Durability of 'Tag-Connect' connectors?

[BreakingOhmsLaw]

I am thinking of using Tag-Connect footprint in a product. Does anyone have experience with them and can give information about the durability of the locking variant?
Looking at the image, those plastic hooks don't instill much confidence. Given the price tag, I'd hate to replace the cable on a regular basis.
https://www.tag-connect.com/product/tc2030-ctx-stdc14-for-use-with-stm32-processors-with-stlink-v3

Any other experience with Tag-Connect also welcome.

[Someone]

Over a thousand cycles and still holding ok here, ham fisted blind monkeys on a production line might well produce shorter lifespans.

[ataradov]

My usual comment with Tag-Connect is that they are pointless compared to a regular connector. They take up more board space and require big mechanical holes complicating routing on dense boards.

They are not much faster to disconnect, and who needs to connect/disconnect on a regular basis where the difference would matter?

If anything, the version with the metal pins and a retaining clip is better, but still way more cumbersome compared to the regular connectors.

And using them in production is just strange, I don't see how it is efficient at all.

[ejeffrey]

What is your use case?  I really prefer the "no legs" version.  I've used those a ton and they are great.  For device programming it's usually enough to just push and hold for the few seconds it takes to load the flash.  The only reason to use the connector with the plastic legs is if you need to do debugging or slow flashing on a device where it's not practical to access the back side to install a retaining clip.  If that use case is a major concern I would definitely reconsider just using a standard rectangular connector.

[ArdWar]

My usual comment with Tag-Connect is that they are pointless compared to a regular connector.

Zero BOM, low wear (clipless one), and to some extent blind mate friendly. But yeah, if you rolls out enough volume that it's worthwhile to delete a header then maybe you should invest in whole panel-sized jig anyway.

[ataradov]

Zero BOM

For production programming test points are also zero BOM. And the same test points can be positioned in a way that allows soldering of a connector for development boards.

Obviously Tag-Connect somehow makes money, so they are being used. But I have looked at 100s of customer projects and I have only seen it used once.

[SiliconWizard]

Well, they are mostly used for production when said production is very "manual" (so low volume or made in ultra-low wage countries) with no test jig. As soon as you have a test jig and matching test points on the PCB, then such a connector is not needed - and for prototyping. Not sure how they manage to sell enough for these markets to make a living, but apparently they do.

[BreakingOhmsLaw]

They take up more board space and require big mechanical holes complicating routing on dense boards.

This is actually a real concern. There's hardly enough clearance between those holes to run the SWD signals out of that footprint, let alone anything else. I'll go for a bed of nails solution instead.

Also, i've just seen the price on the adapter board for the STLINK. 28 Euros for a 20x20mm board with two pin headers. I can literally make 10 of those myself for that price. Hey, I'm okay with a markup, but getting ripped off is something else.
Once the patent on that thing expires, someone is making these connectors for a few bucks.

[BreakingOhmsLaw]

What is your use case?  I really prefer the "no legs" version.  I've used those a ton and they are great.  For device programming it's usually enough to just push and hold for the few seconds it takes to load the flash.

Flashing followed by a mechanical calibration procedure where both hands ares needed. Hence the version with clips.

[mikeselectricstuff]

I can see that they make sense where you need a temporary connection for more than a few seconds, or if consequences of  misconnection are significant ( e.g. programming OTP devices, re-erase time for SPI flash).

For connections of a few seconds for programming, then handheld spring-probes for programming work fine.

If however you need to connect for longer, then they look like a good way to connect one or more boards and leave them to get on with whatever is needed.

Seems like a good product niche that fits inbetween handheld probes and a full-on bed-of-nails jig.
They also do an edge-connecting version that connects to castellated edge pads with a retaining hole.

Also don't underestimate the value of being able to specify an off-the-shelf solution that doesn't require any engineering work to make, or replace when it eventually breaks.
That alone probably makes it a viable product.

[bookaboo]

We use the no leg version for production programming along with a PCBite https://sensepeek.com/pcbite-20
We cable tie the tag connect to the weighted end of the probe (tip removed). It's the perfect weight for an operator to just set on, hit program and do another task while they wait.
Might be just me but I've found bed of nails to be universally just trouble.

I'm pretty sure we have used the same tagconnect on over 5k devices, probably more, though I have seen pins wear out too.

[rsjsouza]

We use Tag Connect for production runs of about 500 units each, but we have been using them for 10k+ units over the timespan of about a decade or so. Minimal replacements were needed.

[ajb]

My usual comment with Tag-Connect is that they are pointless compared to a regular connector.

A lot of connectors won't stand up to very many mating cycles.  50mil ribbon cable connectors are particularly prone to developing loose connections IME, especially when used with shrouded headers.  So Tag-Connect do have that advantage over a lot of other common connectors. 

Seems like a good product niche that fits inbetween handheld probes and a full-on bed-of-nails jig.

I'd agree with this.  The ability to use the same footprint connector with or without the retention feature (you can get little rubber retainers that the connectors without the latches will push into, so you don't need to use the larger latch footprint) for both long-term debugging as well as quick production programming is also a nice advantage.

[Njk]

Once the patent on that thing expires, someone is making these connectors for a few bucks.

Indeed

[tszaboo]

What is your use case?  I really prefer the "no legs" version.  I've used those a ton and they are great.  For device programming it's usually enough to just push and hold for the few seconds it takes to load the flash.

Flashing followed by a mechanical calibration procedure where both hands ares needed. Hence the version with clips.

Typically you use the clip version for the FW development team and the clipless for the production team. Or clipless if the board iv very small. I typically include regular test points to a board where there is space, that way a test JIG can be made for it.
But here is the kicker, test JIGs are expensive, with hidden costs. You spend days designing, assembling tuning them, so that's easily 1000+ EUR cost for a company for a very simple bed of nails tester, and 10000+EUR for something more complicated Ingun or similar test solution.
Compared to this, you can just buy two of the test-ok cables for ~100 EUR, and replace when necessary. That will be after a few thousand boards. Sure it will not do testing, only programming, but sometimes the functional test is enough.

[Psi]

They have good lifespan as long as you're not an idiot and bend the thing a lot while it's inserted.
I've seen pins break on them, but it's pretty rare.

I've only ever used the smaller kind that has 6 pads and 3 pin small alignment holes. I've not used the larger one that has those bigger holes for a spring latch to hold onto.

I've programmed 1000's of  boards just holding the tag connect on the board for a few seconds.

They works well for rapid connect/program/disconnect.  But a jig with pogos works well too. 

Their only real negative is they're perhaps a bit overpriced.

[djacobow]

I'm with Ataradov, I just don't feel the love for TC. My company is fond of them. For production, we'll do something like a customer fixture with pogos in delrin, with the expectation that fw devs will use a Tag Connect to contact the same pins. The problem is that TC is just fiddly and if you bump your DUT or whatever on your desk, you can lose connection and not know it, etc.

I much prefer a TH IDC footprint and that's what I'll use on my own designs. For dev, I'll solder a connector, and for production I can also use pogos. In fact, I like pogos into PTH much better than pogos into pads, as the former are somewhat self-aligning and the contact is an annulus rather than a point.

[Doctorandus_P]

Their only real negative is they're perhaps a bit overpriced.

It's a personal thing, but I have moral objections to paying USD30+ for a connector that costs 50ct to make.

If you fool yourself into thinking you just want one connector, then it seems to make economical sense to use them, but they will probably need replacement after some time, and once you start using them, you probably put them in multiple products or programmers too.

With today's world of 3D printing and easily available PCB services it's quite easy to design your own and make a batch of them. There are also quite a lot of available commercial alternatives of test clips based on pogo pins and some clothes pin alike spring contraption. I have also seen clips for SOIC IC's used to clamp directly to a PCB instead of an IC.

[tszaboo]

Their only real negative is they're perhaps a bit overpriced.

It's a personal thing, but I have moral objections to paying USD30+ for a connector that costs 50ct to make.

If you fool yourself into thinking you just want one connector, then it seems to make economical sense to use them, but they will probably need replacement after some time, and once you start using them, you probably put them in multiple products or programmers too.

With today's world of 3D printing and easily available PCB services it's quite easy to design your own and make a batch of them. There are also quite a lot of available commercial alternatives of test clips based on pogo pins and some clothes pin alike spring contraption. I have also seen clips for SOIC IC's used to clamp directly to a PCB instead of an IC.

Two tighs about this:
Tag-connect is a small family business, I've met them on an exhibition, really nice people, what they achieved is a success story, and I don't have a problem supporting them. This is not a large faceless connector manufacturer, that doesn't answer your emails because your busness makes less than a billion.
And it is not 50 cents to make these parts. Each tag connect is an overmoulded connector assembly to a cable. The tooling cost is several thousands of dollars, and I just don't see how they would automate it. They have multiple tools to make the parts. And then you need to be on top of things, so your cables are avaliable at Digikey, Farnell, have stock, and customers are going to buy like 1 at a time.