MTTF of circuit

nsaspook said:

Honeywell might have some that will work but maybe not very inexpensive. https://www.electro-tech-online.com/custompdfs/2010/08/1917-1.pdf

Click to expand...

Well, that's the problem, initially it was $0.5 per Hall switch, now, configured as a vane interrupter, it is about 10 times as much, which would kill our project.

kjennejohn said:

I ran the MTBF calculator on this.
Part | Fail Rates per Million Hours
---------------------------------------------------------------
relay, mercury wetted contacts | 6.27
transistor, small signal, silicon | 5.73?!
resistor, ceramic compound | 1.13

I used MIL-217E-1 Reliability Prediction Stress Method, benign ground use, at 25 deg C.
I was amazed at the transistor failure rate! Damn near as bad as the relay! You might as well
save the money and leave the transistor out of the design!
kenjj

Click to expand...

ah MIL-217, that brings back memories ... I used to do that on paper with the spec in hand! use a bigger transistor. the failure rate is based on junction temperature and duty cycle, at least in the real thing. the formulas make assumptions that are bypassed in the real spec if you know the junction temperatures and the programs might be assuming full load.

a true application of MIL-217 includes environment including ambient temperature, heatsinking, atmosphere, and vibration as well as the type and quantity of solder joints as well as complexity of ICs.

Dan

You probably have as good a circuit as you are going to get. Your problem is you have 12,000 of them. The relay is on par with a single transistor so adding a complex ic won't do you much good. How good are the mechanics? How do you get power to them.
Lets say each part had a life of a million hours. 5 parts gives you 200000 hours per assembly. divide that by 12000 and you have a failure every 17 hours. Is that good enough?
My bet is you will do 10X better than that if failure is defined as the led doesn't blink at all.
Again, you have some risks that may be far worse than the intrinsic failure rate. Loosing 25% of the lights due to a manufacturing defect (lets say a broken seal on the relay) would be a lot worse for the project.

A search in Digikey turned up this photointerrupter from Digikey for 17.5 cents each in 10,000 quantities. You would need to add an NPN transistor stage to invert the signal so that the LED turns on when the gap is interrupted.

There likely are other photointerrupters available for a similar price.

crutschow said:

A search in Digikey turned up this photointerrupter from Digikey for 17.5 cents each in 10,000 quantities. You would need to add an NPN transistor stage to invert the signal so that the LED turns on when the gap is interrupted.

There likely are other photointerrupters available for a similar price.

Click to expand...

I would be very interested in seeing which one you found...

Othello said:

I would be very interested in seeing which one you found...

Click to expand...

Sorry, I didn't realize that there was a link already hidden in there. Thanks

I think the link "this" should be highlighted and in bright red (like I show) so it is easier to see.

audioguru said:

I think the link "this" should be highlighted and in bright red (like I show) so it is easier to see.

Click to expand...

Well, it wasn't. On my screen it is barely discernible.

This discussion has been very helpful to me. I started out with a reed switch, and while I was aware, in a general sense, of the existence of Hall switches and photointerrupters, I had not known how small and how cheap they had gotten over the years.

After this last suggestion I will run a few tests with photointerrupters and see how they work for us. The reliability seems to be better than reed switches in any case.

Photointerruptor, outdoors. San Francisco????

ronv said:

Photointerruptor, outdoors. San Francisco????

Click to expand...

On the north side of a building, operating at night only, in an installation which extends from 20 to about 170 feet in the air. It is a public art project, an error here or there wouldn't matter.
I will prototype the set up and let you know how it goes.

Certainly, the default highlight color for links is not very noticeable. Changing the highlight to red as Uncle $crooge noted is a good idea.