Low level signals in ribbon cable

Wiring a mixer is a long, tedious and error prone process but at least you can use twin screened cable to look after all the low level signals. One way to simplify wiring is to group a bunch of signals (say four balanced ones) and shoot them down 12way ribbon cable. This is probably OK for line level signals but what about low level signals like mic inputs. A complete no-no or doable?

Cheers

ian

IIRC, some mixer manufacturers used ordinary flat cables for input sub-board connection to a channel strip.

Some used twisted flat cables also.

moamps said:

Some used twisted flat cables also.

Click to expand...

Many thanks, I had forgotten about twisted ribbon.

Cheers

Ian

It's all situational...

Thats why we build prototypes, and why the later revisions are often different.

JR

In my somewhat chaotic days at Cadac, final test, I asked Barry Porter why he only sent signals down one leg of the balanced mix bus. He said one reason was the power consumption of all the extra 5532s, the other was that be lost only 6dBish CMRR driving single legged into the balanced mix bus receive amp, but was still excellent compared  to unbalanced mix bus designs

JohnRoberts said:

It's all situational...

Thats why we build prototypes, and why the later revisions are often different.

JR

Click to expand...

No doubt. I was hoping to draw upon the experience of those who have trod this path before me so as to save me a few revisions.

Cheers

Ian

Hi,
It's example about some manufacturer standard

Cheers,
Pascal

ruffrecords said:

No doubt. I was hoping to draw upon the experience of those who have trod this path before me so as to save me a few revisions.

Cheers

Ian

Click to expand...

I have used ribbon cables for buses inside consoles, but there was little/no voltage inside adjacent wires. If concerned you can interleave grounds.

I joke that consoles are the hardest simple circuits to design. In isolation each circuit block is simple, but something happens when you throw 30+ inside a single chassis.

JR

JohnRoberts said:

I have used ribbon cables for buses inside consoles, but there was little/no voltage inside adjacent wires. If concerned you can interleave grounds.

Click to expand...

At present I crimp regular twisted screen pairs into a Molex connector so for most of its length the signal is fully protected. But hand crimping is a PITA, boring and error prone. I am looking at (trying out) some AMP three pin IDC connectors as the next step -got to be  quicker and less error prone than hand crimping. Beyond that the next step change would be four channels in a 12 way ribbon - but this means routing the balanced input signals of four modules  (channels) along the backplane PCB to a common 12 way IDC. I am already concerned about the length of unscreened signal track this creates so I might initially limit the technique to line inputs and leave the mic inputs wired individually. I guess I am going to end up iterating.

I joke that consoles are the hardest simple circuits to design. In isolation each circuit block is simple, but something happens when you throw 30+ inside a single chassis.

JR

Click to expand...

Many a true word etc.

Cheers

Ian

The new trident  88 consoles and following like do the following. There is a rear input board with  jacks on it and traces that feed a ribbon cable.  they do 8 channels of mic, line, insert snd and ret. , direct out,  monitor path,  and a special mic insert(which bypasses the mic pre in favor of whatever pre is plugged into that connection) all on a ribbon cable going into a mother board which has ribbon connections to join motherboards together for each bucket and has 8 multiple connectors to plug the channels in.  I never seen that as an issue.

pucho812 said:

The new trident  88 consoles and following like do the following. There is a rear input board with  jacks on it and traces that feed a ribbon cable.  they do 8 channels of mic, line, insert snd and ret. , direct out,  monitor path,  and a special mic insert(which bypasses the mic pre in favor of whatever pre is plugged into that connection) all on a ribbon cable going into a mother board which has ribbon connections to join motherboards together for each bucket and has 8 multiple connectors to plug the channels in.  I never seen that as an issue.

Click to expand...

Do they do ALL that lot on ONE ribbon cable? or is it arranged as a channels worth on one ribbon or is it all mics on one ribbon, all lines on anther etc?

Just seeking clarification.

Cheers

Ian

ruffrecords said:

Wiring a mixer is a long, tedious and error prone process but at least you can use twin screened cable to look after all the low level signals. One way to simplify wiring is to group a bunch of signals (say four balanced ones) and shoot them down 12way ribbon cable. This is probably OK for line level signals but what about low level signals like mic inputs. A complete no-no or doable?

Cheers

ian

Click to expand...

It depends very much on your expectations in terms of X-talk, which is the only real issue.
French broadcast test was done feeding one connection a nominal +4dBu signal into the Line input, then checking X-talk in the mic channel at max gain with 200 ohms load; X-talk was to be less than -40dB at 10kHz. All multi-pin connectors (EDAC, Socapex, Harting...) failed the test! No need to say ribbon cable was not conceivable.
I would route the mic input separately, but all the line level connections could be in a single ribbon cable. You may alternate signal connections with a ground guard.

abbey road d enfer said:

It depends very much on your expectations in terms of X-talk, which is the only real issue.
French broadcast test was done feeding one connection a nominal +4dBu signal into the Line input, then checking X-talk in the mic channel at max gain with 200 ohms load; X-talk was to be less than -40dB at 10kHz. All multi-pin connectors (EDAC, Socapex, Harting...) failed the test! No need to say ribbon cable was not conceivable.
I would route the mic input separately, but all the line level connections could be in a single ribbon cable. You may alternate signal connections with a ground guard.

Click to expand...

Interesting test. Were the line and mic inputs on the same or different channels? If they were on the same channel you would also be including the isolation of the line input when the mic input is selected on that channel.

i am beginning to think mic signals are too precious to trust to anything other than twin screened cable but, as you say, line levels will probably be OK.

Cheers

Ian

Crosstalk happens, but is managed most critically inside broadcast consoles that can have wildly different signals happening side by side in the same platform at same time.

You probably don't want to know crosstalk figures at 20 kHz between even respected console channels.  :

Life is a simultaneous equation with multiple variables... pick the most important to you, or your customers. The paper (spec) chase can affect your sanity, as optimizing one may degrade another, not to mention cost. Costs that get multiplied 35x or 50x can add up quickly.

JR

ruffrecords said:

Interesting test. Were the line and mic inputs on the same or different channels?

Click to expand...

Tests were done in several cases; line and mic of same channel, but also of different channels. For radio games, they don't want any X-talk between a channel that carries sensitive info and the players sends. Actually, the most common case they want to avoid is one hot signal polluting the candidates' channels. This hot signal can be a radio mic , a telephone insert or tape playback.
We had to replace all the mic input EDAC connectors with XLR's. The typical 2pF inter-capacitance of EDAC's  induced -92dB x-talk into 200r, which, with the max gain of 65dB, was more than 12dB off the -40dB mark.

i am beginning to think mic signals are too precious to trust to anything other than twin screened cable but, as you say, line levels will probably be OK.

Click to expand...

Better safe than sorry...

Cat5 cables have different twisted turns per pair to reduce xtalk IIRC.  Maybe too custom for audio.

Some things I've noted in the past.

Many times when I see ribbon cables used they always put a grounded wire between all the signal cables, I'm guessing to create some sort of ground plane that screens the signal strands to some extent.

Harrison used screened ribbon cable in their series 10 & 12 consoles.  This is ribbon cable that has a copper mesh on one side of it that is grounded.  The Harrison consoles are very quiet.  I have looked around for this cable and it is not so commonly available now & is expensive.

A few years ago I was at the Audio Note factory in Hove to sell them some valves.  I got a brief tour round the place.  I was looking in one of their preamps which used ribbon cables to join the connector pcb to the main board.  What I noticed was that they ran the cable flat against the chassis.  My thoughts were that with the chassis being grounded & the ribbon cable being so thin, the cable is effectively screened by the ground plane that the chassis creates.

I've seen ribbon cables sandwiched between grounded adhesive copper tape.

https://uk.farnell.com/advance-tapes/at526-copper-33m-x-25mm/at526-copper-shield-tape-25mm/dp/1766743?st=copper%20tape

mrclunk said:

I've seen ribbon cables sandwiched between grounded adhesive copper tape.

https://uk.farnell.com/advance-tapes/at526-copper-33m-x-25mm/at526-copper-shield-tape-25mm/dp/1766743?st=copper%20tape

Click to expand...

Thanks for that. It made me suddenly realise you can get conductive braiding:

https://uk.farnell.com/wurth-elektronik/30508/sleeving-cond-nylon-per-m/dp/1909676?st=conductive%20sleeving

Cheers

Ian

That copper tape is expensive. At least from McMaster.  So is the spray stuff.  I don’t imagine it would be much of a cost savings over other cable.  I picked up some braided shield surplus but it’s also pretty expensive to buy new.