>David Duffy wrote:
>
>>The reason I was going for a high side driver was that this unit is to
>>replace
>>a manual switch firing unit that uses a common ground system on the
>>firing rails (right name for them?). Each rail has 20 outputs on it with
>>a D25
>>connector. I suppose I could use the existing rails but the +/- markings on
>>each channel would be back to front . (Actually red/black spring terminals)
>>Getting 100 channels in a small space is going to be difficult & expensive.
dwayne wrote:
>Makes perfect sense, then.  You have a matrix system and you really don't
>want to have to go into the firing slats and turn all those steering diodes
>around.

Yes, of course, I hadn't thought of the diodes in the rail being wrong way
- yet...

>What exactly is your goal?  I'm assuming that your existing system fired
>only one channel at a time?  How many channels at a time do you want to
>fire?  How are they being fired?  Is it a simple sequencer or do you need
>random access?

The system has 2 groups of 100 outputs. ie. 2-way common-switch and 100
probe firing points - funny how pyro's still like bed-of-nails systems. :-)
The customer asked for a sequencer to start with but maybe with a view to
having it PC controlled at a later stage.

>Blue Smoke out of Calgary does something similar - they have 16 channels on
>each of 16 common lines for a total of 256 channels.  They can fire any or
>all of the 16 channels associated with each common line but can't fire
>channels that span multiple commons in the same instant.  Not much of a
>problem for fireworks shows.
>
>Just how many high side drivers do you need?  20?  How long is your firing
>pulse?  You might consider doing it the 'brute force' method - TIP32
>transistors with small emitter resistors and 2n4403 transistors doing the
>classic 2 transistor current limit.  If you limit your fire pulse duration,
>you don't need to worry about heatsinking the transistors.
>
>You will want to have monitoring to detect a shorted transistor.  This is
>simply having pull down resistors at the rail outputs and ensuring that
>there is no voltage present when the high side driver is supposed to be
>off.  Disable the ground (common) lines if there is a fault.
>
>You really should use relays for the ground switching.  This gives you the
>mechanical air gap that a lot regulations require and if you do wind up
>firing multiple channels simultaneously, they can handle the required
>current.  I'm quite partial to the T90 series from P&B as well as the EPE
>series from NEC.  The EPE relay is really nice - 2 fully independent SPDT
>30 Amp relays in a package about the same size as the T90.  I have welded
>the contacts on those, however, when a screwdriver fell onto the PCB and
>shorted out the current limit stages.  Again, fault detection is a good
>idea.  In my units, a fault at the relay disables the current limit drivers
>and a fault in the current limit drivers disables the relays.

These are all good points to know.

>Your choice of 24V is good.  That is the only regret I have with our system
>- the client insisted that 12V was good enough when we were commissioned to
>design the system some 10 years ago.  Now they use it with 400 foot long
>drop lines (outdoor stadiums) and the low voltage limits the number of
>devices they can fire on each cue.  A future re-design will jump that to
>24V.  I had considered doing a 300V capacitor discharge version at one time
>but never did get around to it.

The same customer has a 12V unit and he complains that it doesn't work well
anymore. He's just gotten used to the 24V unit firing a few in igniters in
series
with no problem. The 12V one can't fire more than a couple in series I think.
Is this a normal thing to do? (string a few igniters in series off 1 channel)
Regards...

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics