negative voltage BJT buffer

Thread Starter

djstar

Joined Jan 26, 2008
39
I have a signal from a particle detector which varies from 0V to -1V. I have to send this signal over a 50 Ohm coaxial cable about 50 cm long.

i have had a look at Emitter follower circuits but i am confusing myself with the negative voltages. The ciruit needs to be powered from 0v and -5. Any pointers in the right direction.
 

Alec_t

Joined Sep 17, 2013
14,280
How accurately does the 0V to -1V swing have to be replicated?
What is the impedance of the signal source?
 

MrChips

Joined Oct 2, 2009
30,706
This is very doable.
A signal that is 0 to -1V is very typical with output of PMT with a +HV bias.
Most often the signal is AC coupled.
You can use almost any supply voltage, positive or negative.
Why do you think that you need a -5V power supply?
I do this everyday and my supply voltage is +5V.
 

Thread Starter

djstar

Joined Jan 26, 2008
39



The source is 50 ohm and it needs to be very accurate. i have attached a picture of a PMT output (from google). The shape will stay roughly the same but the amplitude and width is not consistent.
 

Thread Starter

djstar

Joined Jan 26, 2008
39
This is very doable.
A signal that is 0 to -1V is very typical with output of PMT with a +HV bias.
Most often the signal is AC coupled.
You can use almost any supply voltage, positive or negative.
Why do you think that you need a -5V power supply?
I do this everyday and my supply voltage is +5V.
That is very correct Mrchips, it is a PMT with a +HV bias. The systems we use run on ECL and NIM so every thing is a negative voltage and to keep with tradition every thing is design with discrete components hence the need for using a BJT. Would you be able to recommend any circuits
 

MrChips

Joined Oct 2, 2009
30,706
What are you hoping to do with the signal? Which one of the following?
  • photon counting
  • timing information
  • pulse-height analysis

The usual reason for ECL output is for timing. If you are already using a NIM crate, the common setup is to use a preamp followed by an amplitude discriminator with ECL output.

If the source impedance is 50Ω, use an RG-58/U cable into a PMT preamp.

See if this pdf article helps you:

http://www.becker-hickl.com/pdf/ampmt.pdf
 

Thread Starter

djstar

Joined Jan 26, 2008
39
What are you hoping to do with the signal? Which one of the following?
  • photon counting
  • timing information
  • pulse-height analysis

The usual reason for ECL output is for timing. If you are already using a NIM crate, the common setup is to use a preamp followed by an amplitude discriminator with ECL output.

If the source impedance is 50Ω, use an RG-58/U cable into a PMT preamp.

See if this pdf article helps you:

http://www.becker-hickl.com/pdf/ampmt.pdf
 

Thread Starter

djstar

Joined Jan 26, 2008
39
Thanks for the reply, the PDF makes a interesting read.

My background is normally digital design using FPGA's so my understanding of the system is we have a PMT detector with "some electronics" i know that is very open but i have never seen these electronics. the signal is then feed into a "in house built discriminator board" where we trigger a comparotr when the signal is over a over a threshold, the pulse is stretched and integrated through a discriminator, then back into a comparator and monostable then out to a FPGA to be processed.

we are now trying to do all the discriminating in DSP within a FPGA rather then discrete analogue components.

I'm designing a board which will except signals from a PMT( and its electronics) and buffer the signal and send the signal blackout and into our existing discriminator boards. The same signal PMT signal will also trigger a comparator at around -130mV and output the pulse over LVDS to a FPGA processing board which will perform the discriminating in DSP. fundamentally its a board which will allow us to perform parallel tests with the existing Discriminator boards and our new DSP based board.
 

MrChips

Joined Oct 2, 2009
30,706
I don't see where the requirement to use a -5V supply comes in.

Usually the PMT signal is AC coupled. This means that it normally sits at 0V and the detected signal goes negative.
What you need is a wide bandwidth amplifier. The signal can be biased upwards into the operating voltage of the amp. From here, the signals can be all positive and fed into your FPGA. You can branch out the signal from the amplifier, AC coupled to feed the existing system.

A possible solution is to use a PMT preamp as your amplifier to drive the two systems.
Cremat makes single module preamps:

http://www.cremat.com/

To answer your original request, a basic emitter follower BJT circuit using a PNP transistor and a negative supply may be what you are requesting.
 
Top