Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[TheBoy]

Mate of mine has a project where they need to test if they can get several hundred amps down a bit of coax for a very short period of time, and asked me if I knew of an automotive relay man enough. Starter solenoid was best I could think of, but as built into most starters, makes it a bit cumbersome/bulky.

Initial test is going to be at 12v (hence automotive relays considered), but ultimately will go up into the kV range if tests are positive, and this is when big money will need to be spent. But for now, we're looking for a cheap-as-chips solution to prove the cable can withstand the huge currents required.

[martin42]

Come on think old school,like this
http://www.ebay.co.uk/itm/like/400398269358?limghlpsr=true&hlpv=2&ops=true&viphx=1&hlpht=true&lpid=108&device=t&adtype=pla&crdt=0&ff3=1&ff11=ICEP3.0.0&ff12=67&ff13=80&ff14=108&ff19=0

[TheBoy]

Come on think old school,like this
http://www.ebay.co.uk/itm/like/400398269358?limghlpsr=true&hlpv=2&ops=true&viphx=1&hlpht=true&lpid=108&device=t&adtype=pla&crdt=0&ff3=1&ff11=ICEP3.0.0&ff12=67&ff13=80&ff14=108&ff19=0

We had, but thought that those crappy old solenoids, which didn't hold up too well at the time, only turned over poxy little engines, thus low rated. Do you know what they are rated at? Looking in region of 500-900A

[hoofing it]

The biggest relay I have found is 100amp which I used in a home made split charge system on my camper van. 

[Marks DTM Calib]

Do they know what they are doing?

A relay or contactor will not give you a controlled pulse period, the currents they are talking will only give at a guess a matter of ms before the core is fried.

It needs a silicon device such as an IGBT

[TheBoy]

Do they know what they are doing?

A relay or contactor will not give you a controlled pulse period, the currents they are talking will only give at a guess a matter of ms before the core is fried.

It needs a silicon device such as an IGBT

Mostly, yes. There is a concern about timing accuracy of relays or electro-mechanical switching, but at this stage, low cost experimentation if needed to see if the coax will handle that sort of current, before proper money is spent. 900A relays are readily available, but are approx. £2k a pop, too expensive for an early trial.

As I understand it, it needs a 20mS test at 500-900A for the first stage of the trial.

[Andy H]

Specialist equipment is available for testing switchgear.

http://www.trtest.com/page.php?cid=21

http://www.megger.com/uk/Products/ProductDetails.php?ID=1563&Description=ingvar

High currents can cause things to go pop in a quite exciting manner. Add in high voltages and it gets properly dangerous.

Kit like this would usually be hired when required.

[Kevin Wood]

I doubt you'll find anything mechanical that can provide a pulse that fast at that current, and I'd be very worried about the contacts welding, especially if your current source can deliver significant continuous power (i.e. you're not just discharging a big cap). Solid state will be the way to go, I'm sure.

You might want to google "can crusher" designs - e.g.:http://www.powerlabs.org/pssecc.htm

[Kevin Wood]

.. and if it only needs to conduct for 20ms with a low repetition rate, some fairly modest semiconductor devices might well do the job provided they are switched hard enough to avoid too much dissipation.

[TheBoy]

.. and if it only needs to conduct for 20ms with a low repetition rate, some fairly modest semiconductor devices might well do the job provided they are switched hard enough to avoid too much dissipation.

As I understand it, this test doesn't involve any duty cycle, its a one off test to prove the cable doesn't fail, as the cable manufacturer (understandably) has no specifications related to this type of use.

[Rods2]

Sounds like an ideal test to do in your new garage. What could possibly go wrong? 

[chrisgixer]

Sounds like an ideal test to do in your new garage. What could possibly go wrong? 

It's ok, he's borrowed a pressure washer, and presumably a hose, so he won't need the fire brigade this time at least. Smart puppy see.

[Andy H]

.. and if it only needs to conduct for 20ms with a low repetition rate, some fairly modest semiconductor devices might well do the job provided they are switched hard enough to avoid too much dissipation.

As I understand it, this test doesn't involve any duty cycle, its a one off test to prove the cable doesn't fail, as the cable manufacturer (understandably) has no specifications related to this type of use.

It will fail. Why would anyone expect otherwise 

[TheBoy]

.. and if it only needs to conduct for 20ms with a low repetition rate, some fairly modest semiconductor devices might well do the job provided they are switched hard enough to avoid too much dissipation.

As I understand it, this test doesn't involve any duty cycle, its a one off test to prove the cable doesn't fail, as the cable manufacturer (understandably) has no specifications related to this type of use.

It will fail. Why would anyone expect otherwise 

It is believed it won't...

[Kevin Wood]

.. and if it only needs to conduct for 20ms with a low repetition rate, some fairly modest semiconductor devices might well do the job provided they are switched hard enough to avoid too much dissipation.

As I understand it, this test doesn't involve any duty cycle, its a one off test to prove the cable doesn't fail, as the cable manufacturer (understandably) has no specifications related to this type of use.

Bank of big capacitors that you charge to the required pulse energy, then you might well get away with using a relay / starter solenoid to discharge them into the load? You'll need some way of capturing the voltage and current waveforms to verify what actually went into the load.

Or.. for a one-off, find a test house who have a facility for this type of testing and pay for the test instead of the hardware?

Not sure where you'd go, but there'll be someone who can do it, I'm sure.