short on mainboard for early 90's mass spectrometer

[cashkennedy]

Backstory: I figured the power supply had went bad after 25 years on this mass spectrometer, because all the voltages were 0 and I could clearly tell the power supply was getting 110v input. But I didn't read the manual in its entirety so wasted my time swaping the power supply from a spare parts machine. Should have read the next line in the manual about testing the power supply which tells you if all voltages are 0 you have a short that has activated the crowbar circuit.... wow where has crowbar circuits been for the last 10 years? Do they not cover crowbar circuits in Chinese universities so everything they make just blows fuses / parts when there is a short?

The short is on the mainboard and is between ground and +15v.

In the pictures I labeled the only areas I know for sure where 15v is present and also the only safety caps that were shorted (~0 resistance between each side)

Are there any other areas you think might be using the 15v, or might be necessary to create 250v using 15v. And what would yall recommend testing as most likely part to fail? I tested the 4 Motorola transistors that have round metal caps and the 3 that are labeled all test identical ( with diode setting / all 6 possible wire orders) , and the 1 that isn't labeled tests as if the 2 other pins were swapped , and I have a similar known good board from a slightly different but similar model that has most of the same parts and the 4 Motorola transistors tested identical.

[stj]

Re: short on mainboard for early 90's mass spectrometer

try and find a service manual or chematics - it will save a lot of time and soldering.

[cashkennedy]

Re: short on mainboard for early 90's mass spectrometer

The mass spectrometer comes with a service manual that had lengthy paragraphs on the function of different sections and guides for testing each board as a whole, no schematics though. I doubt HP made further in depth service manuals for these, but I could be wrong.

All the National?? (2 N 's on top of each other?) and PMI chips and 1 BB chip all have shorted caps over them (top side of chip if looking at board in the way that most text labels are readable) So did a bunch of chips in the early 90's just run directly on 15v power? And thus a bunch of chips on here could be in theory the source of the short since a lot of sections would have 15v power?

I have a almost identical board, so I usually just check the corresponding part on the other board before id ever have to solder anything. I don't really want to spend a full 4 hours testing every part on the board cause at that point my employer would probably rather just pay the $800 or so we could probably haggle someone down to on a used board.

[budm]

Re: short on mainboard for early 90's mass spectrometer

I will not be surprised if one of those yellow bypass caps for those ICs is shorted, you may be able to lift just one leg off the board but no matter what it will be time consuming.

[stj]

Re: short on mainboard for early 90's mass spectrometer

those chips are probably op-amps.

[goontron]

Re: short on mainboard for early 90's mass spectrometer

You could put power to it and wait for one of the tantrum capacitors to fail. Though, Not recommended for somebody with less than 10 years dealing with this stuff.

[cashkennedy]

Re: short on mainboard for early 90's mass spectrometer

The bypass caps are being used between the vcc of the ic's and ground? Thus those IC's are running on 15v? And you think its more likely that a bypass cap would fail then one of the ic's themselves?

Also are those 4 motorola transistors / and the resistors next to them enough to create 250v power from 15v? or should there be more parts involved in that ?

[R_J]

Re: short on mainboard for early 90's mass spectrometer

What I have done sometimes is use a 9volt battery and connect it to the +15 volt line for a second at a time and check for a component getting warm.

Those transistors can produce 250 volts but I'll bet the current is really low

[cashkennedy]

Re: short on mainboard for early 90's mass spectrometer

Testing it with a battery seems pretty easy. 1 Mississippi long second?

Not that I would mind pulling up 1 leg of ~ 12 caps.

[R_J]

Re: short on mainboard for early 90's mass spectrometer

You can lift the caps, I usually use a weak 9 volt, I keep the used ones that are usually around 5-6 volts for things like this. I borrowed a Thermal camera once to identify a short on a board turned out to be a shorted microproccesor ic.

[stj]

Re: short on mainboard for early 90's mass spectrometer

your -250v is going to be generated by the invertor circuit on the top-right.
the transformer and 2 large transistors Q5 and Q6.
check those for shorts.

[cashkennedy]

Re: short on mainboard for early 90's mass spectrometer

9v battery didn't work, used a old one from a smoke alarm that was down to 8v, and a fresh 9.5v one but neither if held for about 3 seconds warmed up anything. I was doing positive on battery to positive 15v point, and negative on battery to ground.

stj you're right that the top right generates a voltage using the transformer, but it generates the 70v used to run the filaments, and there is a 2 prong socket right above the transformer that goes to the filaments.

The area with the transformer that generates the 70v does use +15v as the input though, I probed around the known good board that is basically like 2 jumpers and an IC different then the broken one, and found all the areas that use +15v. The area that generates 250v doesn't appear to even get +15v directly even thought he manual clearly said it was created using the +15v unless its coming through the line of 6-8 ic's that someone thought were op amps.

What type of capacitor are the 3-4 larger biaxial gray colored ones that are mostly 22uf (c8, c10 for example in the top right), do those ever fail? Is KEMET the brand?

[stj]

Re: short on mainboard for early 90's mass spectrometer

the kemet 22uf are axial electrolytics.

the 250v may be coming from the aluminium box or the black brick on the other side.
what's in he aluminium box?

[budm]

Re: short on mainboard for early 90's mass spectrometer

Can you upload the manual you have? And where is the connector for the 250V (DC or AC?)?

[R_J]

Re: short on mainboard for early 90's mass spectrometer

The big black box is a dc-dc converter http://www.ergpower.com/DC-DCconverters.html
E915 suggests +15 volts input, so it could be bad?
The -5.015ct means output 500 volts 15ma, CT= center taped = +250v & -250v

[stj]

Re: short on mainboard for early 90's mass spectrometer

o.k. well it could be shorted.

[budm]

Re: short on mainboard for early 90's mass spectrometer

The big black box is a dc-dc converter http://www.ergpower.com/DC-DCconverters.html
E915 suggests +15 volts input, so it could be bad?
The -5.015ct means output 500 volts 15ma, CT= center taped = +250v & -250v

I did the search earlier using that ERG Incorporate E915-5.015CT, it shows as HP P/N 1826-1823 but I did find any more info than that. So that can be the possibility, it will easier to remove the part or disconnect the 15V line feeding it to see if the shorts goes away than trying to remove one cap at a time.

[cashkennedy]

Re: short on mainboard for early 90's mass spectrometer

I took pictures of the only pages of the manual that I think will help you, there actually are some decent schematics there, but they don't cover everything. I think I got on the wrong track on what was creating the 250v because of where the test points are on the board, and I saw that +15v was input to the dc/dc converter for 250v output.

I also attached a picture of what's hidden under the silver box (this is the area where the super sensitive signal comes in so is probably under the box to prevent any static / interference.

I cant figure out any way to remove the dc/dc converters cover, I think you have to just unsolder it from the board. There are 6 leads below it. The one for +15v input is pretty easy to deduce, and probably the other at the bottom is ground. the 2 at top appear to be the outputs, and im not sure what the 2 on the side are for. I will use some braid to get the solder out of that joint tonight, and if I cant get a good disconnection pull the whole dc / dc converter off.

[R_J]

Re: short on mainboard for early 90's mass spectrometer

You won't get the cover off the dc-dc converter, its potted in epoxy.
If you go to the ERGpower site and click on "Please Click Here to find our DC-DC Series nomenclature" it will give you the pdf on these dc-dc's and near the end (page18) 21 in adobe is the pinouts for the E900 Series
there are two outputs, each outputs 250 volts, connect the minus to ground of one output and get +250 volts. connect the Plus output to ground of the other output and get minus 250 volts.

[cashkennedy]

Re: short on mainboard for early 90's mass spectrometer

Man yall are good. Appears the dc/dc converter from erg is bad. I couldn't get the power leg to sufficiently desolder, so I thought about it a second and remembered from the erg part sheet that yall linked that it recommends a inducer on the power leg. So I just pulled the inducer, and now the board no longer has a short. From what I can tell after the inducer there is just 1 kemet cap and the dc dc converter. (The rest of the 15v parts still are connected to the 15v rail, I checked, and ground is no longer shorted to 15v)

It also looked a little like the dc dc converter had been replaced before, but there are a couple of ic's on this board that look that way (flux residue on all the pins). A lot of people used to repair these since they are traditionally 3000+ dollar boards.

I had another one of these old mass spectrometers fry a power supply today. I'm glad I have a spare parts machine...

So are DC/DC converters just generally a really weak link / a lot more prone to shorting then other parts? I just figured it was just Vizio and their retarded desire to put 30 dc dc converters on their original tv models