Hi guys,
My problem is somewhat out of the audio line, but the basic question is still solid state in nature.
Im working here in Kuwait as a contractor, fixing Army electronics.
One of the gadgits is a power controler for 18k BTU air conditioners. It takes in any kind of 220 and outputs three phase 220 ac at 61hz. (don't ask me about the 61 hz I just don't know)
Anyhow Ive got a problem that's been beating me up.
Ive posted this to the AMRAD TACO's (www.AMRAD.org) list and waiting for a reply. Below are clips of the note I sent them.
....................................................
Hi all,
Got another cute one.
One of the tricks to getting these Army motor controlers to work is to match the output darlingtons as a set.
There are three dual darlingtons that make up the output.
The p/n is
Fuji EKV71-050
or
PRX KT524575 S19-A5 (the second gp might be a date code)
The tag goes on to read 75A at 500 V.
Right now we 'do' this matching with a Fluke mulit meter in diode test mode.
The bear is these are not just darlingtons, there are some un detailed resistors and diodes, I can send a photo if anybody asks.
Was thinking of doing a better matching job on our new and growing collection of used working transistors with a curve tracer.
I have nill experience with these tools, but know of some sharp techs who use them with good success. Guess it's time to learn.
Did a web search and found two 'curve tracer kits'.
The first is the gizmo that most of us make with a filliment transformer and some resistors. Ive got a Huntron here, but can't seem to do more than the basics.
This link will take you to a guy offering a fairly expensive kit. ($600 USD)
http://www.fullnet.com/u/tomg/gooteect.htm
Anybody have this kit?
Know the reputation of the company?
Think it's worth the effort?
Ideas and coments welcome.
In other news....
Been fighting with three units on a vexing problem. They give me 'over current' shutdowns after a 40 min to one hour run. Always when closed up and hot. On one unit there was some sensivitity (it would shut down) when I touched the 'ground buss' with my finger.
I got desperate and made a few changes.
Will let you know if the first test article runs for four hours w/o failure. It's 17:10 and it's been cranking since 14:20. If it's still working by 18:20 I will have beaten this trouble.
The fix required that I cheat.
Seems the case ground buss for this digital board comes from one point on the analog card that feeds it. By the time that ground get's to the far side of the digital board it's traveled almost 20 inches.
When I put the scope from the logic board ground to case ground, I got a fair amount of 'hash' for a reading.
I measued the DC difference from the case to a ground point at the far end of the digital card and it came out to about 77mV. When I measured the current, I got a reading of 12.8 mA.
So number one (motor controler) is cranking, number two is in the 'on deck circle', and number three is under the bench waiting for the first to pass, and the second to be put on the motor tester.
These are old units (made in 93) and oxygen has taken it's toll.
Im hoping that this beats this problem down.
I don't want to look at these anymore. I want them gone.
Later guys.
.....................................................
For what it's worth, the 'cheat' worked, and my boss approved. So with a random sample of 'one' working, Im going to test the next two directly.
So any help on this curve tracer kit, or if one of you know of another offering, please let me know.
Many thanks
Jack Crow in Kuwait
My problem is somewhat out of the audio line, but the basic question is still solid state in nature.
Im working here in Kuwait as a contractor, fixing Army electronics.
One of the gadgits is a power controler for 18k BTU air conditioners. It takes in any kind of 220 and outputs three phase 220 ac at 61hz. (don't ask me about the 61 hz I just don't know)
Anyhow Ive got a problem that's been beating me up.
Ive posted this to the AMRAD TACO's (www.AMRAD.org) list and waiting for a reply. Below are clips of the note I sent them.
....................................................
Hi all,
Got another cute one.
One of the tricks to getting these Army motor controlers to work is to match the output darlingtons as a set.
There are three dual darlingtons that make up the output.
The p/n is
Fuji EKV71-050
or
PRX KT524575 S19-A5 (the second gp might be a date code)
The tag goes on to read 75A at 500 V.
Right now we 'do' this matching with a Fluke mulit meter in diode test mode.
The bear is these are not just darlingtons, there are some un detailed resistors and diodes, I can send a photo if anybody asks.
Was thinking of doing a better matching job on our new and growing collection of used working transistors with a curve tracer.
I have nill experience with these tools, but know of some sharp techs who use them with good success. Guess it's time to learn.
Did a web search and found two 'curve tracer kits'.
The first is the gizmo that most of us make with a filliment transformer and some resistors. Ive got a Huntron here, but can't seem to do more than the basics.
This link will take you to a guy offering a fairly expensive kit. ($600 USD)
http://www.fullnet.com/u/tomg/gooteect.htm
Anybody have this kit?
Know the reputation of the company?
Think it's worth the effort?
Ideas and coments welcome.
In other news....
Been fighting with three units on a vexing problem. They give me 'over current' shutdowns after a 40 min to one hour run. Always when closed up and hot. On one unit there was some sensivitity (it would shut down) when I touched the 'ground buss' with my finger.
I got desperate and made a few changes.
Will let you know if the first test article runs for four hours w/o failure. It's 17:10 and it's been cranking since 14:20. If it's still working by 18:20 I will have beaten this trouble.
The fix required that I cheat.
Seems the case ground buss for this digital board comes from one point on the analog card that feeds it. By the time that ground get's to the far side of the digital board it's traveled almost 20 inches.
When I put the scope from the logic board ground to case ground, I got a fair amount of 'hash' for a reading.
I measued the DC difference from the case to a ground point at the far end of the digital card and it came out to about 77mV. When I measured the current, I got a reading of 12.8 mA.
So number one (motor controler) is cranking, number two is in the 'on deck circle', and number three is under the bench waiting for the first to pass, and the second to be put on the motor tester.
These are old units (made in 93) and oxygen has taken it's toll.
Im hoping that this beats this problem down.
I don't want to look at these anymore. I want them gone.
Later guys.
.....................................................
For what it's worth, the 'cheat' worked, and my boss approved. So with a random sample of 'one' working, Im going to test the next two directly.
So any help on this curve tracer kit, or if one of you know of another offering, please let me know.
Many thanks
Jack Crow in Kuwait
Hello,
I have stacks of these if you need them
Fuji EKV71-050
PRX KT524575 S19-A5 - Sounds like a Powerex device.
I have got equivalent charts at work and it would be easy for me to find out.
Regarding the curve tracer, I think it's probably best to shoot Enno Borberly an email. He might be able to suggest cheaper alternatives. http://www.borbelyaudio.com
Or we could could do it at my company.
Recently seen an 80KW digital motor spindle drive come in for repair. These units have regen, so when the drive stops powering the motor, all the current goes back into the drive. Talk about back EMF!. Some power industrial electronics around could make Krell look like complete and utter amatuers when it comes to power!
Kevin
PS. I did once have the wacky idea of using old very high power mosfets for 3 phase equipment (IGBTs are used nowadays) and making a 3 phase "God of Zen"
Christ, if I had the design abilities of someone like JLH, Marcel van Gravel, jcx and others, gods knows what I would come up with. Imagination and creativity matter as much as designing.
I have stacks of these if you need them
Fuji EKV71-050
PRX KT524575 S19-A5 - Sounds like a Powerex device.
I have got equivalent charts at work and it would be easy for me to find out.
Regarding the curve tracer, I think it's probably best to shoot Enno Borberly an email. He might be able to suggest cheaper alternatives. http://www.borbelyaudio.com
Or we could could do it at my company.
Recently seen an 80KW digital motor spindle drive come in for repair. These units have regen, so when the drive stops powering the motor, all the current goes back into the drive. Talk about back EMF!. Some power industrial electronics around could make Krell look like complete and utter amatuers when it comes to power!
Kevin
PS. I did once have the wacky idea of using old very high power mosfets for 3 phase equipment (IGBTs are used nowadays) and making a 3 phase "God of Zen"
Christ, if I had the design abilities of someone like JLH, Marcel van Gravel, jcx and others, gods knows what I would come up with. Imagination and creativity matter as much as designing.
Kevinbd said:
Erno's article in Audio Amateur is, by now, a bit dated (but it will certainly work!).
If you have to do it in a pinch -- EDN just had an article a few weeks ago on an NPN transistor matcher -- but this was for low power devices.
Maxim (www.maxim-ic.com) has an I vs V curve tracer which runs off the parallel port -- it's on their website -- http://www.maxim-ic.com/appnotes.cfm/appnote_number/253
Dr. Steber wrote an article in Circuit Cellar a few years ago regarding use of the sound card as an I vs V tracer and semiconductor component testere -- I will privately mail you his email address (fwiw, he also wrote a great article in QST and QEX on measuring impedance with your sound card.) If I can find the Circuit Cellar article I will send you a PDF.
Unless you like programming in Visual Basic, I would suggest that the USB data-acquisition cards would be a bit of a stretch.
I would really like to buy a curve tracer for output type devices. Something that can handle your typical 240/9240 mosfets or bipolar equivilents. basic power amp stuff.
It would be nice if it could also do the lower current items for matching input stages.
There are a number of used tracers on the market, but i have no idea what specs are relevent when comparing models.
Can someone suggest a model to purchase??? When i get around to buying this, i imagine i could spend up to $1000.00 or so...
It would be nice if it could also do the lower current items for matching input stages.
There are a number of used tracers on the market, but i have no idea what specs are relevent when comparing models.
Can someone suggest a model to purchase??? When i get around to buying this, i imagine i could spend up to $1000.00 or so...
Hi
Go for a Tektronix 576 these are one of the best curve tracers ever made. They will test upto 200 W and voltages upto 1500 v.I have one at home they are very handy.Here's a link to a picture of one.
http://www.fabsurplus.com/equip_owned/5809.html
Regards
Anthony
Go for a Tektronix 576 these are one of the best curve tracers ever made. They will test upto 200 W and voltages upto 1500 v.I have one at home they are very handy.Here's a link to a picture of one.
http://www.fabsurplus.com/equip_owned/5809.html
Regards
Anthony
Jackinnj your answer here.........ctxt
J,
Now that looks promising. It's something I can build with what's on hand here in Kuwait and some mail order parts from CONUS.
The kicker is test voltage. These transistors are rated for 500v, and are run in the 350 range.
Printed the Maxim article and will be reading it on my day off Saturday.
Thanks again
Jack Crow in Kuwait
J,
Now that looks promising. It's something I can build with what's on hand here in Kuwait and some mail order parts from CONUS.
The kicker is test voltage. These transistors are rated for 500v, and are run in the 350 range.
Printed the Maxim article and will be reading it on my day off Saturday.
Thanks again
Jack Crow in Kuwait
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.