Re: P6137 scope probe repair


Tom Lee
 

Lawrance and Roy, and a couple of others:

//Warning: Rant mode on. You'll want to power up your shields, 'cause photon torpedoes are headed your way.

What the hell's wrong with helping Jared *fix *his probe? Posts with vague or even explicit suggestions that his endeavor is doomed to fail are unhelpful, and technically just plain wrong. Adding irrelevant and sometimes wrong trivia, though perhaps interesting, just adds noise and distortion to the discussion. Here's my nth attempt at raising the SNR and actually point Jared to a solution.

It's been previously noted -- multiple times -- that probes use special resistance wire. If you've studied transmission lines at all, you will know that lines need to be terminated in their characteristic impedance in order to prevent reflections.

Probes are magic because they allow one to look at nodes whose impedances are uncontrolled, somehow without producing artifacts due to the inevitable reflections that must be occurring. How is this even possible? John Kobbe and Bill Polits had to solve this problem, or else Tek's efforts at making high-bandwith scopes would be the equivalent of inventing the phonograph without also inventing phonograph records. Kobbe and Polits understood that preventing reflections from misterminations is impossible. But they also understood that resistance is good at providing damping. In a 10:1 probe, you have a large attenuation anyway, so adding damping is relatively non-intrusive. Kobbe stripped some ordinary coax, connected resistance wire to the center conductor. As he pulled out the original lossless conductor, it pulled in the resistance wire. They tried it, and it worked. Thus was born the high-bandwidth probe. See US patent 2,883,619 (1959).

If you read anything at all of what I wrote to help Jared, you'll note that I made explicit mention of how his chances are a function of where the break is, the correct amount of damping being a function of length, among other factors. But contrary to what you implied, it's not a hyper-sensitive function of length. If you chop off an inch from the end of a 3ft cable, that 3% length reduction isn't abruptly going to trash the probe. Why would it? Suggesting that shortening would "most likely produce a semi-worthless probe" is a prime example of one of my pet peeves: Asserting an opinion without technical foundation (I'm guilty of it myself, but at least I'm aware that I have this problem). Worse, it's completely wrong. It only adds noise and distortion to the discussion. The "tuning" you speak of is not a sharp function at all. You can remove a small bit, and the effect will be a small bit. Think logically: Do you really believe that removing a nanometer instantly trashes the probe? Ok, how about a micron? A millimeter? When you acknowledge that maybe the deterioration is an analog, not binary, function of length, you will have made a cognitive leap forward. And if you have no clue at which point the amount removed is too much, that's a sign that you ought to keep your fingers away from the keyboard. The resistance per inch of probe conductor is typically  under an ohm. Jared has narrowed the general region of the break to about an inch from one end. If you think the removal of an ohm's worth of loss out of a 100 or 200 ohm total is going to produce a semi-worthless probe, show me your calculations or measurements that support such an assertion.

Better yet, how about not asserting unfounded discouraging things at all, and just let Jared try to fix his probe, instead of telling him he's going to fail? Or to "go buy another probe", as some Mr. Obvious suggested (do you really think that Jared is so thick-headed that he needs someone to tell him that)?

Next, your trivia about T-coils makes it sound like many probes use them. As far as I am aware, the total number is two: P6047 and the oddball 1kilohm P6048. If you are aware of any others, I would very much like to know their model numbers so that I can update my spreadsheet. T-coils are very much the /exception/, rather than the rule, in probes.

//Rant mode off. Photon torpedoes reset. For now.




--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 1/17/2021 06:29, Lawrance A. Schneider wrote:
On Sat, Jan 16, 2021 at 10:43 AM, Roy Thistle wrote:

The "coax" is not normal 50 ohm coax... on high performance/high frequency
probes.
It is usually a kind of "lossy" coax, which uses a shaped mono-filament
resistance wire for the center conductor. That wire is somewhat brittle
(compared to copper)... and shatters or cracks... when bent, mutilated, or
stapled.
After I wrote my missive about finding the break with a nanoVNA, I thought about the problem further. Last year, I started to explore why Tek Probes use t-coils - boy oh boy was that a fun project. And I then remembered the above. Why am I bothering to note this. I discovered the Probe is 'tuned' within and shortening the lead and reconnecting would most likely result in a semi-worthless probe. I believe Alan Wolke has a video about the wire within a probe. The wire looks like a mess.

larry



Join TekScopes@groups.io to automatically receive all group messages.