Pulse generator 150ps 1Vpp

[luky315]

I made a Pulse generator with a pulse width of about 150ps and an Amplitude of approximately 1Vpp. It uses the Step-Recovery Effect of a HF-npn-Transistor and a shorted-out Stub to shape the pulse.







Unfortunately (for you) the documentation is in german (https://engineeringarts.wordpress.com/2012/07/24/pulsgenerator-150ps-1vpp/), but I try to answer all questions.

[Chipset]

Send one to dave so he can test it with his 13ghz scope.

[luky315]

I have only one handmade Prototype.
The measurement is made with a LeCroy WAVEMASTER 8620A (20Gs/s) and verrified with an Agilent 86100B (Agilent 86118A 70 GHz Remote Sampling Module), so there is no need to use a slow13GHz Scope

[Lukas]

Can you post the documentation anyway, there are some German members, as well.

[Chipset]

It being in german isn't really a problem, pasting it in google translate usually yields readable enough results.

[Time]

Ugly board, beautiful pulse!

[robrenz]

Doesn't a pulse for this type of measurement need at least a short section reasonably steady state voltage before it drops?

[ejeffrey]

I can't speak for everyone else, but when I use a pulse generator I typically want either the shortest pulse possible or the fastest step possible holding the 'high' value for at least 10s of nanoseconds.  I can't think of a time I have wanted a pulse with 250 ps rise and fall times but a 1.5 ns hold time at the top.

[luky315]

This Pulse Generator ist used for Cable testing, not for BW-measurements.
But can also be used to test the BW of a scope if you use the Cursors and not the internal
Risetime-measurement-function.

[robrenz]

This Pulse Generator ist used for Cable testing, not for BW-measurements.
But can also be used to test the BW of a scope if you use the Cursors and not the internal
Risetime-measurement-function.

That makes sense. Thanks

[T4P]

13GHz slow? And what's that POS i see before me anyway  (LeCrap)

[robrenz]

i wonder how he did the vias stitching and ensure connectivity... neat.

Looks like copper eyelets to me.

[luky315]

The Vias are made with a Bungard "Bel Favorit" hand press. They are small (d=0,8mm) copper rivets.

[robrenz]

This Pulse Generator ist used for Cable testing, not for BW-measurements.
But can also be used to test the BW of a scope if you use the Cursors and not the internal
Risetime-measurement-function.

That makes sense. Thanks

I thought about it some more and now I disagree. Here is how I am trying to wrap my head around this.
1. for BW testing we are using a relatively instantaneous risetime to measure how high a harmonic content the scope can digest.  A risetime of zero would contain all the odd harmonics to infinity.
2. If you have a pulse generator that outputs a pulse with no defined flat at the top and you calibrate it's risetime by measuring it on a hypothetical scope that has infinite BW. You could use it because you know the actual risetime or the dV/dT.
3. When you use your less than perfect scope to measure a "zero risetime" pulse with a flat top, the bandwidth required to measure the peak voltage at the steady state drops to a level any scope can handle. That makes sure the voltage used in your dV/dT has relevance or is as accurate as your scopes vertical.
4. When you use your less than perfect scope to measure a "zero risetime" pulse without a flat top, the bandwidth required to accurately measure the voltage on either side of the peak remains infinite. That means the voltage shown and used in your dV/dT has no relevance as it will be much lower than that shown on the infinite bandwidth scope.
5. Even with the flat top you still need to know the actual risetime of the pulse to add to your scopes measured risetime to do an accurate BW assessment.

This is pure opinion and needs to be verified by those who are more knowledgeable.