Latching circuit (D-FF?) for LED Switch Track Indicators

I'll have to give it a look, I'm thinking I may want to produce some boards from a design I'm working on.

"ground" in this case is just a local circuit ground, not the real "water pipe" ground.

Thanks, Rich – Even if I end up not using ExpressSCH and ExpressPCB for this project, I know that they will come in handy in the future.

 

John,

Regarding the additional diodes – I suspect they will work, and I will try them if nothing else to satisfy my curiosity, and anyone else’s who has been reading this thread.

 

As far as the caps: I debated their location, but remembered that early on you had recommended ". . . I'd just add some .1 caps across the input resistors to snub any spikes that come through. . . ", and so I added them across those resistors. I'll try them instead from the pins to ground first, and if it doesn't work, I'll try the diodes.

 

However, adding the diodes would take the component count up to 21 for two turnouts (one Dual-D chip), which would make the total per board for 12 turnouts 126 plus a few more per board, and I need four of those, or well over 500 for the project. It is not the cost, as those components are but pennies, but it is a lot of work even going the PCB route considering that eventually I will replace all of this with the sensors that indicate the actual position of the turnout points. These sensors would act as switches and require no latching circuits.

 

By the way, I did disconnect the turnouts to see if the solenoid coils had any effect, and it made no difference. Still, I consider the field collapse feedback a potential problem, and if this happens to be another source of problems, I will definitely wait until I have the point sensors developed.

 

I need to work on some other things on the layout for a few days, so I may take a short break from this.

 

THANKS!

Alex

 

 

I'm thinking that maybe a new approach might be called for.

 

A bulletproof way to do this is with Atlas Snap Relays, they just parallel with the switch relays.

It seems to me that using Bruce Chubbs C/MRI system would be simpler.

Cheaper too?  More details would help...

John,

 

I moved the caps (IC pin - cap - ground) and it works perfectly on the bench, but still acts up in the control panel whether the turnouts are included.

 

The Atlas Snap Relays would work nicely, but at $10 each it would mean $460.

 

Would sure like to learn more about the "Bruce Chubbs C/MRI system", though.

 

Alex

I have to think that the issues you're experiencing are noise related.  You're still bringing the noisy AC signal right into the board.  Did you also relocate the diodes as I suggested?  Driving that HC gate with signals out of it's operating range would cause all sorts of odd issues.

 

If you stick with this technique, I'd be tempted to totally isolate the AC from the switches from the digital circuits.  Something like this Quad Opto-Isolator would do the trick for four inputs.  FWIW, I've frequently found the need to isolate stuff with an opto-isolator when I'm putting circuitry into a locomotive or operating car.

 

 

 

Just google bruce chubb c/mri. Or computer model railroad interface

I looked at Bruce Chubb C/MRI, you must be kidding, right?  You're trying to kill flies with a steamroller!  All Alex is looking for is switch position feedback, not to automate the whole layout!

John,

" . . . isolate the AC from the switches from the digital circuits".

 

Yes, it may indeed come to that. That's what we did (or had to do?) for all the solid state controls in machines that operated on AC (480/240/120). I will be looking at this again this weekend.

 

-BUT-

Last night I came up with an easier way (from what I had done with my custom made 'mini-micro' switches) to sense the position of the points using a totally non-contact approach. If I did not have all the turnouts installed and ballasted already, the new scheme would be a cinch to install. What I am trying to do now is install it as an add-on feature, with the turnouts in place.

 

Will report back - soon -.

 

Thx!

 

Alex

 

PS. I briefly looked at the C/MRI, and quickly reached the same conclusion as you did.

I've done a number of things with control circuits in AC powered devices, many times I end up using opto-isolation, it just makes things sooo much easier.

Just an update as follow-up.

 

I will be using the non-contact switch track points-position-detectors to drive the control panel LED's, so I do not need the latching circuits we have been discussing.

  

However, the instability of the latching circuit bugged me, so I finally took the scope to the control panel last night, and looked at the Clear and Set inputs simultaneously.

 

The input signal to either Clear or Set from the toggle switch was clear: +5v with no signal, and then down to 0v when the toggle switch was actuated, as desired.

 

But the other input, which should have remained at +5, had about a 3v AC signal present when the toggle switch was actuated for the ‘other’ input. This, of course, would confuse the FF, and result in unreliable outputs.

 

There are several ways to get around this, including using a DC supply to power the switch tracks (turnouts), but I am going to let it rest and concentrate on the actual position detectors.

 

Thought you may like to know, and thank you for your help.

 

Alex

 

Thanks for looking, I was curious why they didn't work.