B-7971 Smart Sockets - Harwin Socket Pins Vs Original Sockets

[Richard Scales]

For a future project I will be using B-7971 smart sockets. I have already produced one device using smart-sockets which went well. 

I now find that I have accumulated lots of the original sockets that were made for these tubes.

My dilemma is - do I go with the Harwin socket pins and scrap the original sockets that I have - or do I solder the tube sockets to the Smart-socket PCB and rejoice at my recycling efforts?

Which solution will be kinder to the tubes?

I would happily use the tube sockets unless the Harwin Socket pins offer an advantage too good to ignore.

Does anyone have any experience of doing this?

I did post it on the smart-sockets yahoo group page but there appears to be zero activity there at present.

All pointers gleefully received.

Kind regards

Richard

[Paul Andrews]

I think any benefits one way or the other would be marginal.

[jf...@my-deja.com]

ISTR that the real sockets are preferred.  The argument I remember hearing is that the pins in the socket "float"; the disadvantage is that you may need to jiggle the tube as you insert it into the socket, but the advantage is that the pins float and do not stress the glass-to-metal seal thereby reducing the probability of creating a leak.  It also means that real sockets will better tolerate imperfect pins.  Conversely, rigidly soldered pins may cause the tubes to fail prematurely.  Of course, in all cases, make sure the pins have been straightened in both cases.

[gregebert]

I stopped using sockets based on my experience with Burroughs 5092 and 6091 tubes. The insertion force is quite high and I have concerns about stress on the glass seal.

I now use socket pins, which really are meant for connectors. They are frequently listed for sale on Ebay, and the same socket pin works for most tubes.

I have personally verified these 

• Nixie - b5092, b6091, b7971, IN-18
• VFD - DT1704, Y1938
• NIMO -  BA0000

If you number each tube and assign it to a specific socket, there will be zero residual stress ( such as with a socket with floating pins) if you place the pins on each tube, then solder in-place. That's because each tube is matched to the socket.

That said, my experience is that even if you swap tubes, the insertion force is so low and there is enough flexibility in the socket pin that you will have no residual stress.

Lastly, you can 3D print pin collars and they look like real sockets.

[Paul Andrews]

I'm wary of using socket pins for my expensive tubes - the whole issue of having the tubes in the pins while they are being soldered. So I moved to Mill-Max sockets for those.

[Nicholas Stock]

I've soldered a lot of tubes (and I mean a lot) including quite expensive ones with socket pins/Harwins and others. I've never had a problem. Still haven't had a B7971 failure (anyone had one of these tubes fail? Serious question....). Just my data point for reference..;-). Just don't let the soldering iron sit needlessly on the socket pin after you've flowed your solder....pretty obvious that one.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/e47239ab-f100-4caf-a11b-7ff68d44d56c%40googlegroups.com.

[gregebert]

The only 7971 failure I've had so far was one of the segments slipping out-of-place and contacting an adjacent segment. That causes 2 segments to glow when either is selected. Nothing related to pins would cause this.

Another thing to be careful about is the IN-18 tube. Most of mine have very soft pins that bend easily, so whatever you use for connecting to that particular tube, be very gentle. They also corrode but that has never caused issues for me because the surface corrosion does not seem to affect conductivity at the low (4-5mA) current. Even 1K of additional resistance due to corrosion would have minimal impact (5 volt drop).

[Dekatron42]

You can also make your own socket-plugs by making a circuit board with the correct pin-diameter and then soldering tube pins so that you get a plug which you can use instead of your expensive tubes when soldering the Harwin or other sockets. To make the pins straight you can either put a few of the circuit boards on top of each other or use a standard socket as a guide - I use tube pin straighteners when I solder the pins so I get them straight. Don't warm the Harwin sockets for too long as the pins in the plug will heat enough for them to get un-soldered, you can remedy that by soldering a few circuit boards on top of each other and also by making a full copper fill on both sides to spread the heat and also use a solder with higher melting point.

An example of what I mean is shown in the photo below (a simple plug to replace the EAA91 dual diode tube in AVO Valve testers with silicon diodes).

/Martin

[Nicholas Stock]

For some further confidence in the process, there were tubes that were actually directly soldered to the PCB (why, I have no idea...cost? excluding the obvious leaded ones like the IN14, IN-8-2 etc...). These include NL84* style and ZM1000's. I would think there's a lot more heat getting to the glass/pin seal with direct soldering vs shielded in a socket pin...

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/a1d369f5-e031-47af-a18e-1466c7c5914a%40googlegroups.com.

[Richard Scales]

Thank you everyone for your input and suggestions.

My first effort was done using the Harwin socket pins and that went OK. having weighed up the pro's and con's and, given that I have accumulated enough of the original sockets, I feel tempted to go with those on the basis that there is that 'float' factor and the sockets will be less rigid than the Harwin approach and I will take that as being more kind to the tubes. I will also assign additional points to myself for re-using the old sockets and not throwing them.

It's not an immediate project and I know that the smart-sockets will look a little weird being adorned by the additional hardware of the socket but I am prepared to give it ago.

... and in a related query, should I take the lack of response from the smart-socket group on Yahoo as in indicator that it is no longer frequented on a regular basis?

... and also slightly related - I am putting together some ZM1350 smart sockets and the original intent there was to use the 'dupont' connectors though I found it really difficult to get them lined up with the ZM1350 pins. I switched to Harwin sockets for those and the ZM1350's slide in/out with relative ease - no real stress. I suspect that the pins on ZM1350's are of a smaller diameter than those on B-7971's.

Thanks again everyone for your feedback on this - very much appreciated.

[Bill Notfaded]

I'd use the socket pins for smartsockets... They're made for them. The real sockets you have are great for many big tubes. I also have a few SK-137 sockets and they're great for other tubes like Z5680M; Z568M ; 7037; 7971; B-7037; B-7091; B-7094; B-7971; B-8033; B-8091; NL-7037; NL-7094; NL-7971; NL-8091; NL-863.

Why not use the smartsockets the way they're supposed to used and I also have NEVER had a B-7971 fail they're rock fkn solid tubes!!! We love them!!!

Bill

[Richard Scales]

I cant really argue with that!

I don't think I'll ever use them for anything else and I do already have the socket pins. I did spend some time cleaning them all up and bending the pins straight - perhaps I'll find a use for them some day. For now I'll continue contemplation but it's looking like socket pins are in the lead!

I too love the B-7971 - if only I had invested when they were super affordable - what was I thinking? I was probably at university at the time and was thinking how exciting all this LED technology was - Doh!

Richard

[jf...@my-deja.com]

On Sunday, September 29, 2019 at 10:50:16 PM UTC-7, Richard Scales wrote:

> ...

> I too love the B-7971 - if only I had invested when they were super affordable - what was I thinking?

> 

I got mine when they were cheaper than the LED alternative: US$1 per tube including sockets, versus about US$5 for a MAN-1.  The disadvantages of the nixies was providing the HV power and the cost of the relatively rare HV driver transistors (I should have gotten the tubes with sockets and driver boards for US$2) compared to direct drive from a 7447.  Naturally, I used the sockets, and used point-to-point wiring  on a perfboard (no pcb) which automatically provided stress relief n the socket pins.  For three six-digit clocks, I have had zero in-service failures after more than 40 years (so I have not yet needed any of the spares).  My only tube failure was when the glass envelope lost a battle with a concrete floor.

[Richard Scales]

$1 per tube - how things have changed!

[Paul Andrews]

Ah yes, but in today's money that's about $16!

[Terry Kennedy]

On Monday, September 30, 2019 at 10:08:48 AM UTC-4, Richard Scales wrote:

$1 per tube - how things have changed!

Here's the ad (original scan by David Forbes):

 

[Richard Scales]

Now, where can I get me some $16 B-7971's ;-)

[gregebert]

I was going to say 'Poly-Paks'.....ad looks very similar, and I remember seeing those for sale by-the-barrel.

Could have.....should have.....would have.....

[jf...@my-deja.com]

...and B-A was generally not the low-cost junque dealer; John Mishna and Poly-Paks were usually cheaper.

[Bill Notfaded]

Wow I was seven years old when that catalog came out and we were wearing red white and blue striped tube socks and driving around on Schwinn Mantra Ray bikes with banana seats and monkey bars!!!

Bill 🤣😂🤣😂🤣😂

[GastonP]

I bet that at that time, the "aproximately 30 transistors and over 100 assorted diodes capacitors and resistors" had more net worth than the 7971s...