SpinCore Radio processor USB board. Here is the manual of the board.
MHz).
fiers to amplify the pulsed signals generated by the RadioProcessor board?
Continuous wave amplifiers can run continuously at their rated power. If th e pulses of RF power that you want don't have a peak power that exceeds the continuous rating, you won't have any problem at all, but the continuous w ave amplifier may be more expensive than you actually need.
High RF power implies large voltage swings, so you won't be able to get mor e power out of the continuous wave amplifiers than their continuous rating.
Designing an amplifier from scratch that can deliver brief bursts of high v oltage and current is an expensive and time-consuming exercise.
There are people around here who might be able to help, but you would proba bly get more help if you specified how long you pulses need to be, and what sort of duty cycle you need.
Most RF amplifiers are designed around a 50R load to feed an antenna transm itting into free space - for ham radio and the like.
If the excitation coils for your NQR have a higher impedance - as seems lik ely, since you will want a solenoid into which you can put your sample cont aining some nitrogen-14 nuclei - this also affects what you might need and could get away with.
Den mandag den 26. december 2016 kl. 15.58.37 UTC+1 skrev snipped-for-privacy@ieee.org :
:
he SpinCore Radio processor USB board. Here is the manual of the board.
0 MHz).
Q
lifiers to amplify the pulsed signals generated by the RadioProcessor board ?
the pulses of RF power that you want don't have a peak power that exceeds t he continuous rating, you won't have any problem at all, but the continuous wave amplifier may be more expensive than you actually need.
ore power out of the continuous wave amplifiers than their continuous ratin g.
voltage and current is an expensive and time-consuming exercise.
bably get more help if you specified how long you pulses need to be, and wh at sort of duty cycle you need.
smitting into free space - for ham radio and the like.
ikely, since you will want a solenoid into which you can put your sample co ntaining some nitrogen-14 nuclei - this also affects what you might need an d could get away with.
the solenoid is generally matched to 50R and used for both tx and rx
you need a pulsed amplifier because it need to ramp up and down and be quiet in microseconds after tx
This isn't regular nuclear magnetic resonance spectroscopy but nuclear quad ropole resonance spectroscopy, which doesn't require a magnetic field. It o nly works for nuclei which have a quadropole moment, and nitrogen-14 does h appen to be one of them. You find nitrogen atoms in most explosives, so the technique is commercially interesting.
There are quite a few people interested in this question
formatting link
talks about using permanent magnets and long excitation sequences to get better signal to noise ratio's.
I'm wondering whether a simple class-D exitation might work.
The simplicity would strictly be in the analog/power part of the circuit - a PMOS power FET tied to a - say - +150V positive rail, and an N-Channel power FET tied to a -say - -150V power rail.
You'd use a cascode driver - more power FETs, possibly in cascode - to turn on the first FET from 30 degrees into the excitation sine wave to 150 degrees and the second FET from 210 degrees to 330 degrees.
Not very complicated.
The complicated - but not all that expensive - bit would be fast digital logic for setting the turn-on and turn-off times for both transistors.
The highest sodium nitrite line at 4.6MHz a period of 217.391 nsec. It has a lie width of 207Hz. The period of a 4.600207Hz cycle is 217.382 nsec so you are going to be wanting to set your pulse timings to better than 10psec.
ON Semiconductor do a chip that is almost equal to the task
formatting link
It does offer 10psec timing resolution, over 10nsec - actually from 2nsec to about 12nsec - so if you did your coarse timing from a 100MHz clock and used the MC100EP195 to generate your fine delays you'd be more or less okay.
In fact you'd want edges 72nsec apart to define a class-D approximation to 4.6MHz sinewave, so you'd probably go for a 200MHz clock and use the MC100EP195 to generate 5nsec fine delays.
The actual delay generated by the MC100EP195 is somewhat temperature dependent, so you might want an even faster clock.
The timing module could be done in ECL on a four-layer board with buried ground and -2V planes. More layers would make it easier.
My feeling is that there are some fast programmable logic chips around that could do much the same job in a single - rather expensive device.
comp.arch.fpga might be the place to ask about that.
Here in UW chem, our older NMR power amps are either ENI wideband amp (like ENI 350 or 3100,) or amateur radio "linear" broadband amps from Henry Radi o.
On ebay I see several used ENI amps for $1,500 and up. In the past I've s een these for under $200, but non-working with dead output transistors (so, buy and repair!)
fiers to amplify the pulsed singals generated by the RadioProcessor board?
Yes.
It appears that the main difference with "pulse" amplifiers is the added pr obe-protecting features, which shut down the output whenever detecting long pulses, CW, or PWM above ~20%.
An amateur radio CW amplifier under $1000 should work fine. (Some of those ham amp kits are under $100!) But it lacks the automatic protection. Jus t make sure to avoid accidental oscillation caused by old corroded BNC cabl es with bad shielding. That, or undergrad students who burn up your match networks with wrong software settings.
On Thursday, December 29, 2016 at 6:33:17 PM UTC+11, snipped-for-privacy@ieee.org wrot e:
:
he SpinCore Radio processor USB board. Here is the manual of the board.
0 MHz).
Q
lifiers to amplify the pulsed singals generated by the RadioProcessor board ?
better signal to noise ratio's.
- a PMOS power FET tied to a - say - +150V positive rail, and an N-Channel power FET tied to a -say - -150V power rail.
rn on the first FET from 30 degrees into the excitation sine wave to 150 de grees and the second FET from 210 degrees to 330 degrees.
logic for setting the turn-on and turn-off times for both transistors.
s a lie width of 207Hz. The period of a 4.600207Hz cycle is 217.382 nsec so you are going to be wanting to set your pulse timings to better than 10pse c.
to about 12nsec - so if you did your coarse timing from a 100MHz clock and used the MC100EP195 to generate your fine delays you'd be more or less oka y.
o 4.6MHz sinewave, so you'd probably go for a 200MHz clock and use the MC10
0EP195 to generate 5nsec fine delays.
ndent, so you might want an even faster clock.
ground and -2V planes. More layers would make it easier.
at could do much the same job in a single - rather expensive device.
There's nothing like sleeping on a problem.
It might make more sense to use something like an Analog Devices AD9913
formatting link
df
to generate a sine wave at up to about 60MHz - that's twelve times as fast as the highest NQR frequency needed - 4.952MHz for RDX - and use a programm able logic device (PLD) to generate the various waveforms needed to drive t he excitation source and the demodulators.
60MHz isn't all that fast. You'd need a fast comparator to turn the - low-p ass filtered up-to-60 MHz sine wave into a square wave to drive the clock i nput to PLD, and you'd need a bunch of outputs from the PLD to generate the excitation pulses and - a little later - the in-phase and quadrature demod ulation pulses.
207Hz line width on the 4.6MHz sodium nitrate line is 45 ppm, so you'd prob ably want to use a good quality and probably temperature controlled 25MHz c rystal oscillator to drive the AD9913 reference input (multiplied up to 250 MHz inside the device to keep the up-to-60 MHz output clean and easily low- pass filterable). The AD 9913 does seem to give you all the frequency resol ution you could need.
Yep. Cut my teeth in the late 60s on a Varian HA-60. Almost 45 years of NMR under my belt as "not a chemist" working for one or another of the major manufacturers. Still doing it. Had a gap for a few years in aerospace electronics and instrumentation but home has always been snuggled up to a big magnet.
ke ENI 350 or 3100,) or amateur radio "linear" broadband amps from Henry Ra dio.
seen these for under $200, but non-working with dead output transistors (s o, buy and repair!)
lifiers to amplify the pulsed singals generated by the RadioProcessor board ?
probe-protecting features, which shut down the output whenever detecting lo ng pulses, CW, or PWM above ~20%.
se ham amp kits are under $100!) But it lacks the automatic protection. J ust make sure to avoid accidental oscillation caused by old corroded BNC ca bles with bad shielding. That, or undergrad students who burn up your matc h networks with wrong software settings.
Thank you Bill very much for your reply.
Could you please check this power amplifier from RM Italy
like ENI 350 or 3100,) or amateur radio "linear" broadband amps from Henry Radio.
ve seen these for under $200, but non-working with dead output transistors (so, buy and repair!)
mplifiers to amplify the pulsed singals generated by the RadioProcessor boa rd?
d probe-protecting features, which shut down the output whenever detecting long pulses, CW, or PWM above ~20%.
hose ham amp kits are under $100!) But it lacks the automatic protection. Just make sure to avoid accidental oscillation caused by old corroded BNC cables with bad shielding. That, or undergrad students who burn up your ma tch networks with wrong software settings.
I can't see any reason why it wouldn't be. The tests described don't includ e any short pulse tests, and it could be that inter-stage coupling might no t respond well to short bursts - you might need to ensure that each pulse s equence had a zero DC content or something equally difficult to guarantee.
What would be more useful would be the circuit diagram - and there are peop le here who would do a better job on that than I would.
The e-mail address I use here is real, and does work. The IEEE spam filteri ng works remarkably well.
(like ENI 350 or 3100,) or amateur radio "linear" broadband amps from Henr y Radio.
I've seen these for under $200, but non-working with dead output transistor s (so, buy and repair!)
amplifiers to amplify the pulsed singals generated by the RadioProcessor b oard?
ded probe-protecting features, which shut down the output whenever detectin g long pulses, CW, or PWM above ~20%.
those ham amp kits are under $100!) But it lacks the automatic protection . Just make sure to avoid accidental oscillation caused by old corroded BN C cables with bad shielding. That, or undergrad students who burn up your match networks with wrong software settings.
ude any short pulse tests, and it could be that inter-stage coupling might not respond well to short bursts - you might need to ensure that each pulse sequence had a zero DC content or something equally difficult to guarantee .
Yup, the one issue I had years ago with a RF amp I built was some inter-sta ge coupling that would ring down when the RF was turned off. That lead to a little bit of "weirdness" when trying to get perfect 180 degree pulses. (Fo r NMR)
George H.
ople here who would do a better job on that than I would.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.