Transient Voltage Test Circuit help

[pigman]

Hey guys,

I am looking into designing a simple test circuit for discharging a higher than usual voltage onto the input of my electronics to test their reactions to Transient Voltages.

Does anyone know of a basic circuit they can link me to or have any suggestions?

The devices I'll be testing are for 24V Automotive/Marine use and it is mostly to ensure the TVS diodes are capable of protecting them without damage.

Cheers

 

[ccurtis]

Effective testing of that sort requires high performance testing equipment to accurately simulate the standard models for the applicable power transients. The usual course is to study those models and design the protection accordingly. After that, the design is tested by contracting with special testing labs familiar with that sort of compliance testing.

However, engineering tests are often performed to avoid multiple trips to the testing labs. And, of course, for hobby circuits, testing labs are impractical.

The following link gives a good idea of the kinds of transients that can be encountered and offers a simple engineering test circuit for the load dump transient.

https://www.electro-tech-online.com/custompdfs/2012/07/an9312.pdf

Here is a link to a simple circuit idea that, although intended to apply a controllable transient load to a power supply output, can be tailored to deliver a controllable voltage transient by switching the power supply in and out across your load through the FET.

**broken link removed**

 

[ronv]

Simulating it might be the easiest. The actual test equipment would be a bear.

 

[canadaelk]

OK, 24V. What amperage? How about a single-line diagram to help you out? E

 

[ChrisP58]

You need to test on the assumption that the source current is unlimited. The actual current should only be limited by the total loop impedance of the device under test and the wiring.

 

[pigman]

hey guys, I'm not sure about the current requirement, I guess it's more about testing the response to voltage spikes from archs and load dumps. I suppose the current requirement will adjust with the ability to change the capacitors to increase or decrease the capacitance, start out low and then raise it slowely to see how the TVS diodes respond and ensure they protect the internal electronics properly.

ccurtis linked the https://www.electro-tech-online.com/custompdfs/2012/07/an9312.pdf has a basic load dump circuit that would probably suit the job, I guess if you guys think this is suitable I'd just need some help with a charge circuit?

Thanks guys

 

[ronv]

One key point. What are you trying to protect? It will make a big difference for example if you are trying to build one device to protect a boat load of equipment scattered about or if you are trying to protect a single device. A little better spec. is in order.

 

[pigman]

Single device, I am just looking for a simple process to put the protection in my little projects through its paces. The project in question here is a linear actuator control device and internally it all runs at 12V via a SMPS so that it auto handles anything between 12-24V however I've had issues with failing input FET and whatnot when the project is taken off the test bench and put into a 24V environment. The only difference from my testing is these Transient Voltages that I can't account for (or didn't know I had to until now). I've implemented a TVS setup to protect with 2 x 1500W 36V TVS diodes and a few input components upgraded but I really want to have a simple charge and release device of some sort so I can hook up my DSO and take some readings on the inside to see whats really happening when it takes a hit from an arch or load dump or similar.

 

[ronv]

What is the maximum input voltage for the problem supply.
What is the maximum current.

 

[pigman]

The SMPS chip is 36V FET is now 80V rated along with the 200u cap on the input of the fet. Current Max of the SMPS is 12A. I used two 36V TVS's next to each other so as to help it clamp the voltage a bit lower / quicker. I may need to use lower voltage TVS diodes but it is getting a bit close to the peak charge voltages of a 24V system and I don't want to cause any serious current leakage through the TVS's

 

[ronv]

Might also want to think about protecting from negitive "spikes" with either a series diode or a clamp diode to ground. Since your problem is probably not load dump (unless there is a bad connection somewhere) it is probably some inductor switching. Maybe a ceramic cap across the big electrolytic for the high frequency stuff.

 

[pigman]

I have a 100n accross the input and the TVS's are bidirectional if that helps

 

[ronv]

I think I would clamp it with a schokky diode so the +24 couldn't go more than .6 negitive with respect to your circuit ground.

 

[pigman]

There is a Shottky on the + for reverse polarity with low dropout, is that what you mean?

 

[ronv]

That should work. Just something so the input can't be driven negitive.

 

[pigman]

kool kool, any ideas on a simple charge circuit to release a test voltage into it?

 

[ronv]

That's a tough one. Maybe if you have a big DC motor you could release it's energy into the protection circuit. The thing is it takes a lot of energy just to charge the cap up to the clamp votage.
Anyway. Maybe something along these lines.
How about the normal environment? Is the noise there all the time?

 

[pigman]

hmmm, interesting, what voltage does that jump to at the peak? I am not sure I completely undersand whats happening here, could you please explain the design a little?
The normal environment doesn't provide consistent results for testing. I can't seem to get it to replicate the issue until some random period..

 

[ronv]

Initially M1 is on building up current in L1. When M1 shuts off the inductive kick from the inductor will go positive. It is limited by the series resistance of the inductor and the time constant with C1 and C2 so they will charge to above 24 volts. The path back to the battery is blocked by D5 so your protection circuit is the only path to short it out. If the voltage rises high enough (Big enough inductor) the voltage will rise high enough to turn on the TVS and short it out. It's really kind of kludgy.
Let me think if there is a better way.

 

[pigman]

interesting mate! my thought process was along the lines of a cap charge circuit from a camera or something (not that I've ever built one) to charge a cap and then a button or fet to release the cap...

I had even thought if it was easy enough I may turn it into a pic controlled unit with an LCD so I could select the charge voltage and it would monitor the charge. But that is me getting ahead of myself. Really I just need a basic test/charge circuit of some sort that I can suss out for testing.