Pulse Head

How about using a fast logic chip in a 50 ohm PCB? Just draw one up and get Alberta Printed Circuits to work on it and the next day you got a PCB. For the 067-0681-01, join the Yahoo Tekscopes group and ask them. (I can't guarantee this, but I think this unit was meant to be connected to the scope's (500 series and so on) calibrator, and sharpens the rise-time. It's basically a tunnel diode in a circuit powered by the square wave from the scope. Since the 500 series has a calibrator that can give 100V square waves, this is more than enough to power such a gadget. )

Also be aware there is no need for a 1ps rise-time for a scope with a

100ps rise time. What rise-time do you need? I'll look in the lab, we have a graveyard of unused Micrel eval boards with SMAs on em. Great stuff.

Ah, look up Jim Williams' AN47 or "The Taming of the Slew" appnotes. Both contain the same circuit, a BJT avalanche pulse generator, using the 2N2369 to be exact. 2N3904 may even work. This type of circuit typically switches in less than 2ns, with 350ps risetimes easy to manage. If that's not enough, you might get some step recovery diodes, which are reasonably easy to use (using them to their fullest potential is only a little more complicated).

The kickinest stuff today involves screaming fast snap diodes, shock lines and GaAs monolithics (including integrated shock lines!), reaching single digit picoseconds. There are other approaches, too, but they're more scientific (like femtosecond laser pulses converted somehow into electricity).

Tim

Palinurus wrote in news:2bydnTC8Sr7xtBHUnZ2dnUVZ snipped-for-privacy@posted.oplink:

TEK TD pulsers needed about 70-100V to drive them.A TEK 106 or PG506 hi- amplitude output was used to drive them.Rise time was IIRC;1 ns.

that 067-0681-01 is just a pot,3 1K 1/8w resistors and a TD inside a nice cast aluminum box with BNC's.

I -used- to have the schematic,when I was at TEK.

I even built my own pulser;if I find it,I'll open it up and copy it down,and post it here or in alt.binaries.schematics.

--
Jim Yanik
jyanik
at
kua.net

You can probably drive it from a standard pulse generator. But TD pulsers make ugly waveforms... a slow rise, then a fast jump, then a slow drool to finish up. I vaguely recall that that particular box may be slow.

A fast logic gate (NC7NZ34) or LVDS receiver (SN65LVDSDCBR), properly grounded and bypassed, will make a clean sub-ns step output without the TD artifacts.

How fast do you need?

John

Jim Yanik wrote in news:Xns9BAA70994D117jyanikkuanet@74.209.136.83:

As far as I recall the pulser was wired something like this

v.pot 1K 1K 1K input BNC>---/\\/\\/\\/------/\\/\\/---/\\/\\/\\--/\\/\\/\\---> BNC | | [TD] | gnd

keep all leads short.Maybe build it on a PCB. (the TEK pulser was point-point wired.)

I can't recall the pot value. adjust pot to where the TD -just- triggers.

--
Jim Yanik
jyanik
at
kua.net

3K output impedance? That doesn't look right. The step at the TD will only be about half a volt.

John

Jim Yanik wrote in news:Xns9BAA726962FD9jyanikkuanet@74.209.136.83:

the TD connects -between- the pot and 1K string. the ASCII "schematic" didn't appear right when I read my own post.

--
Jim Yanik
jyanik
at
kua.net

=A0 =A0 1K

=A0 =A0 =A0 =A0 =A0

.

If you reverse the "input" and "output" BNCs, it will work much better ;)

I have need to generate an extremely fast pulse to characterize the rise time of oscilloscopes, and am disinclined (read "unable") to spend the $Ks to buy a full-fledged scope calibrator. I notice that there are occasionally tunnel diode pulse heads available used or surplus, but nobody seems to know what kind of drive signal they require. One Tek part currently on eBay, 067-0681-01, even Tektronix seems to know nothing about. (Thanks a lot, Tek.) Is any information available? I wouldn't be averse to building one, if I had a reliable design. You can probably drive it from a standard pulse generator. But TD pulsers make ugly waveforms... a slow rise, then a fast jump, then a slow drool to finish up. I vaguely recall that that particular box may be slow.

A fast logic gate (NC7NZ34) or LVDS receiver (SN65LVDSDCBR), properly grounded and bypassed, will make a clean sub-ns step output without the TD artifacts.

How fast do you need?

John

-     About 1 ns. This is just to characterize the rise time of yer average, garden-variety scopes. We used to have such a thing, as part of a Fluke 5500, but a nuby managed to blow it up, and management doesn't want to spring for repairs. As to the logic gates mentioned above - would I need to know anything special about the geometry of the layout? It's rather out of my line.

Except that you can't push a decent edge through a pot.

Most of the TD pulsers I've seen had a circa 50 ohm resistor between the TD and the output.

One nice thing is the little Tek box that has a blocking oscillator, pretrigger delay line, < 30 ps TD step generator, and a battery inside. They show up on ebay now and then. GR 874 connectors. I have a couple around here somewhere... 067-0513-00.

It shouldn't be hard to booger an old S52 pulse generator head to work standalone. It has a < 25 ps TD step generator with SMA output. The old Tek 1S2 also had a nice TD pulser inside, if you can find a junker.

John

I have one of the 067-0513-00 pulsers here. As John says, it has a 30 ps risetime, 0.4V output pulse. It has no input; generates its own pulse, complete with pretrigger output. Runs from a 22V battery, or build your own small power supply. It only draws about 1.5ma, so almost any 22V source can run it. If you're interested in the documentation, I'll post it on A.B.S.E.

--
Dave M
MasonDG44 at comcast dot net  (Just substitute the appropriate characters in the 
address)

Life is like a roll of toilet paper; the closer it gets to the end, the faster 
it goes.

There is another way, used before Dr. Esaki discovered and explained the diode that bears his name; a magnificent discovery and discourse and advancement in solid state physics. This other scheme is also solid state; it is a small variation of the relay. Use a mercury-wetted reed relay; put in the middle of a coax enclosure and engineered to retain a 50-ohm profile on all transitions between solid core to relay and back, the system can achieve sub-nanosecond transistions. If you make a window thru the outer enclosure so as to view the contact, and charge an open cable on one end to a high voltage, when the relay closes, a coincendent pulse of light comes out of that window (PW=2*linelength) and can be used to characterize PMTs; called a Huggins Lamp. The other end, of course, gives a nice electrical pulse of the same width and transistion speeds that can be used to trigger the scope viewing the PMT output. Tek made some mercury-wetted reed pulse generators (no Huggins lamps AFAIK).

e

ed

the

s e

ps

If you go back 35? years Scientific American Amateur Scientist (back pages) did a picosecond pulse (rise and fall) generator employing a double sided PCB unetched as the charge store and the mercury wetted relay as the short generator. Should work with transistor in avalanche mode

What's a SN65LVDSDCBR? I know you like the 65LVDS2 -- running on 3.3V? With a say 22-ohm output resistor for a 50 ohm coax and termination (Zout looks about 27 ohms - see ds fig 9 and 10)?

complete

Schematic would be nice. I modified mine to run on three 9-volt batteries.

John

characterize the rise

Build it on a piece of copperclad, as tight as possible, something like this:

ftp://jjlarkin.lmi.net/BB_fast.JPG

Or buy one of these. I'll give you the SED discount.

John

characterize the rise

yer

What's with the html?

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
 I love to cook with wine     Sometimes I even put it in the food

Document posted to A.B.S.E.

--
Dave M
MasonDG44 at comcast dot net  (Just substitute the appropriate characters in the 
address)

Life is like a roll of toilet paper; the closer it gets to the end, the faster 
it goes.

Tek 109 pulser, vertical mercury-wetted resonant reed, < 250 ps risetime. Yup, I have one of them, too.

John