[Chris] over at the 23B hackerspace had a bit of a problem – a project required the use of a very old rotary encoder with a mil-spec connector. While it might be possible to simply buy one of these mating connectors on Digikey or Mouser, that’s not [Chris]’ usual MO. He has a nice 3D printer, and this connector is basically a cylinder with some holes. How hard could printing out one of these connectors be?
The dimensions for [Chris]’ first attempt at creating a mating connector came from Solidworks’ “Sketch Picture” command where an image can be superimposed over a model and the 3D features created from that guide. If it worked, it would be far too easy, and the printed model didn’t fit at all.
This failure led [Chris] to page through MIL-STD-1651, a portly tome of 200+ pages covering every circular connector possible. After 20 minutes of scanning the specs, [Chris] found what he was looking for: the correct specification showing him where all the pins and holes should go.
After some fine modeling in Solidworks, [Chris] had his very own custom printed Mil-Spec connector. Sure, he ate up more time than it was worth for one connector, but now that he has the STL file, he can print out as many as he needs.
The SICK DGS25 isn’t an old encoder, they’re a current production item I believe. I normally use the DFS60B, but the DGS series is more robust.
Looks like HaD made the assumption that the encoder was old because it had a weird (to them) connector.
All of my larger servos have MS connectors as well as any machine tool at work, They are probably one of the most common weather tight multi-pin connectors in existence. I keep a stock of them as I find them for “just in case”.
We normally order the versions with M12 fittings (still good for IP67 or 68) or sometimes flying leads, but yes the MS connectors are still very popular in industry.
Those connectors can run $100+ a piece, and lead times can be insane. Use http://quicksearch.dla.mil/index.cfm to find any and every mil-spec you can fathom… And some you can’t fathom. The connector should always have the mil-spec number printed somewhere on it. Better yet, avoid mil-spec connectors if you can. Even more fun: Some suppliers of mil-spec connectors make other connectors that are very much like mil-spec, but have extra/different features to make them more versatile.
Also most are cadmium plated.
Mmm Cadmium Green. Isn’t the coating used to reduce mold or something?
Talk about going around your elbow to get to your thumb. The amount of time spent to make this connector was well worth the 20$ to buy it. I do understand that there is a need to justify the purchase of a 3d printer, and I do appreciate that quality of his efforts.
However, I think an easier route to build this connector could have been milled from plastic/rubber stock and the pin pattern transfered via ink and drilled out exactly where they were needed.
Or just unscrew the mil-spec connector (four Phillips heads) , solder on some flying leads, fill the hole with silicone and attach whatever suitable connector was readily available and easily sourced later on.
Messing around for days or weeks reproducing mil-spec connectors based on spec sheets from the ’50s, for one piece, seems like a waste of time. Go around the pothole and move on, instead of building a concrete machine to fill it as you drive over.
Yes. Absolutely right. Or cut off the connector and connect it up any way you want (my first choice). Or look thru the AMP and Amphenol catalogs and source something compatible for $40.
You do realize you’re currently on hack a day. :D
I don’t think he does. Most likely got lost on his way to walmart.com (sneaky bing ads!)
I’m sorry. Flying Leads? Is that the same as a Pigs Tail?
just wires hanging out the back
How long do you think all of this took him? If he’s familiar with the software necessary, then (excluding print time) he may have spent an hour or two on this. Soldering leads onto those pins, filling with silicone, and connecting the other ends of the wires to another connector, finding another connector that works (maybe wait for it to arrive after ordering), that would also probably take an hour or two. They’re just different ways of arriving at a similar result.
Yep, $7 on ebay:
http://www.ebay.com/itm/Amphenol-18-1S-Female-10-Pin-Plug-1-ends-2-long-/171098667712?pt=LH_DefaultDomain_0&hash=item27d6467ac0
10 pin amphenols are very common.
“However, I think an easier route to build this connector could have been milled from plastic/rubber stock and the pin pattern transfered via ink and drilled out exactly where they were needed. ”
Your definition clearly does not match mine. That’s why it’s great that there is a range of tools out there.
If he’s 3D printed it, it’s not going to be to any sort of mil spec (except perhaps certain grades of cheese), whereas drilling some holes in a lump of something properly solid, or soldering & potting, is going to give a far more robust and better sealed connector. Does it matter? Only if the thing is subject to moisture, dirt, vibration, heat, cable tension, etc.
Or just buy a connector for $20 and get on with life.
Exactly, why in the world would someone spend a week trying to interface to a wierd spec when it can easily be modified to something better? his “M.O.” seems to be do it the hard way.
I don’t like to pay $20 for a connector either. There are too many much more interesting components that can be bought for that price. Personally, I like to find one kind of inexpensive connector, buy it in bulk and standardize my projects on it.
That thought fits well with the idea some people have mentioned about taking the encoder apart, removing the pins and either hanging wires out the hole or hacking some other kind of connector into it. I’m sure that encoder and connector being mil-spec are very well built though. I wouldn’t want to hack it up either. Even though his 3d-printed mate for it is not mil-spec who knows, he may want to re-use the encoder in a different project with an actual mil-spec connector some day. Don’t wreck it!
So, yah, if this is a one off, never to be repeated I too would just buy the $20 connector (with much grumbling). He has all his files though. If he wants this connector again he can just print more. As many as he wants. Any time of day. No waiting for shipping. You can’t buy that kind of convenience for $20!
AND
If he wants to he can share his files on the net so that everyone has that kind of convenience!
“So, yah, if this is a one off, never to be repeated I too would just buy the $20 connector (with much grumbling). He has all his files though. If he wants this connector again he can just print more. As many as he wants. Any time of day. No waiting for shipping. You can’t buy that kind of convenience for $20!
AND
If he wants to he can share his files on the net so that everyone has that kind of convenience!”
This! +1
“Shit, it doesn’t fit all the way” That’s what she said! Anyway he knows how to print the connectors now and since he has MIL-STD-1651 he can also print (hopefully) any other connector he likes, so all in all time well spent.
Nice article, except he didn’t mention how he plans to make the actual contacts, which seems like a crucial step in making a functional connector :P I hope there’s a follow-up.
Or how he plans on making the connector backshell…
Contacts shouldn’t be a big challenge – they are by far the cheapest part of a MIL-spec connector, most people that work with such connectors buy lots of extra contacts because if you need to rework the connector – it’s the contacts you’ll be replacing.
There is, of course, the question of whether this 3D-printed insert will work with existing contacts – his testfit only ensures no significant interference with the pins of the mating connector, not whether any contacts in the insert will mate correctly.
Even if you have an insert and contacts, a backshell that will properly interface with the threads of the mating connector isn’t going to be cheap, and you’re almost surely not going to be 3D-printing that.
Back in those days, when 3D printing wasn’t hyped, one simply took a junk piece of this common white polyamide stuff, lathed it to the right diameter, drilled 5 holes in it and cut the slot with a sharp knife or thin saw blade. Finished
Much faster and no 3D printer required. Polyamide is also much tougher than ABS.
Yeah, no 3d printer required but how about the lathe?
3D print the Lathe ;-)
http://www.everyspec.com/MIL-STD/MIL-STD-1600-1699/MIL_STD_1651_1571/
Print plastic cylinder – coat tips of plug with ink – press tips of said plug against plastic cylinder – drill holes with pillar drill. 5 minutes tops? :) That said the result here is probably nicer and definitely reproducible. I’m just thinking the complexity is comparable to trying to kill a cockroach by pelting it with bananas :o)
This is what I’ve done before: Take some female pins, cover in heatshrink, push em into the male connector, then carefully put some epoxy between the pins to hold them together. Once it’s dry, take them out then use more epoxy to make it more robust
“now that he has the STL file, he can print out as many as he needs.”
…he needed one.
I chuckled at this thought. great work anyway.
a little spray wax and hot glue does the same thing in less than 15 seconds.
I am wondering, if it was so hard finding the specification… couldn’t he just take a picture, open it in a photo editor, add a layer, paste identical circles over each pin, make a circle around the hole and then remove the original photo layer?
Getting the scale right might be a little tricky. Just place a ruler next to it when taking the photo.
He did exactly this for the first iteration, and then found it was not accurate enough for a good fit…
no, why didn’t he take a picture with a camera?
with a ruler in the picture, directly from above.
it’s not that his first method was flawed, it’s that his first picture was wrong!