5
\$\begingroup\$

Apologies if this seems like a duplicate of this question but the eventual "answer" to that seemed somewhat anecdotal at best - and there's been a LOT more of these used since 2012 so things may have moved on.

Basically I'm looking at a very similar situation - specifically mounting a Raspberry Pi CM4 module using the two Hirose DF40HC(3.0)-100DS-0.4V(51) connectors. (Manufacturer web page)

Edit: These connectors do NOT have any locating pins or other features, they are what's specified (and used by) the Raspberry Pi foundation for their CM4 and its IO board, so obviously they are being aligned in mass production by them.

My experience with prototype units using these connectors (professionally hand soldered) was that a small misalignment made the module hard to plug in and it concerns me that that's putting undue stress on the connector or the PCB traces.

Best I can find in the datasheets are a line that says "Keep connector warpage within 0.02 mm in connector longitudinal direction" which is pretty darned tight, although later they claim "Guide ribs utilizing dead space ensure 0.33mm of self-alignment" but don't really specify in which direction(s).

So - the question is, is there some proper way to ensure alignment in production (reflow) or do these things really self-align accurately enough when placed on the standard footprint and run through an oven?

\$\endgroup\$
4
  • 1
    \$\begingroup\$ Many such connectors have locating pins that insert into non-plated holes to hold them in position. \$\endgroup\$
    – Hearth
    Commented Dec 11, 2023 at 16:41
  • 1
    \$\begingroup\$ Yes but these ones don't. That's part of the problem. \$\endgroup\$
    – John U
    Commented Dec 11, 2023 at 16:42
  • 3
    \$\begingroup\$ If your design is sensitive to misalignments, I would suggest that you use connectors with such alignment features. \$\endgroup\$
    – Hearth
    Commented Dec 11, 2023 at 16:44
  • 6
    \$\begingroup\$ If we're using the Raspberry Pi CM4 module we have to use connectors that match the ones on the module, we don't have a choice. \$\endgroup\$
    – John U
    Commented Dec 11, 2023 at 17:04

4 Answers 4

8
\$\begingroup\$

My experience with these and other similar connectors is that reflow (as opposed to hand soldering) does indeed provide excellent alignment.

As long as you have the correct pad placement, and the connectors placed roughly right on the solder paste, then there is a very strong self-aligning action when the flux and solder melts due to surface tension.

This is missing with manual soldering.

If there are other components that can't be reflowed (very uncommon), then reflow the board with the connectors and then hand solder the remaining parts.

Whilst I personally prefer connectors with alignment pins for mechanical strength, this is the same process which means even very high density BGAs align perfectly, time after time.

Given that some of these BGA pitches are as small as 0.4mm, with final misalignment of less that 0.1mm causing massive issues, you get a feel for the abilities of the self-aligning process. However, it is worth noting that pad and mask layout is absolutely critical, and you need to follow manufacturer guidance, in particular whether to use solder mask defined or non-solder mask defined pads.

(For what it is worth, I've used a pair of very similar Samtec connectors for attaching high-density DSP mezzanines to a parent ATCA board for telco use. We've had tens of thousands of these units produced, and never an issue with alignment. Having said that, my preference today for 20Gpbs+ ultra-highspeed use, would be the newer Q-Strip connectors which have a row of midline high-current power/ground blades, both for better power deliver/shielding and also absolutely positive location.)

\$\endgroup\$
3
  • 3
    \$\begingroup\$ Thanks for an informative answer based on real experience! It will never catch on round here ;) \$\endgroup\$
    – John U
    Commented Dec 12, 2023 at 11:23
  • 1
    \$\begingroup\$ I've always viewed it best to learn for other people's experiences if possible, but you're right that it's an outlier approach ;-) \$\endgroup\$
    – colintd
    Commented Dec 12, 2023 at 11:28
  • 2
    \$\begingroup\$ +1 for the Q-Strip from Samtec - they are the best and allow for some misalignment. \$\endgroup\$ Commented Dec 12, 2023 at 12:49
4
\$\begingroup\$

Evidently, it is possible, or, some are doing it. I notice the Waveshare Mini Baseboard,

enter image description here

(image obtained from: https://www.amazon.com/Raspberry-Standard-Color-Coded-Evaluating-Integrated/dp/B095CSRWXS?th=1 Mfg page: https://www.waveshare.com/wiki/CM4-IO-BASE-A (no affiliation))

But that doesn't mean it's a good idea.

These are rigid connectors, meaning any positional mismatch of connectors on the two boards is likely to cause faulty connections -- compressing or stretching the contacts themselves, and by how much, depends even on the exact temperature and expansion rate of both.

The most precise alignment I can think of, would be a self-aligning process. Let's talk it through, to see how it would(n't) work.

You could try stacking the boards together, with one set of connectors pasted but not soldered. Send both boards -- the assembled and positioned stack -- through reflow. Either one board serves as positional reference, or both melt, and in either case the solder joints are aligned perfectly with no stress on the connectors. Mind, even getting them together in this way, is already a challenge; even with precise placement and an alignment jig, paste smear and risk of shorts is a risk here. Let alone bumping anything else, if you have nearby components and doing everything in one reflow operation.

We already have one immediate problem: the connectors are, at the very least quite soft, at soldering temperature. The plastic may indeed melt. Normally this isn't a problem, as the plastic has a gel-like consistency that retains its shape. But now we have two connectors wedged together, retaining the force of all those spring contacts. They may bulge apart, or lose planarity (on the soldering face), resulting in loose contacts, weak or broken solder joints, etc.

The problems only continue from there. The connector positions are frozen in place when the solder solidifies. If the boards have identical composition, they can have equal thermal expansion rate, and retain dimensional accuracy down to room temperature. But in general, boards will be made on different processes, from different materials, and almost certainly won't have equivalent copper density and component choice -- note that having a bunch of BGAs on one side increases stiffness of the board in the area, changes the overall expansion rate, etc. On dimensional scales of 10s µm, everything matters. The board surface itself warps in various ways, due to local differences in expansion (like components and copper areas). Displacements of this scale, multiplied by the elastic modulus of the board, result in forces of -- I would guess, in the ballpark of single to tens of newtons, at the connectors. Which is a lot for a little connector like that.

And even if your boards were fabbed on the same process, same materials, the fact that stiffness varies with copper density, board thickness, exact material condition (the material itself may vary enough from batch to batch to matter), means you'll have a hard time getting them to track on the cooldown from soldering to room temperature.

So what can we do about it?

Stiffness is the key. The looser the boards are positioned apart from the connectors, the less side force bears on them. A relatively thin-limbed 'U' shape board could be used, with a large gap between connectors, allowing flex between them, and thus more positioning error. Still maybe not a lot, but it's a start. We can compound this further by using much softer board material -- this type of connector is very popular with flex boards, which can be bent around corners as well; they're especially useful in compact electronics like cameras and cellphones.

But this isn't very convenient for rigid board-to-board applications, and the recommended approach (use just one connector) is best for that case.

\$\endgroup\$
4
\$\begingroup\$

Your question cannot be answered by anyone on this site.

This is a very good question to raise, and is a concern you should have. You should talk about this with your favourite assembly house and verify with them whether they can meet the tolerance that the manufacturer requires. The tolerance from the MFG side by the way, is also something you should ask the MFG; not something easily found in the datasheet usually.

I can offer you my anecdotal experience, I worked on a project where whe had to put 6 connectors in an area of around 10x10 cm, for a total of a few hundred pins.

After obtaining the required tolerance from the connector manufacturer, we sent the query to the assembly house, and they asked us to put several fiducials in the connectors area, in order to minimise the error of the pick and place machine.

You have one additional variable here, which is the CM4 module. You can try to ask to the Raspberry Pi Foundation about the tolerances they can guarantee on their boards, they should be able to provide an answer to that. In our case we were producing both boards, so in the end the assembly house was able to guarantee that they were able to meet the specification from the manufacturer.

EDIT

I just opened one of the files from the page you have provided:

enter image description here

So there you have it, the manufacturer already provides you with the required tolerances, which are impossible to meet from any reputable assembly house - any non reputable one might just lie, and still not meet them.

Link to the document: https://www.hirose.com/en/product/document?clcode=CL0684-4151-0-51&productname=DF40HC(3.0)-100DS-0.4V(51&series=DF40&documenttype=Guideline&lang=en&documentid=0001442210

\$\endgroup\$
7
  • 2
    \$\begingroup\$ You've got a better eye than me, I did not spot that in the documentation! +/-0 is a fantasy tolerance though, if that were true no-one would be able to make any boards using these and clearly (with many many thousands sold) the CM4 using these must be manufacturable. \$\endgroup\$
    – John U
    Commented Dec 11, 2023 at 17:30
  • \$\begingroup\$ Of course it is manufacturable @JohnU, the problem is that the connector mfg is not taking responsibility for this. Unless you find an assy house that can give you ±0 tolerance, you will need to take responsibility for this. You seem to think that the Rpi Foundation is aligning them somehow, but they are not. They are lucky enough that it just works, and they decided to take this risk - or maybe they are unaware such a risk exists. Since you asked yourself this question, you probably do not want to take this risk, for one reason or the other, and therefore should drop the Rpi CM4. \$\endgroup\$ Commented Dec 11, 2023 at 20:53
  • 2
    \$\begingroup\$ I doubt the Pi foundation decided to mass-produce these based on "Meh, close enough, send it!", they may not be splitting the atom but they're not bad at engineering these things. \$\endgroup\$
    – John U
    Commented Dec 12, 2023 at 11:15
  • 3
    \$\begingroup\$ @VladimirCravero The Rpi version of the story is that they didn't disregard/overlook it per se, Hirose changed the datasheet to recommend against doing this after Rpi finalised the CM4 design. Source \$\endgroup\$
    – Jack B
    Commented Dec 12, 2023 at 18:06
  • 1
    \$\begingroup\$ @JackB that's a great insight to hear. I guess I won't be using hirose any time soon then... That's not a great move. \$\endgroup\$ Commented Dec 13, 2023 at 9:20
0
\$\begingroup\$

While I do agree with Hearth that you would be best served with mechanically aligning connectors, and if your SOM vendor doesn't have these, to seriously consider a different vendor¹ with a different module with different connectors:

Alignment, even in "home" assembled PCBs is relatively OK, if you manage to do a somewhat (not even very) consistent application of solder paste and reflow them – with a 60€ hot air station, or, better, a repurposed used pizza/baguette oven, and the thermocouple that came with your multimeter to monitor the temperature. However, that experience only applies to single connector boards with some "screwhole" tolerances. I don't think as someone who "consumes" compute modules instead of standalone boards, you should even care about achieving this accuracy yourself – it's up to your board manufacturer. Talk to them! I don't expect such connectors to be "professionally hand soldered"; these are picked, placed and reflow soldered in one mechanical process.


¹ I admire the Raspberry Pi foundation for what they've done for the pervasive availability of application processor embedded boards. I technically like none of their products very much, and I'm regularly very surprised that an industrial user would want a TV settop box multimedia SoC to drive their automation. So, I do think you will find good alternatives to RPi CMs, especially since during the covid chipocalypse, RPis were just as unavailable, if not worse, as other embedded platforms, the availability argument has suffered a lot.

\$\endgroup\$
6
  • 1
    \$\begingroup\$ Show me an industrial vendor with a better overall offering than the Raspberry Pi though - in terms of support, documentation, software, availability and capability for the price they are ridiculously hard to beat. We're not opening anyone's parachute with this, we need a powerful module that can drive an HD touchscreen with a GUI and the CM4 is superb value in that application. \$\endgroup\$
    – John U
    Commented Dec 11, 2023 at 17:27
  • 1
    \$\begingroup\$ that's not hard. RPi foundation support is pretty shoddy, their software SDK architecture is more than questionable for someone with an automotive/openembedded background, availability of Raspberry Pis over the last 3 years has been abyssimally bad, the central SoCs have no long-term manufacturer dedication, and you're typically not buying 5 of such modules, but more like 50,000, where prices become negotiable. You'll find that I linked multiple alternatives that will play out to be in the same price range. \$\endgroup\$ Commented Dec 11, 2023 at 17:31
  • 1
    \$\begingroup\$ which documentation, by the way? Compared to most SoC, the datasheets you get for the BCMxxx on the RPis are a bad joke. \$\endgroup\$ Commented Dec 11, 2023 at 17:37
  • 2
    \$\begingroup\$ The Pi documentation is very good, very clearly written, and the OS support is great - this is not my first SoC/SoM project and the Pi is not the only device we evaluated for this application. You're assuming we're buying 1000's of modules, we're not, my project is not your project so maybe my criteria are not your criteria? Also THAT'S NOT THE QUESTION! The question is "we're producing a board with these connectors on, how best do we do that?" \$\endgroup\$
    – John U
    Commented Dec 12, 2023 at 8:52
  • 1
    \$\begingroup\$ That's the question you asked me in the comments. I just answered to that comment of yours there. \$\endgroup\$ Commented Dec 12, 2023 at 9:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.