Page 1 of 2

Snap on programming header like the OEM's do

Posted: Fri Jul 05, 2019 2:22 pm
by kb1gtt
Can we find a lower cost version of a connector similar to the below?
https://www.digikey.com/products/en?keywords=TC2050-IDC-ND

I know many OEM's use the spring pins for programming. Can we do the same, or can we do similar?

Re: Snap on programming header like the OEM's do

Posted: Fri Jul 05, 2019 2:41 pm
by kb1gtt
It's a bit hacky, but also much lower cost.

Re: Snap on programming header like the OEM's do

Posted: Fri Jul 05, 2019 2:53 pm
by Abricos
It's not cheap but very easy to make ...

Re: Snap on programming header like the OEM's do

Posted: Sat Jul 06, 2019 2:23 pm
by pk386
What your looking for is Pogo-pin's and sleeves.

https://www.digikey.com/products/en/test-and-measurement/test-probe-tips/622

your best bet would be to 3D print a test fixture for the PCB and mount the pins in the test fixture.

What quantity of boards are you looking to program?

Re: Snap on programming header like the OEM's do

Posted: Sat Jul 06, 2019 4:53 pm
by kb1gtt
I fear 3D printed stuff likes to break when flexed. The layers tend to shear.

We do not really have much QTY. There has been some thought of trying to increase qty, but I'll believe it when I see it. So lets guess like 50 a year or something low like that.

What is prompting this is that MRE doesn't like a thru hole header for the programming port. It currently uses an SMT header, but that could be prone to fatique issues. So the design currently has small via's in the SMT pads to help hold those pads to the PCB. There is some debate about if the solder will get sucked into the via's, and cause connection issues. There is a small-ish chance that the connector will need a touch up of solder if this does not work as expected. So the solution could be to remove the connector altogether and do it like the OEM's with spring pins. We aren't concerned about the cost of the header, we are more concerned with how to program it reliably.

Re: Snap on programming header like the OEM's do

Posted: Sat Jul 06, 2019 5:33 pm
by Abricos

Re: Snap on programming header like the OEM's do

Posted: Sat Jul 06, 2019 10:28 pm
by AndreyB
I think we are not on the same page regarding WHY we need SWD header.

As of today I perceive rusEfi not-a-product ready - while MRE is amazing hardware, rusEfi is not amazing software see also https://github.com/rusefi/rusefi/blob/master/misc/Strategy_2019.md

At the moment I believe that SWD is the most convenient way for END USERS to update firmware which they would need to do often. So, we need this for end users as of today in my opinion, not for factory programming.

Re: Snap on programming header like the OEM's do

Posted: Sun Jul 07, 2019 12:58 am
by kb1gtt
How about 3D printed adapter similar to that spring clip with something like the below.
https://www.digikey.com/products/en?keywords=ED8177-10-ND

Lower cost option from ebay found here
https://www.ebay.com/itm/10pin-4-5mm-height-double-row-pitch-2-54mm-pogo-pin-connector-30pcs/332840987592?hash=item4d7ede6bc8:g:xoYAAOSwCsRbxApX

Also worth noting, the above posted Tag-Connect $40 connector has a KiCad5 native footprint, so that is easy to implement if we decide to implement that style of connector, it is easy in KC. If it's in KC, then it has to be a popular product :)

I understand development vs production as users today are going to need the SWD programming abilities. So it has to be considered as part of the normal use package. I understand that it will be difficult to do better than the programming dongle + soldered on header. I would still like to know how much it would cost to get a pogo connector option. To me it appears we could obtain this pogo style connector for about $5.

Re: Snap on programming header like the OEM's do

Posted: Sun Jul 07, 2019 8:06 am
by mck1117
Since something like this would be qty < 10, that's well within the range in which I'm willing to CNC mill plastic/aluminum tooling to hold the boards and align them with programming pogo pins. Since the quantity is so small I don't really care what the per-unit cost is.

Also, @kb1gtt, 3d printing is a LOT better than it used to be. I've used both a Prusa mk3 (printing PETG - not ABS) and a Formlabs Form 2 (liquid resin + laser stereolithography) and they both make very very good parts. For example, the radiator is held in my car with parts from the Prusa.

Re: Snap on programming header like the OEM's do

Posted: Sun Jul 07, 2019 9:19 am
by kb1gtt
As well for those who don't have access to a 3D printer, services like the below are great. They are reasonably low cost, you can print plastic's, metals, ceramics, and all sorts of things. The quality of the printed parts here is far better than any home owned printer I've seen. It usually takes about a week to get the item printed and at your door step.
https://www.shapeways.com/

Is there enough interest? Should I design a clamp? Perhaps I can make a clamp which will work if you do not install the SMT header. AKA choose to install the header, or choose not to. That's your choice.

Re: Snap on programming header like the OEM's do

Posted: Sun Jul 07, 2019 10:02 am
by Old Grey
kb1gtt wrote: ↑
Sat Jul 06, 2019 4:53 pm
I fear 3D printed stuff likes to break when flexed. The layers tend to shear.
That's only if you are doing to wrong. Like if you print with ABS on a non enclosed printer it can de-laminate.

You can increase the number of perimeters, go full solid infill, decrease layer hight so each layer if squashed harder to the other, print it perpendicular to the stress axis, ie: print longitudinally if it breaks axially, add ribs to increase the 2nd moment, etc. There is a whole YT channel dedicated to 3D printer material strength - https://www.youtube.com/channel/UCiczXOhGpvoQGhOL16EZiTg/videos

I have pogo-pins somewhere, so I might look into when I have time.

Re: Snap on programming header like the OEM's do

Posted: Sun Jul 07, 2019 11:49 am
by AndreyB
That's a weird option - only lot of 30 pcs for $75 available. Where are my 5 pcs for $13?

Re: Snap on programming header like the OEM's do

Posted: Sun Jul 07, 2019 3:32 pm
by kb1gtt

Re: Snap on programming header like the OEM's do

Posted: Mon Jul 08, 2019 2:19 am
by Old Grey
If it's low volume it shouldn't be that hard to 3D print, with some tweaks. ;)

I printed this with 1.4mm holes but had to drill them to 1.5mm, which might be too lose. Maybe with some tweaking you could a straight 3D print, and push the pogo in with a soldering iron so the plastic melts a bit and glues it.


Re: Snap on programming header like the OEM's do

Posted: Mon Jul 08, 2019 5:12 pm
by kb1gtt
Sweet, you need to get some of the style like I posted in that e-bay link. Those can be pushed from the back, instead of via friction around the round part. They have a bit that protrudes and allows an electrical connection. I think those e-bay ones could be put into an IDC header on a ribbon cable/ Then the entire header could be put inside a 3D printed block. The ribbon cable could be connected to the typical low cost programmer we have been using.

Re: Snap on programming header like the OEM's do

Posted: Mon Jul 08, 2019 7:40 pm
by Old Grey
I was thinking yesterday that you could just make a breakout PCB and stack and glue 3 on top of each other, and then I found this.
Image

I'm not sure the stub on the back of those pins will reach in an IDC, but I think they can be soldered to a PCB as a THT part.

EDIT - Now I know what to search on Goo ;)
https://www.google.com.au/search?client=opera&biw=1440&bih=798&tbm=isch&sa=1&ei=YJ4jXfCaF5bd9QOUj47ABQ&q=Pogo+Pin+programmer+adapter&oq=Pogo+Pin+programmer+adapter&gs_l=img.3...12237.24375..25724...2.0..0.204.2243.0j13j1......0....1..gws-wiz-img.......0j0i7i30.dnlm3iTpj5E#imgrc=_

Re: Snap on programming header like the OEM's do

Posted: Mon Jul 08, 2019 9:46 pm
by kb1gtt
The goal would be to connect the below programming adapter to a PCB via pogo style connector. AKA IDC connector desired.
https://www.adafruit.com/product/2548

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 12:59 am
by Old Grey
Seams like they do make a 0.5mm pogo, so I suppose you could just slide it in a female header if that is what is on the programmer. Other wise it's a PCB breakout with pogo on one end and the IDE connector on the other.

0.5mm and 0.68mm pogos.
https://www.ebay.com.au/itm/100-Pcs-Spring-Test-Probe-Pogo-Pin-P75-B1-P75-E2-P100-E2-P50-Q1-P75-LM2-P50-B1/391922331636?hash=item5b40642ff4:m:mbZJQRWDN-lBnLLLQBfTp-A

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 1:06 am
by kb1gtt
Sweet, how do we make that into a PCB mounted connector? Would it be a 3D printed thing, or some how a small PCB? That's a sweet find.

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 10:26 am
by Simon@FutureProof
I was going to suggest we get hold of some individual pogo pins and push them into a 3d printed holder.

Regarding 3d printing strength if done right I have managed to use it for suspension prototypes that could hold the weight of the car in order to do geometry checks. Things have advanced a lot and provided the printer is good, the filament if half decent and the user knows how to design for printing and set his machine the results are very different to what they used to be.

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 12:57 pm
by Old Grey
Yeah I've only seen bad 3D prints when they were done wrong, like the high temp filaments, ie: ABS etc, delaminate if not done in an enclosed printer that can keep it hot.

I'm not really sure what configuration you want, but I was thinking of just pushing the pins into the female IDC connector of this adaptor.
https://www.ebay.com.au/itm/10-Pin-Convert-to-6-Pin-Adapter-Board-USBASP-USBISP-AVRISP-Programmer-USB-STK500/263121501906?hash=item3d43437ed2:m:mTo61hzRT35N_o82XbQEdtA

The only problem is those 0.5mm pins have a sheath diameter of 0.68mm, and I don't know if it will fit in a IDC header hole.

EDIT - Just found some on EEVblog
https://www.eevblog.com/forum/projects/pogo-pin-adaptor-for-programming/

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 6:20 pm
by kb1gtt
How do you connect that connector then use both hands to run the PC? I like the pins crammed into the IDC, but how do we keep the pins in contact with the pads while not touching it any more? I think this would be a great place for a 3D printed thing. Perhaps push the connector onto the board, and spin some part of the it like 30 degrees, which then some how locks it down. Or perhaps the 3D printed thing could have a flip latch, such that you press it down then flip the latch into position.

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 6:56 pm
by Simon@FutureProof
That was my thoughts earlier today while at work, best if we had something that holds the programmer in place.
Although I am left to think that it is a lot of fuss to replace something that already works ok if we are careful.

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 7:33 pm
by AndreyB
OrchardPerformance wrote: ↑
Tue Jul 09, 2019 6:56 pm
I am left to think that it is a lot of fuss to replace something that already works ok if we are careful.
Yes!

Also while 2x5 header we only use 4 or 5 pins altogether.

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 9:11 pm
by kb1gtt
Also allot of fuss to replace the ECU when you can buy one, or even just drive with the one that's already there. I think our addictions are often hard to explain. It would be cool if we made a low cost pogo connector, but it's also not critical. We have a solution, but lets see if we can make it better.

Hmmm, what about a magnet on the other side of the PCB to keep the pins pressed on the connection?

Re: Snap on programming header like the OEM's do

Posted: Tue Jul 09, 2019 9:53 pm
by Old Grey
If it's that 6 pin header row in the middle of the Frankenso, it's going to have to be a robust 3D print to clamp 80mm away on it's own. One of those cheap plastic clamps in my video with more reach would make it easier.
Image

I think the usual method with hand held pogo is to do all the setup on the PC, hold the pogo down, and just press start with the other hand.

The magnet is dependant to the spring tension of the pins, number off, and the protrusion. The one I made needs a lot of force, but the 2nd I made had less protrusion and is a bit easier.
There are some magnetic ones on aliexpress -
https://www.google.com.au/search?q=pogo+pins+aliexpress&client=opera&hs=aMM&source=lnms&tbm=isch&sa=X&ved=0ahUKEwiXvIvS8KjjAhW66XMBHfNZAIMQ_AUIECgB&biw=1440&bih=798#imgdii=tiJMnTqffka_ZM:&imgrc=cZhe0slZRYOikM:

EDIT
If it is that header you could make one like mine that clips on the right-hand side of the PCB, and the use a T pin that goes in the oval slot in the middle of the PCB that you turn 90ΒΊ to lock the other end down.

Re: Snap on programming header like the OEM's do

Posted: Wed Jul 10, 2019 5:18 am
by mck1117
You could totally print a flexible clip that has snap-in jaws that release when you pinch together on the top of it.

Something like this (excuse the actual envelope drawing):

Image

Re: Snap on programming header like the OEM's do

Posted: Thu Jul 11, 2019 2:19 am
by kb1gtt
Should have been a pizza box.

I like the sketch of the 3D printed adapter.

This is the most preferred programmer. It's the same pin out as what is on MRE.
https://rusefi.com/forum/viewtopic.php?f=4&t=381&start=48

Re: Snap on programming header like the OEM's do

Posted: Fri Jul 26, 2019 10:08 am
by kb1gtt

Re: Snap on programming header like the OEM's do

Posted: Sat Jul 27, 2019 7:23 pm
by AndreyB
What is the problem statement again? I have SMD male header soldered to MRE v0.2 (with vias) and I think it is pretty sturdy. I am not aware of any reason to look into pogo or edge or else at this point, i believe male header with IDC cable is a simple, sturdy, cheap and efficient solution.

Open question at the moment is if we need straight or mirrored pinout on MRE board. Are all IDC female to female cables wired the same? Between male on MRE and male on ST-Link, I believe that the default IDC cable would require some kind of mirroring.