How long is too long for a new set of ICs?

[Frankyc2003]

Due to a slight reshuffle of my system, I am looking into moving my amplifier closer to the speakers (which is a good idea), but that unfortunately means that ICs will have to get considerably longer. Probably in the region of 4m…

Any experiences here about such lengths? Recommended or to be avoided???

All input welcome as usual!

Thanks all

:D

[Yomanze]

4m is really not a problem if you're using good quality screened, low capacitance, interconnects. How many pairs do you need?

I'd be looking at: http://markgrantcables.co.uk/shop/index.php?main_page=product_info&cPath=43_1&products_id=216

[Frankyc2003]

That's good news Neil,

I am looking at 3 pairs.

:eyebrows:

[Firebottle]

Copied this from the Whest spec: High current, Class-A discrete output stage.

My guess is it`ll walk it, I bet the output impedance is commendably low.
If you can try it with some inexpensive I/Cs you should be able to prove it.

:cool: Alan

[Frankyc2003]

'inexpensive'
that's pretty much become my middle name!!!:)

[The Grand Wazoo]

I'm was the king of long interconnects!
Use well screened wire & you should be fine.
I don't have too many long runs in use any more - just 2 x 7 metre pairs between a reel to reel machine and preamp - all is good!.

[Frankyc2003]

Well, that's reassuring…
I was getting all worried for nothing then!!!

:lol:

[Mark Grant]

4 metres is no problem at all.

I have made many much longer than that, the longest was around 20 metres and the customer was very happy with it.

[hoopsontoast]

I just bought 2x 3m Ixos 804 Video Component Cables off ebay, gives me 3 pairs of 3m Stereo interconnects for £20.

http://www.ixoscables.com/cgi-bin/ixos-cables/XHV804-300.html

Old stock Video Component cables are excellent value!

[Frankyc2003]

I am thinking DIY…

What recommendations you have in terms of terminations?

Van Damme seems to be offering a good range of reasonably price cable...

[Reffc]

I've just supplied a 5.5m set to one customer and am working on a 10m stereo pair for another customer, so it's no problem as long as screening is good. All I'd advise first is to check output impedance of your source and input impedance of your amp. For non capacitor coupled sources and preamps (ie all SS gear) use the rule of thimb of output to input impedance should be 1:10 and you shouldn't have a problem. To be safe 1:20 is better.

The only time you have to be careful is valve pre to power as most valve preamps use capacitor coupled output stages and whilst output impedance may be quoted as say 500 Ohms, this will be quoted (mostly) at 1KHz. As frequency rises, so does impedance from capacitor coupled output stages and if the cap is too low in rating, you can see 3KOhms or more by the time frequency has risen to 10KHz (these are actual figures supplied to me by a manufacturer of a particular valve amp). In such cases, long ICs are to be avoided.

[Frankyc2003]

Hi Paul,
Basically my 3 sources are Solid State,

1 arcam rDac
1 Whest 2
1 Graham Slee

all going to my Valve integrated… Ariand pro 845

[Reffc]

Then 20m would be fine before you noticed anything much in SQ drop-off ;)

[Peter Stockwell]

And what's the limit with "passive" preamps ? Does it depend on the pot resistance ?

[Reffc]

Same thing applies. As long as output impedance is relatively low, you can get away with a fair few metres on passives using CDP/Digital sources but may have gain issues over longer runs. I always think it's best to keep passive output leads as short as you can but several metres is no problem. You can use a buffer stage after a passive to drive longer lengths if needed and to ensure output impedance is kept low.

[Arkless Electronics]

I've just supplied a 5.5m set to one customer and am working on a 10m stereo pair for another customer, so it's no problem as long as screening is good. All I'd advise first is to check output impedance of your source and input impedance of your amp. For non capacitor coupled sources and preamps (ie all SS gear) use the rule of thimb of output to input impedance should be 1:10 and you shouldn't have a problem. To be safe 1:20 is better.

The only time you have to be careful is valve pre to power as most valve preamps use capacitor coupled output stages and whilst output impedance may be quoted as say 500 Ohms, this will be quoted (mostly) at 1KHz. As frequency rises, so does impedance from capacitor coupled output stages and if the cap is too low in rating, you can see 3KOhms or more by the time frequency has risen to 10KHz (these are actual figures supplied to me by a manufacturer of a particular valve amp). In such cases, long ICs are to be avoided.

:eek::eek::stalks: NO NO NO!!!!! The exact opposite is the truth!

Oh and not all SS gear has direct coupled outputs by a long way! ;)

[Reffc]

:eek::eek::stalks: NO NO NO!!!!! The exact opposite is the truth!

Oh and not all SS gear has direct coupled outputs by a long way! ;)

In which case I have been misinformed Jez. An amplifier manufacturer confirmed that their valve pre was cap coupled and that output impedance rose with frequency. (don't shoot the messenger :lol:) That being the case, I would have thought that cable capacitance and higher output impedance will start to become problematical for longer IC runs. If that isn't the case, I stand corrected.

[hal55]

Most certainly agree that screening is essential. I need a long interconnect to get from phono pre to preamp, and the best is a ebay 5m screened PC-OFC cable designed for component video duty. Betters my other longish unscreened cables easily, including the ultrathin 80s version Mogami patch cable that I used to use. The odd thing is that cable used to sound great (back in the 80s) but now it sounds a touch hard and ragged (read: not friendly to strings). Can there be that much more HF crap in the air today than thirty years ago?

Hal55

[Stratmangler]

Can there be that much more HF crap in the air today than thirty years ago?

Hal55

I take it that you're connected wirelessly to your router?
Multiply that by the number of households nearby.
Do they have microwave ovens?
Then there's the SMPS that come as standard with pretty well all household appliances. Nowt wrong with SMPS per se, cos there are some very, very good ones out there, but the number of decent ones that don't inject a load of crap back into the mains are probably in the order of less than 1%.

So the power socket you plug the kit into is the main carrier of electrical interference into your stereo.

[Arkless Electronics]

In which case I have been misinformed Jez. An amplifier manufacturer confirmed that their valve pre was cap coupled and that output impedance rose with frequency. (don't shoot the messenger :lol:) That being the case, I would have thought that cable capacitance and higher output impedance will start to become problematical for longer IC runs. If that isn't the case, I stand corrected.

Capacitive reactance falls with increasing frequency whilst inductive reactance rises with frequency ;)

In the case of non power amps, virtually all valve amps are capacitor coupled This has no effect whatsoever when it comes to driving cables etc. SS amps can be direct coupled and indeed often are when designed around + & - split power supplies. Single ended SS gear running from just one supply polarity, such as Naim, are as capacitor coupled as valve amps.

When it comes to driving cables (and indeed anything else) a low output impedance is what we want. Valve amps tend to have higher output impedance than solid state but if they use a cathode follower at the output (most do) then it should still be pretty low at around 200 - 800 Ohms. This is low enough to mean that no cable length likely to be used domestically would be a problem. Lets say we want to drive 10 Meters of average coaxial cable which has a capacitance of 100pF/M and that the output impedance of the amp is 500 Ohms, this give a -3dB point at 318KHz ! Not exactly an issue :)

It only becomes a problem with some of the valve amps that don't use a cathode follower and have a much higher output impedance, even then though it takes quite some length of cable to cause a problem in most cases. With an output impedance of 10K and the same 10M of cable we would get a -3dB point of 15KHz. 10K is very high though! Use 5M of cable and we're up to 32 KHz....

The worst case likely to be found domestically is with a 50K passive pot when its at half volume. This has an output impedance of 12.5K.... This is why it's best to use 10K or 20K pots for a passive. Of course that hypothetical valve phono stage with no cathode follower and an output impedance of 10K would not be at all happy driving a 10K pot!! Horses for courses.

[Reffc]

:doh: Of course it does Jez...I must have been half asleep! If it did't then crossovers wouldn't work!

So it's more a problem where cathode followers aren't used...gotchya.

Where passives are concerned, I've always got better results with short ICs but that as you've pointed out, can partly be down to the source used and whether it's happy driving the total load/lengths of cable. I built the buffer amp for that reason, so any passive can then drive long IC runs.