Studio switching solution?

[mhennessy]

I was reminded of this thread at work earlier today, and realised that we haven't heard a peep out of Andy since he started the thread. I do hope this won't be one of those threads where we never get answers to all the questions it raised. And I do hope Andy is OK.

red16V, thank you for chiming in - I was beginning to think that I was some sort of alien species . Surely broadcasting isn't the only engineering discipline where reliability and safety are crucial and logic and common sense must be used (most of the time!)?

If this is to switch off the power to studio loudspeakers, then I covered that in an earlier post: the audio feed to these would be switched via an SLS (studio loudspeaker) relay if that functionality wasn't already provided in the desk. If you interrupt the power, you have to hope they do that cleanly (no pops and clips) and quickly (as some speakers continue to play for a few seconds after the mains has been pulled, and some take a few seconds to come on) - and also hope that they'd be reliable when used this way. I wouldn't assume that any of these points would be true - and what if the speakers get swapped in the future?

Incidentally, if the existing interface box is the same as the Canford Audio unit I linked to earlier, then the GPIO wiring is via an XLR cable. This squarely addresses the earlier points about modifying the existing installation, as no mods are needed:

To add a second one of these Canford Audio boxes, you simply make up a short XLR cable with 2 male plugs and 1 female plug. Once on site, you unplug the existing XLR from the box and join it to the female on your new lead. Then you plug the male leads into the existing Canford box and a new Canford box. Job done in easily less than 30 seconds - much less than the playing time of a bit of music (or even a trail or jingle if you're feeling lucky!). And 100% reversible in the same time. And by appropriate wiring of the XLR adaptor lead, you can make the new Canford box have whatever logic you like (on when mic open or on when mic closed) - such is the magic of GPIOs

Naturally, if the funds aren't available for another Canford box, then the same point applies - an interface box with a GPI is much easier to make than all the left-field ideas put forward so far.

Even if the existing interface box isn't a Canford one, it will have a plug and socket arrangement. Worst-case, it'll be screw terminals - but those are usually done on a Weidmuller block or similar, so will unplug - and a 3-way lead could still be pre-made as described above.

Andy, I still have a broken Canford box on my bench - if it helps, I'd be willing to reverse-engineer the low-voltage GPI side of it, so that you could base a DIY box on it?

Cheers,

Mark

[mhennessy]

OK; curiosity got the better of me...

Here's the schematic. E&OE, naturally. Unmarked diodes are 1N4002, and I haven't yet determined what the zener voltage is.

The fault with this one was the mains transformer (O/C primary). This would have continued to work if it was being supplied with a DC voltage externally, but as the installation it came from made use of the internal power supply, it stopped lighting the lamp. FWIW, this one was blue rather than red, for rehearsal...

The circuit is pretty simple. The name of the game is obviously to light the LED in the zero-crossing detection opto-coupler. An external voltage between pins 2 and 3 (the standard "hot" and "cold" in a 3 pin XLR) will do that; alternatively, shorting pins 1 and 3 has the same effect because of the internal PSU.

The mk2 version of this presumably uses a different opto-coupler, as it only requires 1mA to work. This one will pass nearly 20mA via the external contacts.

Note also the snubber across the triac, which causes LED bulbs to flicker or light when they should be off (and might interfere with attempts to sense the field). The mk2 has a link you can cut to remove that effect, which presumably removes the snubber from the circuit.

Looking closely at the diagram, I'm not sure if you could stack these in series to get the reverse logic that I think is required - I'd have to think harder about that one, but given the unexpectedly high operating current, that plan might not work. Parallel is fine, naturally. The usual way around that is to install a relay in a DIN-rail socket, then you've got lots of contacts to make use of. But obviously, until we hear back from the OP about all this, I'm speculating wildly. But hopefully it's of interest to someone out there

[GrimJosef]

Originally Posted by mhennessy

... Surely broadcasting isn't the only engineering discipline where reliability and safety are crucial and logic and common sense must be used ...

I'm going to stick my neck out here and suggest that broadcasting isn't even top of the list as far as engineering reliability and safety requirements go. I once had an external safety role on a project inside AWE Aldermaston. They worried quite a bit about safety and reliability (you'll be pleased to know if you live in the south of England). And I imagine that the people who make autoland systems for passenger aircraft and the monitoring kit used in neonatal units also pay those subjects at least a passing regard .

Isn't the application of logic pretty much written into the definition of engineering ? If it's not then IMHO it should be. Common sense, though, is trickier. It's surprising how often the phrase is used precisely when there is no agreed common position. If we're not careful it can then become a euphemism for "the way we happen to do things here".

I'm curious too about how Andy sorted his problem out. Perhaps he just looked at the suggestions that appeared in the first 24 hours or so, implemented the easiest one and moved onto the next job ?

Cheers,

GJ

[mhennessy]

Originally Posted by GrimJosef

I'm going to stick my neck out here and suggest that broadcasting isn't even top of the list as far as engineering reliability and safety requirements go.

I'd be really quite worried if broadcasting was even in the top 10, frankly. Hence the comment

But when all's said and done, and even with a good week in-between the start of this and now, I still can't get my head around the "logic" of proposing to detect the presence of an electric field or a current - or indeed the presence of light - in this scenario where the proper (as in easiest and most reliable by a country mile, not to mention industry-default, so suitable hardware readily available, but even if you're building something, it's still the easiest) solution obviously exists in whatever desk is presently in use. I'm all for having a bit of fun, and was actually quite amused by the photocell suggestion, but was disappointed that no-one suggested using a Raspberry Pi with the camera module to watch the position of the faders. It would have been no less crazy

I really don't have a problem with questioning "the way we do things" because often that's how things move forward. A large part of my job is dealing with fresh recruits who question almost everything - just like I did when I joined all those years ago - and it's genuinely good to be challenged. And occasionally frustrating on the odd occasions where I have to concede that the answer is basically tradition...

But GPIOs have been the way we happen to do things for so very long for very good reason. In the time we've been using GPIOs many other forms of control have been and gone - some finding applications elsewhere, admittedly - but the good old GPIO really can't be beat in terms of simplicity, reliability (and ease of troubleshooting when things do break). Presently GPIO-over-IP is the new kid on the block, and definitely has some merit in applications where the whole of the audio mixer is virtual - running on something roughly the size of a Pi and dealing with the audio sources as AoIP streams, and perhaps not even located in the same part of the country as the studio housing the physical control surface - but for traditional consoles, and many other bits of key equipment in current broadcast infrastructure, traditional GPIOs are still very much still around.

Just by way of example, a computer-based radio playout solution that I cover uses the Quancom USBOptoRel16 to interface to the desk and other infrastructure. This product isn't broadcast-specific; it's a general-purpose industrial GPIO interface, but is well-suited to the task. For the main playout product, all 16 ins and all 16 outs are used - fader-starts are obvious, also they are used to light "ready" lights on the desk when the audio is cued and ready to go. It deals with RDS signalling, it detects when the studio "on air" light is on (so it knows whether we need to pay royalties on the music being played or not because it's a rehearsal or a pre-record, to be reported later at playout). It can even control a matrix for network switching. There's more, but you get the idea... Either way, it's an impressive bit of integration, that would be far more difficult to do any other way - and if you did do it another way, it would be much harder to diagnose when faults occur.

Anyway, I'm waffling on in the hope that it's perhaps entertaining to those who might be interested in how broadcasting works - given that this is a radio forum - and I also hope that the information I've given is of use to someone in the future, even if Andy can't do it this way for whatever reason. Obviously there are many other branches of engineering that also use this technique with equally good results, and certainly we use it in many other areas besides radio studios

ATB,

Mark