Dupont Headers for Wire Wrap

[PotatoBox]

Hi there,

I'm planning on doing quite a large 68020-based project and obviously prototyping on a breadboard is not feasible for something like this, and I was wondering if it is possible to use the dupont headers that commonly ship with arduinos for wire wrap?

I am planning on designing a PGA breakout board that would insert into a perf board and let me wire wrap to a PGA socket.

[jpanhalt]

I have only done a few wire wraps, but I have watched others do it in the 1970's.  My impression is that the pins are usually longer than typical header pins.  The pins one sees advertised for wire wrap seem to be about 0.50x" long.  Aries offers two versions in its 12014 wire wrap pins, 0.5" and 0.360" (https://www.arieselec.com/product/12014-pin-line-collet-socket/ ).

"Typical" header pins are about 0.230" long, but a variety of sizes are available up to an inch or more.  Gold or silver plating were at one tine specified, but solder would probably suffice.  Pin size seems to be 0.025" square, which is the same for pin headers and wire wrap pins.

Of course, you could try a shorter pin and see how it holds, but headers with longer pins are readily available.

[BILLPOD]

Good Morning  PotatoBox, From back in my wire-wrap days, I remember the pins had square corners, to cut through the insulation of the wire.

[CountChocula]

As the other folks mentioned, make sure you use pins that have a square profile—the sharp corners bite into the wire and help secure it. You can buy headers with longer pins for cheap on AliExpress, but even the regular ones will work fine if you solder them “upside down” so that the shorter end is mounted onto the board; the longer end is good for at least two or three terminations. Make sure you buy a good wire wrapping tool and especially good quality wire with kynar insulation… the cheaper 30AWG has poor insulation that doesn't strip well and will give you all kinds of headaches.


—CC

[tooki]

My understanding is that wire wrap posts have precise, sharp corner edges, whereas on regular headers the edges may be less defined.

Good Morning  PotatoBox, From back in my wire-wrap days, I remember the pins had square corners, to cut through the insulation of the wire.

Not quite. You have to strip the wire before wrapping it. The sharp edges bite into the conductor itself, resulting in a gastight junction. The few turns of insulated wire around the post are for strain relief.

[PotatoBox]

My understanding is that wire wrap posts have precise, sharp corner edges, whereas on regular headers the edges may be less defined.

Yes this is my main concern, since they are the same width dimension wise and are squared off, but they aren't actually designed for wire wrapping, so I wasn't sure if they'd work or not.

[jpanhalt]

It's easy to be concerned and buy a specialty product.  But the latter will be more expensive and may not be any different, except for a descriptive distinction.  It seems the need for corner sharpness is not well documented and some people use common headers of sufficient length. 

https://forum.arduino.cc/t/wire-wrap-male-headers-or-better-way/62692
https://www.eevblog.com/forum/beginners/wire-wrap-headers/

This author specifies a maximum radius of 0.003":
https://www.tayloredge.com/reference/Instructional/General/wirewrapping.pdf

I know nothing about that author, but have attached a sketch of a 0.025" inch square with two corners radiused to 0.003" .  I don't have a radius gauge that gets that small, but those look pretty round  compared to the typical square-pinned headers I use. 

I suspect, but have not tried it, that 0.230" is a bit short for good wrap + strain relief for more than one wire.

EDIT: I am referring to what's on Google.  There may well be specifications written when wire wrap was more common that addressed corner sharpness.
EDIT2: I just looked microscopically at some ordinary Sullens pinheaders from DigiKey (gold flashed), and the corner radius is nowhere near 0.003" on the body.  At the pointed end, it looks to be close to that value.

[BILLPOD]

Not quite. You have to strip the wire before wrapping it. The sharp edges bite into the conductor itself, resulting in a gastight junction. The few turns of insulated wire around the post are for strain relief.
[/quote]

You have awakened my wire wrapping memory, and now I remember stripping the wire.  The cheap wrapping tool I had, had a slot on the end for stripping.  But I did accurately remember the requirement of sharp corners on the pins.

[PotatoBox]

It's easy to be concerned and buy a specialty product.  But the latter will be more expensive and may not be any different, except for a descriptive distinction.  It seems the need for corner sharpness is not well documented and some people use common headers of sufficient length. 

I think I will just buy the specialty product, since designing and getting a PCB manufactured for a PGA breakout board will probably be more expensive and take more time than just getting enough SIP wire wrap sockets to make up all 114 pins on the 68020 and just socketing it that way.

[joeqsmith]

Hi there,

I'm planning on doing quite a large 68020-based project and ...   for wire wrap?
...

I wouldn't recommend you head down this path but you may find some of the following helpful:

https://www.eevblog.com/forum/beginners/how-many-of-they-younger-crowd-have-ever-done-wire-wrap/msg1200500/#msg1200500

https://www.eevblog.com/forum/buysellwanted/lfs-cheap-wire-wrap-tool-and-sockets-(uk)/

https://www.eevblog.com/forum/chat/do-people-not-wire-wrap-anymore/

Last time I used WW was to build a transient generator that I use to test multi-meters.   Looks like it is nine years old now and still working fine:
https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg753409/#msg753409

[PotatoBox]

Hi there,

I'm planning on doing quite a large 68020-based project and ...   for wire wrap?
...

I wouldn't recommend you head down this path but you may find some of the following helpful:

What am I getting myself into? Wire wrap seems to be sufficient for that type of project and seems to be flexible enough. The only other option is to order PCBs periodically as I make major revisions to the design, and minor revisions can be done with cutting traces and putting on bodge wires, but that seems slow and expensive.

[joeqsmith]

What am I getting myself into? Wire wrap seems to be sufficient for that type of project and seems to be flexible enough. The only other option is to order PCBs periodically as I make major revisions to the design, and minor revisions can be done with cutting traces and putting on bodge wires, but that seems slow and expensive.

Take the time to read through the links I provided you.  If you decide to go ahead with it, consider posting a new subject the projects area about it.   I would certainly be interested in following along. 

That one board I showed in the video was the largest boards I came up with which I would say was actually fairly small.  That board clocks the FPGAs at 100MHz.  Trivial today using a PCB but with WW, there's no controlled impedances, coupling....    I did get it to work and showed it running in a second video. 

Time wise, that last board I made for the transient generator was all done by hand.  I wouldn't be surprised if I had more than 40 hours into it.  I didn't really take a lot of time with it and it was a mess.  Consider that is a tiny board with few pins. 

I'll look forward to seeing your workmanship.   

[PotatoBox]

What am I getting myself into? Wire wrap seems to be sufficient for that type of project and seems to be flexible enough. The only other option is to order PCBs periodically as I make major revisions to the design, and minor revisions can be done with cutting traces and putting on bodge wires, but that seems slow and expensive.

Take the time to read through the links I provided you.  If you decide to go ahead with it, consider posting a new subject the projects area about it.   I would certainly be interested in following along. 

That one board I showed in the video was the largest boards I came up with which I would say was actually fairly small.  That board clocks the FPGAs at 100MHz.  Trivial today using a PCB but with WW, there's no controlled impedances, coupling....    I did get it to work and showed it running in a second video. 

Time wise, that last board I made for the transient generator was all done by hand.  I wouldn't be surprised if I had more than 40 hours into it.  I didn't really take a lot of time with it and it was a mess.  Consider that is a tiny board with few pins. 

I'll look forward to seeing your workmanship.

I will examine my options and hopefully make a thread about it in the near future

I also need to look into logic analyzers for a project like this. I know nothing about them so this is unseen territory.

[MarkT]

Firstly wire-wrap is horrible to do, plays havoc with the eyes staring at 100's of sharp points all the time.

The pins must be hard-square profile with pointed tips and gold plated (to ensure a gas-tight weld), the correct wire and tool must be used (wire wrap tools are high precision and very expensive).  Wirewrap wire is a specific thickness, coating and insulation material, all the mechanical properties matter for reliable creation of joints.  Long pins are standard to allow upto 3 wires to any pin, which is pretty much a requirement for many circuits.

Wire-wrapping typically adds a lot of stray capacitance to circuits, unsuitable for modern high-speed logic for instance.  Its hard to make a ground-plane with wire-wrap!

PCBs, breakout PCBs and breadboards, IDC connectors and ribbon cable, etc are much to be prefered - these days that's the easy and cheap option - that's why wire-wrap died out I think.

However with the correct equipment a wire-wrapping robot is probably still the most reliable way to make circuits, well, low speed circuits, nothing beats a gas-tight weld, and tin-whiskers unlikely to be an issue with everything on 0.1" pitch.