Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

High Energy Load dump/Voltage Clamp

Status
Not open for further replies.

mrmin

New Member
Hi All,

Ok, I'm new here...
I need to sink about 800 joules of energy using a clamp/(maybe crowbar) overvoltage setup to protect FETS. The FETS are 30V(D-S) and I need to clamp anything above 27Volts. I've got my eye on a varistor from littelfuse, but I really think the best device to clamp the pulse (1 second pulse!) is a crowbar using a zener and voltage divider on n/p channel FET to ground..

The math indicates that I need to sink at least 200-300 amps for 1 second and I need to protect the FET from unknown ESD/TVS as well.

But I might go for a hybrid zener/tvs diode/varistor solution...

Anyone done this before?
 
Last edited:
I need to sink at least 200-300 amps for 1 second
So the I-squared-T curve for the device needs to be rated at better than (300^2)*1 = 90,000 amps-squared-seconds.
 
.... Devices / Suggestions ?

I know its a "high energy" setup, but I need to protect some expensive components... and the regulator concept wont work... (long story..here goes..some of it).

The voltage source is a 25V battery capable of 200 amps continous..for a limited amount of time. The load(brushless motor) is driven by a microcontroller(via sensitive FETs) the braking effect requires something to dissipate the power, which uses the battery as the "load" (which struggles to eat this amount of energy). If the braking effect cannot sink the current then the voltage will rise which will eventually destroy the FET's.

Come on guys.. I've looked at all sorts of "load dump" devices and TVs diodes, varistors, etc... I was even looking at a maxim IC to control an external n channel FET.... Anyone done this?
 
Are you trying to "dump" 200A of current @ 27V which is too much for the battery or are you trying to dump 200A into the battery?
 
Well. I dont want to heat the motor windings by using a crowbar-apc ... I need to dissipate it somewhere, so battery is the standard design(to save space) - its kind of like regenerative braking on modern electric cars. The problem is I'm trying to use cheap batteries and they have high internal resistance so they cannot withstand much more than 10-50Amps... So if I can help the "poor" batteries and sink the excess power (of course it would be nice to do something useful with it, but I cannot even store that much energy) ......
 
Fact is: A good battery will temporarily handle this "load dump" event. A bad one(high resistance) will not be able to and goodbye sensitive components.... I have found 20mm 100 joule MOVs which I can parrallel 5 or so up to probably drain this, but I wanted to know what you guys think.. maybe a fet and zener would be more elegant...maybe the switch on times will not be good..currently I'm thinking a hybrid solution is best...
 
a neat trick I have been developing and still am but has showed good promise is to but a large capacity fet or IGBT that dumps though a small resistance directly across the supply lines. Trigger the Fet with a voltage comparator that switches it on above the 27 volts but switches off agian at 27 volts. This works as a simple but so far very good voltage clamper.
Some snubber caps are probibly going to be needed too. both at the motor and the battery end. High current switching could cause random noise in the circuit.

I am developing a solid state control system for running the electromagnets at a friends salvage yard. I have one of his 250 volt 50 amp electromagnets to actualy test with.
when the power shuts off I have to dump 250 volts at 50 amps for up to 4 seconds to someplace or it kicks the volt spike up to the point it will jump over 1 inch gaps and arc out! (how many joules is that?)
Or blows the 1200 volt IGBT's to pieces. which usualy happens first!
I cant just feed it back to the generator.

Like I said this application seems to be doing well. However it will be next summer before the snow melts and I can actualy run it in the yard to do real world application testing.

I could be wrong but would it not maybe cheaper to use higher voltage rated components on the high current side? or a better quality battery? :confused:

I have a lot of friends that always ask me questions about how to get away with using bad or marginal parts that are not up to the task for doing what they need. I typicaly tell them throw it out and use the right part for the application. :p

I had a friend that wanted to set up a portable jump start unit using old riding lawn mower batteries because he could get them for very cheep. He had an older diesel truck with weak batteries. Thats what the jumper unit was to be used for. :p
I told him to just spend the $250 and get new truck batteries. Why lug around 20 crappy lawn mower batteries and cables to do the same job as one good battery in the first place. :confused:

Just my thoughts. :D
 
Last edited:
Hey that sounds neat. The MAXIM ic's have that already built in, the two comparitors and drivers for external power transistors... I think they only have 80v input though... (maybe you can regulator it from 250v ??)...

The 250v x 50amps is 12500 watts total...That is 3125 watts / second... which is 3125 joules for 4 seconds. You have a much worse problem than me :), but we have a similar problem.

Join the club.
 
Actually the 800 joules is partially absorbed by the batteries, but in the case of the cheap batts, the rest destroys the fets.
 
Like I said, still in the testing phase! My dump resistor is a 1 ohm 1 kw wire wound with a 10:1 surge rating.

why are you running 30 volt fets so close to the voltage limit?
I work with many large power switching sytems and the standard design rule is a minimum 2:1 switching device rating over its actual applied application. 27 volts into a 30 volt fet seems like asking for trouble!
There is no head room for error!
Being a brushless motor driver, the fets are most likely set up in an H-bridge configuration. Plus being fets are voltage controled, switching them out for higher voltage rated units should not give too many problems.

I also got to know what uses a 200 amp 27 volt brushless motor? And can I get one too?
 
well unfortunately I'm not into making tiny little boxes and and you probably know these units come with 36 smd fets already mounted... I'm not going to unsolder all this... they may be running on the limit, but you hardly find small fets that will deliver this power without heating up a lot. anyway... I just want to help them along and prevent a lot of smoke and smelly plastic.
 
RC toy?
If it is its a waste of money. just get a bigger controller. The people that design the RC stuff know how to make their products "designed to fail" very well.
Cheap manufaturing cost with high resale!

I used to rebuild RC power boxes for easy cash back in college my first time around. Very easy cash!

The RC nutters were dumber than a box of rocks when it came to what they actualy were buying. All name no product!

I had an easy cash market for fets. I found out what types they used and the companys that build the controllers charged $8 - $10 per fet for rebuild kits.

I bought them from Digikey for about 90 cents a piece in lots of 100. Then Sold them to the RC nutters for $5 each at the RC car tracks! I could pull a good $250 profit for one pocket full at track in a good weekend.

Never owned a RC car either.
Had lots of RC friends and got cheep meals every time I showed up too! I will give you a pair of 90 cent fets for an $8 all you can eat dinner any time.
 
Ya well, at least it got me back into electronics. I've been ordering a lab power supply, microcontrollers and prototype boards again...Hey, it has not changed in almost 15 years. I'm starting to build my own bl controllers and want to see if they work on 1:1 eventually... gotta start small.

Well -- The rc car brushless stuff is coming up just like the hybrid and tesla autos. Kind of a pet hobby of mine. I never actually race it, just experiment on it. These things can burn 2-5 KW of power nowadays.

BTW: Looking for a "ok" scope. Any recommendations? (+-50mhz...dunno my needs yet.)
 
Hi All,

Ok, I'm new here... Dont worry I know a few things. :)
I need to sink about 800 joules of energy using a clamp/(maybe crowbar) overvoltage setup to protect FETS. The FETS are 30V(D-S) and I need to clamp anything above 27Volts. I've got my eye on a varistor from littelfuse, but I really think the best device to clamp the pulse (1 second pulse!) is a crowbar using a zener and voltage divider on n/p channel FET to ground..

The math indicates that I need to sink at least 200-300 amps for 1 second and I need to protect the FET from unknown ESD/TVS as well.

But I might go for a hybrid zener/tvs diode/varistor solution...

Anyone done this before?
You know a few things? Good, start by replacing the FETS that should not be there in the first pace with 100V parts.

Instead of telling us what you think you need tell us what you are trying to do with what hardware.

If you are trying to do regenerative braking, run your 12V motor on a large capacitor with a motor controler that can run on 200V. When the cap is fully charged dump some power into a dissipation resistor.

Dan
 
Last edited:
Hi Dan,

Thanks for the advice, but if you read above then you should see that I dont really want to do that. Its a band-aid solution - I know - but I thought some of you could throw some advice in.... and yes - I should really replace the FETS, but that will mean a 5 kilogram controller.....

Not so small either...

and size, weight and all that counts.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top