Power Designs 1570 HV Supply - Photos

[c4757p]

As promised, here's a mini-teardown (only removed the cover) of a Power Designs 1570 3kV power supply.

[Vgkid]

Very cool, keep us updated on its workings. We have a bunch of 3Kv Kepcos that are collecting dust at school.

[c4757p]

Yep, I'll follow up as soon as I get the replacement tube and some capacitors. Hopefully everything works. Those two massive caps test fine at low voltage (capacitance, ESR and leakage), but I'm not able to test them at working voltage (and I'd rather not be near them fully charged, anyway...) They appear to be PIO (one is a Sprague Vitamin Q) - good quality PIO, but still PIO. Perhaps I'll bring it up slowly on a variac. I'll have to take a look at the circuit and see if that would make the control circuitry grouchy. I refuse to replace them - each replacement would cost not insignificantly more than I spent on the whole unit!

I have a hunch from what I see on the PCB and what I've read that it could be run with one tube out at half output current, but I won't bother trying it because I don't want a 3kV mistake.

[SeanB]

Note as well those caps are oil filled and the oil is PCB based. If they are weeping they need replacing.

[c4757p]

Luckily they seem to be fine. Pretty sure I'd just give up on this if those caps were bad - I can't even find replacements anywhere, and I can't imagine what they would cost if I found them...

[c4757p]

Just realized this thread is in the wrong section 

[c4757p]

A bunch of new capacitors later, it works a treat!

Question: Is failure of neon lamps typically caused by operational hours, or by total life, or mechanical failure? The reason I ask is that out of three neon indicators, two were dead. There are many more being used as voltage clamps throughout the circuit, and I'm wondering whether any of them should be replaced as well, or if the fact that they run with a very low duty cycle means that they can expect much longer lives than front panel indicators.

[bilko]

A bunch of new capacitors later, it works a treat!

Question: Is failure of neon lamps typically caused by operational hours, or by total life, or mechanical failure? The reason I ask is that out of three neon indicators, two were dead.

see here
http://www.nutsvolts.com/uploads/magazine_downloads/NeonLamp-Information.pdf

Edit: inc. quote

[uoficowboy]

see here
http://www.nutsvolts.com/uploads/magazine_downloads/NeonLamp-Information.pdf

I don't see anything there that answers the question, but I may have missed it.

c4757p: I suspect your neons are OK - but you could always remove them from the circuit and test them! I believe neons are mostly degraded by use, not just time, so I would think they'd be OK.

[bilko]

I don't see anything there that answers the question, but I may have missed it.

c4757p: I suspect your neons are OK - but you could always remove them from the circuit and test them! I believe neons are mostly degraded by use, not just time, so I would think they'd be OK.

Apologies, should have quoted, c4757p was asking about NEONS. The article answers his questions

[c4757p]

I still don't see anything in that article that talks about failure mode, unless I'm missing it?

[bilko]

I still don't see anything in that article that talks about failure mode, unless I'm missing it?

No it doesn't say failure mode of a neon is because ...

But it does explain about typical neon currents, over voltage effects sputtering. I guess you just skimmed the article

[c4757p]

I understand what causes them to fail due to operation... I just wanted to know if there are also common failure modes (leaky glass/metal seal, etc) that can cause them to age even when they're not conducting.

I actually read the article very carefully trying to find the information in it.

[bilko]

I understand what causes them to fail due to operation... I just wanted to know if there are also common failure modes (leaky glass/metal seal, etc) that can cause them to age even when they're not conducting.

I actually read the article very carefully trying to find the information in it.

Neons don't usually fail from non use, in fact there are records of neon lamps being in operation for over 40 years. Yes, sure if the glass leaks from faulty construction, or the leads corrode, but that should be obvious to look at.

Is the glass darkened ?
Does the circuit supply AC or DC to the neon, I'm assuming DC if it is being used as a current clamp. In which case the electrodes wear down a lot quicker. You can check, if both electrodes glow it is AC

Edit: Just re-read your post. If the neons are being used as indicator lights, they will eventually wear out, refer back to the article I quoted and look at operational hours in the data sheets, some are as low as 2000 hours

[c4757p]

No, the glass isn't darkened. I'd replace them immediately in that case. Now that I've thought about it for a bit, I'll probably just end up replacing them anyway, they're cheap as hell and the things they protect are not. I've got a dozen or so of the "standard" ones. There's one weird fat one; either I'll leave it alone, or I'll test to see when it breaks down and try to find one that matches. Too bad they don't have part numbers...

Assuming they haven't already failed, I don't think they operate at all during "normal" function. I haven't played around with this all that much, but the only glow it gives off is from the tube heaters. I haven't gotten around to mapping out the PCB yet, so I'm not 100% sure how the circuit works. They may strike during transient events, or maybe they're just failsafe components in case something else doesn't operate correctly (the tubes become leaky and arc over?).

Edit: Just re-read your post. If the neons are being used as indicator lights, they will eventually wear out, refer back to the article I quoted and look at operational hours in the data sheets, some are as low as 2000 hours

Ah, I don't care about the indicators. They don't allow things to break when they fail, I can just replace them when they stop lighting up. Maybe in a more professional environment it'd be important to make sure the HV warning lamp lights when the HV is turned on, but at home it's just me, and I can just look at the switch...

[bilko]

No, the glass isn't darkened. I'd replace them immediately in that case. Now that I've thought about it for a bit, I'll probably just end up replacing them anyway, they're cheap as hell and the things they protect are not. I've got a dozen or so of the "standard" ones. There's one weird fat one; either I'll leave it alone, or I'll test to see when it breaks down and try to find one that matches. Too bad they don't have part numbers...

If the glass isn't darkened they should be OK
They're quite easy to test, PSU above striking voltage and series resistor is all you need. It might be worth doing the test to map out the characteristics because not all neons are the same, different strike voltage, hold voltage, min, max current e.t.c.

[Alexei.Polkhanov]

As promised, here's a mini-teardown (only removed the cover) of a Power Designs 1570 3kV power supply.

I found annotation under one of photos that looks like a mercury switch or a neon lamp -  what is "neon fuse"?

[c4757p]

A neon lamp in a fuseholder

[Alexei.Polkhanov]

A neon lamp in a fuseholder

Why? From where it located this lamp must be used for some sort of feedback, like temperature compensation  or something like that. Anyone can shed some light on that? Looks like there is some interesting design trick behind it.

[c4757p]

Nope, just overvoltage protection. It's dark during normal operation, and flashes if the output is shorted.

[SeanB]

Neons do degrade with time ( over decades) as the radioactive material in the electrodes ( there to start the discharge at a predetermined voltage instead of having to wait for a naturally occurring particle to initiate) gradually decays, but they will work for a century or more if not struck. Failure is generally the glass seal failing letting nitrogen in ( no glow) or the electrodes sputtering until a conductive path is present between electrodes ( needs a long time at a current over 1mA to do this) but they will still work even if the glow discharge is nearly invisible in the glass from regular sputtering. Operating on DC only the anode will suffer erosion so it will take a long time. You will notice a rise in glow voltage as they wear, it is usual.

[c4757p]

Seven months later, I have a schematic. Sort of...

In case anyone is interested, here is the schematic for the very similar HV-1544. I'm considering scanning in the rest of the manual as well.

[bronson]

I found a nice HV-1544 at my local electronics recycler, might have saved it from a grisly death.

Haven't found a manual for it though...  c4757p, I'm hoping you scanned yours?

Any fun things to try with this beast, beyond just insulation testing and spark gapping?

[c4757p]

PM sent.

I use it to power actual HV circuits sometimes - things with tubes and whatnot. But it's also nice for insulation testing.

[bronson]

Scanned (finally) and submitted to the KO4BB and BAMA archives.  Man, time flies when a pet project has no deadline.

Available here until they appear there: https://u32.net/docs/Power%20Designs%20HV-1544%20Manual.pdf