DC coupled 2.7 GHz Active Probe Project - Now Available!

[luudee]

Dear Lasmux,

have you considered printing the enclosure from conductive ABS ?

Kind Regards,
rudi

[lasmux]

I did consider a conductive enclosure. I didn't go for this as I was concerned about increased capacitive coupling to ground on the signal pin, which would negatively affect the return loss on probed lines. I really want to keep that 0.7pF input capacitance, which was very difficult to achieve.

Another issue is that the enclosure is a very small print, and it has to interface with a real PCB. When I was prototyping I found that if the print wasn't dimensionally accurate enough, the PCB just wouldn't fit, especially around the probe tip socket area. Conductive ABS is normally printed via an FDM process which doesn't have the precision needed. Not to mention that FDM prints don't look as aesthetically pleasing as SLS prints.

[tggzzz]

I did consider a conductive enclosure. I didn't go for this as I was concerned about increased capacitive coupling to ground on the signal pin, which would negatively affect the return loss on probed lines. I really want to keep that 0.7pF input capacitance, which was very difficult to achieve.

Another issue is that the enclosure is a very small print, and it has to interface with a real PCB. When I was prototyping I found that if the print wasn't dimensionally accurate enough, the PCB just wouldn't fit, especially around the probe tip socket area. Conductive ABS is normally printed via an FDM process which doesn't have the precision needed. Not to mention that FDM prints don't look as aesthetically pleasing as SLS prints.

Ignorant suggeston: could you use a conductive paint? I assume you could mask off the areas where you want to keep capacitance low.

I've no idea about how well paint would or wouldn't adhere to the plastic, but it would be in the protected interior. The sintered nylon I've used is relatively rough (compared to SLA!), which might help.

[lasmux]

I did consider a conductive enclosure. I didn't go for this as I was concerned about increased capacitive coupling to ground on the signal pin, which would negatively affect the return loss on probed lines. I really want to keep that 0.7pF input capacitance, which was very difficult to achieve.

Another issue is that the enclosure is a very small print, and it has to interface with a real PCB. When I was prototyping I found that if the print wasn't dimensionally accurate enough, the PCB just wouldn't fit, especially around the probe tip socket area. Conductive ABS is normally printed via an FDM process which doesn't have the precision needed. Not to mention that FDM prints don't look as aesthetically pleasing as SLS prints.

Ignorant suggeston: could you use a conductive paint? I assume you could mask off the areas where you want to keep capacitance low.

I've no idea about how well paint would or wouldn't adhere to the plastic, but it would be in the protected interior. The sintered nylon I've used is relatively rough (compared to SLA!), which might help.

That could work tbh. The coating would have to be on the outside though as the enclosure at the front of the probe is not split into two (it's in one part) so you couldn't mask off the area near the front of the probe which is most susceptible to noise. As it's on the outside it would need to be reasonably aesthetically pleasing when applied. I've never tried to spray paint 3D printed parts before though. It sounds... time consuming...

[joeqsmith]

Concerning printing conductive plastics and coatings, I was looking into this as a way to produce low cost waveguide parts.  The video shows most of the end results.  The last attempt was to use a high content copper filament then bake out the plastics.   

https://www.eevblog.com/forum/rf-microwave/experimenting-with-waveguides-using-the-litevna/msg4646872/#msg4646872

[tggzzz]

I did consider a conductive enclosure. I didn't go for this as I was concerned about increased capacitive coupling to ground on the signal pin, which would negatively affect the return loss on probed lines. I really want to keep that 0.7pF input capacitance, which was very difficult to achieve.

Another issue is that the enclosure is a very small print, and it has to interface with a real PCB. When I was prototyping I found that if the print wasn't dimensionally accurate enough, the PCB just wouldn't fit, especially around the probe tip socket area. Conductive ABS is normally printed via an FDM process which doesn't have the precision needed. Not to mention that FDM prints don't look as aesthetically pleasing as SLS prints.

Ignorant suggeston: could you use a conductive paint? I assume you could mask off the areas where you want to keep capacitance low.

I've no idea about how well paint would or wouldn't adhere to the plastic, but it would be in the protected interior. The sintered nylon I've used is relatively rough (compared to SLA!), which might help.

That could work tbh. The coating would have to be on the outside though as the enclosure at the front of the probe is not split into two (it's in one part) so you couldn't mask off the area near the front of the probe which is most susceptible to noise. As it's on the outside it would need to be reasonably aesthetically pleasing when applied. I've never tried to spray paint 3D printed parts before though. It sounds... time consuming...

I'm presuming you don't want anything conductive near the tip, but you do want a conductive shield away from the tip.  I'm also presuming that, say, a 90%/0% coverage in the right/wrong places is better than a 0%/0% coverage, so "missed" coverage isn't too critical.

Hence, could you insert a plug, possibly 3D printed to the right shape, then (spray) paint the interior, then remove the plug? I have no advice about making contact between the paint and the circuit board.

There is also conductive silver-plated copper heatshrink https://ph.parker.com/us/en/series/electrically-conductive-heat-shrinkable-tubing.

There are also microwave absorbent foams, designed to minimise cross-coupling and cavity resonance https://ph.parker.com/us/en/series/microwave-absorber-materials

For the avoidance of doubt, I am not recommending Parker; they are simply the first products I came across They also do conductive paints and epoxy.

[lasmux]

@joeqsmith
Thanks, that's very intereting!

I'm presuming you don't want anything conductive near the tip, but you do want a conductive shield away from the tip.  I'm also presuming that, say, a 90%/0% coverage in the right/wrong places is better than a 0%/0% coverage, so "missed" coverage isn't too critical.

Yes exactly. The important thing is to keep the copper away from the signal tip and the first ~1cm of the probe body. This also applied to the PCB layouy. There are parts of the PCB where there is no ground plane under the signal traces to reduce ground coupling.

Hence, could you insert a plug, possibly 3D printed to the right shape, then (spray) paint the interior, then remove the plug? I have no advice about making contact between the paint and the circuit board.

There is also conductive silver-plated copper heatshrink https://ph.parker.com/us/en/series/electrically-conductive-heat-shrinkable-tubing.

Possibly could use a plug. On reflection, I think a spray solution will be too difficult to implement reliably, without also ensuring there's no conductive paint that accidentally finds itself in the wrong place. I think I'd have to spray directly into the 'cavity' where the plug is, and the pressure of the spray may allow paint to get where it shouldn't be, despite the plug. I might also look into a paint 'dip', where I can dip the back end of the enclosure to a controlled dip. Interesting about the conductive heatshrink.

[tggzzz]

@joeqsmith
Thanks, that's very intereting!

I'm presuming you don't want anything conductive near the tip, but you do want a conductive shield away from the tip.  I'm also presuming that, say, a 90%/0% coverage in the right/wrong places is better than a 0%/0% coverage, so "missed" coverage isn't too critical.

Yes exactly. The important thing is to keep the copper away from the signal tip and the first ~1cm of the probe body. This also applied to the PCB layouy. There are parts of the PCB where there is no ground plane under the signal traces to reduce ground coupling.

Hence, could you insert a plug, possibly 3D printed to the right shape, then (spray) paint the interior, then remove the plug? I have no advice about making contact between the paint and the circuit board.

There is also conductive silver-plated copper heatshrink https://ph.parker.com/us/en/series/electrically-conductive-heat-shrinkable-tubing.

Possibly could use a plug. On reflection, I think a spray solution will be too difficult to implement reliably, without also ensuring there's no conductive paint that accidentally finds itself in the wrong place. I think I'd have to spray directly into the 'cavity' where the plug is, and the pressure of the spray may allow paint to get where it shouldn't be, despite the plug. I might also look into a paint 'dip', where I can dip the back end of the enclosure to a controlled dip. Interesting about the conductive heatshrink.

All valid concerns, of course.

The design of the plug would be critical, of course. Would something like cotton wool be sufficient?

I'm sure there are conductive paints, so it may be possible to just coat part of the interior with a brush rather than a spray. A little mechanical experimentation is in order

Can I suggest starting a separate topic on the subject? I'm sure many people here have experience with generic EMI/EMC mitigation techniques/technology - and they may not be following a long thread about a scope probe.

[KungFuJosh]

Conductive paints are a gamble. You can do them with a brush instead of a spray, which is better, but... still a crapshoot. Many conductive paints require multiple coats to get any real conductivity. They also require specific conditions to apply them, and usually require 24 hours between coats, also depending on climate conditions. It's a lot of work and likely some headaches that may or may not pay off at all. Even if you get it good at first, the paint can eventually flake off and lose it's conductivity as it ages. It might still be better than nothing, but the labor involved doesn't seem like it's worth it to me.

If conductive heat shrink would work, I would try that first. If it works, cool.

[tszaboo]

@joeqsmith
Thanks, that's very intereting!

I'm presuming you don't want anything conductive near the tip, but you do want a conductive shield away from the tip.  I'm also presuming that, say, a 90%/0% coverage in the right/wrong places is better than a 0%/0% coverage, so "missed" coverage isn't too critical.

Yes exactly. The important thing is to keep the copper away from the signal tip and the first ~1cm of the probe body. This also applied to the PCB layouy. There are parts of the PCB where there is no ground plane under the signal traces to reduce ground coupling.

If you look at the design of the N2795, there is a hard plastic part and a soft silicone rubberized part, where the MK1 bionic grabber is supposed to handle the probe. I'm going to guess that this part is shielded for sure.

[Dennis Frie]

Thanks for a great product. A 2.7 Ghz active probe at that price-point is amazing.
This is now my go-to probe to verify signal integrity of digital interfaces. Compared to the Siglent SAP2500 we have at work, this is an amazing price-performance ratio.

[Mechatrommer]

Conductive paints are a gamble. You can do them with a brush instead of a spray, which is better, but... still a crapshoot. Many conductive paints require multiple coats to get any real conductivity.

i have MG CHEMICALS 841WB-150ML NICKEL CONDUCTIVE COATING bought from digi, its worth it sure better than nothing, but you need airbrush spray tool and compressor, tubing etc... using brush is a no no as conductive particles will just stick to the brush, not the enclosure to be coated. and EM protection is in term of dB, more layers, more dB attenuation, maybe 2 or 3 layer is ideal, more than that you wont get further noticable protection. if you tight arse, 1 layer can give you some protection already. ymmv.

It's a lot of work and likely some headaches

yes, painter's skills are really needed to be applied here, covering/masking from spray here and there etc... and conductive particles is special type, we need to keep stirring spray tank on each spray pass otherwise they will settle to the bottom very quickly and clog the spray tool.

If you look at the design of the N2795, there is a hard plastic part and a soft silicone rubberized part, where the MK1 bionic grabber is supposed to handle the probe. I'm going to guess that this part is shielded for sure.

most likely yes, my agilent 1141a has silver coat on inside of enclosure, its durable and scratch resistant type, unlike my water based coating type. ymmv.

[lasmux]

Thanks for a great product. A 2.7 Ghz active probe at that price-point is amazing.
This is now my go-to probe to verify signal integrity of digital interfaces. Compared to the Siglent SAP2500 we have at work, this is an amazing price-performance ratio.

Hi Dennis, thanks! I'm glad you're enjoying using the probes and are making good use of them It really brings me joy that they are providing a useful service!

@Mechatrommer, I've tried a couple of attempts at painting with a conductive paint, by hand. It's not worked very reliably, so I'm not planning on using that method going forwards. It's just not worth the time to be honest given the low price of the probe. I'll leave it to the customer to use copper tape if they are so inclined. I'm working on a new version of the datasheet with instructions on applying copper tape properly.

I only have one probe remaining in stock. I am in the process of ordering/manufacturing more. I've unfortunately been hit by part shortages up till just a week ago, but they seem to be resolved now. My PCBA manufacturer is about to start kicking into gear in the next few days.

[KungFuJosh]

@lasmux,

It might be easier to go the other way- use a conductive material for the case (or part of it) and then insulate it.

Thanks,
Josh

[lasmux]

@lasmux,

It might be easier to go the other way- use a conductive material for the case (or part of it) and then insulate it.

Thanks,
Josh

Yeah, this is something I've been considering. I originally discounted the option, but I think it could be done. Perhaps a conductive body, but with just the tip around the inputs as a non-conductive part.

Probes are now out of stock. I've put up an interest form on the product page so you can get notified when they go back in stock.

[KungFuJosh]

@lasmux,

It might be easier to go the other way- use a conductive material for the case (or part of it) and then insulate it.

Thanks,
Josh

Yeah, this is something I've been considering. I originally discounted the option, but I think it could be done. Perhaps a conductive body, but with just the tip around the inputs as a non-conductive part.

I think that option is 100x easier than applying shielding reliably otherwise. You might want an insulated coating on the whole outer body, but even a good powdercoating might be sufficient and would like nice. You can have a powder coater insulate both sides, and mask the contact points you need to keep the shielding solid.

[Mechatrommer]

either way, the shielding need to be connected to the circuit ground.

[nctnico]

I don't know how the PCB design is currently, but it could help to put all signals on inner layers and have as much ground on the top & bottom layers as possible.

[lasmux]

either way, the shielding need to be connected to the circuit ground.

Yeah for sure. It would be connected at the SMA side-launch connector, as the copper tape shielding method is.

I don't know how the PCB design is currently, but it could help to put all signals on inner layers and have as much ground on the top & bottom layers as possible.

That's not really feasible with this design.

I've had another look, and a conductive enclosure will really push the BOM price up, and that would push the probe away from the ethos of great performance at a very low price. I think the method of using a deliberate application of copper tape is cheap and easy to apply if the user thinks the noise pickup is too high. Again, this would only be the case where the noise level is high enough and measured signal small enough simultanously.

[KungFuJosh]

I've had another look, and a conductive enclosure will really push the BOM price up, and that would push the probe away from the ethos of great performance at a very low price.

By how much? If it's not too significant, and if it accomplishes the goal of improving the probe, I'm pretty sure people would happily pay it.

[Mechatrommer]

and i suspect, switching to shielded design from open space no hand condition will throw away the probe's output profile requiring another series of R&D, profile tuning and/or pcb respin drama.. ymmv.