Rotary contact switch concerns

[Radio_Dave]

I'm in the process of restoring a 45 year old, homebrew, electric clock. It uses rotary, sliding contact switches to change the time (please see attached photo).

I have two concerns. Firstly I'm worried about premature wear (and interference), because of arcing, as the contacts (both made of brass) rotate. Secondly, I have a feeling metal is going to be deposited, onto the tufnol, leading to shorts.

Is there anything, simple, like changing the type of metal used for the sliding contacts or some sort of lubricant that will help?

TIA
David

[MikeyPP]

Is there a sprung loaded plunger** within the post/handle or is just a solid post/handle ?

Brass is soft as you realise , there's bound to be some wear depending on the friction caused by the above?

** If not could be an idea

[Radio_Dave]

Yep, You're spot on. They're 1/8" diameter and spring loaded.

David

[MikeyPP]

45 years old

As it's lasted this longn without any real obvious/appreciable wear, the design can't have been that bad, why shouldn't it last some more without redesigning ...
Unless you're thinking of constantly adjusting it every day ?

[M0FYA Andy]

I'm confused as to what these switches are doing. Are they just used to set the time manually, or are they rotating all the time?
What mechanism actually measures the passage of time?
Andy

[Radio_Dave]

Originally Posted by MikeyPP

As it's lasted this longn without any real obvious/appreciable wear, the design can't have been that bad, why shouldn't it last some more without redesigning ...

Hi Mikey, It was made in 1967 as something to do with an university degree and it has not been used (or finished off) since that time.

Originally Posted by M0FYA Andy

I'm confused as to what these switches are doing.

Hi Andy, The switches are driven by a synchronous motor. From left to right, the first switch displays minutes on nixie tubes, the second one display the tens of minutes and the third one does the hours.

Regards
David

[Herald1360]

Looks like Meccano spindles and wiper elements...... ingenious, but as you surmise, probably not great for longevity. Did (would) the nixies go out briefly at digit change time? Or was there some latch and release circuit employed?

[Radio_Dave]

Hi Chris,

It's early days and I haven't really paid much attention to the speed of the digits changing but it should be quite quick.

If you look at the attached photo you'll see the other side to these switches and how they are moved. The switches don't move in real time (slowly) but are moved, quickly, by levers. The other, clever, part of their design is the 1/8" gap between the fixed contacts and the 1/8" diameter of the moving contact, theoretically, leaving no gap in the transition from one digit to another... Unfortunately, I suspect, it's this large, flat, surface area of the moving contact that's contributing to the arching.

David

[ronbryan]

Have you tried a series C-R snubbing network between each fixed contact and the moving contact? This technique is used to reduce arcing in relay contacts and could be tried on a few adjacent contacts without too much effort.

If you know the voltage and current you are switching, the C and R values can be calculated as per this link http://www.industrologic.com/mechrela.htm

Otherwise 0.1uF and 47 to 100 ohms come to mind.

Alternatively, maybe you could devise a means of electronically switching off the applied voltage during the switching process by deriving a strobe signal from the motor shaft.

Ron

[John M0GLN]

Originally Posted by Radio_Dave

Hi Chris,
If you look at the attached photo you'll see the other side to these switches and how they are moved. The switches don't move in real time (slowly) but are moved, quickly, by levers.
David

Its looks a bit like a simplfied Geneva mechanism.

John

[Radio_Dave]

Hi Ron. Thanks for your suggestions. Unfortunately I don't want to, drastically, alter the original design. I did hope somebody would suggest a way of making the, easily replaced, moving contacts sacrificial (or something like that) but if I have to, I'll live with it's shortcomings and only use it occasionally.

Hi John, I've never heard of a "Geneva mechanism" but after searching about it on Google I can see the similarity.

Thanks
David

If they are make before break it shouldn't be a problem, I also suspect the double digit phenomenom won't happen either, the striking voltage being higher than the maintaining.

[John M0GLN]

Originally Posted by Radio_Dave

I did hope somebody would suggest a way of making the, easily replaced, moving contacts sacrificial (or something like that
Thanks
David

Would it not work electrically if you replaced the existing spring loaded contacts, which are presumably brass, with a carbon brush material?

John

[Radio_Dave]

Originally Posted by merlinmaxwell

If they are make before break it shouldn't be a problem

Hi MM, That's a very good point! Unfortunately the contacts have a tenancy to lift their front edges as they rotate, reducing their contact area under the required 1/8". The designer of this clock made a note of this problem (I have the design notes) and he toyed with an idea to file a V into the moving contact so it would always make a contact across the, required, 1/8". Unfortunately, the ability of the round contacts to rotate spoilt this idea. You've got me thinking though and I'm wondering if should make some square moving contacts with a V in them .

Originally Posted by jcaines

Would it not work electrically if you replaced the existing spring loaded contacts, which are presumably brass, with a carbon brush material?

Hi John, As mentioned above I have the design notes and it was the original idea to use carbon wiping contacts. However, it seems that the carbon left a, conductive, trail in a very short space of time!

Thanks
David

[Herald1360]

It does look like this served its original purpose well, it sounds like the designer learned from it that perhaps it was not the best way to control the Nixies, ingenious though it is. There is always going to be a longterm wear issue with any sort of wiping contact mechanism. Maybe the carbon brush idea, but with narrow radial slots between the fixed contacts would cure the conductive path problem?

Tricky to think of anything obvious without throwing lots of (cheap) modern technology at it!

[John M0GLN]

Originally Posted by Herald1360

Maybe the carbon brush idea, but with narrow radial slots between the fixed contacts would cure the conductive path problem?

That should be an improvement, it would make it like a dynamo commutator, also are the spring loaded contacts and the fixed ones both made of brass? If so making one of them, presumably the easiest to modify, the sprung one, from a different metal should help with wear, perhaps a steel ball bearing if the gap was not too big?


John

[Dekatron]

My mind finally worked again and I now remembered that the magazines "The Radio Constructor" and "Electronic Engineering" published designs like this during the 1960s. If I remember correctly there were special motorized "counting" units that could be bought and used for each digit. I also know that "Practical Electronics" in the March issue from 1965 has the same kind of clock on page 324 - 327. I have that magazine somewhere, but I do not know exactly where right now so if I don't send you a copy of that article in the near future (if you can't find it at a local library) please remind me again.

[glowinganode]

Originally Posted by ronbryan

Have you tried a series C-R snubbing network between each fixed contact and the moving contact?
Ron

Neon bulbs and capacitors don't really mix, you could end up with a set of disco lights!
I think using a carbon brush would be beneficial, though is arcing really a problem? I'd expect the contacts would only be handling a few mA's. Leakage between contacts could be more of a problem.
Rob.

[Radio_Dave]

Originally Posted by Herald1360

Maybe the carbon brush idea, but with narrow radial slots between the fixed contacts would cure the conductive path problem?

Thanks, that's another good idea for me to consider

Originally Posted by jcaines

perhaps a steel ball bearing if the gap was not too big?

Hi John, I'll try this idea first, as it's the easiest. Hopefully the switch will move quickly enough, so the delay between numbers isn't too bad.

Originally Posted by Dekatron

I also know that "Practical Electronics" in the March issue from 1965 has the same kind of clock on page 324 - 327. I have that magazine somewhere

That's really interesting. The designer mentions an, inspirational, 1965 magazine article about a nixie clock using wafer switches to drive the nixies! Please don't go to any trouble on my behalf but please remember me if you ever bump into it in the future.

Originally Posted by glowinganode

I'd expect the contacts would only be handling a few mA's. Leakage between contacts could be more of a problem.

Hi Rob, As mentioned earlier, I haven't had it running for any length of time, so I haven't seen any physical evidence of damage yet. However the other problem of interference, picked up on the radio, is very evident.

Thanks
David

[Peter88gate]

Hi Dave,

I built the clock mentioned by Dekatron from the March 1965 issue of Practical Electronics in 1965 and it is still running today!

It relies on using standard rotary M-B-B switches to operate the nixies and I have had no real problems with contact wear. The "minute" switch required a 10 way switch and for this I used a 30 way Painton Winkler switch wired in blocks of 3 to give the 10 ways. The "tens of minutes" used a 1 pole 12 way switch wired in blocks of 2. I did have to replace the latter switch about 30 years ago.

The switches are driven by a 1 rpm motor using a spider arrangement similiar to your clock. I assume the nixies are GR10M (or their equivalents Z520M/ZM1020/GN4). The total current drawn by these is only about 9mA so there is no trouble with arcing.

If Dekatron can't find the article, I think I should still have it but it would require a trip up into my loft!

Peter