Help with Glassman 40kV, 15mA psu

For Short Term Learning Discussions ONLY. This area is for CURSORY questions and connecting with other users ONLY. ALL technical contributions need to be made in the appropriate forums and NOT HERE. All posts are temporary and will be deleted within weeks or months. You should have already search the extensive FAQs in each of the forums before posting here as your question may already be answered.
Rex Allers
Posts: 570
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Help with Glassman 40kV, 15mA psu

Post by Rex Allers »

I never worked out all the complicated circuit interactions of the feedback signals to control the output on the supply. Today I looked at the P-COMP device and it does seem to be the clamp that limits the output as both the KV and mA outputs approach their max (would be 40 KV and 15 mA for this supply).

So maybe you can check some voltages and see if you can see this tripping too early, and why.

Here's what I worked out today.
------------------------------------------------
Some EW circuit notes

The gain of the I-Follower amp, U8/A, produces the I-monitor signal. This should be +10 V when current feedback from the multiplier indicates 15 mA output. So gain adjusted for,
Imon voltage = I (in mA) / 1.5

The gain of the V-Follower amp, U9/A, produces the V-monitor signal. This should be +10 V when voltage feedback from the multiplier indicates 40 KV output. So gain adjusted for,
Vmon voltage = V (in KV) / 4

These monitor voltages can be seen on the backpanel TB1 strip.
TB1-4 is V-monitor, TB1-7 is I-monitor.

The P-COMP device, U6/B, appears to be the clamp that limits power output as the I and V both approach their max (would be 10 V on both monitor levels). I'm guessing P-COMP means Power Compare or Power Compression.

The device is an op-amp but it is being used as a comparitor. The (+) input is a DC level that is set by R103 pot. Doing the math, the level can be set between 8.5 and 9.5 volts.

The (-) input to the comparitor is the average of Vmon and Imon as,
(Vmon + Imon) / 2

When that (-) input level goes above the DC level on the (+) input, the comparitor will trip, clamping the output.

From the supply specs, we know that with V maxed at 40 KV the I output should be clamped to 12.5 mA. At these levels the comparitor input would be,
(Vmon + Imon) / 2 = (10 + 8.33) / 2 = 9.165 V

That's around the middle of the R103 setable range for the DC level, which makes sense.

If this P-COMP comparitor is tripping too soon for some reason it could be why your output current isn't getting up to 12.5 mA.

As mentioned, the Vmon and Imon signals are availble to be measured on TB1.
The output of the P-COMP comparitor can be measured on the main board circular test pad E31. This can be found on the circuit board about 1 inch below the pot R5, and can be seen on the diagram I sent earlier.
Rex Allers
Matt_Gibson
Posts: 505
Joined: Thu Sep 23, 2021 10:36 am
Real name: Matt Gibson

Re: Help with Glassman 40kV, 15mA psu

Post by Matt_Gibson »

Rex,

I owe you a beer for all that (scotch actually).

I took some measurements and got the following:

V mon: 9.6
I mon: 3.8

Raising either beyond these levels knocks current down and I have to adjust it back up.

Looks like I’m running around 70% of what I should be doing?

I’ve already maxed out the P comp trimpot to get current to this level. I don’t know what would need adjusting at this point.

Any harm in popping the P comp op-amp out and trying?

Edit: Forgot to add that e31 shows 12.98v no matter what I adjust (I/V).

-Matt
Rex Allers
Posts: 570
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Help with Glassman 40kV, 15mA psu

Post by Rex Allers »

Matt,
Not sure what to make of that info. I believe the output of P-COMP (E31) should be high when not clamping and go low to clamp. You say E31 is always high ~13V so it doesn't seem to be clamping. I would expect it to go close to 0V. But you say that adjusting the DC level on the (+) of the op-amp comparitor does have an effect on the supply current. I can't understand that.

I don't think you can pull the U6 chip. First, not sure about the result of floating the E31 level. Second, its a two op-amp chip. The other half seems to be one way to shutdown the switcher chip. Can't figure out what it is doing, though.

Without understanding all those control circuits I can't suggest what else to look at.

-Rex
Rex Allers
Matt_Gibson
Posts: 505
Joined: Thu Sep 23, 2021 10:36 am
Real name: Matt Gibson

Re: Help with Glassman 40kV, 15mA psu

Post by Matt_Gibson »

Rex,

I’ll watch E31 while operating to see what happens when the supply knocks the current down…I’ll also check the value on the input pin to see if it’s somehow getting more than 9.5v (I jacked the trimpot all the way up).

I’m having trouble tracing out where this signal goes and how current is actually adjusted down. It seems that it touches many of the circuits before making its way to the PWM chip.

I reached out to Glassman, but never heard back. I think they really want me to spend big money on a refurbishment…

-Matt
Rex Allers
Posts: 570
Joined: Sun Dec 30, 2012 3:39 am
Real name:
Location: San Jose CA

Re: Help with Glassman 40kV, 15mA psu

Post by Rex Allers »

I just took another look at the schematic and I got the function of the P-COMP device wrong. It's not a comparitor, it's an integrator. So the output won't be bang bang as I was thinking. It will be a level.

With this new point of view I think I now understand basically how the control circuits work.

-----------------
The signal line that goes to pin 2 of the 3524 switcher is what controls the PWM of the switcher outputs.

The I-AMP, V-AMP and P-COMP are all integrators. The DC bias on the (+) input sets the desired control level. For the V-AMP and I-AMP these levels come from front panel V and I pots.

The V-AMP sets the main operating control going to the PWM line through R100. This will strive to set the output so the V feedback matches the V set pot.

The I-AMP output connects to the PWM line through a diode. If the I feedback starts to exceed the I pot set point, the output will go negative, being able to pull down on the PWM level through diode D66, overriding the level from V-AMP.

The P-COMP works basically the same as the I-AMP. Its output connects to the PWM line through D68. If the combined average of the V and I feedback signals exceed the set point from R103, then P-COMP will pull down to limit the output.

There are a lot more circuits, but I now feel comfortable that this describes the main feedback signals and how they control the output.

I think most of my earlier description is ok, except for the part where I though P-COMP was a comparitor.

------------
How to debug your situation is another matter. If the output of either I-AMP or P-COMP is lower than the output of V-AMP then that is what is limiting your output. Unfortunately, only P-COMP seems to have a test point. You would have to measure at the amp chip output pin, I guess.
Rex Allers
Matt_Gibson
Posts: 505
Joined: Thu Sep 23, 2021 10:36 am
Real name: Matt Gibson

Re: Help with Glassman 40kV, 15mA psu

Post by Matt_Gibson »

I poked around and everything looks like it should. Nothing is exceeding the 9.5v input…

While poking around, I saw that the trimpot for the PWM frequency was turned down. I turned it up such that the red glue/marker looked like it had originally been set. This went from 29khz to 40khz. This wound up getting me up to 7.5mA, up from 5mA!

Looking over at my power meter (I know they aren’t all that accurate) I was now pushing 550w, up from 430w or so. So, a “100w” improvement. I let it run like this for a few minutes…

Unfortunately, I could hear some popping noises (much like corona discharge) from near the multiplier and then a sudden loss of current and the dreaded “smell”. I shut down and waited a minute before turning back on. I was able to bring it back up, used my probe to see I was near 38kV, and then the noises started up before loss of current and the smell.

I pulled the multiplier and took a look. No signs of damage or anything, nothing tested as shorted (like before). Trying it again does the same as before. Not sure what’s going on, but I’m guessing the Glassman isn’t happy at all.

An interesting note: The other identical Glassman on eBay (5x what I paid for this one) had the seller telling me that these are custom designs from a flight simulator. There have been tweaks and mods made to them for whatever reason.

No telling what the hell has been done (turning down the frequency) to these that could be causing these problems.

Ugh!$!$!
Post Reply

Return to “New User Chat Area”