voltage regulator +/-

could you give me keywords for the circuit? or is it pretty vague? if I had a name, or some keywords, I might be able to google it.

 

Log in or Sign up to hide all adverts.

"H bridge" and "complementary" are good keywords. The H bridge is a pair of complementary push-pull power output stages just like a power audio amplifier. They can be linear or switched.

I don't know how much current you want, Cato. Be aware that if you stick with linear, you are going to have to use heavy heatsinks if you are pushing an amp or more, and if you use the H-bridge setup, you are going to need to heatsink and electrically isolate four transistors. It's a bit easier to design the circuit using bipolar transistors when you are doing linear. You feed one side of the H-bridge with a voltage that mirrors the other, swings low when the other swings high. Power transistors need to be fed at least 200 milliamps of base current to control 10 amps because their current gains run around 50. You need a driver transistor or transistor pair to power each side with that requirement. Do not cross-couple an output to an input to invert the voltage. Use something else like an inverting op amp circuit. It works better, makes a more stable circuit, and helps keep kickback voltages further away from anything sensitive.

 

Log in or Sign up to hide all adverts.

google this entire phrase:
sourcebook of electronic circuits

edit
that is the name of the book i have
has something like 3000 circuits in it
it's old though copyright 1968 ( gasp )
it even has tube circuits in it

edit
you can also google
schematic

 

Log in or Sign up to hide all adverts.

It's a good idea to use snubbing diodes across each output transistor as well.

 

Diodes that absorb the kickback voltage can heat up really badly. I have in an experimental circuit put resistors in line with the diodes on the theory that they would reduce the kickback voltage and reduce the wear and tear on the diodes but haven't done the detailed testing to see how well that works.

 

You do have to use diodes that can handle the abuse. Me, I'd rather blow a bucks worth of rectifier diodes than 20 dollars in output transistors.

Please Register or Log in to view the hidden image!

Don't know about a resistance in series with the diodes. Caps in parallel might work better. Sometimes you just have to try and see.

 

I am not that great with circuits yet, I still am not sure how to make an h-bridge vary the voltage, positive and negative, by using a single knob (potentiometer).

 

Kevin, the idea of resistors in series with the diodes is that it will increase the resistance times inductance and reduce the peak current and voltage fed back to the diodes by increasing the time that it takes to discharge the inductor. It's not quite as simple as the inductor being like one side of a capacitor, but the idea is very similar. Make the stored energy take longer to discharge to soften the impact. Even stretching the fall time of the collapsing magnetic field by a few microseconds might make a huge difference in the heating of the diodes. This also damps ringing oscillations.

Some items that you try to feed AC current through are going to kick back like crazy. There is a lot of stored energy to deal with. Some time I am going to try to work other schemes but that's what I have for now.

Cato, take some time to learn how inverting and non-inverting op amp circuits work. If you can, set them up with voltmeters and watch what they do. For the circuit that I have in mind, you need an inverting unity gain amplifier and a non-inverting unity gain amplifier.

 

Cato, to vary an H-bridge using one potentiometer, you need to invert the voltage that you feed to the other side, so when one side is at maximum, the other side is at minimum.

It's not complicated if you are used to it. If you want to make it a little simpler, build yourself a power supply that produces a positive and a negative DC voltage at the current you want. Your choice if those two are regulated, but they can be fixed. I don't think that you will notice any difference in performance if you use regulators.

Then bridge your potentiometer between the positive and negative voltages and apply the voltage from the wiper to the input of your complementary output stage. It can be built just like an audio output stage except that you do not need to match the complementary pairs of transistors. Hang your load between that and ground. The complementary pair act as voltage followers. The one on the positive side conducts only when the input voltage is positive, and the one on the negative side conducts only when the input voltage is negative.

 

I had an idea for solving that back emf problem.

what if you wired two diodes in parallel (facing opposite directions), and put a resistor in series with the one that will pass the back emf? the resistor will have no effect when the circuit is flowing properly, but will drop any rweverse voltage.

thought of it last night while I could not sleep =]

 

the check is in the mail let me know if it helps

 

kind odd that it does not show where the output would be. I will figure it out later =], no time at the moment.

 

thats a good question.
it would have to be one of the resistences.

the original article would probably explain it but alas i don't have it

 

I uploaded the schematic to my folder on my friend's site so people can view it:
http://www.killerkittykats.net/quik/cato/

 

Cato, that schematic is obviously not what you would want to use. It is about forty years old. It is written in the style that was considered to be "educational" and is not a practical example.

I probably can't spend enough time to get you a good schematic until the weekend. I definitely have ideas that will work, along with some that I have actually tried.

 

yeah, I was just hoping to get ideas from it that I may use a different design.

so far, the h-bridge idea sounds good, but I started back to school, so I have not had as much time to work on it. if you could show me a schematic, that would be awesome

 

Here is a schematic that gets most of it, but you need a way to invert the signal. If you use a dual potentiometer and fed the plus and minus inputs from each half, with one wired backward to the other, it should work quite well. You really want to use a regulated supply to the potentiomer to prevent feedback. I have more in the works. Thought I was going out today but that didn't happen, so I did this part of the schematic.

Schematic

I have two each of Q1 through Q4 because what you have is two of the same circuit. The 4.7k resistors can be reduced to 2.2k if the total gain of your driver and output transistors is to small. You need a total gain of about 10,000 to run it with the 4.7k resistors.

This would also work fairly well as an audio output stage. The op amps are fed back from the output so they exactly track the input voltage. This eliminates the dead spot but it does generate high frequency transients which need suppression circuitry like grounding the outputs through bypass capacitors, which I did, and some high current inductors, about 10 to 100 microhenries,which I didn't. But this is the H-bridge configuration from my own personal experience. When I finish you will see that while it may look a little complex, it is a lot less nerve-wracking than a design could get. Also, this will work very well from a single-ended supply like a 12 volt battery. It has inherently low quiescent current drain. Like I said, it's not quite ready. It needs DC bias for a zero output and a few more things. Also, I am having to do this without the benefit of a breadboard.

 

your link is dead =]

 

it works for me

 

metakrons circuit
http://f6.grp.yahoofs.com/v1/kPvGQz...VCjEYl_BxOkTG8qedD5WtZSNsksRtwKWLfgg/out4.gif