1
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab

Hi everyone, I recently completed the construction of a spectrometer that studies electron transfer events in inorganic molecules. I am trying to further improve my design however, mainly one major flaw.

I'd like to control where the output of my photomultiplier (PMT) tube is going using a set of BJT transistors as we currently have to manually plug the BNC into the appropriate device. Given the complexity of the setup, it's not feasible to expect workers in the future to be able to do this.

I am somewhat unsure as to how BNC/coax cables operate and would like to ensure that my circuit will work before I invest the effort into building the switch box.

The TTL logic to each gate is to come from an NI USB 6008 IO box.

\$\endgroup\$
6
  • \$\begingroup\$ BNC/coax cables are just cables and connectors, they can be used to carry a wide variety of signals. What matters is the nature of the signal coming out of your photomultiplier in terms of frequency and amplitude, because that's what you need to switch. Do you have any information? \$\endgroup\$
    – Finbarr
    Apr 26 '17 at 7:35
  • \$\begingroup\$ Basically we are exciting a sample using an Nd:YAG laser - the decay of which generates a current spike which forces a trigger event on an oscilloscope downstream. Since the laser is pulsing at 10 Hz, we essentially have a waveform. \$\endgroup\$
    – James
    Apr 26 '17 at 7:43
  • \$\begingroup\$ @David frequency and amplitude was the question. Clearly it will be a waveform! \$\endgroup\$
    – Andy aka
    Apr 26 '17 at 7:44
  • \$\begingroup\$ The frequency of the waveform will always be 10 Hz so as long as the sample is capable of electron transfer - the amplitude however varies strongly on the sample (the amplitude is proportional to what energy level the electrons are relaxing from). \$\endgroup\$
    – James
    Apr 26 '17 at 7:49
  • \$\begingroup\$ Got it. Thank you! \$\endgroup\$
    – James
    Apr 26 '17 at 8:05
1
\$\begingroup\$

It will not work (reliably) like this.

The transistors turn on when their base voltage is below their emitter voltage. But here the emitters are DC coupled to the PMT output. So for this to work the PMT output needs to have a DC-component at the TTL high level.

I would consider:

  1. Looking for a relay/switch with BNC connectors
  2. Seeing if CMOS switches like CD4066 will suffice
  3. Redesign
\$\endgroup\$
1
  • \$\begingroup\$ Excellent - the feedback I was looking for. So I have some CD4051 ICs laying around. Is it possible to perhaps use this DIP to gate the signals? \$\endgroup\$
    – James
    Apr 26 '17 at 8:18
1
\$\begingroup\$

You should bear in mind that the BNC connectors and the coax cables used in these kind of applications generally have impedances of either 50 Ohms or 75 Ohms, so the apparatus you're working with is likely to have input and output impedances to match. CMOS switches like the 4066 and 4051 typically have ON resistances in excess of 100 Ohms, which are therefore likely to have quite a significant effect the signals you're measuring.

As well as the logic signals, the 6008 has a +5V output capable of giving 200mA, so I would use this to provide power for a small signal relay with a changeover contact and use a simple NPN common emitter switch to drive the coil from a digital control line. Don't forget the protection diode.

\$\endgroup\$
1
  • \$\begingroup\$ The impedance on all our cables is 50 Ohms. I'm actually going to look into signal relays as well - I don't think I have any laying around here. I actually like this suggestion as I have a ton of unused channels on my NI box. \$\endgroup\$
    – James
    Apr 26 '17 at 9:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.