Hi all;
this is a general question involving photomultiplier tube ops in general. I have two older RCA 7265 PMT tubes (14 stage bialkaly photo cathode rated for 2500 kv operation). I am attempting to revive these tubes for a neutron scintillation detector. The tubes came with sockets equipped with a dynode divider network and a single discrete device amplifier card. Brandname on the board is FMC, INC. Anyway I have spent several weeks breadboarding and testing these devices and all I get is 60hz noise with no response other than possible noise generated with movement. I don't see anything like darknoise current or photoresponse at all. I am currently using a 1.1 kv positive power supply and have hooked up in both cathod positive and cathod negative mode with no change in output. I do have the vintage RCA PMT manual with all sorts of info on connecting the device but this sort of revival work is not covered in the book. I will close with several questions. Do PMT's die (totally without response) or do they just loose their amplifying power with time. What is the electrical characteristic of a tube that has been exposed to excessive light levels while powered. Thanks for any help the group can offer. Larry Upjohn.
PMT operational characteristics
-
Larry Upjohn
- Posts: 135
- Joined: Wed Jan 03, 2007 3:06 pm
- Real name: Larry Upjohn
PMT operational characteristics
Larry Upjohn
-
Richard Hull
- Moderator
- Posts: 15037
- Joined: Fri Jun 15, 2001 9:44 am
- Real name: Richard Hull
Re: PMT operational characteristics
Naturally, you have the tubes in absolute, unequivocal, ink black, darkness.
Given this, and the proper polarity and potentials on the preamp board's low voltage supply line, you shouldhave a dark current.
Testing the tube's response to light is a bit more tricky without an attached scintillator.
14 stages is a huge and super amplification factor tube! That dynode string must have 12 or more resistors and a few capactiors hanging on the socket pins.
PMTs rarely, if ever, "go bad" in the classic sense of total failure unless they are left in light and powered up. (the ultimate no-no)
I would think you have something hooked up wrong. Note* dark currents are very small. How are you attempting to measure this?
All PMTs must be electromagnetically and electrostatically shielded or hum, 60 cycle, will be picked up. Finally, a tiny light leak to the tube even in total darkness would show up as hum due to the AC mains powered external light sources.
Richard Hull
Given this, and the proper polarity and potentials on the preamp board's low voltage supply line, you shouldhave a dark current.
Testing the tube's response to light is a bit more tricky without an attached scintillator.
14 stages is a huge and super amplification factor tube! That dynode string must have 12 or more resistors and a few capactiors hanging on the socket pins.
PMTs rarely, if ever, "go bad" in the classic sense of total failure unless they are left in light and powered up. (the ultimate no-no)
I would think you have something hooked up wrong. Note* dark currents are very small. How are you attempting to measure this?
All PMTs must be electromagnetically and electrostatically shielded or hum, 60 cycle, will be picked up. Finally, a tiny light leak to the tube even in total darkness would show up as hum due to the AC mains powered external light sources.
Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
-
Larry Upjohn
- Posts: 135
- Joined: Wed Jan 03, 2007 3:06 pm
- Real name: Larry Upjohn
Re: PMT operational characteristics
Richard;
Tubes have nice mumetal shields have been masked off with vinyl electrical tape. Anode output goes to an FET opamp configured in Current to voltage mode with gain of 10-100. I am using AD648 dual opamp chip with a slewrate of 1.4v/microsecond. Tube output is terminated into a 50ohm divider to match the coax patch to the amp. Output looks like badly matched 60hz sinwaves with or without the tube in the socket. I suspect that it may be ripple from the HV supply. These supplies are small DC-DC encapsulated units requiring 28vdc in and putting out 1.1kv. Don't have a HV probe and haven't looked at ripple yet. Again these PMT's were removed from equipment some where in the early 70's and I have stored them since in the dark well cushioned to protect tube integrity. Thanks again for your response! Dynode divider is made up of 1.1megohm carbon film resistors with bypass caps on the last 4 stages. There are also two grids listed in the tube. G2 is common with Dynode 13 and G1 at the photocathode end is common with an exterior ring just above the base of the tube that contacts the mumetal shield. I have tried several ways to bias it but really reluctant to have the shield and ring at HV potential while breadboarding. I have access to some surplus xp2012's and xp2014's of much newer vintage. Just wondering how much more effort to put into the RCA 7265 units. Thanks again for your fast response! LRU.
Tubes have nice mumetal shields have been masked off with vinyl electrical tape. Anode output goes to an FET opamp configured in Current to voltage mode with gain of 10-100. I am using AD648 dual opamp chip with a slewrate of 1.4v/microsecond. Tube output is terminated into a 50ohm divider to match the coax patch to the amp. Output looks like badly matched 60hz sinwaves with or without the tube in the socket. I suspect that it may be ripple from the HV supply. These supplies are small DC-DC encapsulated units requiring 28vdc in and putting out 1.1kv. Don't have a HV probe and haven't looked at ripple yet. Again these PMT's were removed from equipment some where in the early 70's and I have stored them since in the dark well cushioned to protect tube integrity. Thanks again for your response! Dynode divider is made up of 1.1megohm carbon film resistors with bypass caps on the last 4 stages. There are also two grids listed in the tube. G2 is common with Dynode 13 and G1 at the photocathode end is common with an exterior ring just above the base of the tube that contacts the mumetal shield. I have tried several ways to bias it but really reluctant to have the shield and ring at HV potential while breadboarding. I have access to some surplus xp2012's and xp2014's of much newer vintage. Just wondering how much more effort to put into the RCA 7265 units. Thanks again for your fast response! LRU.
Larry Upjohn
-
Richard Hull
- Moderator
- Posts: 15037
- Joined: Fri Jun 15, 2001 9:44 am
- Real name: Richard Hull
Re: PMT operational characteristics
Sounds like all is right. What are the supply voltages for the FET preamp or is a low voltage supplied from the 28 volt HV block. Op amps and the like usually work off a positive and negative supply. I have several surplus tubes with op amps and they need +/- 15 volts. I naturally assume you have the amps powered correctly.
With the hum ( I don't know your amplitude), I would look for an open ground, ground loop or a power common not carried through.
Richard Hull
With the hum ( I don't know your amplitude), I would look for an open ground, ground loop or a power common not carried through.
Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
-
Larry Upjohn
- Posts: 135
- Joined: Wed Jan 03, 2007 3:06 pm
- Real name: Larry Upjohn
Re: PMT operational characteristics
Thank you Richard and Geo for the input. I have been avidely reading these postings for over two months now and the interchange has been of great interest and learning for me. I went back to my schematics for the current dynode divider network and discovered that my configuration has the photocathode and the first dynode at the same potential. My literature all says that their must be at least a 5-10% voltage drop from PC to DY1 for function so hopefully I have not damaged anything. I have been using a 2mhz Tek 5110 to monitor my preamp output and obviously now I will also need to use something on the order of 100mhz for better observation as well. My preamp is powered +-15vdc from a separate powersupply (analog) on the op-amp prototyping breadboard. As soon as I get home tonight I will implement these changes hopefully with productive changes. Thanks again, Larry Upjohn. .
Larry Upjohn
Re: PMT operational characteristics
Hi Larry -
Just to chime in late... I think your first dynode potential would be a show stopper. Nothing from the photocathode is probably getting past the potential barrier, if they are at the same potential - so you would get zero output.
Since the photomultiplier tube has electron multipliers for current amplifrication, the dynodes are progressively more positive as you go away from the photo-cathode. So the last dynode is positive, and the photo-cathode is negative. I think neither end of the PMT is connected to a ground, for noise immunity.
Also, the gain of the multiplier is extremely non-linear with regard to voltage. An older tube, might need a bit more voltage to show gain. But I wouldn't do that first.
First you should be able to get some noise free output from you final amplifier. A good test circuit can be built around a 555 in auto- triggered mode... to produce a square wave... Doesn't need to be a high rep rate... the front edge of the square wave will sort out the amplifier's frequency response. If you let the oscillator drive into a few megohm load, you will get a sub micro amp pulse train. Couple that into your amplifier and you should be able to verify all the signal processing circuitry.
When you're happy there, it's onto the PMT and supply.
Check the HV power-supply. PMT supplies even back in the 70's were high frequency units, so that filtering was easier to accomplish... because clean, sweet DC is absolutely necessary if you want a photocurrent signal, no interference. So check the supply.... ripple needs to be in the PPM levels... no kidding.... vanishingly small, and you will need enough current for the divider string... (Max voltage divided by 15.4 megs, PLUS enough additional current so that photocurrents do not cause the supply voltage to sag.. which would change the gain.)
Fortunately a good capacitor or two...and preferably an RLC filter... will do what you need.
That last item is about fluorescence of the tube's glass. This can persist forat least a whole day, especially if the tube has seen daylight or fluorescent lamps. I would use at least aluminum foil backed up by the electrical tape, to be sure there's no light in-leak. A mm thick metal plate would be more rugged, but the foil-tape will work nicely.
Cheers.
Dave Cooper
Just to chime in late... I think your first dynode potential would be a show stopper. Nothing from the photocathode is probably getting past the potential barrier, if they are at the same potential - so you would get zero output.
Since the photomultiplier tube has electron multipliers for current amplifrication, the dynodes are progressively more positive as you go away from the photo-cathode. So the last dynode is positive, and the photo-cathode is negative. I think neither end of the PMT is connected to a ground, for noise immunity.
Also, the gain of the multiplier is extremely non-linear with regard to voltage. An older tube, might need a bit more voltage to show gain. But I wouldn't do that first.
First you should be able to get some noise free output from you final amplifier. A good test circuit can be built around a 555 in auto- triggered mode... to produce a square wave... Doesn't need to be a high rep rate... the front edge of the square wave will sort out the amplifier's frequency response. If you let the oscillator drive into a few megohm load, you will get a sub micro amp pulse train. Couple that into your amplifier and you should be able to verify all the signal processing circuitry.
When you're happy there, it's onto the PMT and supply.
Check the HV power-supply. PMT supplies even back in the 70's were high frequency units, so that filtering was easier to accomplish... because clean, sweet DC is absolutely necessary if you want a photocurrent signal, no interference. So check the supply.... ripple needs to be in the PPM levels... no kidding.... vanishingly small, and you will need enough current for the divider string... (Max voltage divided by 15.4 megs, PLUS enough additional current so that photocurrents do not cause the supply voltage to sag.. which would change the gain.)
Fortunately a good capacitor or two...and preferably an RLC filter... will do what you need.
That last item is about fluorescence of the tube's glass. This can persist forat least a whole day, especially if the tube has seen daylight or fluorescent lamps. I would use at least aluminum foil backed up by the electrical tape, to be sure there's no light in-leak. A mm thick metal plate would be more rugged, but the foil-tape will work nicely.
Cheers.
Dave Cooper