin. In the case of most (but not all) components providing XLR outputs that kind of adapter should not be used, for the reason @mike_in_nc stated. And the same considerations apply in the case of an XLR-to-RCA cable.
Following is an excerpt from what I had said when we discussed this issue via PM in November 2018:
An adapter that shorts pin 3 to pin 1 would be appropriate when adapting an RCA output to an XLR input, and would be appropriate and sometimes even necessary with **some** designs (especially some tube-based designs having transformer coupled outputs) when adapting an XLR output to an RCA input. However in many cases, especially cases involving solid state designs having low output impedances, such a short could cause adverse sonic effects, buzzing, or even damage, eventually if not sooner. I once diagnosed a severe buzzing problem that occurred with a member’s CD player which was caused in exactly that manner. When he went to an adapter cable which left pin 3 open everything worked fine.
You subsequently received a response from the designer of your Grandinote preamp stating that pin 3 should be left open (unconnected) when adapting its output to an RCA input.
And chances are that would also apply to the OP’s situation, but not necessarily. For a firm conclusion to be reached he should either ask the manufacturer of the component providing the balanced output (as I see Mike suggested just above), or let us know the specific make and model of the component and we may be able to determine the answer.
I am going through the same labyrinth I have a before-amplifier ( not to be confused with the preamplifier) (It is a dox drive rack) which I’ll be using as an external crossover. It has female XLR inputs and male XLR outputs. My question is what is the best way to convert RCA to XLR (cd player to dox) and XLR to RCA. (dox to amplifier)
-typically on the second half Im wondering which would be preferred ? 1. xlr cable from the dox to either: A. Cardas Audio Male RCA to Female XLR Adapters plugged onto my amplifier (McIntoish 2105) B. A transformer. (SO max}
What I like about A. Is there is only one set of cables versus B. Two sets of cables.
depending on my speaker choice Ill need need to be bi-amping or tri-amping… so cost is a consideration x2 or x3…
My brain isn’t intended to be used in complex matters such as this...
Ghentaudio manufactures nearly every possible configuration of cable you can imagine. Decent quality and great service. I have 2 pairs of RCA to female XLR , one going from DAT player/recorder and the other from a CD burner to a Schitt Freya preamp in a little system I have in my home office and 1 pair of XLR male to RCA in the main system going from CD to pre. No grounding issues. No adapters.
even if it is stated that pin 3 should be left open, that’s an invitation to residual captive flutter on the open pin, which will make the highs simultaneously darker and more extended.
The flutter of residual capacitance on the open pin will modulate the return line, as it will be read as signal and thus dealt with by the balanced function itself. And thus micro mangle the transient data of the positive pin’s throughput proper... Basically mis-timed and mis-shaped transient information.
Which will be heard, even if it is very very small, as the ear works solely on interpreting transients and their placement in time and grouped as harmonics in each micro level placement in time.
That’s how we hear. We only process about 10% of the audio signal and all we process is the transients and micro transients, in level and timing placement. So the tiny flutter of an open pin three, is going to produce a signal that is heard. People have various learned and inherent hearing skills so some might not hear it but I think most will.
Maybe not ground it but tie it down through high resistance of some sort. Cut out any possible micro-capacitive flutter, if one can.
Basically, it depends upon how pin 3's open condition is dealt with by the manufacturer. How sensitive it is to potential modulation. That we see many circuits, even if they are resistively loaded (even when unplugged) , they can have caps or plugs put on them, to block stray signals and such from occurring, or interfering with the open condition.
As Bob and Mike indicated earlier the best approach to connecting a balanced output to an unbalanced input is widely considered to be with a Jensen Iso-Max transformer.
A suitable transformer, given that you are using a two-channel amplifier (as opposed to monoblocks) would be the Jensen model PI2-XR. It is available here for about $270 in a configuration having XLR inputs and RCA outputs. (That page shows a PI2-XX, having XLR connectors for both inputs and outputs, but you can specify RCA output connectors in the options menu near the bottom of the page). Further details can be found at the Jensen site.
A transformer such as that would eliminate the possibility of ground loop issues, and it would provide essentially the same noise rejection as a balanced interface would provide. Or perhaps even better noise rejection, because its bandwidth limitation of about 55 kHz at -3db will tend to filter out RFI that may be present. It would also eliminate whatever audible significance the "micro-capacitive flutter" effect that Teo referred to may have. (Although I would not expect that effect to be significant if an XLR-female to RCA-female adapter leaving pin 3 open is placed at the output connector of the dbx unit, and is used in conjunction with an RCA-to-RCA cable. I suppose it might be significant if an XLR cable is used in conjunction with an XLR-to-RCA adapter at the input of the destination component, due to capacitive coupling that may occur within the cable. And there certainly may be an issue if pin 3 is left open on an XLR **input**; when adapting an RCA output to an XLR input pin 3 should be grounded).
Regarding the need for additional cables with a Jensen transformer, as you can see in the manual for the PI2-XR the cables on its output side must be very short (no more than 3 feet long, and preferably less, depending on the capacitance of the particular cable). So given the minimal length they shouldn’t cost a great deal, and whatever sonic effects they may introduce should be less than if they were longer.
Regarding the use of adapters or adapter cables at the outputs of your dbx unit, to expand on my previous comment a bit the only situation in which it is necessary to ground pin 3 when adapting an XLR output to an RCA input is when the signals on pins 2 and 3 are "floating" relative to circuit ground (i.e., the voltages between those signals and circuit ground are not controlled or defined, or putting it another way the signals are essentially "isolated" from circuit ground). That is not the case in most designs, and I suspect it is not the case with your dbx. So I think it likely that connecting pin 3 to pin 1, either directly or via the 10 ohm resistor you referred to, won’t do any good and conceivably could do harm, eventually if not sooner. But without a schematic or other detailed knowledge of the specific design I can’t say that for sure.
NEVER short pin3 to pin1 on a balanced XLR output.
This would have been a perfect topic to post in the "pro audio" section.
First let's explain the XLR pins : 1 : ground 2 : hot (+) normal signal 3 : cold (-) inverted signal
A XLR output is always a male XLR and an XLR input is always a female XLR. An easy trick to remember this is just to think of how a microphone looks like : it has a pins -> male XLR and it is a audio source -> output
Although it is not recommended you can make adapter cables from RCA to XLR and vice versa.
The rules are simple : - RCA ground to XLR ground (pin 1) - RCA center (non-inverted signal) to XLR hot (pin 2) - on the MALE XLR cable part ONLY!!! Short cold (pin 3) to ground (pin 1)
The short is to avoid unwanted signals into pin 3. If you leave it floating it can cause unpredictable input (noise).
Pin 3 on the female XLR of a XLR to RCA cable is left floating otherwise you would short the output and risk damage (you will "overload" the output). Balanced output stages are mostly opamps (some use tranformers) and nearly all of them are protected against shorts, but protections can fail and overloading of the cold(-) signal can negatively influence the hot(+) output signal.
Leaving an output floating is not problem and if you think it is I have one question : do you shorten all the unused RCA outputs on your devices? I guess not.
PS : normally you don't connect XLR chassis to GND (pin 1) but in case you have any hum (ground loop) try connecting those. Some professional devices (especially amplifiers) have a switch for that purpose (ground lift, which factory defaults on "lift").
@danip, Though I am not an EE, I think the OP should be aware that not all equipment with balanced interconnects abides by the AEC standard. Therefore, in an abundance of caution, I suggested using the Jensen Transformers. Almarg suggested contacting the manufacturer for a schematic of the wiring just to make sure it abides that standard. If it should, your advice would be sound. Bob
NEVER short pin3 to pin1 on a balanced XLR output.
It’s never a good idea to say "never" :-)
As my previous posts should make clear, I agree with your post when it comes to most designs. But there are exceptions.
You referred to balanced output stages utilizing op amps, and a notable exception to the general guideline of not shorting pin 3 to pin 1 when adapting a balanced output to an unbalanced input applies to op amp-based output stages that are "cross-coupled," which thereby provide outputs that are essentially floating with respect to circuit ground. See pages 11 and 12 of the following reference, in which such circuits are referred to as "active floating sources." And note the statement on page 13 that "Grounding one output line at the driver, which is REQUIRED [emphasis added] to guarantee stability of most "active floating" circuits, degenerates the interface to a completely unbalanced one having no ground noise rejection at all."
That is further confirmed in the Rane document that was referenced earlier in the thread by Rauda1. See diagram 6 on page 7 of that document.
Also, while I’m not specifically familiar with the designs of the output stages of Atma-Sphere’s balanced tube-based preamps, Ralph (Atmasphere) has explicitly stated in past threads here that connecting the balanced outputs of his preamps to unbalanced inputs without grounding pin 3 will result in a huge hum. Which can be expected, given that the balanced outputs of his preamps, although not transformer coupled, are nevertheless floating with respect to circuit ground. And consequently if pin 3 is not grounded the signal on pin 2 will be received by the destination component with an uncontrolled and undefinable voltage relative to pin 1.
What needs to be known when going from an unbalanced/RCA output (such as an active electronic x/o) to a balanced/XLR power amp input (the XLR wired in accordance with AES File 48: Pin 1 ground, Pin 2 non-inverted signal, Pin 3 inverted signal)? Not in terms of impedance considerations, only in regard to the inter-connect RCA/XLR plug wiring connections.
One way to do that is to simply insert an RCA-female to XLR-male adapter which connects XLR pins 1 and 3 together (as most such adapters do) into the XLR input connector, and use that adapter in conjunction with an RCA-to-RCA cable.
Alternatively, a cable having an RCA connector at one end and an XLR connector at the other end could be wired up correspondingly.
Another approach, which utilizes a two-conductor shielded cable, is shown as diagram 17 on page 8 of the Rane document that was linked to earlier.
Finally, a suitably chosen Jensen transformer could be used to convert the unbalanced signal to a balanced pair of signals. "Suitably chosen" would take into account impedances and other parameters.
@almarg Thanks for the info never seen this kind of driver output on the devices I serviced (or did I miss that???). I am pretty sure that everything I laid my hands was ground referenced. Transformers were a very expensive rare option (available on the better DSP or mixing consoles).
Thanks for the paper i'll give a full read when I have more time.
Anyway, If the OP wants the schematics he can just ask the ppl at DBX, I have never had a decent pro-audio/light brand refuse to hand out schematics or service manuals.
NEVER short pin3 to pin1 on a balanced XLR output.
@danip If the equipment supports the balanced line standard then this should be no worries at all. But it is worthy to note that most high end audio equipment and a fair amount of studio gear doesn't support the standard.
The standard (AES48) calls for the outputs of an XLR (IOW, male XLR) to not reference ground. There really aren't many ways to do that! An output transformer is one way, and obviously if one side of the output secondary is tied to ground (pin 1) its no worries whatsoever. Our preamps are another way this can be done as they have a direct-coupled (tube) output (for which we obtained a patent) that supports the standard, so they too can have one side tied to ground with no worries.
If an output is at risk of damage, this is because its being shorted out, which means it is referencing ground and so does not support the standard. When the circuit references ground, it defeats one of the purposes of using balanced lines, which is to prevent ground loop hum and noise. This practice also allows the interconnect cable to impose an artifact on the the sound, as the shield is being used as a signal circuit return, allowing noise to intermodulate with the signal. Now many studio products that don't support the standard try to get around this problem by having a low output impedance, which swamps the noise, but it should be noted that **any** device that has an XLR output should also have a low output impedance.