Thanks for ur answer

 

Is this just computer simulation or are you going to build a real circuit?
Real circuit:
You will need a signal source. Do you have a nS signal source?
C1 will slow down the signal. I think it should be removed.

 

hi meena,
This is what I see in LTSpice, using a 2N3629.
E
What is the LED type.?

Update:
Corrected images.
1st image with generic NPN
2nd with 2N3629

 

I'll just add,
if your drivng a 200 ps pules , for a blue led your going to need a good 3 plus volts.
obviosuly you will not see a ps pulse of light,

Taking a step back.

If this was nicely defined logic , driving a nice matched impedance line, into a nice resistive load,
then its hard to make a square 200 ps pulse ,

Just think about the volts per second rise / fall time you need, say with a 10 ps rise time of 1v, you need 100000 volts per micro second rise time. Which is quiet interesting .

If you think of what your driving, is Highley inductive / capacitive depending at what voltage your at.
its a very horrible load to drive into.

In your real circuit, your going to be very constrained by the layout you use. No wires for instance.

Good luck,

 

hi meena,
This is using a generic NPN and LED
Stepping the Ctune.
Shows the LED current and Voltage.
E

 

Dear Eric,

We can't make V(vp) pulse in picoseconds ?

Probably not with the circuit parameters you have chosen. Do you suppose the guys that wrote the paper have some kind of "secret sauce"?
The simulation models or the real components they have just might be different enough if they claim they can do it. Maybe you have misunderstood or misinterpreted their claims.

 

hi nnena,
This the LTS sim asc file I used.

Consider it is the LED current that is important.
E

 

If I was going to try to duplicate this result I would be thinking about using 0402 resistors, controlled impedance microstripline connections and components, and a very carefully chosen NPN transistor characterized on a VNA. Even then the results might be disappointing. In this kind of circuit, unrecognized or ignored parasitic elements are everything.

I would say that @ericgibbs did a bang-up job of getting close to the mark.

 

hi,
In my commercial laser rangers I used a triple junction 25W laser diode, which was driven by a ZTX451 transistor in avalanche mode.
Best I could get was about a 5nSec rise time on a 15nSec pulse,, maximum range 5000 mtrs.

If you build the circuit, consider if there is any eye safety requirement.

E

 

Why are you only using a 5 volts supply, whereas the paper you linked clearly mentioned that the circuit relies on the avalanche characteristics of a transistor?
Meaning several tens of volts?

From the paper;
“A voltage exceeding the collector-emitter breakdown voltage of the transistor is applied to the VCC terminal of the circuit.”

 

Why are you only using a 5 volts supply, whereas the paper you linked clearly mentioned that the circuit relies on the avalanche characteristics of a transistor?
Meaning several tens of volts?

From the paper;
“A voltage exceeding the collector-emitter breakdown voltage of the transistor is applied to the VCC terminal of the circuit.”

Sounds like "secret sauce" to me.

 

Eric, your simulation seems to be using an ordinary 0.7V diode instead of a 3V blue LED.
A 3V blue LED will not conduct 0.5A with only 0.8A across it.

I saw the 2N2369 extremely fast transistor in the article. its max allowed Vce is 40V. The article used a supply up to 24V.

 

hi agu,
I just used the generic LED and NPN transistor in order to show a simulation for the TS.
With such a low Vcc of 5V, the circuit is not viable, requires at least 24v and decent Blue LED.

I tried the sim with a 2N2369, it was useless, it maybe a poor model limitation.

Interesting to know what is the TS's practical application or is it just a sim.??

E

 

What will such a brief LED pulse be used for? What will you use to detect any response?

 

What will such a brief LED pulse be used for? What will you use to detect any response?

time of flight measurements are all the rage. 200ps should bring 60mm resolution if the trigger to emission time is very consistent.

 

Instead of relying on an RC constant and current variations in the LED/transistor, perhaps an "old school" approach might be applicable here .... since light travels at just under 1 ns per foot (11.8 inches per nano second), your looking at a distance of 2.36 inches for a 200ps window. If you send a pulse on a wire and have two taps located 2.36 inches apart driving a XOR (probably needs to be a descrete XOR to handle the speed), then theoretically you would have a 200ps pulse on the output of the XOR. Note: because of propagation delays in components and the tapped signal wire itself, you will need to initially adjust the tap locations, but it will at least be a starting point.

 

Here is a nice article from Nature...
Only helpful if you are willing to have a chip made but you can try with discrete parts but the capacitance at will be a challenge. A 200pSec pulse is equivalent to the high time of a 2.5GHz square wave.

https://www.nature.com/articles/s41598-017-07138-3

 

See

 

I separated the current through the pn junction and the barrier capacity. D1 is a pn-junction. D2 is a barrier capacity.

 

hi Alex,
Neat solution.
How much effect does the R2/10k Bootstrap have on the Pulse Rise Time.
E