Who makes mid-mount right angle terminals like this? (updated with project pics)

[ajb]

I just received a PCBite probing kit, and the system works very well.  I'd like to make a couple of different probe heads for it, but I haven't been able to figure out who makes the terminal they use to attach to the arms.  It's an M4 thread (which matches the studs on the ends of the arms), about 11mm long, and solders into the middle of the board (soldered only on one side).  I've seen plenty of terminals that mount on top of the board that would work, but a mid-board solution like this would be nicer--both lower profile and sturdier than a conventional SMT solution.

Any ideas?

[free_electron]

looks like something PEM would make.  pemnet.com or these guys : https://lugsdirect.com/SMT-Surface-Mount-Technology-PrintedCircuitBoard-PCB-WireTerminalLugs-IHI.html

[ajb]

Yeah, I came across both of those while searching, but didn't see anything like this at either one. 

[free_electron]

to be honest, i don't like that construction. i would have made the nut slotted so the board slides into it and is soldered from both sides. This looks too fragile to me.
and pinheaders ? i would have used MCX or some other small, light coax pigtail to SMA or BNC. run the coax along the gooseneck.

[ConKbot]

How big is the production run of this?it looks like an RF  connector drilled and tapped with a thread. I've used either MMCX or SMP connectors before with pretty much the exact same proportions.
Edit:
Something like molex 0734150969, ram a 3.3mm bit though, tap it M4, belt sand off stubs of legs if desired.

[ajb]

to be honest, i don't like that construction. i would have made the nut slotted so the board slides into it and is soldered from both sides. This looks too fragile to me.
and pinheaders ? i would have used MCX or some other small, light coax pigtail to SMA or BNC. run the coax along the gooseneck.

Meh, I'm not especially worried about the mechanical robustness here, it's never meant to leave the bench.  They do at least have a line of vias in the solder pads FWIW.  They make oscilloscope probes that do have a mini coax connector, but for general purpose probing I like pin headers as a good common denominator.  These will often get used with logic analyzers or serial adapters or other things where pin headers are an easy solution.  I even made some scope probe to pin header adapters that I often use for basic probing and will absolutely use with this system.  Obviously there are times when you need a direct probe contact with nice short ground spring to see what you're looking for, but there are lots of times when you don't and pin headers with flying leads are just fine. 

How big is the production run of this?it looks like an RF  connector drilled and tapped with a thread. I've used either MMCX or SMP connectors before with pretty much the exact same proportions.
Edit:
Something like molex 0734150969, ram a 3.3mm bit though, tap it M4, belt sand off stubs of legs if desired.

I might want to make....I dunno, a total of 20 alternate probe thingies ever?  We actually just took delivery of a fancy new CNC lathe at work (dual spindle, Y axis, live tools, etc) that could crank these out by the thousands from raw bar stock, but unless these things are like $30/ea it'll be a lot more expensive than buying them once you factor setup time, tooling, and machine time that isn't spent on what it's supposed to be making 

[eugene]

Am I correct that the part you're looking for is just a hunk of metal with an internal M4 thread? And you only need something that is functional but does not need to be identical to the one that Sensepeek uses?

Can you be a little creative and use something from your local hardware store or McMaster?

[Ian.M]

Slot the sides of a M4 threaded hex spacer?

[ajb]

Yes, this is purely about getting an M4 threaded hole close to the plane of the PCB.  The ones Sensepeek use look like they'd be ideal, so I'd love to find the same ones if they're available off the shelf somewhere.  A brass M4 standoff would be a good backup option--I wouldn't even worry about slotting it, probably, just make a notch in the PCB a little bit narrower than the corner-to-corner distance of the hex. That should get the centerline close to the surface of the PCB and allow good fillet formation. 

[thm_w]

https://www.digikey.ca/en/products/detail/keystone-electronics/7768/4499348
A 2mm PCB would almost fit in the center of these. But I think the plain brass M4 standoff is better idea

You have a CNC lathe but no manual mill there? It would not be too much work to set the standoff in a vice with endstop, cut, flip standoff, cut again, repeat for each part.

You could also set the hex standoff right in the center of the PCB, and solder on both sides. Fully relying on the solder to form the mechanical strength. Again you'd want some vice/jig setup to sit it right in the middle.

[T3sl4co1l]

Y'mean like edge-mounting the terminal?  Hmm, no I guess not, that would be 3.5mm on that one.  But that'd be another possibility still, I suppose -- and stronger than the OP part for the above mentioned reasons.  So, if you don't mind using 3.2mm board, or can afford to stack another 1.6mm piece on top (and solder them all together, and don't mind that it's off center; or, use 0.8mm pieces top and bottom to keep it centered; or use whole-ass metal instead of clad PCB stock, easier to solder?), that might be doable.

For my part, I haven't seen that exact part before.  And yeah, PEM is probably the go-to for that sort of thing; or a few others, I want to say Wurth has a broad selection of PCB hardware too?

I did flip through Keystone's catalog, and realized they don't have anything like that, solder nuts or anything, at least that I could tell.  So, just brackets and terminals.

Other ideas: take a thicker kind of nut, hacksaw (or slitting saw or end mill if you want to be fancy / need to be precise) slots in the sides, and solder into a slot, preferably with castellations so it solders all around (i.e. metallized board edge in the slot).  Or starting with internal threaded rod stock, or standoffs.  Could even do without the slots, if you don't mind the solder joint being that much thinner.  It will be fiddly to solder, you'll need a jig to hold it in place, and then some means of soldering it; squirt some paste in there and reflow is fine, or for hand soldering, use hot air anyway, or just to assist, or use a heavy duty soldering iron, or a regular one plus patience.  No big deal.  Brass is easy to solder, steel is passable, or preferably either one tin plated; avoid nickel plating, it sucks to tin, and, aluminum and stainless are right out of course.

Heh, could even have metallized ceramic standoffs doing this.  Would be very custom, but definitely something that could be made.  Not sure what good the extra insulation or temperature handling would be, when it's already mounting into an insulating material with a low-melting filler....

Tim

[ajb]

I needed to order some other boards anyway, so I threw together a design for an unmodified 6mm hex standoff and sent that off yesterday.  A 6.16mm wide notch should hold the standoff such that the axis of the hole will be coplanar with the pogo pins, which works out nicely and will probably have zero impact on how well this all works.  Boards should be here in about a week, then we'll see how easily it solders in that position.  I think I have a decent idea for how to fixture it while soldering. 

You have a CNC lathe but no manual mill there? It would not be too much work to set the standoff in a vice with endstop, cut, flip standoff, cut again, repeat for each part.

Who needs a manual mill when you've got a seven ton 30HP VMC    (A smaller mill is actually on the list now that we have more space and the big machine is busier with production--TBD if it'll be manual, CNC, or hybrid.) But if the unmodified standoff idea doesn't work, yeah, I can mill some notches in it and try again.

And yeah, PEM is probably the go-to for that sort of thing; or a few others, I want to say Wurth has a broad selection of PCB hardware too?

Wurth was my first thought, I'm using their SMT terminals on a current project, and they do have a decent selection, but nothing quite right.  The right-angle options are all above-board and top out at M3 in SMT.  Their press-fit options go up to M10 (!), which I could probably solder, but at that point something like one of the Keystone-style stamped parts would be a lot easier.

[T3sl4co1l]

Another idea: use two boards as right-angle risers, to support a solder nut in front and behind.  Notch them so they fit into the board slot, and use board edge metallization (castellation) on both joints.  Everything self-aligns reasonably well, and just pushing it together is all the clamping it should need.

A single right-angle board wouldn't be very strong, especially without edge metallization (maximum just four thin solder fillets each side holding it in), but if you don't need much, that'll do.  Also need not be much bigger than the nut, basically just adding to its width to get the solder joints.  Downside: have to cut the notches either deep enough so the mill radius clears the main board, or dogbone the corners at the root of the notch, so it can fit flush.

Tim

[thm_w]

I needed to order some other boards anyway, so I threw together a design for an unmodified 6mm hex standoff and sent that off yesterday.  A 6.16mm wide notch should hold the standoff such that the axis of the hole will be coplanar with the pogo pins, which works out nicely and will probably have zero impact on how well this all works.  Boards should be here in about a week, then we'll see how easily it solders in that position.  I think I have a decent idea for how to fixture it while soldering. 

Good point the pogo pin axis is above the board anyway. Please post photos once you've tried it.

[ajb]

I managed to finish all of my important work with about 90 minutes to spare before starting an overdue vacation yesterday, and I was determined to get this whole idea together and see if it works before I left the lab. 

TL;DR: soldering an unmodified standoff in a notch in the board works great, the overall idea doesn't work yet, but I have a couple of options for addressing the problems. 

In the end I didn't bother with any special fixturing.  I just held the boards horizontally in a little Panavise and the standoffs just settled nicely into the notch.  These are 6mm standoffs with a 6.16mm notch, so the bore of the standoff is about .6mm off the surface of the board.  There are ~2mm wide pads along the sides of the notch with a handful of vias for soldering.  The notch is also well sized to rotate the standoff by 30° and center it in the board, so the 2mm wide pads are repeated on both sides of the board to allow a fillet on both sides in that case, although that would require some sort of fixturing.

The standoffs I had were zinc plated brass, and the SAC305 wire/paste I have didn't wet well to the plating, so I used a sanding stone to remove it from the two sides to be soldered.  I smeared a little paste onto the raw brass surfaces before placing the standoffs to help prevent oxidation before the paste reflowed, then used hot air with a small nozzle aimed into the threaded hole of the standoff so all of that mass could come up to temperature first.  Worked pretty well, although syringing the paste onto the pads would probably produce better results than just globbing it on with a pick like I did. 

(Apologies for the not-great pictures, I was in a bit of a hurry.)



I actually did a soldering test the other day and then did some destructive testing.  With fillets on one side like this, I was able to break the standoff from the board, but it took all of the copper with it, and only after applying enough force that I have zero concerns about the durability of this method for any reasonable use--especially for a piece of benchtop probing equipment.  If you were concerned about it, the underside of the joint leaves a couple of little trenches between the standoff and the edges of the PCB that could be backfilled with epoxy, or the board could be edge-plated and soldered more extensively I guess.

Of course as soon as I submitted the order for these boards I realized I should have designed them differently.  I originally planned to solder pogo sockets in two places with a slot in the board between to allow for a spacer to hold them in place, but that's really not necessary for strength and increased the width of the whole board.  So after soldering the standoffs I chopped the boards in half before finishing the assembly. 



The black plastic pieces around the pogo sockets are laser cut from 3mm delrin.  The holes were laser cut undersize and opened up with a drill to a nice press fit for the pogo sockets, which are all assembled together before soldering to the board to make sure they end up nice and straight.  In addition to the 50mil header shown, there's a footprint for an FFC connector on the other side.  The traces are length matched to the FFC connector just because, and there are shorted 0402 footprints for optional series components. I couldn't immediately find shrouded 6-pin 50mil SMT headers that fit the standard shape IDC plugs, so will need to look around a bit more for that.

The companion to the little probe head is a breakout board, which is reversible to break out either an FFC connector or a 50mil header to a 100mil header. That also gets a laser cut piece of delrin, with a C-shaped bit sized so it slides onto the gooseneck part of the PCBite arm and then drops down around the thicker boss on the base of the arm to hold it in place. 

With everything assembled, here's the whole setup:



My standard approach to test points is 1.5mm round SMT pads on a 2mm grid.  For most of what I do, this is a good compromise between density and accessibility.  It's a useable pitch for bed-of-nails fixtures,  it's fairly easy to get a probe tip onto or solder bodge wires on, being SMT it only occupies one side of the board, and for longer term probing (like on development specimens) I can solder 2mm headers on.  With something like this whole probe-and-arm setup it would be even faster and easier to tap into a bunch of signals at once. 

Now for the bad news.  All of the pogo pins I have on hand are fairly stiff, so the PCBite arm head doesn't have enough weight to appreciably compress even one of them, let alone level out all six.  This does not bode well for the ten-position version I designed but it's also something I expected might be an issue.  Since the pogos are socketed, I'll see if I can find some real pointy ones with nice soft springs and see if that solves the problem.  If not, I have some knockoff Noga-style arms that I'm confident can be made to work with this kind of probe head.

Here's a closeup of the probe head.  It just so happens that the way I soldered this standoff on ends up with the threads clocked such that the PCBite logo on the arm is aligned to the board surface which is a cool accident!



And here's the breakout board attached to the base of the arm.



I sent off another PCB order a couple of days ago, and threw in a revision of this and the 10-position board before I even assembled this one.  The revised design gets rid of the unnecessary double-soldering strategy for the pogo sockets, which slims the whole board down to just under 12mm wide (so you could conceivably put a few of these right next to each other to probe as many signals as needed--and you can remove some pogo pins from their sockets to probe other numbers of signals), reduces the copper-to-outline clearance a bit to improve solderability of the standoff, increases the size of the vias in the standoff pads to provide even better anchoring, and adds a pair of similar size vias (untented on the bottom) to each of the pads where the pogo sockets are soldered to better secure them as well (although I'm not particularly worried about them as they are).  Those should be in once I'm back from vacation and I can build a new rev, and try out some of the other improvements.  If it works out I'll happily share the designs.

So, not quite there yet, but I'm happy with the standoff solution and the overall design and concept, and with some revision I think these can become some really useful tools.  I can imagine using this technique to make other probe heads, like little active probes, or even shielded two-position versions at a range of spacings for better passive 'scope probing.

[T3sl4co1l]

Neato!

For clamping in the mean time, would a counterweight do the trick, or say, something mounted to the bench (much heavier weight? magnets? tape? do it near the edge of the table and use a clamp?) and add a tab to latch/clamp the arm down against the board?

Tim

[ajb]

I thought about adding some weight to the head, but given how much force it takes to compress these pogos I worry about the whole arm flopping over by the time I add enough weight to get a reliable connection.  I have a couple of ideas for adapting this to the standard indicator dovetail mount that the single-knob arms have, so I may just go that route and leave the PCBite arms for the standard single point probes.  I'm not in such a hurry to use this setup that I can't wait to get that worked out, and either way I won't mind having a standard M4 mounting thread on everything.

Plus it will be a nice exercise in patience to try to keep from going in to the office to play with this while I'm on vacation.