Author Topic: The 1986 'coat hanger' mod - what and why  (Read 7524 times)

amillar

  • Full Member
  • ***
  • Posts: 181
  • Karma: +15/-0
    • United Kingdom
The 1986 'coat hanger' mod - what and why
« on: October 03, 2017, 01:12:33 PM »
Another bit of history that came out of a recent thread - what is the difference between pre and post 1986 611s and backplanes.

It's all about preventing hum pickup by balancing the mix bus inputs. In 1985 we built two desks for SWF in Germany. One was (relatively) easy, a 48 channel for a radio studio. The other was a 64 channel L shaped for a TV studio, I think the first L shaped we'd made, certainly the biggest. The TV studio was a fantastic installation, for their equivalent (for UK readers) of "Later with Jools Holland" or "Whistle Test". Two separate studios with a control room in the middle so six bands could be set up at once. And it hummed and buzzed. Really, really, badly. A whole crowd of us flew over with Colin to work out what was going on. It was a truly fantastic installation, probably the best I've ever seen (remembering I'm BBC trained). Huge external patchbay using Lemo connectors, must have cost a fortune, fantastic separate earthing pretty much down to the earth's core. And the German engineers kept saying "our old mixing desk, it did not do this" - the old desk was a Studer with something like 12 channels, if a Russian tank had hit it I would feel sorry for the tank.

Useful lesson: we stuck oscilloscopes everywhere but could not see the noise anywhere. Eventually we asked the Germans if they had a spectrum analyser, of course they had a top-of-the range (for 1985) piece of kit there, and finally we could see it. Just showed that ears are very, very good at hearing buzzing way below the noise floor. And now we could see it we could work out that most of it came from the three phase studio lighting rigs.

By a lot of trial and error we worked out that the biggest problem was where we had run the busses in ribbon cables across the joint of the L shape. Foil screening around them made no difference at all, which showed it was magnetic rather than electric field pickup - a nightmare to get rid of. Colin wanted to cover the desk in mu-metal, I had to explain to him that it would be vastly cheaper to give SWF their money back and and also pay for them to buy a Neve!

Back in Witney (the old SSL factory after we got too big for Stonesfield and before we bought Begbroke) I built two HUGE coils to generate magnetic fields with a desk inside. Fortunately we had a client who had ordered a desk but hadn't finished building the studio, so I had a guinea pig to play around with. And at this point there was a real bit of luck. When the mixamps (both in the 651 and 611) had been designed, there was a theory that making the inputs balanced would make them quieter, even if one leg was strapped straight to earth. Actually this theory isn't quite true, but it did mean that all the mixamps were ready to take a balanced input. So provided the wiring in the desk was changed to be balanced - provided the two inputs to the mixamp were fed with inputs of the same wiring length run closely together and with the same resistance distribution through the desk - magnetic pickup would be cancelled out. It didn't matter that the signal was only fed into one leg, that makes the signal to noise ratio 6dB worse than it could be, but no worse than it was before the mod. And we were trying to do this very quickly and cheaply, so we weren't about to start slinging in loads of extra opamps. (An issue I might write about another time, in 1985/86 the 4000 was seen as a 'dead' product, after all surely the digital team were about to release the digital desk, so little money could be justified on it. That's also how I, as the most junior engineer in the company, came to be doing all this fun stuff!)

The lower backplane is pretty simple, so it was easy to get a new card designed to add the extra balanced quad and 32 channel buses and the 'balancing' resistors to ground. By pure dumb luck the wiring to the 651 was all ready to take balanced inputs. The 11 (group amp) card on the 611 needed a tiny mod to make the track mix amp properly balanced for perfection, just a trim pot to the R6 position. Although in practice I found that just changing the buses made the balance good enough that the pot wasn't needed. I see that I did leave this pot in when I designed the 294 G series group amp card - really I should have turned this into a proper balanced mix amp then, I don't know why I didn't.

The upper backplane would have been a pig to redesign, remember this was the days when circuit boards were designed by putting strips of sticky crepe tape onto film! (Including, of course, the 611, a fantastic PCB layout job by a very talented guy.)  Again by dumb luck (or maybe this had been half thought about) the aux busses already had their own 0V connection to the 651, so rather than fitting the balancing resistors to the backplane I bodged them to the 611.

Now, at this point my memory of the mod gets a bit vague, perhaps someone could help - does the 'coathanger' connect to these aux resistors on the back of the 611? What I do remember was that there was one last bit of hum that I couldn't get rid of until I ran the earth by a devious route round the 611. Since then I've gone on to help write a European EMC standard, and sit on several international EMC committees, and I still wouldn't be able to explain how that worked! (Although I do know a couple of people with Doctorates in physics who would be able to build a computer simulation to work it out - given enough time.) Sometimes with noise and earthing you do just have to make slightly educated guesses...

Anyway, a crew of us went back to SWF at Baden-Baden (on my 26th birthday), fitted new backplanes, modified all the 611s on site, and it worked perfectly. Wherever you set the studio lighting - about 50% to 75% up had been the worst - there was not a scrap of buzz to be heard. A bodge, but a very effective bodge! A very angry client turned into a very happy client, because we were very open about what we found and they could see how much effort we had put into fixing it. And since we knew 4000s were only going to get bigger and bigger, with more L and eventually U shapes, it made sense to introduce the mod to all desks from then on.

Some long lasting memories...
  • For the rebuild we flew across a team of women from our PCB assembly line, brilliantly skilled (especially the two Sarahs). We did the rebuild while the studios were closed over the summer, they also had studio tours running - which they used to divert to come and see "the English women engineers"!
  • The Germans also used to gather in the break times to watch us buy a carton of milk and put some of it in our tea, they thought we were completely mad.
  • The extraordinary German equivalent of Father's Day when the customer told us to stop trying to solve the problem at lunchtime and come and get very, very, 'merry' at a bar on the top of a hill in the Black Forest.
  • Landing and taking off in the SSL jet at Baden-Baden airfield, which really wasn't big enough - needed a phenomenally sharp bank to avoid overflying the military airfield next door - at least that's what our pilot (who had a wicked sense of humour) told us!
  • The massive Hemholtz coils I built in the middle of the production test facility at Witney while working on the mod, with the test department also thinking I'd gone completely mad...

Of course, the million dollar question is: can pre- and post- mod parts work together? I think - remembering this is a very long time ago now - modded 611s (with a pot on the 11/294 card and resistors by the aux busses) will work in an old desk, but an old 611 won't work in a new desk. But I may be wrong - i could probably work it out if anyone needs to know!

Oh, I also had to fix a huge noise problem with the special 'DIN' plasma bargraphs we designed for SWF. Fortunately I don't remember us selling any more of this version. Plasma bargraphs are horrible, horrible, horrible things. For anyone who doesn't know them, you have to bang the top of the desk regularly to keep them working, and they give very nasty electric shocks.

Cheers,

Andy
« Last Edit: October 03, 2017, 01:33:02 PM by amillar »
co-designer and project manager G series analogue 1987
channel strip designer J series 1992-93
design "caretaker" 4000/6000 1985-93
analogue team leader ARC/Bertha 1988-92

Clip6

  • Sr. Member
  • ****
  • Posts: 284
  • Karma: +8/-1
Re: The 1986 'coat hanger' mod - what and why
« Reply #1 on: October 03, 2017, 04:33:28 PM »
 8)

Matt Sartori

  • Administrator
  • Hero Member
  • *****
  • Posts: 733
  • Karma: +21/-0
    • London
    • Solid State Logic Fanatics Database
Re: The 1986 'coat hanger' mod - what and why
« Reply #2 on: October 03, 2017, 05:25:45 PM »
amazing...
Thank you for the insight!

Matt Sartori.

oyanliz

  • Full Member
  • ***
  • Posts: 197
  • Karma: +1/-0
Re: The 1986 'coat hanger' mod - what and why
« Reply #3 on: October 04, 2017, 09:44:37 AM »
A pleasure to read this post

Best regards.

Oihan

xmax

  • Sr. Member
  • ****
  • Posts: 284
  • Karma: +10/-3
Re: The 1986 'coat hanger' mod - what and why
« Reply #4 on: October 14, 2017, 06:00:53 AM »
By far the best post I have ever read on a forum, thank you. Norman Druce
(The guy everybody told switch mode power supplies would not work on
E/G desks)

https://www.atomicinstrument.com/

Druhms

  • Newbie
  • *
  • Posts: 42
  • Karma: +0/-0
    • Panama City
    • Rockin' Room in P.C. Florida
Re: The 1986 'coat hanger' mod - what and why
« Reply #5 on: October 17, 2017, 09:10:46 AM »
Absolutely cool to hear the stories that have shaped how we get our jobs done!
Thanks so much,
JJ

amillar

  • Full Member
  • ***
  • Posts: 181
  • Karma: +15/-0
    • United Kingdom
Re: The 1986 'coat hanger' mod - what and why
« Reply #6 on: October 22, 2017, 10:28:10 AM »
Awwww...thanks guys  :) :) :)

Do let me know if there's any other history on the (analogue) side of the 4k/6k/9k/ARC you'd like to hear about. (The 8k was designed by the service department so I wasn't involved in that at all.)
co-designer and project manager G series analogue 1987
channel strip designer J series 1992-93
design "caretaker" 4000/6000 1985-93
analogue team leader ARC/Bertha 1988-92

marcmozart

  • Sr. Member
  • ****
  • Posts: 483
  • Karma: +29/-0
    • Frankfurt/Germany
    • Mozart & Friends
Re: The 1986 'coat hanger' mod - what and why
« Reply #7 on: October 22, 2017, 04:09:49 PM »
How about - "ALL OF IT" !!  ;)

Awwww...thanks guys  :) :) :)

Do let me know if there's any other history on the (analogue) side of the 4k/6k/9k/ARC you'd like to hear about. (The 8k was designed by the service department so I wasn't involved in that at all.)
1992 SSL 4048 G-Series
Mix Engineer Blog
http://www.mixedbymarcmozart.com

amillar

  • Full Member
  • ***
  • Posts: 181
  • Karma: +15/-0
    • United Kingdom
Re: The 1986 'coat hanger' mod - what and why
« Reply #8 on: October 22, 2017, 04:36:37 PM »
 :)
co-designer and project manager G series analogue 1987
channel strip designer J series 1992-93
design "caretaker" 4000/6000 1985-93
analogue team leader ARC/Bertha 1988-92

Clip6

  • Sr. Member
  • ****
  • Posts: 284
  • Karma: +8/-1
Re: The 1986 'coat hanger' mod - what and why
« Reply #9 on: October 22, 2017, 05:35:43 PM »
Hey Andy.

Feel free to ramble on about anything. You have any pictures from back then?

Could you talk about the changes from E to G and the design of the 292 eq.

The x4 comment is a now a classic.

Thanks for being here. 8)

amillar

  • Full Member
  • ***
  • Posts: 181
  • Karma: +15/-0
    • United Kingdom
Re: The 1986 'coat hanger' mod - what and why
« Reply #10 on: October 22, 2017, 06:05:30 PM »
No problem...give me a day or two and I'll try to come up with a ramble on a new thread about how Paul Frindle and I came to spend 1987 in a garden shed, things to do with a solder sucker when bored, and why Dire Straits "Brothers in Arms" album sounds really good through a G series EQ...

Sadly, I've ridiculously few photos - just shows how times have changed in the digital age.

Cheers, Andy
co-designer and project manager G series analogue 1987
channel strip designer J series 1992-93
design "caretaker" 4000/6000 1985-93
analogue team leader ARC/Bertha 1988-92

Clip6

  • Sr. Member
  • ****
  • Posts: 284
  • Karma: +8/-1
Re: The 1986 'coat hanger' mod - what and why
« Reply #11 on: November 02, 2017, 03:32:32 AM »
And you were saying........Dire Straits.......we are listening ;D

amillar

  • Full Member
  • ***
  • Posts: 181
  • Karma: +15/-0
    • United Kingdom
Re: The 1986 'coat hanger' mod - what and why
« Reply #12 on: November 03, 2017, 01:10:31 PM »
Sadly had to some of that pesky work stuff :) ... haven't forgotten
co-designer and project manager G series analogue 1987
channel strip designer J series 1992-93
design "caretaker" 4000/6000 1985-93
analogue team leader ARC/Bertha 1988-92

Clip6

  • Sr. Member
  • ****
  • Posts: 284
  • Karma: +8/-1
Re: The 1986 'coat hanger' mod - what and why
« Reply #13 on: May 27, 2018, 10:22:31 PM »
Andy where are you.....come back!!!!

amillar

  • Full Member
  • ***
  • Posts: 181
  • Karma: +15/-0
    • United Kingdom
Re: The 1986 'coat hanger' mod - what and why
« Reply #14 on: June 11, 2018, 03:48:10 PM »
Sorry - the last few months have been ridiculously busy at work, and still are - I will try and come back!!!
co-designer and project manager G series analogue 1987
channel strip designer J series 1992-93
design "caretaker" 4000/6000 1985-93
analogue team leader ARC/Bertha 1988-92