Re: Scope Probes off Ebay

On 4/8/24 00:11, Cursitor Doom wrote:

On Sun, 07 Apr 2024 12:08:43 -0700, John Larkin <jl@997PotHill.com>
wrote:
 

On Sun, 07 Apr 2024 18:14:25 +0100, Cursitor Doom <cd@notformail.com>
wrote:
>

On Sun, 07 Apr 2024 09:57:13 -0700, John Larkin <jl@997PotHill.com>
wrote:
>

On Sun, 07 Apr 2024 14:51:29 +0100, Cursitor Doom <cd@notformail.com>
wrote:
>

I never learn. I bought a used "Agilent 500Mhz probe" off Ebay for 50
quid as the highest bandwidth probe I thought I had was a 150Mhz Tek
one. Anyway, I have a large selection of old probes lying around so
decided to check to make sure the "Agilent" one was genuine. Not
surprisingly it turns out it isn't. I haven't calculated what it's
real bandwidth is. I've established it's not as sensitive as the
150Mhz one and that's all I need to know. Whilst I was going through
this palarva, I tested a old probe I came across that I've never used
before and was amazed at the improvement in signal I got with it. I've
just Googled its part number and it turns out it's a 3.5Ghz passive
probe! I never even knew I had one so fast. I would never have ordered
the "500Mhz" one if I'd known I had this forgotten-about one already.
So the fake's going back for a refund and I won't be ordering  any
more probes from anywhere in the forseeable future.

>
The HP54006 is a 6 GHz probe, into a 50 ohm scope.
>
https://www.dropbox.com/s/lxq4ujmkvo755uy/HP54006_probe.zip?dl=0
>
They show up on ebay now and then. There's one now, asking $150.
>
You can make your own with some of those Caddock resistors. The
Caddocks have some magical recipe. And unlike a fet probe, they are
hard to damage. I use them to probe 7 ns 1400 volt spikes in my
Pockels Cell driver.

>
Funny you should say that, but the 3.5Ghz probe I mentioned (a Tek
P6056 to be precise) has a fragile resistor assembly in the tip
according to the datasheet and it's easily damaged by rough handling.
I'd be very interested to know what the secret sauce is in the
Caddocks and why they're more robust than whatever Tek used in the
P6056.

>
And you can do a GHz at least with a 1-cent axial or mini-MELF or 0805
resistor on the end of a coax. 450 ohms makes a 10:1 probe. Fast
circuits are often low impedance circuits and don't mind a 500r or 1K
load.

>
Many years ago when I was somewhat impecunious, I used to improvise
like that, but these days I prefer to just buy whatever I need ready
made. Some of those top-end RF patch cables can be ruinously expensive
to buy ready-made, but what are you gonna do? No matter how good you
are with terminations, you'll never emulate the quality standard of a
properly made, high quality patch lead. When you're as ham-fisted and
half-blind as I am, it's a no-brainer to buy 'em ready-made!

>
Amazon has some crazy cheap coaxial jumpers and SMA and SMB connectors
and adapters. All the ones I've got so far have been fine. For bench
work of course, not production.

 What do you mean by "fine"? How did you go about characterising them
and what were you looking for specifically?  IME you get what you pay
for and cheap connectors are very seldom worth it. Buy cheap, buy
twice as they say.

Not necessarily. I used to buy phase-matched sets of SMA cables
from Huber & Suhner. Then, one day, because H+S did not reply to
a new request for a quotation, I got them from JYEBAO in Taiwan
via a French representative. They were cheaper *and* better.
Jeroen Belleman