Relay driver question....

[amptramp]

There are a couple of things to watch out for in this design.  The capacitors from the transistor bases to collectors act to slow down the turn-on and turn-off of the associated relay.  As the driver tries to bring the base voltage up, the collector will come down and is capacitively coupled into the base and thus, the turn-on will be delayed.  As the transistor turns off, the rising voltage coupled from collector to base will tend to hold the transistor on.  The diodes across the relay will also tend to slow down the turn-off of the relay since the current going through it will be diverted through the diode.  There is also a time constant from the flip-flop input which is AC-coupled to the driver.  If this is too fast, the driver will turn on for too short a time to drive the transistor fully on.

[Processaurus]

Excellent point.  That time would change, too, as the battery ran down and the voltage the coils were seeing got lower.

[R.G.]

You're both right - it is a quick and dirty, works mostly, good-enough-for-effects circuit. Making a guaranteed, works-everytime circuit can take more time -and information about the specific relay.

Some assembly may be required.

And you though Christmas was easy. 

[Processaurus]

You're both right - it is a quick and dirty, works mostly, good-enough-for-effects circuit. Making a guaranteed, works-everytime circuit can take more time -and information about the specific relay.

Some assembly may be required.

And you though Christmas was easy. 

Definitely not!  Effects design is a lot more fun and forgiving than something like medical life support system design. Speaking of forgiving, I read an interesting tidbit looking up different debounce circuits for this, and that is that TV remote control designers don't worry about having bulletproof fast debounce, because studies have shown people will unconsciously just keep pressing the button until it works. 

This looks like a big circuit, that would frighten people off, but it is a 1 chip (and a cheap, common one) solution that's quite minimal/elegant.  It opens up the door to use much more interesting switches, I just ordered some of these illuminated vandal proof switches to try a pedal, for example:



If the low battery iffiness bothers people, a 78L05 (the petite 5 volt regulator) in place of the aforementioned 140 ohm dropping resistors to feed the coils' + pins on the also aforementioned NEC 5 volt latching relay would sturdy up the operation at varying battery voltages.  Also, changing the 10K resistors to Vcc in the RC combo that sets the pulse length to be larger will give a longer pulse, if there is a need.  Too long being much preferred to too short in this application.

[Processaurus]

Doh!  Those switches are about a millimeter too shallow to be able to switch them with your foot reliably.  Every 5th or 6th press doesn't quite push them down enough.  Have to save them for building an elevator.  I guess I'll use the small bear "soft touch" footswitches, unless there is something else interesting out there that foots the bill.  I really like the big Lehle buttons, but I'm sure they're a custom deal.  Very Euro-civic.

[Processaurus]

Just got some of the aforementioned 5v NEC relays, and adapted the Geo circuit with a 130 ohm resistor (closest thing to 142 ohm) in series with +9v and each coil, the circuit works down to a supply of 6v.  Not bad operation for when the battery starts to go.  As an experiment I also tried a 5v regulator feeding the coils directly, and the circuit only worked with a supply down to ~5.5v; seems you can't run the batteries appreciably lower with the regulator...