Hi,

I am trying to produce voltage pulses with very short rise time (<15 ps), medium width (20 ns) and amplitudes up to 15 V. As a promising candidate I selected photoconductive switches and ultrafast photodiodes. However, I have limited knowledge about electronics at all and thus, even less knowledge about these devices. Do you think it is possible to produce pulses like that by applying a bias voltage to the switch/diode and send a laser pulse with approx. 20 ns width to the photosensitive area?

I already found photoconductive switches at Hamamatsu and Laser Quantum that might work. In particular, I am talking about the G4176-01 and the Tera-SED3, respectively. However, I cannot imagine how to connect these switches. For the Tera-SED3, one would have to use the ground as the output but how to deal with the grounds than without violating the impedance?

I appreciate any help, advise, new insights and ideas. Also, if you know another supplier of photoconductive switches or photodiodes that might work for my purposes, I would be very grateful if you share these with me.

 

The first thing to do is check the data sheets for the devices that you have chosen and see if they can operate as quickly as you need.

 

hi 92,
I used avalanche mode transistors to generate 30-50nSec period pulses, rise time less that 1nSec, to fire LD163 dual laser diodes for a 5kMtr range finder.
I might be worthwhile checking the avalanche option.

http://www.aholme.co.uk/Avalanche/Avalanche.htm

E

 

I was in contact with the company, from the datasheet side everything should be fine. However, they were not able to give me exact information on how to use these terahertz emitters for my purpose. Attached, you can find a sketch of a photoconductive switch. Usually the - electrode is the ground. However, I have to use the - electrode as an output now. Since the switch is connected with a BNC cable, I have no idea how to deal with the grounds now and how to get my voltage pulse (which is in the ground line of the BNC) to the sample.

 

Hi ericgibbs,
thanks for the idea!
I think that the rise times of avalanche mode transistors is still to long. I strictly have to be below 20 ps. But I will look deeper into that. Did you build it yourself or are there commercial ones available?

 

Do a preliminary assessment on the numbers.
20ns is 50MHz (25MHz for 50% duty cycle). Hence you are looking at 250MHz bandwidth minimum.
15V with 15ps rise time is 1000V/μs slew-rate.

This is all in the realm of optical communication devices.

Are you concerned with the driver circuitry for the emitter or the receiver circuitry at the photodiode?

 

Hi MrChips,

I am not sure if I really understand your answer, but let me try to reply the best I can.

I think I need much larger bandwidth because my rise time is approx 50 GHz and the length 50 MHz. If I would go for 250 MHz bandwidth, I would cut all the high frequencies responsible for the fast rise time.

I haven`t found any working device for me up to now, but maybe you can show me one or two examples.

And your last sentence confuses me a bit. I am talking about the circuitry at the receiver photodiode. I want to bias the receiver and when light hits the optical input, the ultrafast excitation of electrons and thus the ultrafast increase in conductivity shall allow me to produce a pulse with very short rise time and several tens of nanosecond width.

 

That is where I was confused with the 15V requirement. That is going to put a heavy demand of the slew-rate of the receiving circuitry.
I would guess that you will need 50Ω output impedance and matching 50Ω coax cable at the receiver.
What output voltage do you really need?

What exactly is your application?

 

I know, it is very hard to achieve, but I saw a paper in which they used a photoconductive switch to produce such voltage pulses. The connection from photodiode/ photoconductive switch to the sample will be done by a 50 Ohm high frequency high bandwidth coax.

The pulses are required for an experiment to measure the ultrafast dynamics of the polarization of a ferroelectric solid.