Video Teardown & Repair of a Stanford Research PS350 5000V DC Supply

[Hugoneus]

In this episode Shahriar repairs a Stanford Research Systems Model PS350 5000V-25W High Voltage Power Supply. The unit continuously displays 2.5kV without the output being enabled and produces no output voltage. Verification of power supply voltages reveals the issue is linked to a disconnected 15V voltage regulator IC. After the repair, the output voltage is verified with both positive and negative outputs. The principle operation of the instrument as well as the Cockroft-Walton high voltage generator is reviewed.

Watch the video here: [40 Minutes]
youtu.be/09otx2Pfphw

More videos at The Signal Path:
http://www.TheSignalPath.com

[Lightages]

Thanks again for the videos! They are always informative, to the point and well worth watching. Can I make one piece of constructive criticism? I know, I am not waiting for you answer!

Almost all of your videos have a weird color cast that can be attributed to using auto color balance on your camera. This usually results in this strange color cast when florescent lights are being used. It would not be perfect but you could do a color balance while filling the screen with a white piece of paper first, or buy a white balance card made specifically made for this purpose. They aren't that expensive.

The other thing is the use of auto exposure. One video it was tracking all over the place with reflections from the aluminum in the shot and then you blocking the light with your hand. Again, there are grey level cards that can be purchased for very little money to help set the exposure, and then lock it. You can also just do a exposure level setting with the back of your hand in the shot and then lock it. You might need t play a bit get what you like.

So again, thanks for all the work you have put into your productions and sharing with everyone else.

[peter.mitchell]

To provide strain relief on the packages attached to the chassis you could mount the devices with the legs facing up, and then wire them point to point with some stranded wire down to the pcb.

Usually the component legs are longer and are formed to provide strain relief in these kind of mounting situations.

[TiN]

I bet Dave is envy your easy repairs This is getting a tradition.
 

[BFX]

Lucky man ... again 

[Radio Tech]

Another great video and repair.
I really like the way you explain things.  I know you said this was your last repair and I hope you don't mean it was the final repair.

Thanks for sharing.

[Len]

I don’t mind that the problem was easy to find and fix because it was an interesting failure mode. Looking at the back of the unit, I noticed the bottom of the case was bent. It looks like that instrument has been knocked around quite a bit.

Can I make one piece of constructive criticism? 

On the plus side, Shahriar has learned that Youtube success can be assured by putting a cat in every video.

[Hugoneus]

Thank you all for your comments.

I am hoping to get some better equipment for filming. I'll see what I can do in the meanwhile to improve color balance and exposure. Thanks for the feedback. So better exposure and more cats, got it.

I have ordered an SHV connector and cable, I am eager for their arrival. It would be interesting to investigate some dielectric breakdown properties of capacitors.

[HighVoltage]

You have a great way of finding faults and I learn something from each of your videos.
Somehow I always need much longer to find the fault in a test equipment and fix it.

Great high voltage power supply.
I am looking forward to see some of your experiments with it.
A big tank you for all your great videos.

[c4757p]

Wow, that regulator/transistor mounting is really dodgy. I would definitely consider drilling a couple new holes a cm or so higher and connecting them to the board with short, flexible wires. They must have had a lot of failures from that.

[edpalmer42]

Wow, that regulator/transistor mounting is really dodgy. I would definitely consider drilling a couple new holes a cm or so higher and connecting them to the board with short, flexible wires. They must have had a lot of failures from that.

Actually, Shahriar should consider moving all three of those regulators.  If the one that broke is moved, the next one in line becomes vulnerable.  However, as Len mentioned, it looked like the case was bent or maybe just a foot was missing.  If the unit was dropped (are there any mangled corners or edges?) that could have caused the transformer to shift violently and snap the regulator leads.  If that was the case, maybe changes aren't necessary.

[TheAmmoniacal]

Some capacitor experimentation could be very interesting!

Run a series of comparisons between rated voltage and failure voltages (different brands?).

See how the capacitor characteristics change over time the more they spark over (ESR, capacitance, dissipation factor, phase angle).

See if they consistently spark over at the same spot or if it continues to burn holes in the dielectric.

Explain why electrolytic capacitors are polarized and why they don't like having reverse polarity applied to them. Would also be fun to see if you could reverse the polarity of a capacitor (have it powered reverse-current limited for a long time).

And if you got some precise micrometers (or calipers), examine the dielectric breakdown voltage of some substances and compare with table values. (PTFE, air, PE, HDPE, 'vacuum', etc.)

(Just some ideas)

[Mark Hennessy]

Rather than using short wires, I would consider forming a strain-relief loop in the leads of the replacement ICs (and consider replacing them all for the same reason. It might just about be possible to do that with the existing mountings, but new holes about 5mm higher than the current ones might be better.

Do the threads belong to screws, or did they press threaded inserts into the aluminium side? If the latter, moving upwards becomes difficult.

[TheSteve]

Shahriar - you have to start showing us the hard to fix stuff too! Hah, just kidding around, I really enjoy your videos. I think you've inspired many of us to hunt for the "good" broken stuff on ebay.
Now when we can expect your cat to start helping with the repairs, maybe getting screwdrivers for you.

[German_EE]

<sigh> my girlfriend wants to know what the cat is called, I think her priorities are elsewhere

[PA0PBZ]

I wonder what the bandwidth is when using it as a voltage multiplier. I'm thinking way below audio frequencies seeing it regulate and thinking of all the caps in there.

[Hugoneus]

<sigh> my girlfriend wants to know what the cat is called, I think her priorities are elsewhere

Haha... The cat's name is Pooch!

[owiecc]

I have the same unit. We tested some film caps to see how they die. We put 5kV on a 1kV cap for a few minutes. There were some sparks inside but the cap survived. No apparent loss in capacitance.

[Hugoneus]

I have the same unit. We tested some film caps to see how they die. We put 5kV on a 1kV cap for a few minutes. There were some sparks inside but the cap survived. No apparent loss in capacitance.

Interesting, I just tried a 250V foil capacitor with 1kV on it. Failed after about 30 seconds or so and became a short. Punch-through failure.

[HighVoltage]

Some polypropylene capacitors have a self healing process.
http://www.wima.com/EN/selfhealing.htm

With your new instrument, you should be able to analyze this behavior.
Would be interesting to see how fast they recover.
May be you can make this visible with a scope connected.

[IanB]

I am hoping to get some better equipment for filming. I'll see what I can do in the meanwhile to improve color balance and exposure. Thanks for the feedback.

I have to say I've noticed nothing that bothers me about the video quality. White balance, exposure, focusing, all seem perfectly acceptable. Even if they are not perfect there is nothing that intrudes or distracts from my enjoyment of the videos.

[c4757p]

I am hoping to get some better equipment for filming. I'll see what I can do in the meanwhile to improve color balance and exposure. Thanks for the feedback.

I have to say I've noticed nothing that bothers me about the video quality. White balance, exposure, focusing, all seem perfectly acceptable. Even if they are not perfect there is nothing that intrudes or distracts from my enjoyment of the videos.

Yes, this. Perhaps I could nitpick the video if I wanted to, but I would gain nothing from it being "better".

[Lightages]

Actually when he was poking around inside one of the instruments, the shiny aluminum was reflecting the light directly into the camera and causing the rest of the image to be very dark. As he moved his hand across the shiny aluminum the exposure would shift rapidly to good and then rapidly to underexposed again when he moved his hand away. The color balance is the typical green cast from florescent lighting too. I did not bring this up to be critical of his great work, but rather to suggest a way to make better quality for almost no effort.

[IanB]

It's odd, but on my screen the video seems to have a yellow cast. To check this I just sampled a few RGB values from some of the whitest things I could find in the video (like paper or white plastic) and they all were equally saturated in red and green (= yellow) and lacking in blue. None of my samples were lacking in red. Ho hum. Maybe color is in the eye of the beholder? 

[BravoV]

Maybe color is in the eye of the beholder? 

https://en.wikipedia.org/wiki/Tetrachromacy#Human_tetrachromats