ebay crate - scint detector

This is bound to get mixed up with things in Electronics, check both. Physics-specific stuff here, mostly.

ebay crate - scint detector

Postby chrismb » Fri Aug 19, 2011 10:19 am

I was idling on ebay a few days ago and grabbed something one of my 'regular' ebay vendors was selling;

crate3.gif


Not tried it yet, I'm told it appeared to work OK on a 400V GM tube.

I bought it mainly because of a post by Alex Klotz on fusor; http://www.fusor.net/board/view.php?bn= ... 1308449638

He says he plugged his CNM 18-1 tube into an Ortec and seemed to have no problem running it.

However, I am not familar with this kind of gear and would like to ask what sort of 'tube load' this is expecting to see. I mean, the corona discharge tubes run with a few 10's uA running through them. What is on the front end of that 'HV' port, is there a resistor and a signal is taken off the port-side of that resistor, does that mean I should add a resistor-and-cap in series to the corona tube? The kit looks good and the vendor has been reliable before, so my only worry now is that I don't blow it up somehow through my ignorance. (Which included never seeing this plug type before - PET 100 - never heard of it before a few days ago).

So, what means of connection d'y'all think I should make between 18-1 and this to get the best out of the tube? (Esp @ Doug, as you've worked with these corona tubes now)
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: ebay crate - scint detector

Postby Doug Coulter » Fri Aug 19, 2011 12:15 pm

I would just use this for the sort of thing intended, it's not going to work on corona tubes that produce millivolts across megohm loads. For those you need a high input impedance, low noise, high gain preamp to get to the volt level. A geiger tube or a phototube gives those volt levels already (in the case of geiger tubes, tens or hundreds of volts - we get 150v pulses out on the standdard counter here). A photo tube might give volts to tens of volts into maybe 10k -- which is what this was designed for most likely. The phototube will also draw some milliamps of current for the divider chain, so that power supply can hack that for sure.

But you're not going to get enough level out of a corona tube, and I didn't see an adjustable gain and threshold, which are needed for those, since they do put out a lot of noise all the time, and you have to threshold that out to get a good count signal. In other words, the SNR of the corona tube ain't that great and you have to reject stuff below a tight threshold, and only accept stuff that's above that to get a reliable count with one. Even then you'll have some errors from noise pileups once in awhile making false counts, and noise subtracting from a marginal neutron hit making you miss a few.

If you had separate power output and signal inputs, you could use it with that two transistor preamp design, though. Joe is working up a board for that one now, along with the standard counter work he's doing, and we'll have them available sometime soon if you don't want to just build one on perfboard as I did. It's a pretty cheap and easy build.

Edit: Checking that thread, yes, he plugged his into an Ortec, designed for gas proportional tubes (weak signal neutron detectors specifically), and of course it worked. That's not what you have here...
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: ebay crate - scint detector

Postby chrismb » Fri Aug 19, 2011 1:26 pm

Fair enough, I thought that might be the case.

Nonetheless, is there any harm in trying it, and/or what should I do to prevent any harms either way to this from any tube I might put on it? Is the HV/detector output short-circuit protected, or if not then what is the internal resistance likely to be that makes this thing 'tick'.

I was imagining that if you run the output via a resistor with a small cap across it, then you limit the current by that resistor yet if the tube 'fires' and you get a potential drop, the cap will pull the detector line down, giving you a count. That way, you can adjust that resistor until you get the signal you want (that is, on any unknown tube that you are probing), thereby avoiding excess load on the HV drive.

Is it not safer to do something like this with an unknown tube and unknown supply?
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: ebay crate - scint detector

Postby Doug Coulter » Fri Aug 19, 2011 1:46 pm

chrismb wrote:Fair enough, I thought that might be the case.

Nonetheless, is there any harm in trying it, and/or what should I do to prevent any harms either way to this from any tube I might put on it? Is the HV/detector output short-circuit protected, or if not then what is the internal resistance likely to be that makes this thing 'tick'.


You could probably put any tube it was designed for on it. You might need additional series resistance for some types (geigers and others) that need a current limit and make big signals anyway.
I have no way of knowing what the internal series R is (heck, the connector for the input isn't even in the picture?) That's what you ask the guy you bought it from about, or google for the manual and schematics. The ones I've looked at use 10k or 100k. That will instantly fry most corona or geiger tubes on the first pulse. The corona tubes need a huge current limit series impedance, they reccomend 40 megs or so, way way more than is going to be in that box, likely.

chrismb wrote:I was imagining that if you run the output via a resistor with a small cap across it, then you limit the current by that resistor yet if the tube 'fires' and you get a potential drop, the cap will pull the detector line down, giving you a count. That way, you can adjust that resistor until you get the signal you want (that is, on any unknown tube that you are probing), thereby avoiding excess load on the HV drive.


If you put a cap across a series resistor, the impedance becomes a short circuit at HF (like a tube pulse) and makes the resistor no longer a current limiter for fast signals, like all tubes produce.
(the pulse is the equivalent of a few mhz during the risetime)
So you don't limit tube peak current -- it can still draw plenty while charging the cap -- and get no benefit from the resistor except at low frequencies, which don't exist in this situation. There isn't that kind of "free" in electronics. The corona tubes like to see about a 40 megohm impedance, though with a small enough cap, you can use 40 megs in the DC supply and about 100k preamp input impedance. Say 47 or 100pf in that case, they can take the extra current it takes to "charge" one that small. But you still have a tiny signal, so you still need a preamp unless you can set that thing for millivolt pulses and I didn't see any knobs for that on the front. Looks like it's looking for volt-level stuff like a scintillator would make into 10k or so. However, with less effort than typing the question, you could hook it to a signal generator through a small capacitor and see what it takes to count it.

Teach a man to fish....

chrismb wrote:Is it not safer to do something like this with an unknown tube and unknown supply?


You seem to not understand some very basic electronics, as your question implies that you think a cap will pass the signal (so it doesn't attenuate like a resistor alone would), without passing the current it takes to make the signal! You can have one, or the other. Not both.

Reactance (works like resistance for computing currents and such) is X = 1/2*pi* F * C (c in farads). A swift risetime is a high frequency component, in this case in the mhz range most likely. Just like with a current limiting resistor -- you need to allow some drop to protect the tube, which kind of negates the signal pass through function of the cap. You just have to get the gain some other way (and in any case, these only put out a few mv and even if you got it all into your box, its unlikely to be enough to make it count.) As mentioned above, these are noisy tubes, so you also need some sort of settable sharp threshold to separate the baseline noise from the real signal pulses. I see no knobs to set up that kind of stuff here.
We've documented a dirt cheap working circuit for these here already...you can mod this to look like that, or build what we know works already.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: ebay crate - scint detector

Postby chrismb » Fri Aug 19, 2011 1:55 pm

Doug Coulter wrote:I have no way of knowing what the internal series R is (heck, the connector for the input isn't even in the picture?)


We might be talking at [two lots of] crossed purposes, here. The 'HV' supply is the connection to the tube. The detection is within the HV supply.

In regards the cap, I'm suggesting to use the cap to hold the current through the resistor steady (viz. no big swings for the cap), so that when the tube fires it pulls the cap and resistor down together.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: ebay crate - scint detector

Postby chrismb » Fri Aug 19, 2011 1:58 pm

Doug Coulter wrote: However, with less effort than typing the question, you could hook it to a signal generator through a small capacitor and see what it takes to count it.

Actually, I didn't think of doing that, so typing the question was useful. ;)
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: ebay crate - scint detector

Postby Doug Coulter » Fri Aug 19, 2011 2:29 pm

Re your first response - in that case the tube still sees whatever peak current the power supply provides through it's own series R, with no additional current limiting taking place in your added series R except for DC (where you don't really need it most likely). The cap acts like a short circuit during the pulse if it's big enough to convey the pulse to the rack unit. So during the pulse, it's as if any additional current limiting you've added with a resistor just doesn't have any effect -- so the tube tries to short out the stiff power supply (10k is very stiff to one of these tubes that want a 40 meg impedance) and will be damaged.

Most times these will have a fairly low series R built in. Depending on if it was for a two wire connection to a phototube and divider string, it's going to be in the 10k range, to limit the DC drop due to the phototube DC requirements. It's way not enough to limit the current to the few microamps the tube needs when it "fires". Too much current messes up the tube, or it may not be able to "shut off" the internal discharge fast enough to prevent damage (heating on the 1 mil wire, or excess dissociation of the quench gas). If you bypass your external resistor with a cap you lose its ability to limit peak current during that fast risetime, and might as well leave the whole mess off there, except for limiting DC current. I seriously doubt it would work with the 40 megs in series alone, as the resulting pulse across the likely 10k load in there would be in the nanovolts and its very unlikely they put that sensitive a preamp into something where it would only be a liability. You can of course, just open the thing up and trace out the circuit and not have to guess. I can't do that for you from across the pond.

Remember, with infinite short risetime, any finite cap takes infinite current to charge it. It's a short circuit. The risetime on these is a fraction of a uS. The Russians say use 40 megs...presumably they know their own stuff. We used a 120pf cap to couple into our preamp, which seems to be small enough -- but you only get a few millivolts during a pulse (even into a scope proble 10 meg load) which likely isn't enough to make this thing count.

Yeah, use a signal generator to test the thing, it's the best way, especially if you can make square-like pulses with one. But this is quite unlikely to work the way it is, you likely need something else entirely. Not built for this purpose, but for a very different one, after all. This was meant to see many-microamp pulses on top of a milliamp of phototube DC current. The tube was meant for a few uamps of DC and makes only a fraction of that more during the pulse.

You could just use this power supply, with a 40 meg series R, and a small cap to a scope probe to test the tube and ignore the counting function. Or modify it to have separate power and signal inputs, and use the HV to power the suggested circuit, then use the preamp to drive the signal input of the counter. That will very likely work. Or wait a little -- our standard counter will have a spare input that will work with our preamp -- and a ccfl is a trivial way to make tube power for these. Or you could simply use this thing with the sensor it was intended to be used with. It's not a match to a corona tube -- not even close. Might was well use a slot car DC motor to try and push a Camaro. Scales involved are all wrong here.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: ebay crate - scint detector

Postby chrismb » Fri Aug 19, 2011 5:02 pm

I've already tried powering these tubes. I've tried endlessly and could see nothing in the signals. It cost me about 2 months of 'project time' last year when I paused my efforts to try to make the things work, and I gave up. I've got as many HV supplies as I can shake a stick at, so I was hoping that a detector being integrated and at the end of a coax that it'd be screened from the incessant RF noise that seemed to be all I was picking up. There again, I might've already had these tubes working all along off my supplies but as I've not got a known neutron source it's tricky to tell!! As Mr Hull wisely pointed out, in not so many words, if you've got an unknown tube go carefully with it, if you've got an unknown detector with a known tube you're in with half a chance to set the detector up right. My best efforts at circuits count as unknown tubes with unknown circuits! All I figured out by myself (with zero info on them then) was that they were drawing a continuous current which, at the time, made no sense to anyone when I asked how to use the tubes, until all this 'corona discharge' stuff has come out recently.

It makes sense that I'll not be using this kit with the corona tubes, but I'll check it out with a sig gen and see if I can figure it out that way.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: ebay crate - scint detector

Postby chrismb » Sat Aug 20, 2011 6:23 am

I've tested this unit with a square wave, and I'm pleased with it!

I coupled the detector to a representative tube but did not apply voltage. I then put a 50cm wire (no shielding) from the sig gen to a 100pF cap, then another 50cm of wire to the 'anode' of the tube. The purpose of this long wire and weak cap was to create a signficant amount of noise.

My scope was connected across the tube, i.e. where the detector was also coupled to the circuit.

The background 'fuzz' registered as around 50mV, top to bottom. The counter counted nothing at all. I switched on 1MHz at 2V indicated on the sig gen knob, with -40dB. The 'fuzz' rose to 20mV on top and the scaler counted 1MHz. I limited the sig gen output so that there was only 15mV apparent on top of the 'fuzz' and the count rate underestimated by a factor of 10. Dropping further, and the counter stopped.

So the answer is - it appears to be able to count at least 20mV at 1MHz on top of 50mV noise.

...update... so I did decide to try it with the russian tubes, but it doesn't seem to be quite enough. So I'll be needing those preamps, guys!!....
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: ebay crate - scint detector

Postby Doug Coulter » Sun Aug 21, 2011 6:14 pm

You'll probably want variable gain too, but that's easy -- all you need is a pot on the output.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA


Return to Metrology

Who is online

Users browsing this forum: No registered users and 2 guests

cron