[KrisBlueNZ]

It would have been better to have them all in a straight line. If you'd used 1/4 watt resistors, nine of them might have gone end-to-end across the space on the board. Then you would just make sure they were raised somewhat off the board. That's how I would have done it. But as long as you have plenty of clearance between the two end connections and you use well-insulated wire you should be alright. Keep the two wires well apart from each other. Don't put them through the same hole.

 

[tryppyr]

Well, I do have ten 3M 0.5W resistors, so if a better configuration is in order I can certainly do it.

I removed R13 and soldered in the test block. I reassembled the system, again leaving HV MULT unplugged. When I powered up this time I tested at the same junction point. As the system was warming up the point measured about 9.5V. After about a minute the voltage rose to 12.5. None of the neons on the HVPS lit this time. Also, there was no ozone smell.

By this I deduce that R13 had failed.

By the way, when I removed it I saw that it said 1%, not 10%,

 

[KrisBlueNZ]

Ah, more progress!

I'm not sure whether you need to change your resistor arrangement. As long as the ends are distant from each other, and the whole setup is not too far away from the board, and the wires are well insulated, you should be alright.

I would expect the voltage at that test point to be around 0V. But if the HVPS is regulating now, we can worry about that later, if at all.

How about the 68k resistor? Is it still getting hot?

Is the heater in the CRT glowing? That is a good indication that the HVPS is working properly.

What about the +48V rail? That's probably the next thing to fix. If it's still wrong, you can try replacing the four small transistors in the regulator circuit.

 

[tryppyr]

Bad news. I just rechecked the work in preparation for the testing of the 48V rail and found that the first test was invalid. I had forgotten to connect the off-board Q1 transistor. Once that was connected all the problems came back. Ozone smell was almost immediate. R14 got very hot very fast. I even saw a faint puff of smoke.

So apparently, R13 was NOT the problem.

 

[KrisBlueNZ]

Bummer. You can put the original R13 back into the board.

It's probably the CA3094A. At least, that seems the next logical suspect. Unfortunately, the CA3094A is obsolete. I found an eBay auction for five CA3094E devices, but you need the A suffix because the power supply is 30V and the non-A-suffix device is only rated for up to 24V. Also the E device is in a plastic DIP package. Can you try to find one? I'll continue thinking about the situation...

 

[tryppyr]

Thanks, Kris. Yeah, I'll put R13 back and seek out a CA3094A.

In the meantime I think I'll also work on getting the other rails of the LVPS up to snuff.

 

[tryppyr]

R13 is back in place now.While I had the HVPS board out I decided to take down the numbers of all the transistors so I could check to make sure they agree with the schematic. Problem is, I don't have any way of cross referencing the Motorola or NS part numbers with what's detailed in the replacement parts list.

Q3 and Q5 are supposed to be MPSA42. The Motorola part number on both is 036-4-575. I'm tempted to just pick up a couple of NTE287 and replace Q3 and Q5. I note that Q5's emitter is on the -2295V line with R13 and also R14. Q3's collector is tied to TP6, and Q3 seems to be involved in the Deep Erase function, which we know shuts down the CRT. Frankly, nothing on the board shows visible signs of failure, but at 99 cents apiece replacing those two would be a cheap test.

Update: I also ordered a replacement HV Oscillator.

 

[KrisBlueNZ]

NTE287 is a replacement for MPSA42 so you should be able to do that. I doubt that either of them would cause that problem though.

What do you mean by "HV Oscillator"?

 

[tryppyr]

HV OSC is the name for the off board Q1 transistor that I forgot to plug in on the first test of the resistor block.

 

[KrisBlueNZ]

OK, there's no need to replace that. If it wasn't working, you wouldn't get any high voltage.

 

[tryppyr]

Thanks

 

[tryppyr]

I hate being so dense, but I'm trying to find the HP part number for the CA3094A. I keep going around in circles. What is its designation on the board?

 

[KrisBlueNZ]

It's A15U1.

 

[tryppyr]

Thanks. I was having a real senior moment. I couldn't stop staring at the A16 parts page.

I spent a fair amount of time yesterday seeking a CA3094A replacement. I spent a LOT of time trolling through audiophile blogs and such. No sites showed up selling CP3094A. However, I did see one site that had DIY plans for building an OTA based on the CP3094 schematic. It was done using 6 of the 2N5087, 7 of the 2N5210, 6 of the 1N4148 and a couple of resistors. It occurred to me to wonder whether that might work for the O-scope. Of course, the DIY version was derided by critics, saying there was virtually no way to get the tight tolerances of the integrated circuit. But the musician that made it was happy with the results.

Still, I don't think that would work for me.

Anyway, I replaced the two MPSA42s on the HVPS. I haven't powered it up since then because I want to see what else I can do before the next power test.

- Greg

 

[KrisBlueNZ]

I saw that article on the discrete CA3094A as well. I agree with the critics that it's not a good idea.

You mentioned CP3094A twice in your previous post. It's CA, not CP. Just making sure that you're using the right search term!

It might be possible to use a CA3080 with some external components, but I'm not sure how to do it. A genuine CA3094A would definitely be the best solution.

 

[tryppyr]

I totally agree. I'll keep looking around. The reason I wanted to deduce the HP part number is because I've found a couple of sites that seems to have the NOS HP parts (still in the orginal plastic packages). I'd MUCH prefer those to the ones removed from other units (i.e. solder on the pins).

 

[tryppyr]

Quick update. I poked and poked with the HP part number (1826-0167). Got a lot of interesting information, but no leads on a NOS part. Did find out this chip is considered important to several naval vessel types, and has been given a FEAT designation (whatever that means).

Anyway, I was unable to locate a NOS chip, but I did find a "Used - Like New" on Amazon for $8 plus shipping... so I ordered that. What the heck, eh.

And yes, I ordered a CA3094AT. Another senior moment, apparently.

 

[tryppyr]

Since I'm basically at a stand-still on this project, while waiting for the CA3094AT to arrive, I've been looking things over to see what else I might be able to replace, on the off chance that something I have readily available might lead to a resolution.

That said, I don't want to arbitrarily replace parts that are working just fine with parts that aren't quite right for the circuit. Hence the following questions.

I see a number of diodes on the HVPS that seem to be replaceable with the 1N4007. These are described as:

Diode-Pwr Rect 400V 750MA DO-29
Diode-HV Rect 1KV 600MA DO-29
Diode-HV Rect 1KV 800MA DO-29

Would the 1N4007 be an inappropriate swap for any of these?

There are some other diodes on the board as well, with the following descriptions:

Diode Switching 30V 50MA 2NS DO-35

Would the 1N4148 be a suitable replacement?

Also noted that there is a specialty diode on the board:

Diode-HV Rect 10KV 5MA 25DNS

This part is obviously going to be harder (and more expensive) to source, so the question here is how likely is this one to be the problem?

 

[KrisBlueNZ]

I wouldn't worry about any of those. Let's assume they're innocent until proven guilty. The last one is OK because if it wasn't, you wouldn't be seeing the neons lighting up.

If you want something to do, how about fixing the +48V rail from the LVPS? Have you tried replacing the four small transistors in the +48V regulator? See post #79.

 

[tryppyr]

Haven't done that yet, but I did remove both boards for additional work, so I'll get on that now.