re: solenoid hacking project

cbmnut_at_hushmail.com
Date: 2002-09-27 16:13:22

Hi all, esp. Rudd, Gideon:

an update to my project.  I constructed the circuit with IPS511
HS switch, and using independant 9v battery connected to vcc of switch,
and common ground to c64, I was able to activate switch without resetting or disturbing of running program of c64.
C64 is connecting to logic input.
There were some sparks even as I connected the solenoid (with switch on), so the ips511 seems pretty robust.
However I want to try this now with a portable cpu, the atmel avr 2313.
My only question concerns the running of the power supply.
Should I do this version 1:
+ ----LM2586-(18v out)--IPS511---solenoid---gnd
batt 12v              |--7805-(5v out)--avr2313--gnd
- ----

or this version 2:
+ ----------lm2586--ips511--solenoid--gnd
batt 12v  |--7805--avr2313--gnd
- -----

In other words, I run two regulators in series (version 1), the 
2586 is a boost switching regulator at 100Khz outputting 18v@1amp.
This rail is now used for 7805 linear 5v regulator, and the 5v
is used to power cpu.
In version 2, the both regulators take power directly from battery.

The difference?  I am looking to reduce noise which could upset cpu.  Some small noise eg on reset pin could reset the running program, or worse noise could damage i/o pins, or otherwise reset cpu, corrupt flash/eeprom, etc.

The solenoid coil back EMF is absorbed by a diode in ips511.  I could also use external diode, however it would never switch on unless it had a lower forward voltage than the in-built diode (correct?).
However there is still the .42v or so of negative current which would reach the 2586 power supply, I guess this is just seen as a brief current draw?.  This would be fed slightly by the output filtering capacitor, and possibly also by the current sensing feedback, which I suppose would cause another charge-pump cycle.
In version 2 it seems safer to me?
I can provide a full circuit.
Thanks!




Get your free encrypted email at https://www.hushmail.com

       Message was sent through the cbm-hackers mailing list

Archive generated by hypermail 2.1.4.