Another question for Frank (or anybody)

gregmav · « **on:** 12 Mar 2009, 05:05 pm »

I have always thought that if you want to run longer cables between preamp and power amp, the suggestion has been to use "balanced" cables/connectors.

My question is: If I want to place my preamp with the rest of my equipment in one part of the room, and then place my power amp near my speakers in another part of the room, can I do that with standard unbalanced RCA terminated cables? If so, what would be the recommended maximum length I could use?

Any thoughts or opinions would be appreciated.

turkey · « **Reply #1 on:** 12 Mar 2009, 05:26 pm »

Quote from: gregmav on 12 Mar 2009, 05:05 pm

I have always thought that if you want to run longer cables between preamp and power amp, the suggestion has been to use "balanced" cables/connectors.

Yes, that's the best way.

Quote

My question is: If I want to place my preamp with the rest of my equipment in one part of the room, and then place my power amp near my speakers in another part of the room, can I do that with standard unbalanced RCA terminated cables? If so, what would be the recommended maximum length I could use?

It can be done. Frank may be able to give a definitive answer, but I'd imagine you could get away with a 25 foot run without too much trouble.

avahifi · « **Reply #2 on:** 12 Mar 2009, 05:54 pm »

Yes, you can do it the way you described, but I suggest it is not a very good idea.

In general, a power amplifier is designed to drive loads, a preamplifier, within very specific limits, is not designed to drive loads.

A 25 foot long interconnect cable will represent a rather large capacitive load to the preamp. The capacitance per foot of the cable times the number of feet. This will do two bad things. First as this "capacitor" gets bigger, it will tend to roll off the very high frequencies. The amount of roll off will depend upon the value of the capacitor into the input impedance of the amplifier. For this reason, all other things being equal, keep the interconnects as short as practical and as low capacitance as possible.

The second bad thing is that the capacitive load has to be charged and discharged ever duty cycle of playback. This requires current. The current requirements to charge the capacitor may exceed the output current capability of your preamp's line stage. When the demand for current exceeds what is available, the distortion goes towards 100 percent. Our Insight and Ultra preamps will do well with this, the pure tube T8 not as well as small signal tubes alone have poor current drive capability and we are doing the best we can here with the relatively high output current 6N1P tubes.

Actually there is a third bad thing. A capacitive load tends to slow down the feedback loop and cause the feedback to arrive late. This can been seen with simple square wave tests - - as the capacitive load increases, the leading edge overshoots. If the load is too great, the unit can go into full bore oscillation. Again, extremely unlikely with AVA equipment which is designed to drive nasty capacitive loads, but the larger the capacitive load, the more the output will deviate from the input, and there is a name for that - - - distortion.

Finally the amplifier is a load driving device. 25 feet of 14 gauge standard two conductor speaker wire is a negligible load for an amplifier as compared to the loudspeaker itself.

So it seems to make sense to let the amplifier drive most of the load, an easy to drive speaker wire, and remove as much load as possible from the preamp which has nowhere near the load driving capability as a well designed power amplifier.

That is the reason I think your scheme is not a good idea.

Best regards,

Frank Van Alstine

P.S. One other thought. The very low level signal to the amplifier is much more sensitive to noise and RFI pickup than the high level signal to the speakers (especially if you twist the speaker wires 3-4 turns per foot). Your idea will maximize noise pickup possibilities.

charmerci · « **Reply #3 on:** 12 Mar 2009, 06:06 pm »

Frank's above response is the reason I don't read audiophile magazines anymore on a regular basis, i.e. smart answers to the audiophiles' audio "tricks".

cdorval1 · « **Reply #4 on:** 12 Mar 2009, 06:48 pm »

Amen.

turkey · « **Reply #5 on:** 12 Mar 2009, 07:49 pm »

The "Wire and Cable Scene" articles in Issues 16 and 17 of The Audio Critic are available for reading, and would seem to indicate that keeping speaker wires short is more important.

avahifi · « **Reply #6 on:** 12 Mar 2009, 08:16 pm »

Lets see, if we should keep the speaker wires short, we need to remember that the speaker wire is essentially nothing more than an extension of the voice coil wires in the speaker (at least in a single driver system). Thus we must now shorten up those voice coil wires folks to follow this advise. Take your drivers apart and unwind at least 500 feet of the 600 feet or thereabouts of the wire wound on your voice coils! Lets really shorten them up for different sound.

I guarantee if you take this advice you will have different sound, if any sound at all.

Franlk

TomW16 · « **Reply #7 on:** 12 Mar 2009, 08:26 pm »

Quote from: avahifi on 12 Mar 2009, 08:16 pm

I guarantee if you take this advice you will have different sound, if any sound at all.

That reminds of a quote years ago that said "If the interconnects are shortened too much, a drastic reduction in sound results" meaning that the interconnects must be long enough to attach the components.

I appreciate the solid engineering that goes into AVA components.

Keep up the great work Frank!

Tom

Wayner · « **Reply #8 on:** 12 Mar 2009, 08:29 pm »

I suppose in a perfect world, all the cables would be really short (they'd also be super-conductors

), but many times with particular room designs, we all can't have our systems all merged conveniently together. Causing the least amount of harm to the overall signal should be the mission. Higher voltage (and current) speaker wires are less prone to interference then interconnects with volts (or micro-volts) traveling thru them. It is a compromise, but it is the best choice for causing the least bit of harm. I have speaker cables running right next to power cords, not because of choice, but because of the situation my systems are in. There is no discernable effect on the signal transfer.

Many manufacturers say that running speaker wire up to 50 feet isn't really a problem as long as the user considers what gauge wire they should use at those distances. 14 or 12 awg would be my choice.

Wayner

turkey · « **Reply #9 on:** 12 Mar 2009, 08:37 pm »

Quote from: avahifi on 12 Mar 2009, 08:16 pm

Lets see, if we should keep the speaker wires short, we need to remember that the speaker wire is essentially nothing more than an extension of the voice coil wires in the speaker (at least in a single driver system). Thus we must now shorten up those voice coil wires folks to follow this advise. Take your drivers apart and unwind at least 500 feet of the 600 feet or thereabouts of the wire wound on your voice coils! Lets really shorten them up for different sound.

I guarantee if you take this advice you will have different sound, if any sound at all.

Franlk

Funny guy.

Doesn't really address the issue though.

turkey · « **Reply #10 on:** 12 Mar 2009, 08:42 pm »

Quote from: Wayner on 12 Mar 2009, 08:29 pm

Higher voltage (and current) speaker wires are less prone to interference then interconnects with volts (or micro-volts) traveling thru them. It is a compromise, but it is the best choice for causing the least bit of harm. I have speaker cables running right next to power cords, not because of choice, but because of the situation my systems are in. There is no discernable effect on the signal transfer.

Yes, and I have RCA patch cords running right next to power cords, for the same reason you do. There is also in my case no discernable effect on the signal transfer.

Quote

Many manufacturers say that running speaker wire up to 50 feet isn't really a problem as long as the user considers what gauge wire they should use at those distances. 14 or 12 awg would be my choice.

Have you read the articles I mentioned?

geezer · « **Reply #11 on:** 12 Mar 2009, 09:07 pm »

Quote from: turkey on 12 Mar 2009, 08:37 pm

Quote from: avahifi on 12 Mar 2009, 08:16 pm
Lets see, if we should keep the speaker wires short, we need to remember that the speaker wire is essentially nothing more than an extension of the voice coil wires in the speaker (at least in a single driver system). Thus we must now shorten up those voice coil wires folks to follow this advise. Take your drivers apart and unwind at least 500 feet of the 600 feet or thereabouts of the wire wound on your voice coils! Lets really shorten them up for different sound.

I guarantee if you take this advice you will have different sound, if any sound at all.

Franlk

Funny guy.

Doesn't really address the issue though.

It does for me.

Zheeeem · « **Reply #12 on:** 12 Mar 2009, 10:15 pm »

Quote from: turkey on 12 Mar 2009, 08:37 pm

Funny guy.

Doesn't really address the issue though.

No, although Frank's first post does address the issue. And indeed, many of the points in Frank's first post - where he talks about RCL issues - are points noted by Aczel. But Aczel leaps to the conclusion that "...the RLC values of interconnects are of considerably smaller consequence than those of speaker cables" that is, to tell the truth, completely looney. The output after ICs is amplified; not so the output after speaker cables. And preamps are not designed to deal with the same sort of loads that amps are.

boycephoto · « **Reply #13 on:** 12 Mar 2009, 10:52 pm »

OK, If I unwind the voice coils in my speakers to see how they sound with short connections, will they still be under warranty?

About 30 years ago I had a pair of Kenwood LO7m mono amplifiers. Kenwood supplied a 50 foot RCA cable with each amp to attach to the preamp, they also came with an 18 inch speaker cable. When I sold the amps I kept the cables. I still use them to connect my OmegaStar EC to OmegarStar EX 440H. The EC sits about 18 feet from the 440H. If the high end is rolled off I can't hear it. The EC seems to have plenty of drive. Now I have made custom base boards in my room. They are 1x6 inch oak that I have routed 2 - 3/4 inch grooves on the back side that the cables sit in. The cables do not touch each other and also are separated from AC lines. The biggest benefit is that I don't see the cables.

I have always kept the interconnects short between the Super PAS 3 and Ultimate 70.

Keep listening and enjoying.
Dave

turkey · « **Reply #14 on:** 12 Mar 2009, 11:37 pm »

Quote from: Zheeeem on 12 Mar 2009, 10:15 pm

No, although Frank's first post does address the issue. And indeed, many of the points in Frank's first post - where he talks about RCL issues - are points noted by Aczel. But Aczel leaps to the conclusion that "...the RLC values of interconnects are of considerably smaller consequence than those of speaker cables" that is, to tell the truth, completely looney. The output after ICs is amplified; not so the output after speaker cables. And preamps are not designed to deal with the same sort of loads that amps are.

The test results that Aczel shows tend to back up his opinion. Frank has given nothing more than anecdotal evidence.

I don't think that whether or not the signal is amplified after the wire or not is relevant. What is relevant is the load seen by the preamp and amp.

Whether or not preamps are designed to deal with the same loads as amps are is also not relevant. What is relevant is whether or not the load of a short or long connection to a power amp is within the output capabilities of the preamp. I'm not seeing anything to show that it isn't.

If you have an AVA preamp, it doesn't apply anyway. I based my original answer on the person having an AVA preamp since he was posting in the AVA circle. Well, unless he has a T8 preamp, and you probably already know what I think of toobs.

Zheeeem · « **Reply #15 on:** 13 Mar 2009, 12:21 am »

Quote from: turkey on 12 Mar 2009, 11:37 pm

The test results that Aczel shows tend to back up his opinion. Frank has given nothing more than anecdotal evidence.

No. Actually not, in either case.

Aczel's test results simply show the performance of his respective amp and preamp under varying load conditions. The results will change with different equipment. Or, to put it another way, his results are valid only for that particular gear and cannot be extrapolated. So, unless you have the same equipment, the results are useless.

Frank actually did not provide any anecdotal evidence. He provided good engineering based explanations of why long IC runs can be a problem (or serious problem) under certain conditions. Frank's explanations can be generalized across a wide range of gear.

gregmav · « **Reply #16 on:** 13 Mar 2009, 12:39 am »

Thanks for all the feedback.

Is it possible to get away with using cables from a preamp to a power amp that are between 10 to 15 ft long without any deterioration in signal quality? I am thinking of purchasing an Insight EC preamp and a Insight Power amp. Would using balanced cables with an adapter to convert to RCA on each end create problems? Frank??????????? Anybody???????

martyo · « **Reply #17 on:** 13 Mar 2009, 12:45 am »

Remember you can email, pm, or phone him too. The phone can be helpful when one answer leads to another question...........

dB Cooper · « **Reply #18 on:** 13 Mar 2009, 01:52 am »

I thought capacitance tended to roll off low frequencies, and inductance tended to roll off high frequencies, that being the reason for the inductors (coils) scattered throughout my Dyna FM3 tube tuner on the heater circuits (to curtail stray RF in the wrong places). But then, I got my EE degree out of a Crackerjacks box.

avahifi · « **Reply #19 on:** 13 Mar 2009, 10:49 am »

Evidence needed? Simply measure the capacitance of your pet interconnect cable. Wire a capacitor of that value across (in parallel with) the output of your preamp. Use a couple of spare RCA plugs to do this easily.

Then measure the high frequency square wave performance with and without the capacitive load at the output. When you see the results I strongly suspect you will go for cable length and capacitance per foot that is the minimum practical.

Regards,

Frank Van Alstine