TD pulse generator

Craig Sawyers <c.sawyers@...>

Hi all

Just received a 067-0513-00 TD pulse generator, bought from eBay for $22.
Of course, there is the big question of whether the TD is alive and well, so
I tried connecting the curve tracer to the GR output connector; staight line
in both directions up to 50mA. Ouch!

The TD is held at the end of a short length of air line by a little knurled
screw. I took the airline out, removed the screw and tapped onto a surface
to get the TD out. Tiny little beast, around 1.5mm diameter and about the
same high, with a flange on one end. Got that carefully connected to the
curve tracer. Again nothing to 50mA - so I just kept increasing the voltage
to see what would happen.

At 90mA - bingo! I've never come across a TD in Tek gear with that large a
peak current. Anyway, I measure Ip=90mA at Vp=0.4V; Iv=10mA at Vv=0.5V;
peak point current of 90mA at 0.85V. With those specs, and the ultra-low
capacitance package I could start to believe the front panel spec of <30ps
switching time.

Next problem was the batteries. There are two 22.5V Energizer 412 batteries
in there in parallel (essentially obsolete), and both were as dead as a
doorpost. Fortunately, they are almost exactly the same size as standard 9V
batteries, so I substituted two of those wired in series. Connect to 7A19
for a quick test - and a nice looking pulse appears!

Neat thing is that this unit has a pre-trigger output, occurring at 60ns
before the TD pulse. It achieves this with a delay cable, so it is not

That gave me a problem when I went across to using a 7T11/7S11/S6 to measure
the rise time - the 7T11 needs 75ns to get going, so the fast rise and fall
are off screen. I could get them back again by using 20ns of cable from the
TD head to the S6, but the rise time is dominated by cable effects (measured
around 150ps). So I used random sampling, which is not as good as
sequential sampling as regards jitter (spec of 30ps as compared with 10ps),
but it was the only way to get a measurement. I used an SMA/GR adaptor, and
then plugged the head straight onto the S6.

Interpreting around the jitter (which was actually more like 15ps), the
total rise is about 40ps. Now the S6 is specced at <30ps, as is the
067-0513-00, so the combined response should be <43ps - around what is

So all in all, this was a good $22 worth (oh, and $32 to get it across the


Join to automatically receive all group messages.