The place to be when you have TEA. Discuss all kinds of test equipment.
Important: Use tags for the type of equipment your topic is about.
Forum rules
Use tags for the type of equipment your topic is about. Include the "repairs" tag, too, when appropriate. If a new tag is needed, request one in the TEAdministration forum.
bd139 wrote: ↑Mon Jun 17, 2024 12:24 pm
Just got an HP E3610A power supply from eBay for a respectable price. Initially looks good.
...
Oh now hang on a minute, no it doesn't. There's a modification, clearly an insane one. There is what looks like a series 1n4148 with the negative lead ...
Some people! I will have to very carefully check this one out and inspect it I think...
Red annulus on both +ve and -ve terminals?
Can it supply the rated current? 1N4148 are 300mA devices, and they show 290mA
bd139 wrote: ↑Mon Jun 17, 2024 12:24 pm
Oh now hang on a minute, no it doesn't. There's a modification, clearly an insane one. There is what looks like a series 1n4148 with the negative lead
Seems like a strange bodge. I'm wondering whether there might be a legitimate reason for doing it. Why a 1N4148 and not a 1N4007?
From the photos it looks clean and not codged about, but you can't tell until you have a good look.
bd139 wrote: ↑Mon Jun 17, 2024 12:24 pm
Just got an HP E3610A power supply from eBay for a respectable price. Initially looks good.
...
Oh now hang on a minute, no it doesn't. There's a modification, clearly an insane one. There is what looks like a series 1n4148 with the negative lead ...
Some people! I will have to very carefully check this one out and inspect it I think...
Red annulus on both +ve and -ve terminals?
Can it supply the rated current? 1N4148 are 300mA devices, and they show 290mA
"Saving up for a dual output PSU". Really?
Well considering the diode fell to bits instantly when I touched it, it tried at least.
This has been repaired as attached. The voltage drop was calibrated out by the seller. Obviously this changes due to the load and how much smoke was coming out of the 1N4148.
IMG_0850.jpeg
Of course now we have problem number two which is the drift! Fortunately not the seller’s doing this one. References are spot on and no drift. U1 is bad. This is the voltage amplifier and is an LM741. Large thermal drift. This is a known issue with these and later ones had LF411CN instead. I have some new TI 741’s so I’ll stick one in and if that doesn’t improve it, I’ll order some LF411’s.
Initially I thought it was 1999 vintage but it’s a bit earlier at 1992. I’ll go through the ECNs and apply them as well if I can be arsed.
Neutral feedback will be left as it’s obvious the seller was trying to hide this.
You do not have the required permissions to view the files attached to this post.
bd139 wrote: ↑Mon Jun 17, 2024 12:24 pm
Oh now hang on a minute, no it doesn't. There's a modification, clearly an insane one. There is what looks like a series 1n4148 with the negative lead
Seems like a strange bodge. I'm wondering whether there might be a legitimate reason for doing it. Why a 1N4148 and not a 1N4007?
From the photos it looks clean and not codged about, but you can't tell until you have a good look.
My only conclusion is a previous owner was a fuckwit.
bd139 wrote: ↑Mon Jun 17, 2024 2:08 pm
The voltage drop was calibrated out by the seller. Obviously this changes due to the load and how much smoke was coming out of the 1N4148.
...
Neutral feedback will be left as it’s obvious the seller was trying to hide this.
If you can demonstrate they calibrated it out, then I'd screw the vendor - and neutral feedback seems unusually generous.
I'm sure you can get better and cheaper at Newbury or Dunstable Downs. Not sure whether I'll take a full car, but I definitely don't have any PSUs to offload.
bd139 wrote: ↑Mon Jun 17, 2024 12:24 pm
Just got an HP E3610A power supply from eBay for a respectable price. Initially looks good.
...
Oh now hang on a minute, no it doesn't. There's a modification, clearly an insane one. There is what looks like a series 1n4148 with the negative lead ...
Some people! I will have to very carefully check this one out and inspect it I think...
Red annulus on both +ve and -ve terminals?
I vaguely remember reading about the logic behind the colors on the binding posts, and it being that the negative terminals were not at ground potential unless jumpered to the adjacent ground terminal; if that was not done they were floating and as supplies could be connected in series to boost voltage (to a point), the negative terminal could be hot, thus the red rings. I personally don’t like it, and think they should have used red & black for the outputs, and green for the ground, but they didn’t consult me.
No idea why in the wide, wide world of sports someone would have installed that diode! Weird.
bd139 wrote: ↑Mon Jun 17, 2024 2:08 pm
The voltage drop was calibrated out by the seller. Obviously this changes due to the load and how much smoke was coming out of the 1N4148.
...
Neutral feedback will be left as it’s obvious the seller was trying to hide this.
If you can demonstrate they calibrated it out, then I'd screw the vendor - and neutral feedback seems unusually generous.
I'm sure you can get better and cheaper at Newbury or Dunstable Downs. Not sure whether I'll take a full car, but I definitely don't have any PSUs to offload.
Ok so this is now repaired and not drifting. As suspected the voltage amplifier opamp was wonky. It was replaced with a TI UA741CP which is just what I happened to have lying around. I'll calibrate it another day but it's delivering the rated current properly now anyway
The culprit...
IMG_0851.jpeg
I'll have a think about the seller. It was an easy enough fix and performance is as expected now. Neutral is fine. If I was a novice or without the diagnostic ability I have I probably would have just returned it.
You do not have the required permissions to view the files attached to this post.
bd139 wrote: ↑Mon Jun 17, 2024 3:25 pm
I'll have a think about the seller. It was an easy enough fix and performance is as expected now. Neutral is fine. If I was a novice or without the diagnostic ability I have I probably would have just returned it.
From what you have said, the seller deliberately misrepresented its operation in two ways: calibrating out the diode, and choosing a current 10% of the PSU's capacity and just under the diode's capacity.
That deserves a rap over the knuckles. At least ask for demand 50% off.
bd139 wrote: ↑Mon Jun 17, 2024 3:25 pm
I'll have a think about the seller. It was an easy enough fix and performance is as expected now. Neutral is fine. If I was a novice or without the diagnostic ability I have I probably would have just returned it.
From what you have said, the seller deliberately misrepresented its operation in two ways: calibrating out the diode, and choosing a current 10% of the PSU's capacity and just under the diode's capacity.
That deserves a rap over the knuckles. At least ask for demand 50% off.
If you get a refund then you can't leave negative or neutral feedback. Money is less important than tarnishing his reputation
bd139 wrote: ↑Mon Jun 17, 2024 3:25 pm
I'll have a think about the seller. It was an easy enough fix and performance is as expected now. Neutral is fine. If I was a novice or without the diagnostic ability I have I probably would have just returned it.
From what you have said, the seller deliberately misrepresented its operation in two ways: calibrating out the diode, and choosing a current 10% of the PSU's capacity and just under the diode's capacity.
That deserves a rap over the knuckles. At least ask for demand 50% off.
If you get a refund then you can't leave negative or neutral feedback. Money is less important than tarnishing his reputation
I've never had to get nasty.
Tarnishing his reputation with a well-constructed thorough complaint is very public spirited - and necessary.
Badly modified with a series diode on the (-) terminal and had serious drift problems. Both corrected: removed diode and replaced U1 so not a write off. But clearly barely as described and this would have been known to the seller based on the current test shown. Packaging not great either. Knocking around inside a much larger box with insufficient packaging. 50/50. Should have been marked as spares/repair and untested.
I suspect the seller acquired this thing with the wacky 1N4148 mod, which would limit its output to about 300mA, so not a very useful power supply. He must have discovered this. He knew enough to replace the 1N4148, but not enough to realise it was nonsense, or sort out the problem with the op amp, if he could use it long enough to notice that.
He unloaded it on ebay with an essentially dishonest description and photos tailored to skirt around the limitation. A few 1N4148s may have been sacrificed in producing the photo.
Giving him neutral feedback, and scathing but accurate feedback, seems just.
Someone else could have bought this, not technically knowledgeable and just expecting a working power supply, and found it died on them straight away. Then they would have returned it and he would have had more problems.
bd139 wrote: ↑Mon Jun 17, 2024 3:25 pm
Ok so this is now repaired and not drifting. As suspected the voltage amplifier opamp was wonky. It was replaced with a TI UA741CP which is just what I happened to have lying around. I'll calibrate it another day but it's delivering the rated current properly now anyway
The culprit...
IMG_0851.jpeg
9118 - that’s after the Danbury, CT plant I used to work at closed, so my hands are clean on that one.
bd139 wrote: ↑Tue Jun 18, 2024 10:33 am
Having done it before I can confirm that a 1N4148 will take 2A for more than 1µs but only once. And it'll glow orange.
Is it then still a 1N4148? Or has it mutated into a blob of mixed badness?
bd139 wrote: ↑Tue Jun 18, 2024 10:33 am
Having done it before I can confirm that a 1N4148 will take 2A for more than 1µs but only once. And it'll glow orange.
Here's that fascination with fire coming through again
bd139 wrote: ↑Tue Jun 18, 2024 10:33 am
Having done it before I can confirm that a 1N4148 will take 2A for more than 1µs but only once. And it'll glow orange.
Is it then still a 1N4148? Or has it mutated into a blob of mixed badness?
Two blobs of mixed badness to be clear. Each bonded to their own lead.
bd139 wrote: ↑Tue Jun 18, 2024 10:33 am
Having done it before I can confirm that a 1N4148 will take 2A for more than 1µs but only once. And it'll glow orange.
Here's that fascination with fire coming through again
tggzzz wrote: ↑Mon Jun 17, 2024 1:03 pm
Can it supply the rated current? 1N4148 are 300mA devices, and they show 290mA
At DC.
Pulsed 20A is possible.
Which datasheet says that?
On-Semi says 1.0A for a 1s pulse and 4.0A for a 1µs pulse - non-repetitive.
Makers have to put out absolute max ratings, but no one would rely on a device being able to do that, and the ratings are quite likely a bit iffy anyway. If you are even approaching it, get something better specified, which in the case of 1N914/6, 1N4148/4448, isn't difficult or expensive. I've seen them used as PSU rectifiers in very low consumption TE, but mostly as signal diodes coming nowhere near their limits.
It was probably shown at 290mA because that was the highest current they could manage without having to remove the cover and replace the dead diode with another 1N4148, again. They didn't even have the sense to bung in a 1N4007.