Scope probe to bnc adapter-type panel mount socket.

[Greg Robinson]

Hi everyone.
Bit of an odd question, I'll start with some context.

I'm putting together a simple curve tracer to use with my scope. Now, I could make the horizontal and vertical outputs 50 ohms, use BNC connectors and coax and 50 ohm feed through terminators on my scope, but that would mean some extra expense - I don't have any 50 ohm feed throughs (not much use for them until now) so I'd need to purchase those, and I'd also need a circuit capable of driving 50 ohms (100 ohms), which would add significantly to my BOM compared to if I were able to say, drive a 1Meg/10Meg scope probe which any old op amp can manage.

So, my plan is to use regular 10:1 scope probes rather than use coax. All good so far.
Now, I do already have probe tip to BNC adapters, but it got me to wondering if there are purpose built 3.5mm or 5mm panel mount probe sockets? I haven't been successful in my searches, but if they do exist, I'm sure someone here would know, though I would not be at all surprised if they don't.
If they do exist, it would just mean I wouldn't need to dedicate more adapters to a single piece of equipment with the potential for them to go missing, etc.

I also welcome any feedback about my intentions, if my thought process is flawed, please let me know. Thanks again for any help.

[tggzzz]

Curve tracers, unlike VNAs, are low frequency and don't need impedance matching. Indeed, that would be impossible since your UUT will have undefined and varying impedance.

The signal source will need to supply whatever current is required by the UUT, so ideally it will have a low output impedance.

You don't want the UUTto be disturbed by the measuring devices, so they need a high input impedance. Usually 1Mohm is high enough, but that will depend on the UUT.

So,  to over simplify, the ideal source is a PSU or high current opamp, and the input can be a scope input connected to the UUT by pieces of wire or coax. No 50ohm terminations needed.

[Greg Robinson]

Thanks tggzzz, but I'm not talking about the probe from the curve tracer, rather the horizontal and vertical outputs from the curve tracer to the scope.

[tggzzz]

So am I. Or isn't the scope part of the curve tracer?

[Greg Robinson]

Ok, then I'm completely misunderstanding what you're saying. How on earth would the horizontal and vertical outputs effect the DUT? There'll be mixers/buffers between them. I'd already been planning for the probe input to have a high impedance, and I wasn't asking about that and certainly not suggesting I would make it 50 ohms. I'm very confused.

[Greg Robinson]

Oh, and curve tracer would be external to scope, horizontal and vertical outputs feeding 2 channels in XY mode.

[janoc]

What he is just saying that you don't need to bother with the feed-through terminators. Just connect it directly.

BTW, if you don't have feed-through terminators (they can be expensive and hard to find), you can also use a regular BNC T and use a normal non feed-through terminator. Those are cheap as chips. It takes more space but it works as well.

[tggzzz]

Ok, then I'm completely misunderstanding what you're saying. How on earth would the horizontal and vertical outputs effect the DUT? There'll be mixers/buffers between them. I'd already been planning for the probe input to have a high impedance, and I wasn't asking about that and certainly not suggesting I would make it 50 ohms. I'm very confused.

So are we.

Draw a block diagram and post it here.

[Greg Robinson]

Ok, hopefully this helps clear some things up. Please excuse the crudity of the block diagram, it's simplified and I'm only in the very early planning stages as yet, not onto design. Horizontal and vertical outputs go to a standalone scope.

[tggzzz]

OK, that doesn't look like a conventional curve tracer, i.e. one that can produce plots including

[Greg Robinson]

Ok, guess it more resembles an octopus circuit then?

[David Hess]

At curve tracer speeds, you do not need to terminate the transmission line between the external curve tracer and oscilloscope inputs.  Even up to 20MHz, no termination is required unless the transmission line is long.

You might want to check out how Tektronix implemented the vertical outputs on their old oscilloscopes.  They use a 950 ohm impedance to drive the 50 ohm transmission line.  Into a 1 megohm input, this typically yields 20+ MHz determined by the cable capacitance and into a 50 ohm input, it yields full bandwidth to 100s of MHz with an attenuation of x20.

[Greg Robinson]

Alright, seems we've got hung up on the 50 ohm termination thing. If you recall, in my first post I actually stated I DIDN'T want to do that.

My main question hasn't been touched on at all though.
Does anyone know if panel mount scope probe sockets (similar to the common probe-to-BNC adapters but without the intermediate BNC) exist?

[David Hess]

Does anyone know if panel mount scope probe sockets (similar to the common probe-to-BNC adapters but without the intermediate BNC) exist?

Printed circuit board versions exist but are expensive; they look like a Peltola socket.  Tektronix used panel mount probe tip sockets occasionally for the compensation output but I think they were custom made for them.

[gcewing]

I think what people are saying is that the sweep rate of your curve tracer doesn't need to be very high, so keep it low enough that transmission line effects in the cable are negligible. Then you don't need to bother with 50 ohm termination or impedance matching or any of that stuff. Just connect it directly to the scope inputs with an ordinary piece of coax.

[Greg Robinson]

Printed circuit board versions exist but are expensive; they look like a Peltola socket.  Tektronix used panel mount probe tip sockets occasionally for the compensation output but I think they were custom made for them.

Thanks David, I thought that would likely be the case, but if you don't ask you'll never find out.

[David Hess]

In theory Tektronix and LeCroy sell the printed circuit board sockets but considering that they are just a bent up piece of metal and essentially expendable, they charge an outrageous price.