Reverse scan cable

[Elia Orselli]

Greetings,

I am wiring an analogue reverse scan board to a Dolby processor and I am confused by Dolby's instructions on how to proceed.

The FB 206, also available on this site, recommends to 3-wire shielded cables so that each channel gets hot, cold and ground reference connected independently, plus a shield drain. This "ground reference" wire should be inside the overall shield BUT not electrically connected to it.

The CP650 manual indeed mentions the same method as the best one. However for "non EU countries" and "if has to be done" also mentions what seems to be commonly done everywhere: a balanced, twisted pairs 4-wire cable is used and the ground reference is connected to the shield.
Screenshot_95.png
I wouldn't mind following the recommended way but I am not sure I have ever seen a cable with a twisted pair plus a single cable inside the same shield.

I notice that Sony on SDDS players seem to follow the same recommended practice as their pinout mentions hot, cold and ground for each channel.
92844888_2671049009883128_3101278395611217920_n.png
I have searched the old forum an I did not find any reference to this recommended way.

Is there anybody that can shed any light on this please?

Thanks!

[Sam Chavez]

BACP and USL always recommended using Belden 8723 that uses two color coded pairs, red/black on left and green/white on right channel. The color coding eliminated confusing left and right channels. The shield or drain wire serves both pairs. We always terminated the shield at the processor end and also at the output of the reverse scan reader to eliminate an RF buzz if need be.

[Dave Macaulay]

When I was installing these, I would run the L/R signal cables each with a 2C shielded cable (like Belden 8451) with the shield connected only at the processor end, and add a reasonably heavy (18ga or so) ground wire connecting the Dolby reader board ground to the processor ground.
Projector and processor are almost always on separate power circuits (and should be) so the utility ground connection between them can go a long way to the breaker panel and back, sometimes to different breaker panels so the grounds travel to and back from the electrical room: slight ground potential differences result, so connecting the shields at both ends can introduce hum and buzz.
You can use 4C shielded (or 3C - but that's not common) using an extra conductor for the ground connection again terminating the shield at the processor end only. I prefer to have an external ground link with normal stp cable.

[Marco Giustini]

Dave,
so are you suggesting that there isn't a single cable that can achieve what is recommended as best practice?
If I understand correctly you would need at least 3 4C shielded cables: two for signal (L+/L-, R+/R-) and one for signal ground (using the pair inside the shield).

[Steve Guttag]

I'm with Sam on this one. We always used Belden 8723 however we also always ran a separate 18awg ground wire to bring the soundhead (which can often be floating...both electrically and on some form of shock mount) to the same potential as the sound processor. This way the shield can do its part to drain any noise and let the ground wire do its part to for potential.

[Marco Giustini]

That makes sense to me. I am just surprised there isn't a specific cable that can do everything under one jacket.

Dave and Steve, what about the ground not being shielded as in the drawings provided?

[Steve Guttag]

The ground not being shielded? That doesn't make sense...something must have changed in translation.

Shield should go to an earth-ground. The purpose of shield is to provide an electrostatic conduit to drain that noise away. The projector side has a questionable earth potential at the soundhead so the shield should NOT tie there as you don't want noise to drain there.

However, with many "active balanced" and unbalanced systems, you need the signal ground potential to be as close to the same as possible...hence we bond the soundhead signal ground to the sound processor ground. Now a proper differential system (e.g. one that may have a transformer) does not need the ground reference as the transformer isolates the signal grounds and the output is inherently differential and thereby effectively eliminates common-mode noise. Good transformers are expensive and bring with them other challenges.

[Sam Chavez]

Steve as right as usual. He should be, I was his first teacher.

[Elia Orselli]

Thank you all!
I think that Marco was referring to the «separate 18awg ground wire»: is it a shielded cable or a plain cable?

[Marco Giustini]

Steve,

I was just reading the FB:

There must be a connection between the ground pins at the Cat. No. 655 solar cell circuit board and the audio common in the Cinema Processor. This connection must not use the shield of the optical input cable, otherwise RF energy can be imposed on the Cinema Processor ground system.
Pin numbers 6 and 9 of each 9-pin D connector ("Projector") on the Cinema Processor allow these connections to be made. The wire that connects either of these pins to the Cat. No. 655 audio ground should pass inside the same shield as the optical input cables and not connect with the shield at any point.

I'm wondering whether an ethernet cable with independently shielded pairs would work as a single cable for the task? Two pairs for the signal and one pair for signal ground. Indeed ethernet cables are not the best when it comes to solder them.

[Steve Guttag]

There is zero reason for having a ground conductor also be shielded. It is merely establishing a zero potential difference. The shield is strictly for noise immunity and hence should drain to the best earth ground. As for our 18awg wire, it is jacketed but otherwise plain. You can use THHN stranded, for instance though we typically used a more flexible cable.

[Marco Giustini]

Thanks Steve,

I am also dealing with a reverse scan setup and the above information is very helpful. I found an old Vic5 RS cable and it doesn't seem to follow that advice - it follows the "if it is necessary" approach pictured in the manual. And personally I am not sure I have ever seen an extra ground cable running alongside the main audio cable to the sound processor. Not arguing, just wondering.

[Steve Guttag]

There, now you've seen one! P6260026.jpeg

The Cell cables (Belden 8723) terminate in DB9 diecast shells where the drain is to the shell. Those green wires going to the Phoenix block in the middle(ish) part of the CP650 tke the soundheads to the same potential as the CP650. No ticks, pops, hums, buzzes either.

I'm also a big believer in metallic conduit (EMT, Greenfield and metallic liquidtight but there is no reason for liquidtight except some like its aesthetics better...I hated it for film machines because the likes of RCA and Simplex soundheads would eventually fill up the liquidtight...which I ended up emptying for digital cinema).

You'd be surprised at how many engineers, technicians, manufacturers don't get the difference between ground and shield. And don't get me started on Ground versus Neutral on AC circuits in the USA.

[Marco Giustini]

I didn't doubt that you were being honest with me. Just wondering why there is a second "if you really have to do it" way, why it says "for non EU countries" and why I never seen a setup like the above!

(very neat install!)

[Sean McKinnon]

A ground should be connected at both ends. A shield may or may not be connected to ground at one end or the other. A shield is like a faraday cage around your wire(s) while a ground is a drain to offer the least resistive path for ground so any noise will be conducted to ground instead of to the signal point. Grounding one end of a shield allows any noise it picks up to also drain but isn’t always necessary or advised.

does that jive with you Steve?