Hi All,
I have been struggling with a circuit design and can't come up with something that will work. I want to turn a single input voltage circuit into a 'universal' input by having a relay make the appropriate connections based on input voltage. I know the relay needs to switch one of the AC wires from one AC input of the bridge [full wave rectifier] to the center tap of the primary caps after the bridge [voltage doubler]. And it seems obvious to me that I should start in the full wave position because otherwise I could voltage double 250VAC which would be destructive.
However, the circuit to detect the AC voltage and change the relay's position is eluding me. I've got several circuits in SPICE but none of them work. Most of the circuits rectify the AC voltage and then AC couple a voltage divider to some circuit which needs to trigger at the appropriate AC voltage, which I'm thinking should be about 160-180VAC.
Unfortunately, nothing has worked for me. Zener knees are too soft, discrete pass transistor regulator is too lossy, I can't get the hysteresis around my transistors to work correctly, etc. I want to keep this circuit simple with just diodes, transistors, and passives. Does anyone have any circuits or ideas they'd be willing to share? Thank you for any help.
I have been struggling with a circuit design and can't come up with something that will work. I want to turn a single input voltage circuit into a 'universal' input by having a relay make the appropriate connections based on input voltage. I know the relay needs to switch one of the AC wires from one AC input of the bridge [full wave rectifier] to the center tap of the primary caps after the bridge [voltage doubler]. And it seems obvious to me that I should start in the full wave position because otherwise I could voltage double 250VAC which would be destructive.
However, the circuit to detect the AC voltage and change the relay's position is eluding me. I've got several circuits in SPICE but none of them work. Most of the circuits rectify the AC voltage and then AC couple a voltage divider to some circuit which needs to trigger at the appropriate AC voltage, which I'm thinking should be about 160-180VAC.
Unfortunately, nothing has worked for me. Zener knees are too soft, discrete pass transistor regulator is too lossy, I can't get the hysteresis around my transistors to work correctly, etc. I want to keep this circuit simple with just diodes, transistors, and passives. Does anyone have any circuits or ideas they'd be willing to share? Thank you for any help.
Hi Evan,
It might be a little easier if you could convert the rectified AC to be more like DC, to feed to your trigger circuits.
A simple RC low-pass filter will tend to give the average value of a rectified AC signal.
If you can use enough of a low-pass to get the ripple-voltage bands to not overlap, you should be "home free".
I don't know how much time-delay you can accept. So that might be a minor issue.
Alternatively, and probably better in this case, you could leave most (or all) of the series resistance out of the lowpass, keeping the capacitance to gnd, and add a "bleeder-resistor" in parallel with the cap, and have a Peak Detector (or, actually, more like an Envelope Detector, since there's a bleeder R), which might enable your circuit to "make the decision" a little more quickly, and probably more easily and reliably.
You might also want SOME low-pass filtering, after the envelope detector, depending on your trigger circuitry, etc.
You'll probably just have to experiment with the C and R values (which is trivially easy, with Spice). But a few uF and a large-ish R in parallel (maybe start with 2.2uF or 3.3uF and 100k or more) should work pretty well, for a low-frequency envelope detector. And you can scale C or R pretty far, either way, once you find an RxC value that works for you.
You are lucky, in this application, too, since you don't have to worry about tracking different frequencies and continuously-varying amplitudes. (The real fun begins when your envelope detector has to work accurately for both low AND higher frequencies, and has to accurately track amplitude changes in minimum time. THAT's when I finally got serious about voltage-controlled resistances; i.e. for the bleeder R.
Good luck!
- Tom Gootee
http://www.fullnet.com/~tomg/index.html
It might be a little easier if you could convert the rectified AC to be more like DC, to feed to your trigger circuits.
A simple RC low-pass filter will tend to give the average value of a rectified AC signal.
If you can use enough of a low-pass to get the ripple-voltage bands to not overlap, you should be "home free".
I don't know how much time-delay you can accept. So that might be a minor issue.
Alternatively, and probably better in this case, you could leave most (or all) of the series resistance out of the lowpass, keeping the capacitance to gnd, and add a "bleeder-resistor" in parallel with the cap, and have a Peak Detector (or, actually, more like an Envelope Detector, since there's a bleeder R), which might enable your circuit to "make the decision" a little more quickly, and probably more easily and reliably.
You might also want SOME low-pass filtering, after the envelope detector, depending on your trigger circuitry, etc.
You'll probably just have to experiment with the C and R values (which is trivially easy, with Spice). But a few uF and a large-ish R in parallel (maybe start with 2.2uF or 3.3uF and 100k or more) should work pretty well, for a low-frequency envelope detector. And you can scale C or R pretty far, either way, once you find an RxC value that works for you.
You are lucky, in this application, too, since you don't have to worry about tracking different frequencies and continuously-varying amplitudes. (The real fun begins when your envelope detector has to work accurately for both low AND higher frequencies, and has to accurately track amplitude changes in minimum time. THAT's when I finally got serious about voltage-controlled resistances; i.e. for the bleeder R.
Good luck!
- Tom Gootee
http://www.fullnet.com/~tomg/index.html
Evan Shultz said:Hi Tom,
Thank you for the reply. This is similar to the scheme that I came up with that successfully works. Thank you for your suggestions and letting me know someone else thinks like me (although, maybe that's not a compliment
Hi Evan,
You're welcome. No problem, Evan. I'm glad to hear that you already got it working. (And I'm sure that that MUST have been a compliment.
For a commercial product...
Hello All,
Would using opto-isolated TRIACs be useful for an autoranging circuit?
The TRIAC, once triggered should conduct until the current passing through it drops below the TRIAC's rated threshold, like after each 1/2 AC mains cycle, so after each 1/2 AC Mains cycle you would need to re-trigger it with the opto-isolated driver (which looks like an ordinary optocoupler.)
This ought to allow the designer to come up with an AC input monitoring circuit that can enable / disable the voltage doubler circuit every 1/2 AC Mains cycle.
Hello All,
Would using opto-isolated TRIACs be useful for an autoranging circuit?
The TRIAC, once triggered should conduct until the current passing through it drops below the TRIAC's rated threshold, like after each 1/2 AC mains cycle, so after each 1/2 AC Mains cycle you would need to re-trigger it with the opto-isolated driver (which looks like an ordinary optocoupler.)
This ought to allow the designer to come up with an AC input monitoring circuit that can enable / disable the voltage doubler circuit every 1/2 AC Mains cycle.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.