The $11185 Connector | Hacker News Reader

The $11185 Connector

bigmessowires.com

63 pointszdw46 comments

The title is very misleading... it's not an $11,185 connector, that's the price of setting up custom tooling for this component. After-which the factory quoted a very reasonable $0.13 per piece.

The author went into this stating he'd be willing to pay for their tooling... then is sticker-shocked when they quote him for so. It's like the author went into this thinking they'd pay a few hundred bucks (for the custom tooling) and get a short fab run (< 100) back out. These factories aren't going to tool-up unless they're about to produce thousands... it doesn't matter if it's PCB's, connectors, action figures, jewelry display stands, etc. It's not worth their time at small scale.

The reason these connectors are difficult to find (if possible at all) is people aren't buying them... so you have to make it worth the factory's time to custom fab them for you. And if they think you're going to be a one-time-run, then you aren't going to get as good of a price as if you signed some long-term production agreement (10,000 a month for 12 months, etc...),

He also didn't go about finding a fab in the best way.. he went to Alibaba which is essentially the Amazon of factories... definitely not the "best price" you can get.

Since you seem to be familiar with this sort of thing, what is the best way to find a fab if not going to Alibaba?

There is no specific place to go to... a lot of it simply comes from working within the industry and learning the key players. My company doesn't buy/sell electronics, but the factories operate more-or-less the same.

Going to Alibaba is like shopping on Amazon. You wouldn't go to Amazon to purchase bulk (> 10,000) pieces of some item... you certainly aren't getting the best deal. Amazon takes a cut (so does Alibaba), and retailers listing there have catered their prices towards end-users (so significant mark-ups).

If it were me, I'd post on electrical engineering forums and ask who people recommend for fab jobs. Sometimes a nice supplier/distributor will tell you who they use for fab jobs (but a lot do tend to guard this).

Alibaba just isn't the best way to get the best price...

Yeah and my cousin Vinny can get you a better deal, too bad you aren't family.

> Yeah and my cousin Vinny can get you a better deal, too bad you aren't family.

I can appreciate your sarcasm here, but it's not constructive and doesn't change the economics of custom fabrication.

These factories don't sit around waiting to get small-time jobs to fab a handful of some custom components.

You make a good point, though. The shell size (DB) is the same and you can purchase shells separately. The problem devolves to getting the contacts (it's a standard item you can buy easily) and molding a custom insert. For small quantities, it seems like he could do it himself. The mold shouldn't be difficult to make.

[edit] Even easier would be finding out what material the insert is made of and getting the connectors milled to shape and drilled for the contacts. Fun problem!

Steve, are you listening :-)

If you've got a DB19F floating around you could even make a jig to hold the pins in place in the exact configuration you need, then simply epoxy them in place.

Alibaba was a reasonable place to look, the other suggestions require insider knowledge, and are significantly more speculative.

Commissioning a custom production run is an insider thing. But, yeah AliBaba was a good place to start if you have no contacts.

A possible analogy:

Would you accept a consulting gig if it was a one-off job, required a 2 hour commute in both directions, and you expected to be onsite for only 1 hour? Maybe...

You might accept if you were really desperate to get any job, or you might accept if your contract pays for your commute time and expenses. But if neither of those conditions are met, you would probably turn it down as it wouldn't be worth your time.

This is over-simplifying this a bit, but it's along the lines of how the factories view this sort of thing.

The author is asking them to spend more time and expense tooling up than they will in actual production (your 4 hour commute round-trip). If they were already tooled for this component (you lived in the same city), then they'd be more open and provide a better rate (you charging just your on-site fee). But since those conditions are not met, they will make you foot the bill for tooling (you charging for your commute and expenses), especially since the likelihood of some other buyer coming along and commissioning this component are very slim (the consulting gig is a one-off).

One thing you can do is go to brokers. These people buy old stock from suppliers but also buy old stock from manufacturers who go out of business or who no longer meed the part.

The parts you get are espensive and don't have the same chains of paperwork and may have been mishandled.

There is a problem that you can end up bidding against yourself and driving the price up.

>One thing you can do is go to brokers.

Those guys will skin you alive if they can. I'd almost rather take my chances making a production run... almost.

And, to be fair, I'm kinda surprised he didn't do this.

If he needs 1,000 connectors, that's $11 per connector. That's a little expensive, but not outlandish.

Obviously, he knows his volumes and we don't. But, he's spending probably spent almost $2K-$3K buying up the worldwide stock of these. How many more does he need to sell before that $11K is reasonable?

He says he recently bought 500 for about a dollar each and that other places had quoted him between $2 to $4 per part.

That variability in pricing is already a bit scary but going from $2 to $11 is something you'd only do in extreme need.

>The title is very misleading...

Yeah... but... I think it's kind of obvious. Not sure where the line is between "enticing" and "bait" but BMOW isn't selling crap, so it's not like you had to tolerate some nasty popup surprise or ad spam or anything.

>he went to Alibaba which is essentially the Amazon of factories... definitely not the "best price" you can get.

Good enough for a first approximation. Maybe he'll at least figure out that he can probably bargain that down quite a bit.

Talk to Sullins Connector Corp directly. http://www.sullinscorp.com/

The last time I had a connector I needed, they took something they had off the shelf, milled it, stuffed it with different pins, and charged me $100 for 5.

I was stunned. I haven't had customer service that good in decades.

Really. Call them. Even better, take a factory tour.

They may not be able to do it, but they have 30 years of designs archived. One of them will probably work.

Except he did. They declined. Maybe if he knew somebody higher up at Sullins?

They declined because, "we are currently phasing out our d-sub line and will be going obsolete in June."

That's basically permanent.

Unlikely. If they are "phasing out" a product line to the point of it being gone in a couple months it means that they only have inventory and aren't manufacturing them anymore. So, even if they have the designs, they no longer have the equipment and materials.

A quick search suggests that they're still made:

http://www.connectworld.net/cgi-bin/rrdata/DB19FS (female)

http://www.connectworld.net/cgi-bin/rrdata/DB19MS (male)

He bought all those ones:

http://www.bigmessowires.com/2015/01/28/in-search-of-the-elu...

DB-25-connector, meet hacksaw and belt/disc sander. Welcome, DB-19 connector!

While this would be a no-go for a 'regular' product, it would be perfectly acceptable for the niche application - floppy emulators for old personal computers - these connectors are used for.

I did pretty much exactly this a few weeks ago to make a DB-23 video cable for an Amiga :]

I worked in the nuclear industry a long time ago. The reason some parts are expensive is the documentation (chain of ownership) and testing involved. Some parts must be flame tested (set on fire and self-extinguish in X seconds), so that 50% of the batch are destructively tested in order to meet specific requirements. It doesn't justify excessive profiteering on a setup or per-part basis, but there is reason for connectors costing $90/part.

Do you have any nuclear industry war stories to share?

I think you might be better off a) finding and contacting an original mfg who still has the tooling, or b) making some kind of adapter board by pulling the pins from a DB25 and soldering them onto a pcb.

I wouldn't be shocked if you paid them and got the first parts and had a problem.

Then they will point to your lack of tolerance or chamfer specs or somesuch and then want more money to modify the tool.

For a completely custom connector, that seems like super cheap NRE to me. I work as an EE and have paid more than that on small run prototype stuff that is simpler to manufacture than a connector (think prototype circuit boards). Connectors require multiple sets of tooling and specialized processes unique to the connector industry. Even getting the tooling for the simplest plastic part is goig to be thousands of dollars.

I have a package of DB25 connectors left over from a product I used to build and sell that has slowly been whittled down over the years. It's because of things like this that I have been reluctant to toss them out. At sometime in the future I think they might be useful to me or someone else. Likewise the bag of 50 Centronics connectors. They have to be valuable sometime, right?

Or maybe I'm just a packrat.

This just reminds me. I wanted to buy a couple of extra allen wrenches for my camera tripod just so I could put them in various bags and hopefully never be without one.

Check Amazon...

1 wrench $6

100 wrenches $12

I now have a box of 100 allen wrenches all the same size.

Give them away at the office as a reward for a job well done.

Funny, did the same thing for a few P-38 can openers.. now I'm never without one... they're just better when the hand crank ones start wearing out.

Presumably the reason that the other factories did not make an offer was that they would have had similar set up costs and already expected that the volume wouldn't be large enough to make this viable...

That's too bad: Steve Chamberlin appears to have taken the post down probably due to unwarranted "UR DOIN IT WRONG" in his page comments.

There are a couple other comments [1] [2] that touch on this. But it does seem like there are some shortcuts that could be taken with DB-25 connectors. Anyone who's still manufacturing them (which may not be many) might be able to modify their tooling to not add 6 pins on one side and not add the protective cowling.

Is it a safe, strong connector? No.

Would it cost $11k in tooling? I have no idea I'm a biologist.

It'll be interesting to see what he settles on.

edit: Ha, he's already doing that: http://www.bigmessowires.com/2015/02/01/designing-a-db-19-su...

[1] https://news.ycombinator.com/item?id=9012293#9012602 [1] https://news.ycombinator.com/item?id=9012293#9012695

Bargain for everything in China. It is the culture. They will haggle over the price of anything. It's actually kind of fun once you get used to it.

I'd strongly recommend looking into 3d printing these. You could pretty easily 3d print a plastic version of the connector itself (and shell, if male), then add aluminum tape to the inside if you need EMF protection. Then you just pop in pins (which you can get at Fry's by the hundred for a couple bucks), and you have a finished connector. A few hundred dollars in setup and you're done, with a unit cost of under a dollar, all in.

None of the low-end 3D printers could hold the tolerances needed. Even a Form 1 isn't that precise.

Connector tolerances are very tight. That's why intermittent electrical connections are mostly a thing of the past.

I regularly print tighter tolerances with an FDM printer (modified Makerbot Replicator 2). D-sub connectors just aren't that tight; if you were printing the pins themselves, it'd be different, but the plastic side of things is not an issue at all.

For connectors where the pins are crimped to wires, then inserted into the connector block, the hole into which each pin fits has a complex form. It has to be shaped so that the pin clicks into place, but isn't pushed out by ordinary connector insertion forces. FDM extruders extrude a thread bigger than the tolerances required for that.

Getting a longer D-sub connector and machining the plastic down is more likely to work. 3D printing isn't magic.

Slight misalignment might actually improve electrical contact as long as it weren't too bad.

http://www.digikey.com/product-search/en?x=0&y=0&lang=en&sit...

All are non-stocked, and appear to be DB25, not DB19.

Those are all DB25s that happen to have a part number starting with "DB-19".

Do you need to support older hardware? Or is there a particular reason you can't use DB-25 and just use 19 pins?

Good story, but I flagged it for the misleading title. Perhaps a less sensational one would be, "Making new DB-19 cables is expensive."

The Chinese are tired of being paid a fraction of what citizens in developed countries make for doing the same work.

This is a blog about how a citizen in a developed country is "shocked" at the concept that other people should be paid a fair wage even if - gasp - they're Chinese.

The phrase "shocked" doesn't occur in the post, so I'm not sure where that term comes from. He does say that the quoted figure is more expensive than he hoped for, which is reasonable, since he goes on to present a number of quick calculations suggesting he can't justify purchasing them at that price.

Reply on news.ycombinator.com