Unbalanced Interconnect length...

[Chops]

Some of you might know that I have upgraded from my Emotiva XPA-2 Gen 2 amp to a beautiful pair of Odyssey Audio Stratos monoblock amps.

Anyway, I've known for some time that these amps do not utilize true balanced inputs, they are just convenience XLR's tapped into the standard unbalanced RCA connectors. That said, many owners of Odyssey amps claim that unbalanced interconnects actually sound better. This is all fine and dandy except for the fact that I need a 30 ft run of cable to reach from the XP-20 to the Stratos amps.

I've never run unbalanced cables that long other than in car audio, and that was usually 17 ft to 20 ft at most, and with an AudioControl line driver up front right off of the head unit to boost the head unit's preamp voltage (usually 2v t0 4v) up to 9.5v rms (13v peak).

Considering the XP-20 has a max output of 7v (nearly 15v or so around clipping which I would never run my system that hard), I think that I should probably be fine running 30 ft of cable in the house.

My question is, what are your experiences running a good length of unbalanced interconnects in your systems? Did you find one brand or series better than the other?


Since the majority of my cables are Wireworld, I think I'd like to keep with them, and most likely upgrade some of them as well along the way.

Many thanks in advance!

 

[DonH50]

30' should be no problem if the runs are not near power lines or other noise sources. I have run 100+ feet line-level signals without problems in the past, and until a year or two ago had two 50' runs to my rear subs with no problems.

Can't help on cable brands, sorry. I use non-WBF-approved crap cables from Pro-Co, Blue Jeans Cable, and so forth. The long ones I have around are from my recording days when I needed to reach the stage from my mixer, or mixer to amps on stage, or some such so they tend to be rugged and well-shielded (braid + 100% foil) but are not audiophile brands. I have a couple pair of double-shielded cables I used when they had to run in the power channels, and that helped somewhat in keeping out coupled noise (mainly from the light controllers, which put a lot of HF garbage on the power lines).

Edit: The Maggies look like 1.x-series, have you considered going up to 3.x? I started with MG-Is but found the larger panels for more bass and especially true ribbon in the 3 series was a major upgrade. Curious - Don

 

[adrianywu]

The most important parameters to consider is the output impedance of your preamp, the input impedance of your power amps and the capacitance per meter of your cables. The output impedance, together with the capacitance of the cable forms a low pass filter, which means high frequencies are attenuated. You can calculate it here: http://www.sengpielaudio.com/calculator-cable.htm
The longer the cable, the higher the total capacitance. It might not matter if your preamp has a very low output impedance, but if not, you will need to choose your cables carefully.
I tend to go in the other direction. I use only balanced interconnects, with balanced differential preamp and amps. My cables are all unshielded, and have very low capacitance. I have no problem with noise, but I don't run very long cable lengths.

 

[Ron Resnick]

I use non-WBF-approved crap cables from Pro-Co, Blue Jeans Cable, and so forth.

So do I (200 feet of Belden 1192A). But I was not aware that our cables are non-WBF-approved!

 

[Ron Resnick]

Hi Chops,

I am sympathetic to this question. All else being equal I would use unbalanced cables in my 50 foot interconnect runs between my line stage preamp and the amps for my panels, and between the line stage preamp and my self-powered woofer towers. (So this is two pairs of cables, each cable 50' long.)

The only reason I currently am not using unbalanced cables is because I believe that VTL pre-amps and VTL amplifiers and Gryphon amplifiers not only use true balanced inputs and outputs, but also are actually optimized for balanced operation.

This is hugely controversial but, based on private discussions and Masters & Makers interviews, I think there is evidence to support the proposition that balanced connections and balanced cables tend to suppress even order harmonics. I like the sound of even order harmonics, so I would rather not have them suppressed.

I just don't know how to evaluate on a net basis what I think would be a uptick in sound quality from single-ended cables versus any detriment to sound quality occasioned by using the single-ended inputs and outputs of all of my components which I believe are optimized for balanced operation.

What should I do? Obviously I should swap out all of the balanced cables for single-ended cables, and see if I hear a difference, and, if I hear a difference, determine which I prefer.

Unfortunately I cannot do this, because while I can easily replace the balanced Belden cables with single-ended Belden cables I cannot do the same for all of the short interconnects which are Cardas Clear Beyond.

 

[JackD201]

If I'm not mistaken unbalanced cables are rated to somewhere in the vicinity of 120 ft.

 

[DonH50]

So do I (200 feet of Belden 1192A). But I was not aware that our cables are non-WBF-approved!

That's a long run! Isn't 1192A a star-quad cable for mics? IIRC, then that is a great choice for a long interconnect, albeit not 100% shield coverage, but unless you are in an extremely high-noise environment it won't matter (plus the star quad arrangement improves LF noise rejection).

A balanced connection should reject common-mode noise, and a differential circuit does reject (or significantly reduce) even-order distortion, but neither should affect the even-order signal from the source. The influence of the cable on even order signals is negligible; the cable only passes along what was sent into it, and has no idea the harmonic content of the source. As a thought experiment, if I send a signal having 1 kHz fundamental and significant 2 kHz (second harmonic) energy, why would the cable suppress the 2 kHz signal? What if I start with a 2 kHz signal, would it get attenuated? (No!) A properly working cable will readily pass all signals in the audio band (and well beyond). What is true IME (as well as many others') is that some folk do prefer some amount of low-order even harmonics (distortion or not), but that has nothing to do with cable topology. It partially explains why single-ended circuits (tube or SS) may be preferred, though I tend to think the amount of higher-order harmonics is more important (a topic well beyond the scope of this thread, and others such as Ralph @Atmasphere have previously discussed in detail).

I think all my posts and stuff (equipment) are non-WBF-approved; I always seem to catch flack coming from the objective side of things. No worries, I rarely post, and try to stick to what I (think I) know about the technical stuff. Preference remains just that, and I still miss my old tube gear, hiss and harmonics included. Well, maybe not the hiss...

If I'm not mistaken unbalanced cables are rated to somewhere in the vicinity of 120 ft.

I am not aware of a "rated" limit; it would depend upon the components (drive and load impedances and signal level, e.g. line vs. mic), the (noise) environment, and the cables. There is a limit for S/PDIF (10 m) and AES3 (100 m) single-ended copper cables sending digital signals, but I do not recall a limit for analog audio (I may have forgotten). That said I tend to look very closely when going over 50'~100' because it is so easy to pass by a noise source like a power run or something and end up with injected noise. I assume yours is a rule of thumb and I tend to agree with it. I have had 20' (or less) runs with noise problems, and 100 m (~300') runs that were noise-free, but in very different locations and systems.

FWIWFM - Don

 

[Ron Resnick]

Thank you, Don!

I have 200 feet of 1192A in total (four cables of 50' each).

 

[microstrip]

If I'm not mistaken unbalanced cables are rated to somewhere in the vicinity of 120 ft.

I can't see how we can have a general rating for cable length - they can have capacitance going from a few picofarad per foot to a few hundred of picofarad per foot.

In practice I have found that, for example, Cardas IC cables are great fo long lengths ~30 feet.

 

[Another Johnson]

3 meters was my limit with typical mid level rca interconnects, but as others have shared, it really is dependent on your gear and your noise environment.

 

[microstrip]

Thank you, Don!

I have 200 feet of 1192A in total (four cables of 50' each).

You will only discover if they are adequate after you try others for comparison. Recently I tried changing a pair of XLR Mogami 2534 15 feet cables connecting the Lamm LL2 preamplifier to the JLAudio subwoofers in a friend system with an auidiophile cable and it was a real improvemet in space and bass definition. It was unexpected - we did it mostly for the fun.

 

[JackD201]

AES/EBU for unbalanced microphone grade cables

 

[adrianywu]

My unshielded balanced interconnects, constructed by myself, have capacitance of 70pF per meter. A shielded cable will have capacitance in the hundreds of pF/m. Even though the corner frequency of the low pass might be over 100kHz, don't forget, the filtering effects of all the cables (and between amplification stages) are cascaded. There is a low pass filter between each stage of amplification, and between each piece of equipment connected by cables. The total effect might not go down to as low as 20kHz, but attenuation of signals up to 100kHz is noticeable, even though we cannot "hear" sounds above 20kHz. That's why I tend to use larger interstage coupling capacitors than usual (manufacturers tend to use the lowest value possible to save cost).

 

[DonH50]

Typical coax runs around 20 pF/ft or about 66 pF/m or less. For example, Belden RG-6 variant 1892A is about 53 pF/m:


Not sure what cable has hundreds of pF/m, that seems very high. Perhaps some low-impedance differential or quad cable?

For a SS preamp having 100-ohm output impedance, 1000 pF (about 19 m of the above coax) yields about 1.6 MHz bandwidth. If you have a tube preamp having very high output impedance of say 1000 ohms, then the -3 dB frequency drops to 160 kHz.

When cascading components, each device will have its own input and output impedance to be considered, and assuming they are active each cable is driven independently. However, you must also consider the bandwidth of the components themselves, which further reduces the end-to-end bandwidth when cascaded.

 

[adrianywu]

Typical coax runs around 20 pF/ft or about 66 pF/m or less. For example, Belden RG-6 variant 1892A is about 53 pF/m:

View attachment 115270
Not sure what cable has hundreds of pF/m, that seems very high. Perhaps some low-impedance differential or quad cable?

For a SS preamp having 100-ohm output impedance, 1000 pF (about 19 m of the above coax) yields about 1.6 MHz bandwidth. If you have a tube preamp having very high output impedance of say 1000 ohms, then the -3 dB frequency drops to 160 kHz.

When cascading components, each device will have its own input and output impedance to be considered, and assuming they are active each cable is driven independently. However, you must also consider the bandwidth of the components themselves, which further reduces the end-to-end bandwidth when cascaded.

Indeed, properly designed coax cables should not have excessively high capacitance, since these are designed to achieve a particular characteristic impedance. The capacitance is a function of the distance between conductors, the dielectric properties of the insulators and the configuration of the conductors. Many exotic "high end" cables have rather extraordinary reactive characteristics and function as filters, i.e. tone controls, which is why some people find one cable "better" than another. The issue of the cascading effects of filters; output impedance/cable capacitance as low pass, and coupling caps/input capacitance as high pass, are very real, esp. with some tube equipment having high output impedance and inadequate coupling capacitance. Maybe that is part of the reason why they are felt to have an outstanding midrange, since they have nothing else aside from the midrange.

 

[Lagonda]

You will only discover if they are adequate after you try others for comparison. Recently I tried changing a pair of XLR Mogami 2534 15 feet cables connecting the Lamm LL2 preamplifier to the JLAudio subwoofers in a friend system with an auidiophile cable and it was a real improvemet in space and bass definition. It was unexpected - we did it mostly for the fun.

You don't have to go audiophile, just going to 2549 instead of 2534 will give you a substantial improvement !

 

[Chops]

Great information here! Thanks so far for the input.

Just for convenience purposes, here's a few specs for my preamp and amps...

Pass Labs XP-20
Frequency Response: 2Hz to 60kHz
Output Impedance: 50 ohms single ended/1k ohms balanced
Output Voltage: 7v single ended/15v balanced

Odyssey Stratos Monoblock
Frequency Response: 1Hz to 600kHz
Input Impedance: 22k ohms


Looks like I/O impedances between the two units are well within safe limits, and I believe the XP-20 has more than enough "drive" to run SE interconnects at a decent length.

 

[DonH50]

Great information here! Thanks so far for the input.

Just for convenience purposes, here's a few specs for my preamp and amps...

Pass Labs XP-20
Frequency Response: 2Hz to 60kHz
Output Impedance: 50 ohms single ended/1k ohms balanced
Output Voltage: 7v single ended/15v balanced

Odyssey Stratos Monoblock
Frequency Response: 1Hz to 600kHz
Input Impedance: 22k ohms


Looks like I/O impedances between the two units are well within safe limits, and I believe the XP-20 has more than enough "drive" to run SE interconnects at a decent length.

The XP-20's 50 ohms driving 50 feet of 25 pF/ft coax yields about 2.55 MHz bandwidth, should be sufficient. Your runs are shorter and the coax should have lower capacitance if you stick to "normal" coax.

 

[Chops]

I just got done speaking with Galen from ICONOCLAST (Belden). I do plan on upgrading all of my Wireworld cables to ICONOCLAST eventually, but for right now wanting to pinch my pennies for the dedicated 20 amp lines upgrade to the room.

So with that in mind, Galen recommended the Blue Jeans Cables' LC-1 RCA cables, so I just ordered a 30ft pair of those. They should be here in a couple of day.

I'll let you guys know how it turns out. Thanks for the input!

 

[Chops]

I was thinking... If this works out as planned, should I also switch my Altair G1 over to unbalanced as well?

I mean, what's the point of running only one piece of equipment balanced if everything else is unbalanced, especially when I can use a 1m or less interconnect from the streamer to the preamp?

Plus, that would just be balanced circuitry having to convert over to unbalanced in the preamp anyway. Am I right in this thinking?