BBC Repeater Unit?

[knobtwiddler]

Hi,

Please see photo attached. Has anyone here come across one of these before? I believe it to be a repeater unit (it was bought as a mystery unit, and I planned to put the transformers into an EQ that needs some), with a PPM that keeps an eye on things. The curious thing for me is that the transformers do not have any buffering, which I would expect to generate insertion loss... So maybe not the world's most efficient repeater, unless I'm missing something?

It has a couple of Ernest Turner PPM meters, with a typical driver PCB. I believe the unit must have been built for the BBC. The transformers have BBC part numbers (starting with the familiar LL prefix - LL106SA + LL63RB). They are detailed in this excellent Beeb document: https://www.bbceng.info/additions/20...Tx_ocr_sml.pdf (page 32)

The unit’s not easy to test, as I/O is via a large D-Sub at the rear. I’ve connected one of the meters to +24V and a stepped generator, and it works perfectly. I haven’t checked the TFs yet.

The meter drivers have their own input transformers, so you have a total of 3 p/channel. I am guessing that the signal being monitored passes through, with the driver card listening in? I haven’t worked out what the pots on the front panel are for…(I connected my osc directly to the pins on the PCB). It also has a latching pushbutton (illuminated) next to each pot.

Is my guess that it's a repeater correct? And why doesn't it have a buffer? To be honest, I only want to find out to satisfy curiosity. I will likely put the TFs into my (rather nice, inductor-based) EQs and maybe use the meters and case for other projects (a repeater isn't something I need to my knowledge!).

The ET meter that I’ve tested is bang on, just like other ET and Sifam meters. I suspect you could mix many up, put them randomly in stereo pairs and the levels / ballistics would match nicely. The standards on PPMs were a lot tighter than what I’ve seen on VU meters!

Thanks!

[knobtwiddler]

A couple of other thoughts: from looking at the looming (haven't had a chance to bleep through it), I am wondering if the switch and pots activate external items, i.e. have no effect on this box. Maybe it's not a 'repeater' unit (which would be odd without a buffer)? Maybe it goes in a rack and interacts with other items? Maybe the transformers work in conjunction with something else that goes to the D-Sub? The drivers have their own input transformers and float from the business signal?

[Radio Wrangler]

Perhaps it's not a repeater of payload signals, but a set of repeater PPM displays?
getting fed a replica signal from some splitter with the gain to make up for its insertion loss and also preventing any fault on the remote PPM feeds dragging down the payload path.

David

[Ted Kendall]

Telephone hybrid, perhaps?

[mhennessy]

Only one picture, but from what I can see it looks very home-made to me. Quite apart from the PSU, that rack case was sold by Canford - popular within the BBC because it was/is very easy to buy from Canford. I remember building things in those cases, and being frustrated with the thin steel walls that made it impossible to properly countersink screws. The aluminium sub-plate was better than nothing, but still, it wasn't a nice case to work with.

As to what the intended purpose was, it's impossible to guess. I don't know if much can be inferred from the giant line transformers, given that the D-type connector will undo the isolation voltage benefits of them. Often, these things are made with whatever is lying about the workshop.

The BBC was held together by these sorts of one-of, bespoke units, made for a very specific installation to solve a very specific problem, built by a local engineer in their spare time. Often just small die-cast boxes hanging in backs of bays, completely undocumented. I've done it myself. Up in the attic I've got exactly that Canford rack case housing a load of CMOS logic and audio/video switching relays, all connected to a remote panel with illuminated push-buttons. That was the controller for an edit suite that was built in 1996 - there was no commercial unit available to do it with the required options (it was a 3 machine edit suite that was also capable of being used for playout should there be a problem in the main playout area). All designed from scratch, based on what the engineer leading the project wanted. Still don't have the heart to throw it away, despite the fact it's utterly useless today!

The PPM driver cards (and PPM specs) are covered here: http://www.bbceng.info/ti/eqpt/ME12_10.pdf

[knobtwiddler]

Maybe it is. Thanks.

There are a few curious aspects to it, not only the potential for insertion loss. Why pots? Were it to adhere to some sort of standard, you'd imagine a switch, with set gains? And why are the pots (as I suspect) part of a separate circuit, that goes to the d-sub and nothing in the same enclosure? What do the pushbuttons do (again, I suspect they have no effect on anything in enclosure, and actuate via the d-sub)?

Someone in the Beeb's tech dept made the PSU on Vero card. Was that normal practice there? It's very neatly made, as is the tidy looming (I will definitely reuse it). I feel a bit bad, pulling it apart for its parts, as someone obviously put graft into making it, but the parts will be useful (especially the TFs, for my EQs), whereas I am not planning on making any outside broadcasts and therefore have no use for a repeater! I consider myself quite lucky to have found something with 1:1 transformers, as I bought it as a gamble, for that reason.

edit - written as Mark was writing, will reply separately.

[knobtwiddler]

it wasn't a nice case to work with.

Indeed. I've never liked them (I bet you Canford still sell them!). Last time I bought was was around 20-25 years ago. I sent it to a local machinist, for the purpose of making it more robust and up to duty on the road. The bill from the machinist ended up at around 3x what the case cost.

looks very home-made to me

If it is, then I believe the person who made it was doing that sort of thing for a living. I don't think it was at this point, and am more inclined to think it was made by the eng dept. of a broadcaster. The only real 'home made' aspect to it is the PSU. The neatness of the looming and machining on the case suggests a pro was involved (but I may be wrong). Although on Vero, the PSU is rugged and the card is braced well - is likely as tough as a commercial PSU.

If all signal pins on the d-sub are transformer isolated, the transformers shielded, and the power supplies also floating, then would the isolation benefits of the transformers be impaired using a d-sub? Maybe I'm missing something? Thanks

[paulsherwin]

Mark is quite correct about BBC engineers knocking up things like this on an ad hoc basis using whatever was in stores at the time. It seems to have been part of BBC culture until the Birt era, with even station engineers in local radio being encouraged to build things when they had time on their hands. One engineer I knew in the 70s was building an entire portable stereo desk, despite his station being mono at the time. I've no idea if he ever finished it.

[Ted Kendall]

Originally Posted by mhennessy

Still don't have the heart to throw it away, despite the fact it's utterly useless today!

There's a lot of that about...

[mhennessy]

By "home made", I obviously meant "in house". Almost everything I make for the BBC (and myself!) is on Veroboard, but they are on-offs, so not worth making a PCB for. Veroboard done right can be surprisingly good.

Again, I've only got the one low-res photo here, and you have the unit in front of you, but I don't think the looming was done to a high standard. It looks like it was done by someone who was "enthusiastic" rather than "experienced". I don't say that to be critical - it's just a clumsy way to describe the difference between a broadcast engineer who has a really wide range of disciplines to someone working on a production line in a commercial manufacturing outfit.

Of course I can't evaluate the metalwork, but there's every chance the front panel was sent out to a friendly local company for drilling and engraving - we did that a lot as our in-house metalworking facilities were very limited indeed. Back then that was easy - today, it's much harder because of the centralised finance systems (usual large company red tape).

Those large transformers were designed to terminate PO lines entering/leaving a building. As such, they were able to withstand higher than normal surge voltages, and the electrical isolation between primary and secondary was higher than for a normal "rep coil". I forget the numbers, but they were certainly greater than you'd get from a D type, hence my comment. They would be installed in racks where the PO circuits came in - I've never seen them inside equipment. Actually, that's a lie - we've put a few into die-cast boxes for use as rep-coils, but they're only used at line level on the bench in the audio lab. They double as door stops. We used them because we had them, not because we needed their high voltage isolation properties. It'll be the same story for your unit.

Thinking some more, and in the absence of a front panel photo I'm having to guess here (hint: you're allowed 5 images per post ), but I suspect it's a rudimentary AMU (audio monitoring unit).

Here's my best guess about how it works, based on what I'd do with that kit of parts, so no guarantees about what was actually done: line level balanced comes in, is converted to unbalanced in those bricks so it can be applied to the volume pots, then is converted back to balanced by the more sensible transformers in the middle of the unit so that it can travel onwards to a power amplifier and pair of speakers. Guess they didn't use XLRs because that thin steel is a PITA to drill out (but D-types aren't too bad to punch or file out).

The signal for the AM20/3 PPM driver cards will be picked up prior to the volume control, though I can't say which side of the giant transformers it will be. If it's prior to them, that again shows that safety isolation was not the reason for using those paperweights.

It's possible the switches are to mute or "dim" the audio, but again, I don't have the unit in front of me. It would be a trivial job to reverse-engineer it, so hopefully you'll be willing to share the results with us

[Hartley118]

Just a thought. Is it a bespoke telephone interface unit or ‘telephone balance unit’ designed for interfacing the public phone network to a studio mixing desk for phone-in programmes? Those mighty LL106 transformers could just be isolating the PPMs from the PSTN line DC - not an obvious choice for the job, but maybe what was available from Stores?

Insertion loss would be compensated with gain from the desk.

Martin

[mhennessy]

Anything's possible, but I doubt it in all honesty.

I've never been anywhere where TBUs were in short supply - they're not expensive, and they enjoy a long service life. And of course they are certified for connection to the telephone networks. You'd be taking quite a risk connecting a DIY unit to a telephone line, even if it's on your PABX rather than a DEL.

[knobtwiddler]

This is the extent of my tests so far:

The I/P transformer is wired 1:1 and secondary feeds straight to PPM driver I/P (which has a transformer on its input as well!). A +4dBu signal registers 5 on the PPM on the left channel, as it ought to.

Right channel had a genuine dry joint, where a wire had been curled around the Vero pin on PPM PCB, but no solder attached... Lead was flapping around inside. I imagine that as wire is solid core, it must have worked in the early stages of its life...(it has rack marks).

On right channel, the PPM latches at 5 when power applied. The pushbutton seems to release it sometimes, in keeping with the 'dim' theory.

Switch and pot for left channel (which works as an AMU) do not seem to do anything.

O/p transformers are 1:1 as per datasheet. When injecting a signal directly to pins they work fine, but I cannot discern any connection to the input transformer, either via bleeping pins or using sig gen. The d-sub pins connect to the O/P transformer pri / secs, but I cannot discern any connection between them and any other internal circuit - they seem to be accessed independently via d-sub at rear.

edit - wiper of pot goes to rear d-sub

[dglcomp]

Talking about Canford enclosures, they do still have a metal work place at Southwell Business Park on Portland, think it might have been a business they acquired.