Has completely taken over from neet PMT

Forum: Analog electronics and circuit technology PMT power supply question





Since I want to do gamma spectrometry, I need a high voltage source with low voltage fluctuations. Ready-made power supplies for photomultipliers are usually very expensive. So I want to build something myself. After some searching, I came across a possible solution on the Linear website. Page 5 http://cds.linear.com/docs/en/application-note/AN118fb.pdf Now my question is, there are -1000V produced, but by changing the rectifier and LT1006 one should be able to achieve a positive voltage? What do you think in general about the solution? Is it good? Are there any better ones?


Uiiiii ..... I think there is a little bit of theory missing. Let it be, it won't work anyway.


Oh D. wrote:> Uiiiii ..... I think there is a little bit of theory missing. Let it be> it won't work anyway. So buy a ready-made power supply from Hamamatsu or something similar? Unfortunately, it costs around 100 €


> Unfortunately it costs around 100 €. But you don't think you can make it cheaper? With a breadboard or something similar is nothing. For 100E, buy.

: Edited by user


Do not get discouraged! The circuit is the Royer converter currently quoted on the homepage. For positive voltage you have to turn the rectifier so that the anodes are at ground, i.e. mirror vertically. You switch the resistor divider to GND and you connect the tap to the FB pin.


Oh D. wrote: >> Unfortunately it costs around 100 € >> But you don't think you can make it cheaper? With a breadboard or> similar is nothing. >> For 100E, buy. Well, the parts for it cost maybe 20 €. I have to have a print made for the evaluation electronics, so the print does not cost me anything extra.


Are you sure your PMT needs positive tension? This is quite unusual, most PMT want a negative HV, this has the advantage that the signal output is on GND and not on 1000V.


Thomas Z. wrote:> Are you sure that your PMT needs a positive voltage? This> is quite unusual, most PMT want a negative HV, this> has the advantage that the signal output is on GND and not on> 1000V. Yes, I had already asked the detector supplier about that. That's exactly my problem, because the negative HV supply is available from 40. € :)


TrollHunter wrote:> Do not be discouraged!> The circuit is the> Royer converter currently quoted on the home page.> For positive voltage you have to turn the rectifier around so that the> anodes are to ground, i.e. mirror vertically.> The resistor divider you switch to GND and the tap> you close to the FB pin. So it doesn't need the LT1006?



Whoa boy, cut the shit! What you are up to is terribly illegal and dangerous! Please don't radiate your neighbors! http://www.gesetze-im-internet.de/bundesrecht/strlschv_2001/otal.pdf If someone notices what you are doing, then you have a really big problem. But most likely you will only radiate your own eggs, so go ahead and do it ...


Jemin K. wrote:> Whoa boy, leave that shit! What you are doing is terribly illegal and> dangerous! Please do not radiate your neighbors! >> http://www.gesetze-im-internet.de/bundesrecht/strlschv_2001/otal.pdf >> If someone notices what you are doing, then you have a really> big problem. But most likely you will only radiate your own> eggs, so go ahead ... haha ​​... you don't seem to know what gamma spectrometry is, it doesn't generate any radiation! https://de.wikipedia.org/wiki/Gammaspektoskopie With this one can measure the "composition" of radiation. You can then see which isotope is causing the radiation (e.g. pointer of an old clock, radon) or (glow stripe, thorium. The high voltage source needs the tube to work! First inform before yelling! PS Since I live in Switzerland, I'm interested the DE radiation protection law one way or another not me.



The PMT should really need a negative voltage. This is the only way to have the output against ground. You should read that again and check it critically. If the power loss is not so critical, you can also regulate linearly instead of the switching regulator. That should give a little less interference and could also be built on the breadboard. For only 40 EUR you can probably buy the negative supply ready-made.


The finished detector is built into the R980 PMT from Hamamatsu: http://www.hamamatsu.su/pdf/etd/r980-01.pdf The data sheet says "max 1250V" and not "-1250V". The Detector has a BNC connection, the resistors are already built into it. So you just have to connect HV and the signal. Since the thing cost me a few 100 €, I really don't want to mess it up and screw it around. The seller sent me pictures of how the part works and how it can measure with an evaluation unit that delivers HV +. Therefore the part needs + HV! [Quote from * seller:] ************ As far as the Hamamatsu C4900 is concerned, that specific HV module puts out negative HV (- HV), so you would need the C4900-50, C4900-51 or similar for positive HV (+ HV) output. *************** [and according to manufacturer's website (Scionix):] *** It is possible to operate a photomultiplier tube in two ways: A. anode at positive potential (cathode at ground) B. anode at ground (cathode at negative potential). For measurements of DC anode current such as in some X-ray applications, option B is the only choice since in the first option the anode must be separated from the follow-up electronics by means of a high voltage capacitor. On the other hand, option A is used for most standard applications since the m-metal shield should be preferably at cathode potential. Option A implies that cathode, detector mass (ground) and shield are all connected together. In option B, the shield must be very well insulated from the detector mass and special construction requirements apply. ***

: Edited by user


Have a look here: http://www.theremino.com/wp-content/uploads/2012/02/PmtAdapter_V3.3_SCH_NoText.jpg A R980 is running on it. > The Detector has a BNC connection, the resistors are already built into it. So you just have to connect HV and the signal. Take a look at the further documentation under Theremino, w.W. the dynode resistances are changed there, i.e. increased in relation to the standard circuit.



The fact that 1250 V and not -1250 V are specified for the PMT has nothing to say. The voltage between the 2 parts of the tube is given. The question is how the housing is connected. The anode to ground is of course also possible. You should see that in the output signal. There would then have to be a transformer or coupling capacitor for the output. The circuit above with the LT1534 could possibly have a problem with drift: the LT1534 does not have such a precise reference and also has a relatively high input current. For the spectrometer application, the voltage should be very stable. I would prefer a linear solution.



Johnny S. wrote: >> For 100E, buy. >> Well, the parts for it cost maybe 20 €. I have to have a print made for the> evaluation electronics, so the print does not cost me anything> extra. I don't know what hourly wage you calculate your working hours with. The 80 € that you save at 0 € / h wages are only a fraction of the costs of the PMT anyway - at least if it is from Hamamatsu ;-) Heinzinger also offers finished HV modules. N / A, like the currently are priced. http://www.heinzinger.com



Lurchi wrote:> The circuit above with the LT1534 could possibly have a problem with> drift: the LT1534 does not have such a precise reference and also has a relatively large> input current. For the spectrometer application, the voltage should be> very stable. I would prefer a linear solution. There are several implementations of the basic concept in the AN-118; For example, you can also feed the Royer Converter directly with a linear current source (such as an N-FET or NPN, more like an N-FET here).


In the meantime I was able to purchase a finished HV module from someone here in the forum.


Johnny S. wrote:> In the meantime I was able to acquire a finished> HV module from someone here in the forum. I wish you good luck with it! It is unbelievable how many hollow heads this topic has brought up here. For those who did not yet know: In this application, the output signal of the PMT consists of pulses that can be coupled very well to the evaluation circuit via a capacitor. With the anode on + HV it is possible to connect the photocathode to ground, because this is where the scintillator is usually glued on, which comes into contact with the shield and possibly also the sample. This results in less noise than if the photocathode is on -HV and the thin glass of the tube has to take over the insulation. In addition, the photocathode can be damaged by the - albeit low - insulation current. P.S .: Johnny S. wrote:> Since the thing cost me some 100 €, that is not little. Better to ask in advance, because some people may still have suitable tubes lying around. Incidentally, I don't think the R980 is a very lucky choice, because this PMT still has considerable sensitivity in the red, while the scintillators usually emit their light in the blue or even UV. Red-sensitive cathodes have a noticeable thermal emission even at room temperature, i.e. higher noise than blue-sensitive ones. The choice of the photocathode naturally also depends on the scintillator used. P.P.S .: The cathode of the R980 is not as bad as I thought. In any case, it has a lower dark current than the Sb-Cs cathode I was thinking of.

: Edited by user

In terms of circuitry, I would monitor the output current of the PMT with an optocoupler. Your pulse evaluation does not see such a DC signal. If anode currents of 20µA or more occur, then something is super rotten, probably incidence of external light. Then it is time to switch off the high voltage until the error has been rectified so as not to endanger the PMT.


Hp M. wrote:> Johnny S. wrote: >> Since the thing cost me some 100 €, >> That’s not little.> Better to ask in advance, because some people may still have suitable> tubes lying around.> The R980 I don't think it's a very lucky choice, by the way, because> this PMT still has considerable sensitivity in the red, while the> scintillators mostly emit their light in the blue or even UV.> red-sensitive cathodes have a noticeable> thermal emission even at room temperature, i.e. higher Noise as sensitive to blue.> The choice of the photocathode naturally also depends on the> scintillator used. I meant the detector, not the tube. So a complete unit with crystal, tube, resistors, mu-metal shield from Scionix. The detector is still on the way to me, but looks similar to this one: https://c2.staticflickr.com/8/7442/11665010476_395be6f119_b. jpg Mine simply has a BNC on the back and is completely silver.



I used the first circuit specified in the AN-118 from Linear for a Dekatron power supply and the whole thing works perfectly. The advantage of the specified transformers is their availability. 800-1000 volts is not a problem, certainly not even more with suitable diodes. However, the proposed double diode cannot absorb these voltages. Noise is really very low and should also be well suited for PMT applications.


click clack wrote:> Noise is really very low and should also be> well suited for PMT applications. The main problem with this application is the constancy of the high voltage. The HV must only drift very little, especially during the measurement period, because this shifts the entire spectrum. The long-term constancy, however, is not that critical, because you have the option of calibrating with a test source before starting the measurement.

Write an answer

Providing an email address is voluntary. If you would like to be automatically informed by e-mail about replies to your post, please log in.

Important rules - read first, then post!

  • Use upper and lower case