Hard Gold or Enig Finish for Edge Connector Fingers

[daveejhitchins]

For the 41 years I've been producing add-on boards, for various Acorn add-ons, I've always used Hard Gold for any edge connectors. However, finding a PCB supplier to actually provide this service at a reasonable price* along with a reasonable price for the PCBs is getting difficult!

* The suppliers that Techbridge Circuits use used to charge by area. Now they want £50 per design for any quantity! Add to that the higher PCB price from these suppliers and the final item price is getting higher than I like.

So - the question is: What do people think of me using ENIG finish on the edge connector fingers? Bearing in mind that an ENIG finish is considerably thinner than a hard gold finish!

IPC-4552 ENIG Specification (2002)
For thickness IPC-4552 stated: The electroless nickel thickness shall be 3 to 6 µm [118.1 to 236.2 µin] The minimum immersion gold thickness shall be 0.05 [1.97 µin] at four sigma (standard deviation) below the mean; the typical range is 0.075 to 0.125 µm [2.955 to 4.925 µin].

Electroplated gold is thicker and harder making it ideal for edge-connector contacts for PCBs which will be repeatedly plugged in and removed. First plate 3 – 6 microns of nickel onto the edge connector fingers and then on top of that 1 – 2 microns of hard gold. The plated gold is not 100% pure; it contains some cobalt to increase the wear-resistance of the surface.

Electroless gold gives excellent solderability, but the chemical deposition process means that it is too soft and too thin to withstand repeated abrasion. Electroplated gold is thicker and harder making it ideal for edge-connector contacts for PCBs which will be repeatedly plugged in and removed.

I think the above details of the 2 finishes give all the information needed. I'm going to order 5 prototype AP5 boards from JLCPCB, who only offer ENIG to see what it's like in practice.

But would love to hear your comments. I'm still leaning towards hard gold! After all these years the plating on my products, that I've come across, are still in good shape . . .

Dave H.

[danielj]

ENIG is rather thin - but then if you think that slogger never bothered with gold on their edge connectors, it's probably better than that!

d.

[daveejhitchins]

ENIG is rather thin - but then if you think that slogger never bothered with gold on their edge connectors, it's probably better than that!

Yes, I've had that in mind while contemplating this question . . .

Dave H.

[danielj]

I have some Apple II cards that I got from JLC with "gold fingers" - I'll pop one in the post if you'd like to have a look at it?

d.

[1024MAK]

The real question here, is how often do members unplug and plug in cartridges, interfaces or expansions?

If the answer is not very often, then Enig may be acceptable.

If the answer is regularly or often, then that’s when hard gold would have the advantage.

Also the other option is to not use gold or Enig at all. Various computer and electronic equipment manufacturers (including Amstrad, Sinclair, Memotech, Acorn with the Master Compact expansion port) did not use gold or Enig.

While others did use gold (Acorn with the Acorn Election expansion port, Commodore, various disk drive manufacturers).

Mark

[sirmorris]

My development zx81 has had ~750 zxpands on and off its HASL edge connector. Now I won't pretend it's in good shape, but it still works.

[Ramtop]

All my production ElkSD-Plus1 cartridges use JLCPCB's ENIG finish with 'gold fingers' and I've never had any reported issues with wear on the contacts.

The prototypes used HASL, and one thing I have found is that ENIG boards need to be kept *very* clean to make a good contact. Presumably becaue HASL is thicker it seems more reliable, but the ENIG finish is a bit temperamental, especially if the cartridge socket is well used and not as tight as it should be.

[daveejhitchins]

I have some Apple II cards that I got from JLC with "gold fingers" - I'll pop one in the post if you'd like to have a look at it?

Thanks for the offer, Daniel - but as I'm having the AP5 prototypes with ENIG I'll have 5 to play with, so I should get a reasonable 'feel' for the finish.

Just as a side note: Any products we made a Filtronic that used connectors the finish for the actual contact area was specified as 30 microns of hard gold. Usually sourced from Samtec!

Dave H.

[cmorley]

'gold fingers' and I've never had any reported issues with wear on the contacts.

"Gold fingers" is a common Chinese term for hard gold on the edge connector only. Hence the checkbox for ENIG and a separate checkbox for "gold fingers" on some mfg's websites.

[Ramtop]

JLC doesn't permit gold fingers without ENIG, but since they don't charge extra for it either I'm not entirely convinced they're doing anything but removing the solder mask between the contacts.

[cmorley]

Removing solder mask in an area is defined by the Gerbers you send to the PCB house. You might also find that the price quoted on the website is not the price you pay after their engineers look at it.

PCBWay website quoted $2/board (200 off) I sent gerbers for but then changed that to $30/board after a human had looked at it - then irritatingly refused to say what the price sensitivity was hence why I never bother with PCBWay.

[Ramtop]

Can't comment on other fabs, but JLC strip everything but the copper/gold finish from the contacts area when you select gold fingers. I found that out by accidentally uploading the wrong gerbers for a board, where I hadn't removed the solder mask and actually had left silkscreen over the top of the contacts. JLC just removed it all and the boards came out perfect.

And I may just have been lucky but I've given JLC many, many orders for boards in a wide variety of configurations and never had them deviate from the quoted price (I know they will do it sometimes, but seems to be quite rare). Given the microscopic margins they make I suspect the auto-quote system has to be pretty good at getting it right to minimise human intervention.