The JFET (J2) is supposed to turn off Q13 when the switch to the battery opens. This cuts power to the load in the green box (D2, D3 and the two resistors). Then the P-channel FET (M1) conducts supplying the load in the yellow box.
Schematic is here.
This all works in simulations but when I breadboard it the JFET pulses on and off at 50% duty at 60 Hz. When I connect my probe between the JFET gate and ground the circuit works correctly.
What's causing this?
Put a 10 Meg resistor from the Gate to ground. you're allowing the gate to float.
-- http://webpages.charter.net/jamie_5"
And mind the reverse Vbe limit on Q12 especially if capacitors are added somewhere.
-- Regards, Joerg http://www.analogconsultants.com/
"Hammy"
** Geeeez -do ya reckon it might be stray coupling of 60 Hz supply hum into the gate of that FET ?
.... Phil
Thanks
Joerg, I didn't even think to check the Vbe, it is at 3.28V when just supplying the 20mA load. When I add the additional loads 8 in total for a combined 80mA the Vbe spikes to 15V on a spice simulation. The
80mA load is represented by R13 and the combined load capacitance estimate of C4 180uf. In the real world the loads represented by R13 and C4 will be supplied from the 7805 through 50 to 70ft run of 22AWG. When I add the DC resistance and a small inductance to simulate the line parasitic's the Vbe doesn't exceed 4V. I'm using U9 to block these loads until the flyback takes over the 7805 input voltage, and the battery is switched out. I also want these delayed because I want the loads supplied by the flyback to start first. The flyback is represented by V2 pulsed source. It'll be supplying 720mA worst case.Schematic with loads.
Now that I think of it another stupid mistake is using the J177 to switch out Q13. The max pinch-off for the J177 is 2.5V. This limits the usable battery range to about 7.5V or else the JFET will conduct as soon as the battery is connected.
I guess I'll resort to plan B use an LM2903 to monitor when the switch is open then turn Q13 OFF. I can always use the other comparator to monitor the battery level and light an LED when it gets to 6.5V. I'll hook up the circuit with the 80mA load switched in and monitor the Vbe if it is a problem I'll try to clamp it with a zener. If that doesn't work I guess I'll replace it with a P-channel mosfet. I didn't want to use a FET because they leak anywhere between 10 to 250uA. But maybe I'm being too anal;)
Can't see J177 in that schematic. But be careful with things such as reverse Vbe. Exceeding it can degrade or destroy a transistor.
Another issue to watch out for during turn-on: If Q9 and Q10 conduct this can cause an excessive base current in Q12 and destroy it.
-- Regards, Joerg http://www.analogconsultants.com/
Sorry I was referring to the first schematic I posted (J177). I omitted it from my second post because it won't work down to about
6.5V battery voltage to switch Q13 off. I'm going to try a comparator.Whenever it senses less then 6V (switch open) at the emmiter of Q12 turn Q13 off . I measured a little over 5V so 6V should be safe.Thanks for the pointers I'll definitely be flicking it on and off and checking what you suggested as well as anything else I can think of.I have some BC640 PNP's in TO-92 a little beefier then 2n3906's I'll swap Q12 out with the BC640.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.