Strip-board layout for detector for SNM 18-1

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

Strip-board layout for detector for SNM 18-1

Postby chrismb » Sun Sep 04, 2011 2:10 pm

Been off line for a few days, came back to laying out the bread-board circuit, that worked for the 18-1 signals, into strip-board.

After 7 re-iterations, here it is:

N-detector-layouts.gif


I keep spotting an error, going back, making corrections, make new errors! Any 'de-bugging eyes' would be very welcome on this, I've spent several hours now, and I am now going cross-eyed over it!

Columns 2 to 8 contain the feedback amp, with Doug's values in but the feedback resistors modded to suit the pnp I have in my bin and give a high input value, which seems to work well for the 18-1.

I've added some 'earth' lines vertically between an upper and lower 0V rail. They separate each 'stage', just to help minimise feedback, FWIW, first one being between column 8 and 9.

The regulator (note, I will be using a LM2931 here, because I found the 7805 sucks up 4mA doing nothing at all, and the whole circuit takes under 2mA) and the comparator stage then go from column 9 to 18, with another 0V line running at the end of that.

The HEF40106 Schmitt trigger is to clean the signal up, as I have put in an 'anti-bounce cap' (in column 19) so the signal then needs 'cleaning'.

I also tested the HEF40106 to see whether it used less power idling with its unused inputs held high or low, or floating. I concluded they should be held high.

I've yet to deal with putting in the trimmer thresholding resistor (column 12) which will need another column or two to make it fit the holes, but otherwise I think I've made the most of the board space!! I've doubled-up on some of the holes to save a little space.
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Strip-board layout for detector for SNM 18-1

Postby Doug Coulter » Sun Sep 04, 2011 2:30 pm

I'm unfamiliar with strip board, but if it's one like those kludge boards with a lot of built in conductors, you'd do one heck of a lot better with a careful perf board layout, and not have the leakage and stray capacity those have. With this much gain and the high input impedance, you're really flirting with making an oscillator here, perhaps a nicely complex one where the feedback is from current spikes drawn by the chips back into the front end over the power wiring, or just mutual inductance in the wiring...

One issue we had with these here is just a tiny amount of corona out in the air (as in off the 40 meg resistor or tube cap) is enough to make it false-pulse. I wound up encapsulating that in quartz and paining the HV part of the connection to the tube with corona dope to put a stop to that. I wound up painting the inside of the shield can as well. Even so, with 1 screw loose out of the 5 holding the can down to the flat copper top of the tube/moderator assembly, I found that it didn't take much EMI to make the thing count with no actual neutrons present. This really needs to be in a very tight shield -- millivolt sensitivity on high impedance means that if a scope probe nearby picks up anything you can see, so will this. The response time of the tube has zero effect on the response time of the preamp for noise, BTW. You can take away some of the HF response (goes to around 50 mhz as is) by putting a tiny capacitor across the NPN collector resistor -- extra negative feedback at high frequencies, shouldn't need to be too big - some pf's.
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: Strip-board layout for detector for SNM 18-1

Postby chrismb » Sun Sep 04, 2011 3:11 pm

That's why there are capacitor breaks right across it, and guard rails at 0V both sides. I have aimed to set it up so that the strip carrying the 'signal', at any point L-to-R, is surrounded with a capacitively coupled DC rail.

I figure it can't be worse than a bread-board, and it worked on the bread board. Worse was to come - I recoupled the output of the Schmitt trigger back onto the HV cable itself (downstream of the 100MOhm resistor), so that it would look like a 'GM tube' type response from the outside, and it worked fine, no feedback upstream.

Incidentally, in a 'fully closed' battery powered system, I have skipped on the transient diodes. Do you think there is any good reason I should include them?
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Strip-board layout for detector for SNM 18-1

Postby Doug Coulter » Sun Sep 04, 2011 3:21 pm

Well, should the HV coupling cap, which has a fair amount of stored energy in it, get shorted at the tube end (for example, the tube arcs or shorts internally), all that energy will blast the BE junction of the first transistor in the preamp, which might not live through that. That's why I put them in. I have also run without them (use 1n4148 if you use any) but that's tempting fate a little.

Capacity to ground on signal lines may not be good at all. If it's across, say, the emitter resistor of the first transistor, it will increase gain at super high frequencies, for example -- it lowpasses the negative feedback in that case. In other places the phase shift induced by the RC might cause other issues.
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: Strip-board layout for detector for SNM 18-1

Postby chrismb » Sat Sep 24, 2011 11:02 am

Well, after dozens of hours of planning and replanning (measure many times, cut once?) I finally built this. (Zzzz... what a dull chore...!!)

P9240801crop_s.gif


I know it's quite simple, but I'm no frequent builder of these things, so I'm very content with my effort to get pre-amp/discriminator/output-buffer all condensed down into a space of 7 x 25 holes, despite it taking several days. (Is it just me, or does it always seems to take an eternity making stuff like this when you nibble a few hours out of several days and you think back and say 'gee, that took weeks!!.)
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Strip-board layout for detector for SNM 18-1

Postby Doug Coulter » Sat Sep 24, 2011 3:01 pm

Looking good! I just use perf board, with or without plated pads, myself, and it goes pretty quick usually. I did the preamp design on the hornyak thread this morning in about 2 hours that way, including rustling around for the parts. But I do this a lot, so I got fast at it. Sounds like you might have laid out a pc board in the same time. If I ever got fast at that, I'd almost always do a PCB, particularly once you get up to IC's. Believe it or not, they are actually easier to make minor changes to than these techniques are. Tracks cut easy, wires are easy to add, extra parts can be "air wired" and so on. But with a perf/solder build, the bent over leads (which are the "tracks") make changing components hard later on - and "cutting tracks" is pretty hard to do cleanly.

That should be way sensitive enough for anything making some real neutrons. A version here counts ~350 cps or so on about 1m neuts second (with the moderator right on the tank)...so you'll see this get out of background with high confidence if you're making any amount of neutrons. The Hornyak I just finished is only about 16.7 cps on the same source -- but it's truly compact, and needs no moderator. That one eats more power to run, too, as I didn't do the multiplier-tap trickery on the phototube dynode chain (going to have to try that soon, RCA thought of it pre WW-II).
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 1 guest

cron