Mee_n_Mac:
And here it is … as expected the circuit, as depicted, doesn’t work. See if your schematic, and actual circuit, agree w/mine.
The transistor could sink a little current since a reverse biased base-emitter junction does leak. In addition the peak inverse voltage of that junction is surprisingly small. I had that last point hammered home once when I noticed a higher than expected current in a circuit.
LTspice might model the leakage pretty well but I am betting it doesn’t model the reverse breakdown.
waltr:
Linear Tech also makes DC-DC Voltage inverters.
If you only need a few mAs at most then the 555 approach works but for any more current a DC-DC converter is the way to go.
Well, I finally managed to track down some tech specs on what I’m trying to replace. I have some vintage computers here. One of the newest is an Apple IIc which has an internal voltage converter unit that takes 15 volt 2 amp DC input and converts it to +5, +12 and -12 using a bunch of inductors, capacitors and some unidentifiable vintage transistors.
I’m fine on the +12 and +5 side. Those circuits are very straight forward.
The -12 is the one that I am working through. The specs I’ve found say that it only needs to draw 100ma on the -12 line, so it may very well be that I can just do it directly with the 555 if I drop the voltage a bit from the +15 to +13 which is still in range for a 555.
From what I’m reading, 100ma should be no issue with a 555 voltage inverter circuit. Would your experience agree with that?
Which 555 circuit, the one UhClem suggested ? And there’s some question in my mind as to what supply voltage you have to convert into -12v. And of course the specs on the output, how much pk-pk ripple is allowed, how much variation w/load current (or supply voltage), how long is allowed from startup to being in spec ? For example the last could be several seconds if you use the aforementioned 4700uF caps. And it may depend on the 555 used. Most had 200 mA ouputs but I think some were less ?
Just for giggles here’s an extension of the above 555 based circuit. That circuit was ‘single phase’; for each cycle of the 555 C3 would ‘charge up’ from the 555 and then dump what charge it could into C4 to make the negative output voltage. If the load empties C4 of charge faster than it can be refilled, then the negative output voltage sags from it’s OL value of -13.7v. What might be done ?
One thing would be to make the circuit ‘2 phased’; that is to add another cap (C5) that’s getting charged while C3 is discharging and then is dumping into C4 while C3 is charging. The added diodes steer the current for C5 just like D1, D2 do for C3. Q1 and Q2 act like switches, alternately connecting C5 to Vcc or ground depending on the state of the 555.
As can be seen the negative output voltage still sags but not as much as before.
(click on to open)
(click on to open)
It the above circuit “good” ? Well there’s one glaring potential problem that I see but I’ll leave the fun to others. Certainly not a huge improvement for the parts added.