Maker Pro
Maker Pro

TO-220 leadbend failure - Update

Y

Yzordderrex

Jan 1, 1970
0
The thread was hijacked and by the time we got to post number 153 it
had morphed into something else which had nothing to do with my post -
Heheh - So I figured to start fresh. I placed OP at the bottom for
reference.


We stuck a couple of controls in the environmental chamber and brought
them down to 20 below. Took them out and walked them downstairs to the
high power lab and plugged in in - Boom. Blew up 2 out of 4. Looks
like the moisture getting into the drive was enough to kill the power
supply. Drive looks clean as a whistle when the moisture gone.

I had mentioned to the group that I had thought contamination was
killing units. One of the other (young & green) engineers would offer
up the fact that units would come back clean in an effort to shoot down
contamination as a cause. So now we will have a meeting on Monday and
I have all the ammo I'm going to need. Clean up the packaging and if
failures continue we will take another look.

Boss is having a leadbend fixture made as well. Thanks for all the
comments that apply I'll sort them out and send them up the chain.

regards,
Bob






I have a small 480v 5w flyback power supply that has had a large number

of field failures. In general the failure rate is about 0.4%, one out
of 250 units. This I can live with. One particular customer in
Northern California has had a 25% failure rate. Most of the units come

back filthy so that there may be an environmental aspect to the problem

as well. I think units are in an air stream which may be blown in from

a rooftop intake. We have beat the crap out of the power supplies here

at the factory and have not been able to kill them. Heated transistors
to 150c, put 1500v dc on the supply, etc and they keep on running. I
am begining to think there may be a stress related problem with the
lead bend. Production is bending a complete 90degree at the drain lead

of the mosfet directly at the package. It looks like the package
ruptures and sends plasma out and destroys the clamp diodes and control

chip and so forth. Destruction is very complete here.

So question is then - has anybody ever heard of failures related to
this type of handling? I will contact mosfet vendor after discussing
here.


thanks,
Bob
 
P

Pooh Bear

Jan 1, 1970
0
Yzordderrex said:
The thread was hijacked and by the time we got to post number 153 it
had morphed into something else which had nothing to do with my post -
Heheh - So I figured to start fresh. I placed OP at the bottom for
reference.

We stuck a couple of controls in the environmental chamber and brought
them down to 20 below. Took them out and walked them downstairs to the
high power lab and plugged in in - Boom. Blew up 2 out of 4. Looks
like the moisture getting into the drive was enough to kill the power
supply. Drive looks clean as a whistle when the moisture gone.

I had mentioned to the group that I had thought contamination was
killing units. One of the other (young & green) engineers would offer
up the fact that units would come back clean in an effort to shoot down
contamination as a cause. So now we will have a meeting on Monday and
I have all the ammo I'm going to need. Clean up the packaging and if
failures continue we will take another look.

Boss is having a leadbend fixture made as well. Thanks for all the
comments that apply I'll sort them out and send them up the chain.

Once upon a time you could get 'hermetic TO-220' IIRC. Look into it. Otherwise
redesign for TO-3.

Graham
 
M

Mark

Jan 1, 1970
0
did you mention something about a rooftop intake?

have you considered lightning?

Mark
 
L

legg

Jan 1, 1970
0
The thread was hijacked and by the time we got to post number 153 it
had morphed into something else which had nothing to do with my post -
Heheh - So I figured to start fresh. I placed OP at the bottom for
reference.


We stuck a couple of controls in the environmental chamber and brought
them down to 20 below. Took them out and walked them downstairs to the
high power lab and plugged in in - Boom. Blew up 2 out of 4. Looks
like the moisture getting into the drive was enough to kill the power
supply. Drive looks clean as a whistle when the moisture gone.

I had mentioned to the group that I had thought contamination was
killing units. One of the other (young & green) engineers would offer
up the fact that units would come back clean in an effort to shoot down
contamination as a cause. So now we will have a meeting on Monday and
I have all the ammo I'm going to need. Clean up the packaging and if
failures continue we will take another look.

Boss is having a leadbend fixture made as well. Thanks for all the
comments that apply I'll sort them out and send them up the chain.

You'd checked normal operation at -20(C?) first?

I think most operating environments specify non-condensing RH.... so a
spritzer might have been just as informative.

RL
 
Y

Yzordderrex

Jan 1, 1970
0
Good Idea.

I'll mention it when we meet again. The sales and service groups seem
to have come around to accepting the fact that it is contamination
causing probs.

We are about to offer an IP64 (NEMA 4 type) so that maybe there is an
opportunity to use this particular installation as a beta test site.

regards,
Bob
 
T

Terry Given

Jan 1, 1970
0
Yzordderrex said:
Good Idea.

I'll mention it when we meet again. The sales and service groups seem
to have come around to accepting the fact that it is contamination
causing probs.

We are about to offer an IP64 (NEMA 4 type) so that maybe there is an
opportunity to use this particular installation as a beta test site.

regards,
Bob

One dead drive I looked at came from a milk powder factory. the heatsink
was an inserted-fin type, 4mm fin-fin spacing. The *entire* heatsink was
one solid block of Al and rock-hard milk powder. moist environment made
milk powder stick, heat baked it. repeat until no airflow at all.
customer eventually sent the drive for service as it kept tripping.

The service guys couldnt get the milk powder out, it was like concrete.
So they replaced the entire heatsink. Other than that, the drive worked
fine.

Cheers
Terry
 
J

joseph2k

Jan 1, 1970
0
Pooh said:
Once upon a time you could get 'hermetic TO-220' IIRC. Look into it. Otherwise
redesign for TO-3.

Graham

TO-3? Who has that kind of real estate nowdays? TO-66 maybe.
 
P

Pooh Bear

Jan 1, 1970
0
joseph2k said:
TO-3? Who has that kind of real estate nowdays? TO-66 maybe.

I doubt that many devices even exist in TO-66 any more actually.

Graham
 
Top