The amp should not care if the input is balanced or single ended.
You will not get the benefit of a balanced preamp, but it will work fine.
You will not get the benefit of a balanced preamp, but it will work fine.
First the XLR output/input voltage is twice that of rca. Second XLR uses differential op amps to cancel noise where rca does not have that feature. Will it work?? YES it will but you will not get all the benefits that XLR has to offer. Remember that XLR was originally designed for commercial purposes (ie long cable runs of 20 to 100ft) hence the doubled output voltage and noise cancellation circuitry.
That being said right now I am demoing balanced vs rca interconnects (same make - model and length) and although I like the XLR connection and if I listen very closely I can hear a slightly wider soundstage. That's it. I prefer rca's for 2 reasons. 1 The gain from XLR is too much (1 meter cables). 2 rca is virtually an industry standard and should I ever go back to all rca I won't have to have the cable re-terminated due to more equipment being rca only (especially reasonable price stuff).
Ralph (Atmasphere), one of the excellent papers at your site describes two of the fundamental advantages of fully balanced internal design, for power amplifiers, as being cancellation (or at least substantial reduction) of some distortion products, and presentation of an essentially constant load to the power supply. Won't these advantages be negated if a single-ended signal is put through a balanced amp?
Also, the benefits that a balanced interface provides in terms of reduced susceptibility to ground loop issues, and rejection of common mode noise, will be negated if an adapter is used, respectively because the signal return path will not be isolated from inter-chassis noise currents, and because of the impedance imbalance between each of the two legs and ground.
The two benefits relating to balanced internal design, I believe, can be retained if you were to choose a fully balanced power amp that has rca as well as xlr input connectors. I would expect that in those cases the input stage into which the rca input is fed would perform a single-ended to balanced conversion, resulting in a balanced signal being processed internally.
All of the benefits described above can be retained if you were to use a Jensen transformer between the preamp and power amp, to perform a single-ended to balanced conversion. See this paper. Note, btw, that the adapter cable shown in Figure 2.1 of the paper is NOT configured in the same manner as an rca cable + rca-to-xlr adapter, and the latter will provide considerably poorer noise rejection than the already degraded noise performance that is shown for the adapter cable.
A lot of users here have reported successfully using Jensen transformers in this kind of application with excellent results and no noticeable adverse effects, although I recall Ralph expressing some mixed feelings about them.
Won't these advantages be negated if a single-ended signal is put through a balanced amp?
No. As soon as the signal is in the amp, it will be in the balanced domain if the amp is really balanced.
Using a Jensen may not be a bad idea- you have the SE preamp drive the Jensen, and then run you balanced interconnect as far as you need to the amp. I've done that a lot and its always worked well.
As soon as the signal is in the amp, it will be in the balanced domain if the amp is really balanced.But if one side of the balanced signal path is processing a music signal, and the other side is processing 0 volts (as opposed to a signal that is the inverse of what is being processed by the other side), how will any distortion products be cancelled, and how will the load on the power supply not fluctuate with the music (assuming the amp is not Class A)?
I'm not disagreeing, just questioning something that doesn't seem to make sense to me.
Thinking a little further about my previous post, Ralph's point does indeed make perfect sense to me, assuming that the input stage within the amp is what is known as a long-tailed pair differential amplifier. That will produce a balanced output in response to a single-ended input, thereby realizing the benefits of fully balanced internal design in the subsequent stages of the amplifier, including the power stages where it matters most.
I assume that is the architecture used in Ralph's designs. So the question becomes whether or not the input stage of all or at least most other "fully balanced" amps can be expected to have been designed in an equivalent manner. I don't know the answer to that question.
One note of caution: Not all Balanced amps are created equal.
I put this very question to ARC's tech team some years back and learned -
ARC's balanced amps (according to ARC) will perform "very poorly" if you try this scheme. I'm not particularly technically oriented, but my understanding is that the lack of noise cancellation in the single ended signal will trip the protection circuitry at less than half the rated output. My 122 WPC VT30SE will output +/- 50 WPC (IIRC) before the protection police shut down the party.
Disclaimer: This posting is based on my recollection of a coversation with ARC's tech guy many years back, so I hope I got it right and that this explanation makes sense to those more knowledgeable of the issues at hand.
If the amplifier is truly balanced, it will not care if the input signal is balanced or single ended, nor will it matter if you use the RCA or XLR for that input.
I can imagine an amplifier that is balanced but does not take advantage of all the possibilities that balanced operation offers, but the idea that using a single-ended source with it might set off the protection circuit seems weird. If that is really true then there is a problem in the amplifier that needs to be fixed; I can't think of a reason why a properly operating amp would behave that way.
Now when you attempt to run a balanced amp with a single-ended source, care must be taken that the hookup is correct. Often its a good idea to make sure pin 3 and pin 1 (ground) are tied together. If this is not done, pin 3 can float and inject noise into the amplifier. This might be why we see a few weird responses in the posts above (although I would have expected that this sort of nuance would have been dealt with, but maybe not...). The Cardas adaptors we have seen do not do anything with pin 3 so they usually need some sort of attention, depending on the setup.
I assume that is the architecture used in Ralph's designs. So the question becomes whether or not the input stage of all or at least most other "fully balanced" amps can be expected to have been designed in an equivalent manner. I don't know the answer to that question.Al, you're exactly correct -- the design of what constitutes a "balanced" input stage, and a "balanced amplifier" circuit, varies wildly among different manufacturers and circuit designers.
There *should* really be two different, independent aspects to "balanced" equipment design -- the first being the rejection of noise as a result of equipment interconnection, the second being any performance benefits/demerits of differential circuit operation. From what I've seen in high-end audio, designers and engineers get these two goals confused the overwhelming majority of the time.
I know you're familiar with Whitlock's papers on the subject of balanced interconnection - these give excellent insight into the issues of balanced input-stage design and how common-mode rejection is dependent on the balance of impedance (NOT voltage!) presented by the line driver and cable/connectors. Basically, the input stage's tolerance to source imbalance is a function of the ratio of the common-mode and differential-mode impedances. That is, for a given differential-mode (signal) input impedance, the higher the common-mode impedance . . . the greater tolerance the balanced input has for source impedance balance errors. This is how a balanced input circuit with a high common-mode impedance (i.e. a transformer) can indeed give substantial rejection of ground noise from an unbalanced source, provided it's wired with a properly-terminated balanced adapter cable.
Where the necessity for balanced voltages comes into play is for the rejection of even-order distortion products -- this occurs by enforcing symmetry in the circuit's transfer function . . . and the majority of designers confuse this goal with that of interconnection noise rejection.
A common example is this notion of a "balanced amplifier" being simply two amps stuck in a chassis wired to different pins of an XLR connector. In this instance (as you correctly postulated) any voltage imbalance between the two amplifier stages will disturb the necessary null for cancellation of even-order harmonic distortion products. And this voltage imbalance will be affected by any impedance OR voltage imbalances in the preceeding equipment, cables, or the input termination resistances . . . not to mention any differences in gain, distortion performance, or bandwidth between the two amplifier stages. It's even common to see this approach on with two otherwise conventional solid-state power-amp circuits, which makes very little sense at all . . . because the differential input stage alone (with high open-loop gain) is very effective at eliminating even-order distortion products, and the load impedance is then effectively halved across the amplifer output, making the output stage much less linear (everything else being equal).
I think it's exceptionally bad form to design a preamp or amplifier in this manner, but it's alarmingly ubiquitous . . . I guess because this idea of "balanced" is so in vogue right now. There's also a common variation on this where the two amplifier stages share a common differential feedback ladder (like an "instrumentation" op-amp), making the common-mode gain unity, and (hopefully) substantially lower than the signal gain. This latter topology can be somewhat successful, but IMO it's still a bad choice to rely solely on this for input-stage noise rejection . . . the noise is still present to a degree through the entire circuit, and can intermodulate with the signal to a degree dependent on the circuit's linearity.
I believe Ralph's amps are based on differential amplifiers for each stage, and this is one (rare) instance where a single approach can work reasonably well for both interconnection noise rejection and even-order distortion cancellation. Here, the tolerance of the amplifier to source impedance imbalances is mainly the choice and tolerance in input termination resistors, and the tolerance of source voltage imbalances is a function of the transconductance of each stage, and the matching of certain circuit elements (i.e. input triodes and plate-load resistors).
So the short answer to the original poster's question is . . . how well the amplifier works with a simple adapter is dependent on how well the amplifier is designed . . . and as always, simply having a brand name with "audiophile-cred" doesn't mean it's well-designed.
Kirk, thanks very much for another of your exceptionally knowledgeable big-picture perspectives, which certainly makes a persuasive case for the conclusion expressed in your last paragraph.
The Cardas adaptors we have seen do not do anything with pin 3It amazes me how so many manufacturers, both pro-oriented and audiophile-oriented, seem to have it backwards when it comes to rca-to-xlr adapters. Rca to xlr-female adapters, which would be used on xlr outputs, should leave pin 3 unconnected, but almost invariably ground it to pin 1. Rca to xlr-male adapters, which would be used on xlr inputs, should ground pin 3 to pin 1, yet here we have an example of a respected high-end manufacturer leaving pin 3 open.
12-16-10: Tmsorosk: I think most of the responders here have missed the Question. Bobheinatz wants to use a single ended ( 2 wire cable ) in a balanced set up , Which will require a 3 wire cable . You can use adapters or reterminate , but you won't get a balanced signal with a 2 wire cable.The discussion has focused mainly on the EFFECTS of putting an unbalanced signal into a balanced amp, so I don't think the question has been missed.
12-15-10: Kirkus: A common example is this notion of a "balanced amplifier" being simply two amps stuck in a chassis wired to different pins of an XLR connector. In this instance (as you correctly postulated) any voltage imbalance between the two amplifier stages will disturb the necessary null for cancellation of even-order harmonic distortion products. And this voltage imbalance will be affected by any impedance OR voltage imbalances in the preceeding equipment, cables, or the input termination resistances . . . not to mention any differences in gain, distortion performance, or bandwidth between the two amplifier stages.I think it's also worth noting that using an adapter to put an unbalanced signal into that kind of design would result in up to 75% of the amplifier's power capability being unable to be utilized, since the voltage swing capability between the two output terminals would be cut in half.
The protection circuit issue with the ARC amp may be weird, but it doesn't mean the amp needds to be fixed, either. I learned of the issue not because I had a problem, but because I called ARC before I even tried the approach. They explained the issue to me as an inevitable result of the amp's design. Now, it may not be broken, but that doesn't mean you have to like the design.
Of course, it is possible that I "misremembered" the explanation, but I don't think so.
I can only tell you what ARC told me.
They said that RCA to XLR adaptors would cause the VT-130SE to suffer "serious" (or words to that effect) performance deterioration including loss of more than half of the power output capability. They weren't talking about my amp, specifically, but rather every example of the VT 130SE. And ARC said that it wouldn't work properly for - as best as I can recall - the reason I cited, lack of noise rejection tripping the protection.
As a non-technical guy, it would be hard for (and silly of) me to try to defend their design decisions (much less try to justify the consequences of those decisions). I take it from your post (and several of the posts preceding yours) that the technically informed/design community types out there do not approve of this approach. I would never argue with you guys.
All I can say is that the amp sounds great - although I've never tried to use it with RCA to XLRs ;-}
Isn't ARC in your 'hood? Don't you Minnesota high end manufacturing types huddle over hot cocoa when the winter sets in?
Seriously, if you do cross paths with any of their tech folks, put the question to them. I might have somehow mangled the "why", but I'm just about 100% certain that they told me that the design will put out less than half the rated power when used with a single ended input signal via RCA to XLR adaptors.
Martykl, yes, we are all usually out on the lake ice fishing every weekend starting about September or so. I'll ask them next time I see them but I heard that Dean was injured up north recently so I may not see him very soon. Apparently he was near a penguin farm when they got loose and he was nearly waddled to death.
I have the reverse situation.
I am using a Krell power amp which only takes XLR inputs and a Mark Levinson DAC which has both RCA and XLR outputs. I was previously using a Krell KRC preamp which has both XLR and RCA inputs and outputs. All my cables are XLR terminated. I am thinking of switching to a Conrad Johnson ET3 preamp, specifically running XLR cables from DAC to preamp and also running XLR cables from preamp to power amp.
Can anyone please tell me which adapters are suitable? I am confused between male and female, XLR to RCA etc.
Also, again the question of sound degradation in this setup - how best to avoid this?
I runRCA IC's from a Wyetech Opal pre (RCA out) to Krell FPB 300cx (only XLR in) and use BAT adaptors (female on RCA side, male on XLR side). Sounds pretty darn good to me, and has for several years now. I had used Cardas adaptors, which also were good but IMHO the BAT's are superior.
The other alternative is to have amp side re-terminated, which I had done on IC's and quite frankly, could not tell a difference from adaptors. YMMV.
Mikey, as you've gathered from the posts above there are a number of possibilities, and the internal design of the specific components can be an important and unpredictable variable.
My instinct, though, would be to simply use an rca cable between DAC and preamp, and to use a Jensen transformer between preamp and power amp. I'm sure that a call to Jensen will get you some good advice as to selection of a particular model.
I would not recommend using an adapter between DAC and preamp, in part because as I mentioned above most adapters short the signal on xlr pin 3 to ground (most equipment can tolerate that on its outputs, but some cannot), and in part because I doubt that it would provide any benefit compared to an rca-to-rca connection (aside perhaps from saving you the cost of a new cable).
If you want to try an adapter at the preamp output instead of using a transformer, using the adapter in conjunction with your present xlr cables, here are two low priced examples of the kind of adapters that would be suitable. Cardas and others probably offer higher end counterparts, although keep in mind Ralph's comment above that the Cardas adapters he has seen do not ground xlr pin 3. When adapting an rca output to an xlr input, you DO want pin 3 to be grounded.
Mikey, yes that's correct, if you want to use your existing xlr cables in both places.
However, I would not do that without first either verifying with the DAC manufacturer that it can tolerate having xlr pin 3 grounded, or obtaining an xlr-female to rca male adapter that you are certain does not connect pin 3 to pin 1 (perhaps from Cardas).
See this thread for an example of a problem that was caused by using an adapter on an xlr output.
Thanks for your feedback.
I want to keep my existing cables to minimise the costs as I won't get much back for them.
My DAC is a Mark Levinson 36. Do you know if this would give a problem?
Otherwise, I would go for:
DAC output (balanced) to Preamp input (single ended): male RCA to female XLR
- Pin 3 is ground, so if the ML 36 is OK with this.
- Pin 3 is open
Preamp output (single ended) to Power amp input (balanced): male RCA to male XLR
- Pin 3 is ground
Would this be alright and would I minimise the potential for sonic degradation?
My DAC is a Mark Levinson 36. Do you know if this would give a problem?Don't know, but I found some specs indicating that its output impedance is "less than 6 ohms." That is worrisome, because it would mean that large amounts of current would be drawn from the output stage by a short to ground, unless some protection mechanism kicked in.
The adapter choices you linked to look good, except that I would suggest that you confirm with Cardas that that particular adapter leaves pin 3 open.
Just a FWIW: using adapters is always tricky and should be regarded as temporary. If you are interested in getting the most out of your gear, take your cables to a tech and have him install the connector on the end of the cable such that it satisfies your needs. The mechanical connectivity of adapters can job the system of bass impact and dynamic punch- I have seen it many times.
That's probably true- unless maybe you have the manufacturer do it. FWIW though if you set up the system entirely around the idea that everything has to be sold, that *can* limit the performance you can get.
Now if the amplifier has a single-ended input, it might be moot in that application. Its my opinion that if you keep the cable short, you are probably better off using a single-ended cable than you are getting a line transformer to convert to balanced. Most transistor amplifiers have differential inputs and don't care if the input is single-ended or balanced.
However beware if the amplifier's balanced input only works if the amplifier is monostrapped. Such amps often do not have differential inputs so it makes a big difference how you drive them!
Mikey, I can relate to wanting to preserve the resale value . . . But having the correct termination for your application pretty much trumps all other cable-quality parameters. Also, the additional mechanical stress that Atmasphere noted from using an adapter is a very real and common issue; one errant tug on a big XLR cable/RCA adapter can wreck the preamp's connector in pretty short order, leading to a much greater loss in value.
But if you're just wanting to try out the CJ preamplifier before making a final decision, here's what I would do. First, contact Krell and see if their power amp will work with an adapter cable. If they say yes, then use the Markertek adapter (pin 3 grounded) and your existing XLR cables. If not, then use a Jensen DM2-2RX with a short RCA cable from the preamp.
Then just use an RCA cable to connect the DAC to the preamp via unbalanced in/out.
And if you keep this setup, then you can upgrade RCA cables as you see fit, and/or reterminate your amp cable.