CUB3 Question

[Phil m]

I've wound a couple of pickups now with my new setup, the CUB3 with a reed switch.

I'm wondering how precise the counter actually is. How do I know if that thing is doubling up or missing any counts while it's working?

When I wind at high speed you see sometimes that the counter jumps a digit or two. Is that because it's got a hard time keeping up with the switch or because it's doubling up?

[Joe Gwinn]

I've wound a couple of pickups now with my new setup, the CUB3 with a reed switch.

I'm wondering how precise the counter actually is. How do I know if that thing is doubling up or missing any counts while it's working?

When I wind at high speed you see sometimes that the counter jumps a digit or two. Is that because it's got a hard time keeping up with the switch or because it's doubling up?

They are quite precise, if correctly set up.

For what fraction of a turn is the reed switch closed? Is there exactly one closure period per turn? Measure this with a voltmeter across the switch, slowly turning the bobbin holder by hand. Voltage will be ~3 volts when switch is open, and ~0 volts when closed.

[Norcal_GIT_r]

It maybe my reed switch, but my CUB3 miss counts when I get over 1700 rpm.

[Pep]

I think that it has something to do with the reed switch or the setup. The CUB3 is made for high speed.

[ken]

I use an infrared system myself, because I thought the reed might not be able to switch as fast as I wanted it to go. If you do use the light system, remember to shade the counter's sensor assembly. My halogen desk light I use on my winder would 'screw up' the counter from time to time.

Ken

[Phil m]

They are quite precise, if correctly set up.

For what fraction of a turn is the reed switch closed? Is there exactly one closure period per turn? Measure this with a voltmeter across the switch, slowly turning the bobbin holder by hand. Voltage will be ~3 volts when switch is open, and ~0 volts when closed.

Hey Joe,

Here's the results of my test. I get approximately 1.6 V on my multimeter for about 320 degrees of the turn of the bobbin holder and 0V for the leftover 40 degrees when it comes close to the reed switch. What does that mean?

[Joe Gwinn]

Hey Joe,

Here's the results of my test. I get approximately 1.6 V on my multimeter for about 320 degrees of the turn of the bobbin holder and 0V for the leftover 40 degrees when it comes close to the reed switch. What does that mean?

That the CUB3 will begin to miss pulses above that speed at which 40/360= 11%= 1/9th of a turn takes 5 milliseconds: 0.005*9=0.045 secs per turn, or 22.2 RPS, or 1,333 rpm. The actual speed may be slightly high, depending on the CUB3, but the CUB3 is rated for 5mS minimum duration pulses.

If the CUB3 is missing many pulses, it will seem to double a pulse when it gets lucky.

So, to go faster, make the reed switch remain closed for a higher fraction of the turn.

[Phil m]

thanks very much. I'll play around with the magnet and the reed.

[Joe Gwinn]

thanks very much. I'll play around with the magnet and the reed.

If you can't get much more than 1/9th of a turn, use two pieces of mild steel acting as polepieces to extend the range of angles where the flying magnet can couple to the stationary reedswitch.

[Phil m]

That's a really good tip. I wouldn't have thought of that. I already played a bit with the reed, it's still at 1/9th but is repositioned differently and already I could wind at around 1000 rpm and the counter didn't have any problems with doubling up/missing counts as before. I'll still try the steel thing though

[Joe Gwinn]

I use an infrared system myself, because I thought the reed might not be able to switch as fast as I wanted it to go. If you do use the light system, remember to shade the counter's sensor assembly. My halogen desk light I use on my winder would 'screw up' the counter from time to time.

Reed switches are fast enough, opening and closing in a few milliseconds.

For example, at 3,000 rpm (50 rps), each turn takes 20 milliseconds. If one has 50% closure time, the few milliseconds of switch delay will be about 20% of the 10-mS closure time, and all will be well.

The issues with reed switches are getting sufficient closure time, and ensuring that there is exactly one closure period per turn, these both being a matter of mechanical and magnetic design. Once built correctly, such setups are very reliable. (Assuming that the glass envelop is held in something soft, like rubber.)

Eventually, the reed switch will physically wear out and need to be replaced, but this will be a very gradual process. Reed switches are good for 10^6 to 10^9 operations, depending on quality of manufacture. If one is winding 10000-turn coils, one will get a minimum of 10^6/10^4= 100 coils per reed switch. In practice, one will get at least 1000 coils per switch.