Using surface-mount RF coax connectors as general purpose test points

[jebiga]

Hello,

I often find myself having to probe PCBs with lots of surface mount components of varying shapes and sizes. Think tall electrolytic caps, lots of varying connectors, debug headers, standoffs, etc. Because of all the "clutter" on the board, it is difficult to measure voltages on the board with a scope. I usually sprinkle around a bunch of test point pads, but accessing the test points with the scope can be difficult when probing 4 signals simultaneously. Furthermore, getting a good ground at the same time can be difficult because of the limited space. I already have a set of PCBite probes, and while they are useful they're still a bit finicky with small test pads.

This inconvenience has me thinking: why not use an RF connector, such as the Murata SWG-series MM8030 SMD connector, for my test point needs? These are cheap (around 10 cents/piece in the quantities that I'd buy), and give me both signal and ground which I can connect directy to my scope by using the Murata SWG probe and an SMA to BNC adapter that Murata also sells. This would allow to me to conveniently probe signals around my PCB without having to deal with test pads or ground clips.

Are there any pitfalls with this approach that I'm not thinking of? I would mainly use this for non-RF probing, such as probing PWM lines, shunt voltages, I2C/SPI/UART busses, and some analog signals. My scope has a 50Ω input so as far as I can tell, I just need to adapt the SWG connector to BNC and I can plug it directly into my scope. No ground clips needed! Will the 50Ω impedance have any issues with the types of signals I plan on probing?

Also, the Murata SWG series is a switched connector, so when I plug in my probe, the signal will be disconnected from the rest of the circuit. This isn't ideal for my purpose, as I would like to see these signals "in action". Is there any issue with modifying the footprint to short the switch out so that the signal is always connected whether a probe is connected or not?

Thank you in advance for any advice you may share.

[srb1954]

Are there any pitfalls with this approach that I'm not thinking of? I would mainly use this for non-RF probing, such as probing PWM lines, shunt voltages, I2C/SPI/UART busses, and some analog signals. My scope has a 50Ω input so as far as I can tell, I just need to adapt the SWG connector to BNC and I can plug it directly into my scope. No ground clips needed! Will the 50Ω impedance have any issues with the types of signals I plan on probing?

Probing logic signals such as buses and general analog signals directly into a 50 probe is not likely to be very successful as the loading will be far too great for most chip types and they will not be able to provide the drive current required. For example, if you have a 3.3V logic signal and want to drive it high into a 50 load your driver ship will need to supply 66mA. Not many logic ICs can safely manage this drive current.

For general purpose probing a 10M scope probe is more appropriate. You can get adapters to go on to the end of the scope probe to adapt to BNC and then another BNC-SMA adapter to go to your Murata cable.

A word or warning though. With a 10M probe on the end of the adapter cable it will make the input of the cable appear capacitive and that extra capacitive load may be enough to cause some op amp circuits to go unstable.

[jebiga]

Thank you for your response and helpful feedback.

I see the issue you're pointing out with the 50Ω impedance causing problems with the lines I'm probing. The driver on those lines is an MCU and like you said it definitely won't drive anything over 20mA, if that. Back to the drawing board on this whole Murata SWG connector...

I have 10MΩ probes, however dealing with grounding them is annoying on a high-density board like mine and so I wanted a test point connector that had signal and ground. Something that would last more than 2 cycles (U.FL, I'm looking at you) would be nice too. As far as I can tell it doesn't seem like anything exists in that category, unfortunately.

Edit: thank you for the suggestion of the probe-end to BNC to SMA to Murata setup. I misread that originally, and now upon re-reading I see that it might actually be a good solution for my usecase. I will investigate further.

[srb1954]

Thank you for your response and helpful feedback.

I see the issue you're pointing out with the 50Ω impedance causing problems with the lines I'm probing. The driver on those lines is an MCU and like you said it definitely won't drive anything over 20mA, if that. Back to the drawing board on this whole Murata SWG connector...

I have 10MΩ probes, however dealing with grounding them is annoying on a high-density board like mine and so I wanted a test point connector that had signal and ground. Something that would last more than 2 cycles (U.FL, I'm looking at you) would be nice too. As far as I can tell it doesn't seem like anything exists in that category, unfortunately.

Edit: thank you for the suggestion of the probe-end to BNC to SMA to Murata setup. I misread that originally, and now upon re-reading I see that it might actually be a good solution for my usecase. I will investigate further.

Tektronix have a PCB mount jack that a standard probe directly plugs into. It simultaneously makes the signal and ground connection and provides mechanical support for the probe.
 
https://www.ebay.com/itm/203787569348

Ridiculously expensive for a little bit of bent metal but you may be able to find something equivalent from another manufacturer.

[Kleinstein]

For logic signals one can reduce the loading by adding a series resistor towards the test points. This makes it a Z0 probe when used with terminated (e.g. 50 Ohm) cable. E.g. with 1 K of resistor one would get a 21:1 probe with moderate loading and good signal quality. If one goes as far as adding connector, why not add a resistor too.
A shunt resistor would often be OK with 50 / 100 ohm loading.

[langwadt]

For logic signals one can reduce the loading by adding a series resistor towards the test points. This makes it a Z0 probe when used with terminated (e.g. 50 Ohm) cable. E.g. with 1 K of resistor one would get a 21:1 probe with moderate loading and good signal quality. If one goes as far as adding connector, why not add a resistor too.
A shunt resistor would often be OK with 50 / 100 ohm loading.

http://jahonen.kapsi.fi/Electronics/DIY%201k%20probe/