EEVblog #306 - Jim Williams Pulse Generator

[HackedFridgeMagnet]

So you're saying at some unknown (to me) voltage, we could get a percentage of 2n2222s to avalanche in a similar way? Would they need to be some old batch do you think or any more modern 2n2222s could avalanche?

[T3sl4co1l]

No need for such a poorly defined transistor (seriously, the world would be a better place if everyone forgot about 2N2222 and picked 2N4401 instead ), it's been my experience that 2N3904 is avalanche capable.  Typically around 90-100V, with a 4.7k B-E resistor.

Tim

[linux-works]

I just saw this thread.  I'd like a kit, also

I'm in the bay area, so if I can be of help, just ask.

[eurofox]

+1 In for a kit MKII version.

eurofox

[kmel]

+1 for a MKII kit

[joetorelli]

Hi,
New to the EEVBLOG. I just found this forum ~3wks ago. I have been reading and watching this blog forum so much my wife thinks I am cheating on her, because I been doing nothing else. Except getting majors headaches from all the reading.
Anyway are these kits still being made? I went to ebay and could not find anything. And this subject seems to have faded off. If they are still being made count me in for a calibrated kit.

Thanks

[e-pirate]

Hi, folks!
I'm currently designing my own version of PCB for this avalanche pulse generator. This board will be placed in aluminum box. BNC connector is a PCB vertical thought hole connector that will be soldered to the board parallel. The 50 ohm resistor mounted vertically right under signal/central wire. This is me very first design of such type generators (and frequency), so I want to know what do you think about ground polygon distribution. I separated grounds of voltage source and the output part. Any ideas how to improve?
Oh, and there is a switch on the right side of the box.

[OldTech]

@free_electron,

Hope you are well soon.

I am interested in two calibrated Picosecond Pulser MK-II units.

OldTech

[EV]

You don't have delay line in your kit? The pulse will be very short and not flat. Look at pictures of reply 214 in this thread. There is delay line.

Hi, folks!
I'm currently designing my own version of PCB for this avalanche pulse generator. This board will be placed in aluminum box. BNC connector is a PCB vertical thought hole connector that will be soldered to the board parallel. The 50 ohm resistor mounted vertically right under signal/central wire. This is me very first design of such type generators (and frequency), so I want to know what do you think about ground polygon distribution. I separated grounds of voltage source and the output part. Any ideas how to improve?
Oh, and there is a switch on the right side of the box.

[e-pirate]

You don't have delay line in your kit? The pulse will be very short and not flat. Look at pictures of reply 214 in this thread. There is delay line.

Well, I don't know how to calculate wave guide 
And I will probably combine ground polygons. I don't think there is a reason to separate them.

[han]

is there anyone ever try the impulse generator on Spectrum analyzer?
sorry if there already the answer in the 31 page before...

[c4757p]

You don't have delay line in your kit? The pulse will be very short and not flat. Look at pictures of reply 214 in this thread. There is delay line.

Well, I don't know how to calculate wave guide 

Google seems to do well for "how to calculate waveguide"...

And I will probably combine ground polygons. I don't think there is a reason to separate them.

Agreed. Unless you have a real reason to force ground currents to be separate, you're just creating extra inductance, which may or may not be a problem for you.

I like it otherwise. Nice, compact design, even though it's PTH. Perhaps you could fit some length of delay line under the battery. Alternatively, use coax!

[free_electron]

This project has been on the backburner a bit.

I will place an order for blank circuit boards this week.

Calibrated kits are bo longer available. I don't have access to the scope anymore.
Now, since we have a good sample spread from the original batch we know what these generators do in terms of rise time. So calibration is not really necessary

I am still in doubt what i should make available.

A) Board + 2 NOS transistors from the same batch.
B) Full kit including case

Board and transistors can be sent in a regular envelope. But , if everyone then has to start digging for all the other stuff .. Its time consuming and wont meet minimum shipping requirements from digikey or others.

Let me do a price calculation for the full kit. Parts prices may have fluctuated.

[alank2]

Awesome news!  I'd prefer one with a case, but I'll take whatever is available and don't mind sourcing parts from digikey if needed.

Please keep us posted!

[gdewitte]

+1 for the case kit.

[e-pirate]

Here is a new layout with a single ground polygon. I will search through Internet how to make a waveguide and probably design another version.

[GeorgeOfTheJungle]

I will place an order for blank circuit boards this week.

I think the name of Jim Williams ought to be on the board!

Jim Williams' avalanche (FAST) pulse generator (See Linear Technology AN47), or something like that, perhaps?

[Chris.M]

Could you please put me down for one Full kit including case.

[H.O]

Excellent news!
I'm (still) in for a kit, whichever you decide, thanks!

[e-pirate]

I think the name of Jim Williams ought to be on the board!
Jim Williams' avalanche (FAST) pulse generator (See Linear Technology AN47), or something like that, perhaps?

Completely agree! Just in respect to Jim.

[fpliuzzi]

Thanks for making these available again. Please put me down for one full kit including the case.
(if you decide to just make the board and two transistors available, that would be fine too)

[linux-works]

+1 kit for me, please.

[cyr]

is there anyone ever try the impulse generator on Spectrum analyzer?
sorry if there already the answer in the 31 page before...

SA and scope below (with some external attenuation).
Note how the first null in the spectrum corresponds to 1 / pulse width.

Also, there is a rule of thumb that most of the frequency content of a square wave is below approximately 0.5 / rise time. That would be 2.5GHz for a 200ps edge, which looks about right.

[han]

is there anyone ever try the impulse generator on Spectrum analyzer?
sorry if there already the answer in the 31 page before...

SA and scope below (with some external attenuation).
Note how the first null in the spectrum corresponds to 1 / pulse width.

Also, there is a rule of thumb that most of the frequency content of a square wave is below approximately 0.5 / rise time. That would be 2.5GHz for a 200ps edge, which looks about right.

Thanks for the effort.
i just remember the Fourier theory when i in univ

[idpromnut]

If the parts + case can be sourced via digikey then i am ok with just the board and transistors.