Electronics > Projects, Designs, and Technical Stuff
"Tail" pulse generator // particle detector simulator
ChristofferB:
Hey all!
As some of you may know from the previous threads I've been pestering you with, I'm messing around with a few gamma spectroscopy, cosmic ray detection etc. projects.
One universal tool I'm lacking is a detector simulator (as calibration with a real detector requires a fairly strong monoenergetic gamma ray source, which is not really an option).
Basically, what I need is a pulse generator to simulate particle detection events; sharp rise, slow decay, entire pulse duration somewhere between .5 and 10 ns.
Pulse looking something like this:
I'd like the pulse event to be triggerable by TTL pulse, so that the frequency can be set fairly easily with a standard function gen (or simple square wave osc).
How would one go about designing something like this? I know the old-school way is to discharge a capacitor through a resistor, switched with a mercury wetted relay.
If I can find one of those that might still be a possibility.
Another option is to just couple a narrow square wave pulse into a charge sensitive amplifier via a small value cap, but I'm unsure if this will ever get very good performance.
Thirdly, I've read about people using a photodiode coupled to a charge amp, with a VERY dim LED flashing it briefly.
What do you think? Thanks in advance!
T3sl4co1l:
Wouldn't a reasonably fast (maybe not 74HC, but LVC or better?) logic gate have enough ramp rate, and a differentiator would do to generate the pulse? Probably needs to be unipolar, so something would be needed to clamp or control the opposite edge so it's at least not objectionable. (An open-collector output would be handy; 74LVC07 isn't too bad, but an RF transistor might do better.)
The differentiator can be tweaked, in general a pulse-forming network (PFN), for whatever waveform is desired. It should probably be attenuated or buffered to remove sensitivity to the load.
Variable pulse height could be controlled by varying supply voltage (over a reasonable range; a level translator or analog switch might be needed to cover a wide enough range), or variable attenuation.
Timing obviously can be generated from a Poissonian timing generator, to get realistic distributions as well.
Tim
David Hess:
T3sl4co1l beat me to it; AC couple a suitable fast pulse generator into a 50 ohm load, simmer, and serve. AC or LVC logic will be fast enough to require careful ground plane construction for a clean pulse shape.
If you need more than 5 volts, then things get interesting.
ChristofferB:
That sounds orders of magnitude easier than what I was thinking! Thanks!
I'll try a 74AC or 74LVC oscillator AC coupled into either my wanted load, or a shaping net/buffer.
Having some binary pi or T attenuators in series would give me a pretty big selection of pulse heights from 5V and down, right?
David Hess:
--- Quote from: ChristofferB on June 08, 2020, 11:44:37 pm ---Having some binary pi or T attenuators in series would give me a pretty big selection of pulse heights from 5V and down, right?
--- End quote ---
That is right. Like T3sl4co1l says, some easy pulse height control can be achieved by varying the supply voltage to the gate from 5 volts (actually a little higher) to 3.3 or maybe even down to 2 volts but this will affect the switching speed.
A simple alternative to varying the supply voltage is to use an open drain output like a 74AC125/74LVC125 where a resistor pull-up or pull-down (or a current source/sink) on the output sets the output voltage. This would allow outputs down to millivolts so a much larger range.
If you need more than 5 volts, then the same open drain output can drive the emitter or source of a higher voltage cascode RF transistor which could potentially get you to 10 or maybe even 15 volts but now developing enough current into a 50 ohm load will be the limitation. They used to make RF transistors packaged to support this configuration with two base connections and one emitter connection but availability was never good and is worse now.
But I suspect an AC or LVC gate operating at 5 or 3.3 volts driving a high pass RC filter, followed by attenuators as needed, will do exactly what you need. The coupling capacitance is low enough that a small trimmer capacitor will have enough range and make it adjustable.
Navigation
[0] Message Index
[#] Next page
Go to full version