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.

Help with dynamic test load

Status
Not open for further replies.
Meaning that with the linear supplies I was loading before, I couldn't see a wave on the scope without the modulating square wave. With the switcher loaded up, I have a wave with no square wave modulation.
 
there was a waveform with the switcher across the sense resistor
I was confused by your dangling participle. :D
I suspect it is a ground loop of some sort. probably not a problem.
 
You do know that it is the nature of switching regulators to have ripple on their outputs, don't you?
 
That's what it is. I was quickly changing the timebase to get a display and read it wrong. The ripple is at 15kHz. Switching frequency.
 
So I'd like to use a single 10 turn pot for adjustment. Slim pickins for a 1M precision pot it seems. If I change it back to 100K I need to go with a lower reference voltage right?
 
So I'd like to use a single 10 turn pot for adjustment. Slim pickins for a 1M precision pot it seems. If I change it back to 100K I need to go with a lower reference voltage right?
Sorry, but I don't remember seeing a 1Meg pot in any of your schematics. Can you post what you have now?
 
I changed the 100k pot to 1Meg when I changed the reference method. Here is a current schematic.
 

Attachments

  • New eLoad 2.png
    New eLoad 2.png
    16.6 KB · Views: 160
I changed the 100k pot to 1Meg when I changed the reference method. Here is a current schematic.
I would do it like this. You can scale the pots up by as much as 20 times (10k each), or anywhere in between, so long as you scale R2 proportionally. This way, you can use any value that you can find.
The full scale trim is a set and forget. One turn should be sufficient.
 

Attachments

  • New eLoad 2_ron.PNG
    New eLoad 2_ron.PNG
    23.3 KB · Views: 133
Oh cool, I'll do that. R6,8 and C2,3 are for stability (at least that's what I was told in one of the articles I read). Are they not needed?
 
Oh, ok. I missed the part about C4,R9 being replacements for those compnents.
I can't guarantee that my solution is better than what you were using, because (1) I don't have a model of your power module, and (2) I wouldn't trust it if I did.:(
 
Hey Roff, I can't imagine they hurt but if I don't need them best to keep it simple. Here's a datasheet for the power module.
 

Attachments

  • KS621K30.pdf
    597.5 KB · Views: 190
Well, damn. I made the changes and ran it at 10A for a while. No problems. Ramped it up to 20A, no prob. Ran it at 27A and the module shorted again. Did find out that the switcher will do 60A for about 2 sec before protection kicks in. Impressive.
Unfortunately I did not get a picture on the scope before it went.
Here are meter readings from the dead module:

E-C 4.3
B-E 43.3
B-C 39.3
Bx-E 4.3
Bx-C .4
B-Bx 39.3
What's excellent is nothing is acting like a diode now. Swapping leads gave the same readings.

Here's a good one for comparison.
E-C FWD 382K REV 8.3K
B-E FWD 97 REV 97
B-C FWD 8.35K REV 358K
Bx-E FWD 54 REV 54
Bx-C FWD 8.23K REV 336K
B-Bx FWD 42.6 REV 42.6
 
Well, damn. I made the changes and ran it at 10A for a while. No problems. Ramped it up to 20A, no prob. Ran it at 27A and the module shorted again. Did find out that the switcher will do 60A for about 2 sec before protection kicks in. Impressive.
Unfortunately I did not get a picture on the scope before it went.
Here are meter readings from the dead module:

E-C 4.3
B-E 43.3
B-C 39.3
Bx-E 4.3
Bx-C .4
B-Bx 39.3
What's excellent is nothing is acting like a diode now. Swapping leads gave the same readings.

Here's a good one for comparison.
E-C FWD 382K REV 8.3K
B-E FWD 97 REV 97
B-C FWD 8.35K REV 358K
Bx-E FWD 54 REV 54
Bx-C FWD 8.23K REV 336K
B-Bx FWD 42.6 REV 42.6

Interesting thread. Seems you must be somehow exceeding the safe operating area of the final power devices. Is your heat sink coupled properly with the devices, using heat sink compound?

Back in the 80s I was often repairing large lambda linear power supplies used in large process control systems, say 30amp. these had to work 24/7/365 and didn't always get installed in the best locations for heat management. Anyway we found that we really needed to load test them for several hours to see if repairs or other weak components might surface, hence the need for an adjustable load tester. They didn't want to spring for the real expensive ready made devices so we had to improvise.

The electric shop had these ancient load banks that used stacks of 2" square carbon plates, the holder could fit as many as 30 plates as I recall. There was a large compression knob and you squeezed down to get the load current you wanted. Trouble was it wasn't a very precise adjustment and hard to meet and hold a specific value. After a lot of experimentation I settled on just long lengths of 10 and 12 gauge wire that would be cut to specific load values. The insulation would get a little warm but not smoke ;) It would change resistance a little with heat but was close enough for our testing requirements and was easy to store away when we were done. ;)

Good luck on your design

Lefty
 
Last edited:
Hi Leftyretro, that's pretty cool. The setup sounds like the old carbon pile voltage regulators. The test load I have now is resistive using a dozen of those little 600W block resistors.

I wondering the same thing about SOA for these. The device and heatsink are not getting too hot, I can keep my hand on them indefinitely. Maybe the junction temp is somehow exceeding its max. Is there something about these devices that makes them unsuitable for this use?
 
Not that long. I had switched to the .01 sense resistor for more current and ran it for less than a minute before I heard the snap crackle pop sounds from the module. I can go ahead and put another module in and try to get a scope picture next time, but I sure would like to figure it out without toasting any more. Would putting a breaker in for protection help or would the damage be done by the time it kicked?
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top