Author Topic: Photomultiplier - Cherenkov detector success! - soldering questions  (Read 4188 times)

0 Members and 1 Guest are viewing this topic.

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Hey all!

I've just received a big box of photomultiplier tubes, including the one i needed - a large square end-on pmt.

The divider pcb soldered on the back didn't suit my needs, so i removed it, unfortunately one of the spot-welded wires on the pins broke, and i decided (foolishly) to remove the rest.

How do I go about connecting the tube now? see attached.

 -> There are still fragments of wire spot-welded on, so a socket is a no-go.

 -> Soldering that thick of a pin that's feeding through a glass envelope feels a little risky but maybe a copper heat wick near the bottom of the pin would allow careful soldering?

 -> Little two-screw screw terminal blocks might work, they become quite small when the plastic is removed.


Does anyone have any experience on this front? I really don't feel like risking my lovely tube so I'd like to hear your advice before doing anything.

Thanks!
--Chris
« Last Edit: February 14, 2020, 08:00:26 pm by ChristofferB »
--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3656
  • Country: fr
  • Analog, magnetics, Power, HV, Audio, Cinema
    • IEEE Spectrum
Re: Photomultiplier - soldering directly on tube pins?
« Reply #1 on: February 08, 2020, 12:54:48 pm »
Chris suggest to include photo of entire tube and model number

Typically these unbased tubes need the dedicated divider board, and the tube wire termination is very delicate and easy to crack the dûment seal.

Can you PM me the photo of tye "bigbox", Iam collecting Photomultipliers since 1968

Kind Regards


Jon
An Internet Dinosaur...
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Photomultiplier - soldering directly on tube pins?
« Reply #2 on: February 08, 2020, 01:08:41 pm »
I would try very sharp, flush side cutters to remove the remains of spot-welded wires. It's likely good for a socket after this, but if not, you can carefully file down the rest (or use sand paper).
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #3 on: February 08, 2020, 01:24:19 pm »
It's a Photonis XP3392/SP5 - used with gamma cameras and scintillators. The datasheet (http://www.hzcphotonics.com/products/XP3392.pdf) specifies a divider very different from what was installed, but the divider board also had some features i couldn't make out without removing it completely, like an optocoupler to short dynode stages / some preamp stuff.

Board divider was :         K  D1  D2  D3  D4  D5  D6  D7  D8  A
                                      4R   R    R    R    R    R    R    R   0.75R              R=2M

Datasheet reccomend      K  D1  D2  D3  D4  D5  D6  D7  D8  A
                                       2  1.5   1.5  R    R    R    R    R    0.5R         


Aside from the XP3392 I also got 3 Hamamatsu R1594 tubes, which I can't find much info on, and a Hamamatsu R268HA in a peltier cooled housing, complete with divider!

I'm liking the soldering with a copper heatsink option more and more - i'm really not comfortable with the mechanical strain of screwing a grub screw into the pin.

a last idea could be to get individual tube socket pins and push on, but they won't get very far from the spot welds.

--Chris

--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #4 on: February 08, 2020, 01:29:24 pm »
I would try very sharp, flush side cutters to remove the remains of spot-welded wires. It's likely good for a socket after this, but if not, you can carefully file down the rest (or use sand paper).

That might be possible, but I'd still need individual pins - the 19 pin pmt socket is pretty exotic.

Edit: Just tried it on two pins, it works but I can hear the vibrations resonating through the entire dynode stack. I'm worried the vibrations might crack a glass seal around a pin. On the other hand, the glass seals protrude almost a centimeter inwards into the tube.
« Last Edit: February 08, 2020, 01:34:28 pm by ChristofferB »
--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #5 on: February 08, 2020, 02:04:01 pm »
okay this isn't the prettiest solution but I think It's the safest. I found some old 9-pin tube sockets where the individual pins are slotted, and the welded wire piece fits in the slots! The pins can be slid over! I'm just going to do this, then solder some tinned wire to the socket pins and to a divider board and not mess anymore with it.

Thanks for your input!

--Chris
--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3656
  • Country: fr
  • Analog, magnetics, Power, HV, Audio, Cinema
    • IEEE Spectrum
Re: Photomultiplier - soldering directly on tube pins?
« Reply #6 on: February 08, 2020, 02:31:19 pm »
Bonjour Very fine pix, nice bunch of tubes

Suggest to not  solder to  the pins, as the pin wire may be hard  to tin.
 in  placing a heatsink, beware to strain on the glass.

The divider résistance values depending upon the pulse response time required so can vary greatly as long as the dynode voltage ratio is maintained

These baseless tubes are often for very low leakage currents so in building a divider don't use a pcb.

Bon courage


Jon
An Internet Dinosaur...
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #7 on: February 08, 2020, 03:16:00 pm »
So maybe this deserves a new thread but on the other hand, very related.

I've designed a divider of the PMT, see attached. This is very new electronics to me, so have i made any grave mistakes?

The application is gamma spectroscopy with a NaI(Tl) scintillator. 

The divider resistance is the same as the prewired base recommended in the datasheet, but the divider that was installed originally was in the order of 20-30 Mohm.
Don't get the higher values from the theory I have read so far.

Either way thanks for the help!

--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15156
  • Country: de
Re: Photomultiplier - soldering directly on tube pins?
« Reply #8 on: February 08, 2020, 03:41:58 pm »
The resistors look Ok, especially if taken from the DS. The capacitors looks a little odd, especially C5.  The larger values does not make much sense there.  I would more like consider a capacitor from D8 to ground, as the output is capacitively loaded to ground.  The current circuit looks a little odd with the capacitors only connecting D8/D7/D6 but no capacitive link to ground.  D8 can see a relatively large current spike (e.g. like the output current) and this is the reason for the additional capacitor(s).

The C1 and R2 part very much depends on how to look at the signal. If longer cables are used one could consider series termination.
I would consider extra diodes or maybe a neon tube for protection.
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #9 on: February 08, 2020, 04:47:10 pm »
Oops, C5 is also supposed to be 10n. My bad.

The capacitors across the final stages is how I've seen it done in litterally all examples, see below, so I just copied that over.


https://scionix.nl/voltage-dividers/#tab-id-1



Maybe I should remove the output parallel RC circuit completely, and rely on cables/external termination as needed, making it more open for experiments.

My first worry is to fry the PMT by overcurrent/voltages, so I'm happy that side of things look good.

--Chris
« Last Edit: February 08, 2020, 04:50:21 pm by ChristofferB »
--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2851
  • Country: us
Re: Photomultiplier - soldering directly on tube pins?
« Reply #10 on: February 08, 2020, 06:10:05 pm »
You can get pins for AMP connectors where a wire is crimped to a box-style contact which is then inserted into a housing.
These are made for .025" square posts, but there may be other sizes.  Also, there are DB-25 style connectors where wires are crimped
to the contact and then inserted into the connector.  These would fit slightly larger pins.  So,measure the pins on the tube and
then find a suitable contact for that diameter.

Jon
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2851
  • Country: us
Re: Photomultiplier - soldering directly on tube pins?
« Reply #11 on: February 08, 2020, 06:12:06 pm »
Oops, C5 is also supposed to be 10n. My bad.

The capacitors across the final stages is how I've seen it done in litterally all examples, see below, so I just copied that over.


https://scionix.nl/voltage-dividers/#tab-id-1



Maybe I should remove the output parallel RC circuit completely, and rely on cables/external termination as needed, making it more open for experiments.

My first worry is to fry the PMT by overcurrent/voltages, so I'm happy that side of things look good.

--Chris
Note that thes positive HV arrangement requires a very high voltage coupling capacitor, which MUST not fail!  If it does, you will have the HV goign right into whatever is receiving the PMT signal.

The negative HV arrangement has the anode at ground, so is a bit safer.

Jon
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15156
  • Country: de
Re: Photomultiplier - soldering directly on tube pins?
« Reply #12 on: February 08, 2020, 08:01:20 pm »
To be on the safe side for the PMT, one can start with a lower voltage, so that the gain is lower.  One point is to avoid higher intensity light.
For use with a scintillator one would later use a relatively high voltage to get high gain.  There are usually 3 regimes:  low gain for linear operation of the PMT,  high gain for pulse analysis like with a scintillator and very high gain for photon counting.
 
The following users thanked this post: ChristofferB

Offline richnormand

  • Supporter
  • ****
  • Posts: 721
  • Country: ca
Re: Photomultiplier - soldering directly on tube pins?
« Reply #13 on: February 08, 2020, 09:07:12 pm »
Have a look at this link for proper design and usage of PMTs and base wiring according to intended usage:
https://psec.uchicago.edu/links/Photomultiplier_Handbook.pdf
It is a copy of the venerable RCA classic about photomultipiers. It has pretty much most of the the answers to your questions.
Well worth a complete read, even if you are familiar with PMTs.

Avoid UV light (fluorescent or white LEDs) as dark current might be high for a few days. For low intensity (photon counting) I alway wear gloves to avoid oil from my hands that could create a leakage path on the base/electronics.
« Last Edit: February 08, 2020, 09:12:10 pm by richnormand »
Repair, Renew, Reuse, Recycle, Rebuild, Reduce, Recover, Repurpose, Restore, Refurbish, Recondition, Renovate
 
The following users thanked this post: ChristofferB

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #14 on: February 08, 2020, 10:29:56 pm »
Oops, C5 is also supposed to be 10n. My bad.

The capacitors across the final stages is how I've seen it done in litterally all examples, see below, so I just copied that over.


https://scionix.nl/voltage-dividers/#tab-id-1



Maybe I should remove the output parallel RC circuit completely, and rely on cables/external termination as needed, making it more open for experiments.

My first worry is to fry the PMT by overcurrent/voltages, so I'm happy that side of things look good.

--Chris
Note that thes positive HV arrangement requires a very high voltage coupling capacitor, which MUST not fail!  If it does, you will have the HV goign right into whatever is receiving the PMT signal.

The negative HV arrangement has the anode at ground, so is a bit safer.

Jon


I was actually planning on using two DC blocking caps in series for that eventuality.

The problem with negative HV is that anything earthed or conductive MUST be kept physically away from the tube, especially the window. Since scintillation crystals usually come in aluminum cans this is impractical for these.


To be on the safe side for the PMT, one can start with a lower voltage, so that the gain is lower.  One point is to avoid higher intensity light.
For use with a scintillator one would later use a relatively high voltage to get high gain.  There are usually 3 regimes:  low gain for linear operation of the PMT,  high gain for pulse analysis like with a scintillator and very high gain for photon counting.


The standard procedure for powering up any HV particle detector is to ramp it up slowly until the count rate stabilizes, I've heard, so it might be worth it to keep it at half voltage for a while to see how it acts.

For the first probe I'm just going to wrap the pmt and scintillator (optical coupling grease between? not solved yet) completely in self-vulcanizing tape.

Once I've verified operation, I'll get some 100x100mm aluminium tubing and make a proper probe with everything wrapped in foam.

Have a look at this link for proper design and usage of PMTs and base wiring according to intended usage:
https://psec.uchicago.edu/links/Photomultiplier_Handbook.pdf
It is a copy of the venerable RCA classic about photomultipiers. It has pretty much most of the the answers to your questions.
Well worth a complete read, even if you are familiar with PMTs.

Avoid UV light (fluorescent or white LEDs) as dark current might be high for a few days. For low intensity (photon counting) I alway wear gloves to avoid oil from my hands that could create a leakage path on the base/electronics.


I did come across it myself, but yeah, it's a pretty great reference! Thanks!

--Chris
--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline richnormand

  • Supporter
  • ****
  • Posts: 721
  • Country: ca
Re: Photomultiplier - soldering directly on tube pins?
« Reply #15 on: February 09, 2020, 03:50:56 am »
@ChristofferB
You mentioned optical coupling in the above post. Yes you need it otherwise the acceptance angle for photons from the scintillator to the PMT surface will be severely limited.

For optical coupling between several 5" PMTs with plastic and 3" PMTs with xtal scintillators I have used:

Temporary: for testing up to a month: KY Gel from the pharmacy.... works quite well, good index of refraction, and water soluble so it is easy to clean 8)
Usual stuff: remote control cars differential oil. Clear and viscous. Good index. Get the highest viscosity you can get. Got mine from hobby and RC shops. Been there for ten years on my two muon detectors. Make sure you have a strong adhesive tape around the scintillator and PMT to avoid leakage over time. :)
Better: is a pro index matching gel. The real stuff is about $70 for a few cc from the likes of 3M and such. This might be a better cost alternative, as an example, https://www.ebay.com/itm/Silicone-Optical-Coupling-Compound-for-PMT-Photomultiplier-Scintillator-Detector-/261918290249 but I have no idea how good it is compared to the industrial stuff.
« Last Edit: February 09, 2020, 03:57:03 am by richnormand »
Repair, Renew, Reuse, Recycle, Rebuild, Reduce, Recover, Repurpose, Restore, Refurbish, Recondition, Renovate
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #16 on: February 09, 2020, 04:12:10 am »
Ha.. KY gel.. how great it would be to get that into a published journal  ;D

I suppose any viscous and clear liquid could really be used, right? Glycerine or microscopy immersion oil might work as well, optically wise, but both probably a little too liquid.

I wonder if the diff oil is a silicone oil or a petroleum based one? Maybe I'll check out my local hobby store.

Yeah I know the Iradinc guy on ebay. I was actually considering getting his complete kits, since that humongous heat shrink he sells is pretty ideal for light proofing.


I'm still working on how to best make a counting setup, since just building a lead castle is a little much. I can't find any sources to whether the entire PMT should be lead shielded or just the scintillator. Either way, a lot of high end scintillation probes also has a mu-metal shield, so maybe I should try making the probe in a 'scope CRT mu-metal shielding tube?

thanks for the advice!


 
--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline profdc9

  • Frequent Contributor
  • **
  • Posts: 334
  • Country: us
Re: Photomultiplier - soldering directly on tube pins?
« Reply #17 on: February 09, 2020, 04:49:32 am »
If the pins are small enough, you might be able to use Nixie tube pins and make a socket PCB to solder them into.  For example:

https://www.ebay.com/itm/100-pcs-NIXIE-TUBE-Socket-Pins-Gold-Plated/153513091292

You can design a PCB that could have holes for the pins and solder the pins into the holes.   Then you insert the tube into the pins.   You could leave a hole in the PCB for the protrusion in the bottom.

Also, maybe plumber's silicone grease would work as a temporary index matching gel if it was a thin layer?
« Last Edit: February 09, 2020, 04:51:44 am by profdc9 »
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15156
  • Country: de
Re: Photomultiplier - soldering directly on tube pins?
« Reply #18 on: February 09, 2020, 10:17:35 am »
Some magnetic shielding of the PMT may be a good idea, but it depends on the tube. I got away without any magnetic shield.  I don't think one would really need Mu-metal. Something like transformer steel should be good enough for most cases. A little fixed residual field should not be a problem. The shield from an old analog scope should be well good enough, if it fits.

The PMT does not directly needs a lead shield, however as the scintillator would be open to the PMT side otherwise the lead shield should extend somewhat over the PMT, up to fully covering the PMT. The shielding part would probably be one of the late stages, but it can help to keep the housing of the PMT small and round.

For the first tests one can even get away without index matching grease - this is more about the last 10%. A thin PE foil could avoid direct glass to glass contact.  I would do the first test with a LED dummy anyway.
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #19 on: February 09, 2020, 12:49:37 pm »
I agree, shielding is secondary, and can always be added later in layers if it turns out to be needed.

A 6 cm long cylinder of copper sheet with a back plate for connectors that can be slipped over the back end of the tube to shield /protect the divider is probably the first thing I will add, also for an electrical safety point of view.

For pmt coupling, I also found some light curing optical coupling glue used for fiber optics and lens assemblies. This would probably be ideal since it holds the xtal onto the tube by itself, and also provide the desired coupling. Probably expensive though.


--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #20 on: February 09, 2020, 02:27:33 pm »
Just found the perfect shielding for the pins/divider - the inner tube of a stainless thermos! There is even room for ~4mm foam padding around the tube!

I'll need to cut out a 40ish mm hole in the bottom for HV and SIG connectors, then bolt it on the back but I think this is pretty much ideal!

--Chris
--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15156
  • Country: de
Re: Photomultiplier - soldering directly on tube pins?
« Reply #21 on: February 09, 2020, 02:42:50 pm »
Stainless steel does not provide a lot of magnetic shielding. So I would prefer normal steel.

For optical coupling I would start with something that is easy to remove, so more like silicone grease, Vaseline or just a plastic foil as a cushion.
 

Offline ChristofferBTopic starter

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: dk
  • Chemistry phd student!
    • My channel:
Re: Photomultiplier - soldering directly on tube pins?
« Reply #22 on: February 09, 2020, 03:04:51 pm »
This is not meant as magnetic shielding, this is just to hold / protect the electronics/user. A magnetic shield could be slid on top. My machining/fabricating capabilities arent the greatest so a favourable shape material to start out with is pretty important.

I think I will go with cling film as a cushion at the very beginning, just to verify that the tube actually works!
--Christoffer //IG:Chromatogiraffery
Check out my scientific instruments diy (GC, HPLC, NMR, etc) Channel: https://www.youtube.com/channel/UCZ8l6SdZuRuoSdze1dIpzAQ
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2851
  • Country: us
Re: Photomultiplier - soldering directly on tube pins?
« Reply #23 on: February 09, 2020, 08:53:51 pm »
The problem with negative HV is that anything earthed or conductive MUST be kept physically away from the tube, especially the window. Since scintillation crystals usually come in aluminum cans this is impractical for these.
Well, actually, this is VERY rarely any problem.  The window is not a problem, anything that is transparent to light is likely to be non-conductive, too.
Generally, the sides of the tube are not much affected by external fields, and usually PMTs are placed in Mu-Metal shields to exclude magnetic fields which can affect the electron paths.

I MUCH prefer the negative HV method for safety reasons.  Also, the 3 KV or whatever caps are somewhat difficult to find.

Jon
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2851
  • Country: us
Re: Photomultiplier - soldering directly on tube pins?
« Reply #24 on: February 09, 2020, 08:55:57 pm »
Some magnetic shielding of the PMT may be a good idea, but it depends on the tube. I got away without any magnetic shield.  I don't think one would really need Mu-metal. Something like transformer steel should be good enough for most cases. A little fixed residual field should not be a problem.
Transformer steel would be bad, as all soft iron is very susceptible to aquiring a permanent magnetization.  Mu Metal is good in that it is much less
susceptible to getting magnetized.

Jon
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf