Driving max current hi and low from a battery

[mramos1]

I have a unit that I need to drop a battery across two terminals. After 5 seconds I need to flip the battery backwards between the terminals. If I do it by hand it works but I want to let a PIC do the work.

I was going to use relays, then I thought maybe FETs might do it. I did see an EDN article with 2 transistors that would drive the ground side to milivolts. Or something like a 1/2 an H-bridge. Did did try with NPN/PNP and just get not get what I wanted. If it can be done with transistors that is fine too.

I have a pic chip running in 3 volts (2xAA so close) and want to drive this lite load back and forth to 0 and +3V. But I want as much of the battery there as I can get to the terminals.

I was thinking a pair of FETs on each of the two connections. Does anyone have an input on this.

Thanks.

EDIT: continued in third post...

 

[Nigel Goodwin]

Looks like an obvious job for a DPDT relay - nice and simple, and minimum voltage loss.

 

[mramos1]

I have one G6HU relay. Now if I had 3 more it would be done

Sorry Nigel, I mentioned only 1/2 the project since I was hoping for a solidstate version and would dup it and drive them with the pic. Relays I will need 4 DPDT.. I can see if RadioShack still sells them.

The project is actually 2 sets of the terminals and they work together.

   T1-1   T1-2    T2-1    T2-2
1   +        -         +         -
2   -        +         +         -
3   +        -         -         +
4   -        +         -         +

 

[Sceadwian]

Sounds like a good usage for a basic h-bridge.

 

[Nigel Goodwin]

Sounds like a good usage for a basic h-bridge.

But more complicated, and more voltage loss!.

 

[Cabwood]

You only need ONE DPDT relay.

 

[mramos1]

You only need ONE DPDT relay.

Think you missed post three, or I missed something..

 

[agent420]

I have a unit that I need to drop a battery across two terminals. After 5 seconds I need to flip the battery backwards between the terminals. If I do it by hand it works but I want to let a PIC do the work.

I'm just curious what this means; what is the application? Perhaps there are other solutions if we knew what the problem was?

 

[mramos1]

I did detail it in the 3rd post. Looks like relays will have to be it for now.

 

[agent420]

I must be higher than usual, because I don't see anything below that gives a clue what you are attempting to do, or what the 'unit' is...

I have one G6HU relay. Now if I had 3 more it would be done

Sorry Nigel, I mentioned only 1/2 the project since I was hoping for a solidstate version and would dup it and drive them with the pic. Relays I will need 4 DPDT.. I can see if RadioShack still sells them.

The project is actually 2 sets of the terminals and they work together.

   T1-1   T1-2    T2-1    T2-2
1   +        -         +         -
2   -        +         +         -
3   +        -         -         +
4   -        +         -         +

 

[Hero999]

I'm confused to, I assume that you just want to swap the terminals round.

You don't need a PIC with your relay, a single 555 timer will be cheaper and more effective as the 200mA output can drive the relay without using a buffer transistor.

 

[mramos1]

I'm confused to, I assume that you just want to swap the terminals round.

You don't need a PIC with your relay, a single 555 timer will be cheaper and more effective as the 200mA output can drive the relay without using a buffer transistor.

I need to change the times on it and not sure on the times yet and some cycles might be longer than others. So 555 is out, I will just use transistors and drive some relays, but was hoping to use just transistors like an H-bridge.

For the table below, T1-1 is terminal 1 pin 1, T1-2 is terminal 1 pin 2, T2-1 is terminal 2 pin one, T2-2 is terminal 2 pin 2. +/- is the battery position I want connected to them.

   T1-1       T1-2        T2-1        T2-2
1   +            -             +             -                 for 5-20 seconds
2   -            +             +             -                 for 5-20 seconds
3   +            -             -             +                 for 5-20 seconds
4   -            +             -             +                 for 5-20 seconds

I am just looking for a good driver to force the + and minus, all the rest works and is done.

I think Nigel said just use a relay.. That might be the only way. I did make a NPN/PNP driver but there was a lot of loss pushing the minus side.

 

[Nigel Goodwin]

I think Nigel said just use a relay.. That might be the only way. I did make a NPN/PNP driver but there was a lot of loss pushing the minus side.

An H-bridge is only trying to imitate a relay - and not doing a very good of it. Relays are cheap, simple, reliable, and perform better than an H-bridge. The only advantage of an H-bridge is if you want to use PWM power control - and even then you can do it single ended, and use a relay for reversing.

 

[agent420]

I need to change the times on it and not sure on the times yet and some cycles might be longer than others. So 555 is out, I will just use transistors and drive some relays, but was hoping to use just transistors like an H-bridge.

For the table below, T1-1 is terminal 1 pin 1, T1-2 is terminal 1 pin 2, T2-1 is terminal 2 pin one, T2-2 is terminal 2 pin 2. +/- is the battery position I want connected to them.

Is there some reason you do not wish to communicate what it is your doing? I get you are trying to switch polarities around, but on what and for what purpose? It's a lite load, but you need as much battery as possible?

Mostly, I'm just curious. But also sometimes other people can provide insight on a project solutions you may have overlooked.

 

[mramos1]

An H-bridge is only trying to imitate a relay - and not doing a very good of it. Relays are cheap, simple, reliable, and perform better than an H-bridge. The only advantage of an H-bridge is if you want to use PWM power control - and even then you can do it single ended, and use a relay for reversing.

Unfortunatly I think it will involved PWM later. It is an electroplating process, for not the relays will do to get it started.

Can a transistor H bridge really get it down to 0volts without tons of extra parts? FET will switch faster, but I do not believe it is required. But the 3904/3906 version did not go to 0 well. Late for work, but will post what I tried later tonight.

Thanks for all the help.

 

[Sceadwian]

I've worked at a metal finishing plant (Zinc / Electrolytic Nickle / Zinc Alloy) for 10 years. What exactly are you trying to do on what scale? It sounds like now that you've said plateing that you're trying to do pulsed electroplating with current reverse, but using batteries? You must be using a tiny cell? Even the Hull Cells we use at work for chemistry testing draw 2-5 amps at 5-10 volts.

 

[mramos1]

Sceadwain:

This one pulls about 300ma. It is Titanium and Zinc with a number of chemicals. It done by hand, the Zinc rods will retore themselves, try to automate it for a friends. I will try relays like Nigel said, then go FET later. I found an EDN circuit that used 2 additional transistos to drive to low millivolts. Just hoping someone might have a simple, low part count circuit.

 

[Sceadwian]

300ma seems extremely low, even for a tiny cell. You must be plateing only a few tiny parts at a time or at a VERY low rate? I'm not even sure if metals will even properly deposit onto a part's low current density areas with that little current. It's a bit off topic but I'm curious about the specifics of the operation, for my own curiosity if nothing else. As I said our Hull Cells draw 1-5 amps and those are just bath analsys cells and they're only 250mls. Please feel free to private message me here on the forums or e-mail me at sceadwian@yahoo.com if you would like.

 

[mramos1]

There is a small amount of surface area, thin rods right now. HE does plan on a larger surface at a later date. I will need to be able to drive a few AMPs then (or now) and thought FETs would do that for sure. But for now. Low power is fine. I will get relays tomorrow and make the board for him.

The power is to reclaim the zinc as the chemical reaction in the cells depletes it.

It is an experiment a chemist friend is working on. And he has has it working by hand right now, but wants to automate it to see how long the zinc will last. Change the times, I will probably add some feedback to the PIC later as well once he tells me what he needs.

He is testing a patent.