Help with Cold Cathode

Thread Starter

metamict

Joined Jul 19, 2017
5
Hi,
I have a Cathodoluminescnece system, which is a scientific instrument. It is circa 1980's technology. This instrument uses a cold cathode to create an electron beam, which is directed towards the surface of a sample. The electron beam will cause certain samples to luminesce. The maker of this instrument was essentially a one man company and that man has since retired and is no longer reachable. The good thing is that the cold cathode portion of the instrument is user-serviceable.

The system I have worked for quite a while but I haven't used it in several years. In starting up the instrument, I have come across several issues and I have rectified all but one. The system works by pumping a vacuum (~50-100 mTorr) on the sample chamber and cathode. Once the appropriate vacuum level is reached, the high voltage is switched on and a dial is used to increase the voltage (up to an operating voltage of 10-15 kV).

The problem: When turning up the voltage, the current increases quickly. By the time the voltage is turned up to a value of approximately 2 kV, the current reaches a limit and trips the high voltage circuit.

Some additional information:
-Without a vacuum, the voltage can be turned up to 15 kV (the maximum value). Only a small current is observed, which according to the manual is related to the current limiting circuit.
-The moment as a vacuum is applied to the system at 15 kV, the current limit is tripped and the HV is turned.
-Based upon the behavior of the vacuum gauge, I believe that the gauge is fine and that the system is pumping down to an appropriate level.
-I have carried out the troubleshooting steps listed in the manual. Based upon the fact that I can turn up the voltage to 15 kV when no vacuum is present, I think that the power supply is working properly.

I have attached a manual for the instrument which includes various troubleshooting options and circuit diagrams.

I fear that this may be too complex of a question to post online, but I thought I'd give it a shot in case anyone has ideas. I will do my best to answer questions and respond quickly. Oh, I'm far from an expert in electronics, though I have some basic knowledge and tools.

Thank you kindly,

Chris
 

Attachments

#12

Joined Nov 30, 2010
18,210
The manual states however that the current should be less at higher vacuum:
That is counter-intuitive for me, however, I tend to focus on the over-current lock-out circuit: Page 4, the schematic.

Apparently R9 is the 1 meg resistance, represented on the schematic as one resistor but explained in the text as any of several configurations of several resistors required for physical length to avoid arcing across the resistor(s). This is a prime suspect. One megohm will limit the 15KV current to 15 ma, which is not panacea, but somewhat of a stabilizer. Failure often occurs only when subjected to high voltages and replacement might be required as the only way to be sure the fault. Of course, any amount of surface contamination can render this part ineffective, and years of not being used invites surface contamination.

What I interpret from the testimony is that excess current does exist and the safety lock-out is functioning properly.
(It might be helpful to read the part number on Q1.)
When the voltage at R5 becomes positive enough, the voltage at R6 fires the SCR called Q2, and that engages the lock-out circuit. The current through Q1 is in the range of 12 ma to develop enough voltage across R6 and the gate of an SCR can be in the range of 3 ma to 50 ma. Knowing the part number of Q1 would give a clue, but I don't expect a failure in that direction.
The capacitors in the HV section are probably ceramic and have a very good reputation.
Please continue measuring voltages and continue the conversation.
 
Last edited:

Hypatia's Protege

Joined Mar 1, 2015
3,219
Based upon the fact that I can turn up the voltage to 15 kV when no vacuum is present
Ooops! Missed that:oops:

The problem: When turning up the voltage, the current increases quickly
That being the case, your problem almost certainly owes to insufficient chamber evacuation or 'outgassing' contaminants (inasmuch as coincident malfunction of the current indicator and current limiting circuits is most unlikely)...

Best regards
HP:)
 
Last edited:

nsaspook

Joined Aug 27, 2009
7,283
Just a WAG:

Sounds like you have might a vacuum leak (external or internal virtual leak from contamination) near the high voltage anode terminal. If your vacuum system conductance is sufficient to pump the vacuum vessel to the needed base pressure at the vacuum pump gauge the gradient of pressure across the vessel at the leak will have a much poorer vacuum level at a small cross section than normal causing a higher electric field gradient across the leak that increases the changes for local ionization of the leaked gasses easily resulting in a avalanche current and short inside the device as voltage increases above a critical level.

 

#12

Joined Nov 30, 2010
18,210
near the high voltage anode terminal.
Just a nit-pick:
In this case, the anode is nearly at earth ground. It is the cathode which is driven by about -15KV.;)
I don't know whether you missed that point or (correctly) deemed it insignificant.
(All of us can think in either direction.)
 

nsaspook

Joined Aug 27, 2009
7,283
Just a nit-pick:
In this case, the anode is nearly at earth ground. It is the cathode which is driven by about -15KV.;)
I don't know whether you missed that point or (correctly) deemed it insignificant.
(All of us can think in either direction.)
Good point, I used the penning tube as the example. What's important is the actual field gradient across a possible leak.
 

Aleph(0)

Joined Mar 14, 2015
597
Just a WAG:

Sounds like you have might a vacuum leak (external or internal virtual leak from contamination) near the high voltage anode terminal. If your vacuum system conductance is sufficient to pump the vacuum vessel to the needed base pressure at the vacuum pump gauge the gradient of pressure across the vessel at the leak will have a much poorer vacuum level at a small cross section than normal causing a higher electric field gradient across the leak that increases the changes for local ionization of the leaked gasses easily resulting in a avalanche current and short inside the device as voltage increases above a critical level.
Nsaspook that's exactly the kind of dreck I went through with TEM but I don't know what to say to OP except for close inspection cuz u can't use ion getters at 100 Torr:eek:

That being the case, your problem is almost certainly insufficient evacuation (inasmuch as coincident malfunction of the current indicator and current limiting circuits is most unlikely)...
What was your first clue, Capt obvious:p!

Quote by HP deleted at her request

HP that's totally correct! Insulation strength is weakest at 3mm of vacuum! Which is a big reason hermetic refrigeration compressor is useless for like _poor man's_ vac pump:rolleyes:

It may be helpful to observe operation of the unit in a darkened room for evidence of corona/arcing...
HP I say OP prolly won't see just 2kv leak but it's worth a try:cool:
 
Last edited:
Virtual leaks, real leaks and outgassing can be a real pain. All it might take is a real screw rather than a vented screw.

Where I worked we only has one Penning type gauge on our glass sealing vacuum manifold. I probably only worked on the gauge twice in 25 years. Then there are the Hasting's TC gages, The capacitance manometer and the Ionization gages.
 

#12

Joined Nov 30, 2010
18,210
HP that's totally correct! Insulation strength is weakest at 3mm of vacuum! Which is a big reason hermetic refrigeration compressor is useless for like _poor man's_ vac pump:rolleyes:
The big warning in refrigeration school was: Refrigeration compressors are not built to operate below atmospheric pressure. The windings turn into a huge cathode and the motor commits suicide.:(
Besides, when you empty the container, the heat of the motor is inside a Dewar flask.:eek:

Repairmen worship 1 atm reading on gauges when manipulating refrigerant.
 

Thread Starter

metamict

Joined Jul 19, 2017
5
Wow - thanks to everyone for all of the responses. Quick and very insightful. It looks like I have some good options to work through. I'm going to follow-up on these and will get back with responses to each of these ideas as I make progress.

This is my first time posting here and I'm really amazed at the helpfulness of this community. Thank you all for taking the time to thoughtfully consider my questions. I'll be in touch with an update soon.
 
HP that's totally correct! Insulation strength is weakest at 3mm of vacuum! Which is a big reason hermetic refrigeration compressor is useless for like _poor man's_ vac pump:rolleyes:
Indeed! Howbeit I believe you mean 3mmHg absolute!:cool: -- FWIW I deleted my (erroneous) post for its inconsistency with Paschen's Law, personal observation, common sense and sanity! -- (FWIW my description of the normalized current vs. absolute pressure characteristic was both 'inverted' and 'transposed' owing to my simultaneous 'confusion' of EMF with current -and- 'Vacuum pressure' with 'absolute pressure' [respectively]:eek::oops:) -- Such are the 'wages' of 'multitasking' on 3 hours sleep in as many days - excuses, excuses...o_O

With marked abashment
HP:oops:
 

Aleph(0)

Joined Mar 14, 2015
597
Indeed! Howbeit I believe you mean 3mmHg absolute!:cool:
HP That's right! 3mm above just total vacuum! My bad:oops:!

FWIW I deleted my (erroneous) post for its inconsistency with Paschen's Law, personal observation, common sense and sanity!
HP I say you need to relax cuz I totally saw through that and I say other ppl did too! HP like you know I worked with re-entrent Vac systems for like two years and even I can still get bogged down by the totally stupid variety of units and reference conventions with no observed consensus on standard in sight:mad:! So I guess just denominations of pascal and absolute pressure in every field of science and trade would be making it too easy:rolleyes:

So anyhow I got your e and deleted quote in my post so now everything is like you say _fine as wine_ which I say had totally more to do with your upside down backwards thinking than sleep deprivation:p!

Wow - thanks to everyone for all of the responses. Quick and very insightful. It looks like I have some good options to work through. I'm going to follow-up on these and will get back with responses to each of these ideas as I make progress.

This is my first time posting here and I'm really amazed at the helpfulness of this community. Thank you all for taking the time to thoughtfully consider my questions. I'll be in touch with an update soon.
Metamict tnx!:) Also when you solve problem plz let us know what was going on even if we all got it wrong:cool:!

This is my first time posting here and I'm really amazed at the helpfulness of this community. Thank you all for taking the time to thoughtfully consider my questions.
Metamict thank YOU! Cuz your courtesy and following up on responses makes you the rarest and best type of first time poster:)!
 
Last edited:
Metamict thank YOU! Cuz your courtesy and following up on responses makes you the rarest and best type of first time poster:)!
Totally agree.

I HATE THE PASCAL. it's just too big of a number.

I like inches, but whatever wizzard came up with 1/8, 1/16, 1/32, 1/64 and 1/128 I'd like to beat him/her with a wet noodle. it's nice to be able to use body part for a foot (foot) and a inch (thumb to knuckle).

Just recently called a manufacturer of bread and said the bread was something like 17/32 thick. I thought the decimal value would confuse her. Turns out she heard of 16's, but not 32 nds of an inch. I guess a ruler marked in 1/10 of an inch tics would really confuse her.

I still don;t have a handle on mm. Especially 227 mm. Again, the numbers are too big.

==

I never really learned how to use an RGA (especially a stand-alone one), but I did fix one or two. Two home-brew systems had RGA's built into their designs and we eventually got a stand-alone one.

One thing that really helped was to standardize on the plug used (twist lock) and put the same disconnect on the 10 or so rough pumps of the same general model we had. When a pump went out of service, it was removed and replaced with a re-built spare. Spares were a good thing.

We were able to etch and gold deposit the contacts on used crystals for thickness monitoring.

So, I have some experience fixing Auger/SIMS machines, electron microscopes, EDAX (Energy Dispersive X-Ray Analysis) systems (mostly crude). Setting up an ebic (electron beam induced current) system and building a Faraday cup and stage was fun stuff. The SEM's I used were crude: ISA Alpha-9; A Jeol T-200, a Phillips something. I never got trained on the Amray and the fancy edax because it had a service contract. Old school stuff - Poloroid camera.
 

Thread Starter

metamict

Joined Jul 19, 2017
5
I'm back and will try to address some of the points made above in an attempt to move this fix forward:

"Sounds like you have might a vacuum leak (external or internal virtual leak from contamination) near the high voltage anode terminal."

Metamict Response: I cleaned the cathode/anode system and replaced consumables (anode/cathode/discharge tube) (all of which is very much accessible). This has not helped. Regarding a vacuum leak: I ordered a new vacuum gauge (a calibrated thermocouple) with a meter that is independent of the system. I'm pulling a vacuum of about 150 mTorr. This suggests to me a slight leak (as the terminal pressure should be less); however, 150 mTorr should be right in the sweet spot to produce cathodoluminescence.

"It sound to me that parts are moving due to the vacuum."

Metamict Response: The anode/cathode assembly is held in place by a removable metal case. The case serves as a shield for x-rays as well as a path for the ground. The screw holes in this assembly fix the location of the anode/discharge tube/cathode. So I don't think anything is shifting much when the vacuum valve is opened.

"Apparently R9 is the 1 meg resistance....is a prime suspect."

Metamict Response: I was told this by someone else. I'm going to look into checking this. Access to the inside of the power supply is, unfortunately, challenging with the way the system was put together.

"Knowing the part number of Q1 would give a clue, but I don't expect a failure in that direction."

Metamict Response: I will try to get that number when I am able to get deeper into the system.

So at this point, I'm going to try to get into the electronics box in an attempt to check some of those circuits.

Thanks - and any other thoughts or suggestions would be appreciated.
 
I'm pulling a vacuum of about 150 mTorr. This suggests to me a slight leak (as the terminal pressure should be less)
Water. Have you changed the pump oil? Checked the oil level? Have you measured the vacuum at the pump? Corollary - Is the pump good?
Does the pump have a backstream preventer? Have you run the pump with th ballast partially open (helps remove water).

Is the system easy to bake?
 

Thread Starter

metamict

Joined Jul 19, 2017
5
Water. Have you changed the pump oil? Checked the oil level? Have you measured the vacuum at the pump? Corollary - Is the pump good?
Does the pump have a backstream preventer? Have you run the pump with th ballast partially open (helps remove water).

Is the system easy to bake?
Thanks for your ideas:
--> Pump oil was just changed with fresh oil.
--> Oil level is fine
--> Pump is fine (the vacuum pumps to an acceptable range)
--> No backstream preventer (there has never been one and the system has worked in the past)
--> System is not easy to bake.
--> Vacuum at the pump is great.
 
Last edited:
At this point, i really think you have a leak which is a real pain to detect it's too big. Acetone won't work.

If you have to, you have to start blocking off ports. Wiggling feedthroughs. This is the range that's really really tough to find. its a BIG leak.

Although I never tried it, I'd tray an upside down can of canned air that uses R-134A as a propellant. See: http://www.engineeringtoolbox.com/thermal-conductivity-liquids-d_1260.html; https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwiWj9OX37TVAhVHTCYKHaDQCbsQFggwMAA&url=https://mycourses.aalto.fi/pluginfile.php/208222/mod_folder/content/0/Refrigerant_R134a.pdf?forcedownload=1&usg=AFQjCNFRxfcOBZghHlgzKrBUCZYNKAePtA

I've used the "air" to cool things, but not leak detect.

https://www.mcmaster.com/#canned-air/=18qreuv

That's ABOUT the pressure the vacuum gets stuck when there is too much water and it's a real PITA to find.

I wasn't the vacuum GURU, but I had my share of finding leaks. We had same UHV systems and one system that used an ION pump.

You also might try Helium or Argon. Helium is wierd. It just gets everywhere FAST. It's lighter than air, so start at the top with minute quantities.
 
Top