problem: CLSENSE shutoff of an LM3524 based circuit under all circumstances

[Mark Sulkes]

Problem with board from a commercial laser (circa 1988) employing an LM3524 whose emitter outputs drive several MOSFETs. The output current of the MOSFETs is monitored through a 22k resistor going to +CLSENSE (pin 4); -CLSENSE (pin 5) is grounded. The problem is that even with zero MOSFET current Pin 4 is ALWAYS above the ~0.2 V shutoff threshold for pin 4 (Pin 4 around 0.7 V with zero current fed back); thus the LM3524 emitter outputs are shut off--circuit doesn't work. The same observations are made with a similarly wired LM3524 on a prototype board. The board in question definitely worked at an earlier time. I suspect that it formerly worked because a circa 1988 LM3524 had different specs on the CLSENSE inputs, such that there was no shutdown with zero MOSFET current using the present feedback arrangement to Pin 4. First, please comment on whether this is the case. With a current LM3524, how should the current feedback be directed to Pin 4, such that the LM3524 is on with zero current and will shut off with the intended maximum current? (See attached file for circuit.)

 

[73's de Edd]

.


Sir Mark Sulkes . . . . . . .

I'm looking at your engineering drawing sheet and its time era on its dating.
Look back to the 1980's and see what might have been your specs on that older unit that might differ.
Tejas Instrumentations is now their big Sugar Daddy, and claim compatibilityof the device.
Refer to this referencing on its page 10-124 upwards and compare against your new TI device referencing.
Up in your 950 V voltage divider sampling and particularly R21 . . .at 10 meg . . . .any chance that it has drifted.
Typically high voltages and high resistance values don't get along too well together . . . long term.

I am short on my prior time commitments now.

http://www.smcelectronics.com/DOWNLOADS/1980-VOLTREG.PDF

73's de Edd




.