Universal Battery Charger, Home Made

[jack11]

I've been trying to make a simple back up charger for when the sun isn't shining and I have to crank up the generator, for 12, 24, and 48 V batteries, and anything in between.

I am sure people have done this before, but I can't find any info here.
I am using a 0-120 Vac, 10 Aac variable autotransformer/Variac, a power diode and/or a diode bridge, and a large capacitor.
I plug the Variac into the generator, connect the rectifier circuit to the Variac's output, and set the Variac's dial to the voltage that produces the charging current I need.
This Variac has a 10A fuse in its output circuit, so it limits the current to 10A regardless of the voltage setting.

It works fine using a half-wave rectifier, but I am not sure the generator likes having to deliver power on half-cycles only, may be bad for the generator long term?
Plus, I am not taking advantage of half the energy in a given amount of time, I could be putting into the batteries.

When I hook up a full-wave diode bridge to the Variac, I run into trouble. I have two bridges, one from a DC welder rated 40A 800V, the other one smaller from some electronics gear. Both bridges act the same.
With the bridge's AC terminals hooked up to the Variac's output, and the bridge's DC terminals hooked up to a scope probe or a DVM only (no load – no current flowing, except very small through the probe/DVM), the Variac's winding starts buzzing when I raise the voltage from zero to about 10 Vac-peak.
When I push it to about 15-20 Vac-peak, the buzzing gets louder and the Variac blows a 10A/250V fuze.
During this time I see a full-wave rectified voltage waveform rising on the scope, then it goes to nothing when the fuse blows.

Can't figure out why it's doing this.
Don't think it's the bridges, I've checked the diodes in forward and reverse directions, seem ok, no shorts or opens.
The bridge ratings are adequate for the charger I need.
There is no load on the bridge, so no current should be flowing, but still the Variac behaves like it was overloaded when it blows that fuse.

Has anyone messed with these universal chargers using a Variac, and run into a similar problem?
How about using a half-wave rectifier with a generator, wouldn't it damage or unbalance something inside the generator when used over an extended time period?

[OperaHouse]

Your scope is likely grounded to neutral so the bridge is shorted.

[Mary B]

Need to use an isolation transformer before the variac. Unless it has one built in, a few do.

[Flux]

I think the others are right but a bit more explanation may help.

A variac is an outo transformer, there is no isolation between input and output. I assume that your scope is earthed  and this is where the trouble is coming from.

With the half wave rectifier the variac neutral and scope earth will be the same point so it works.

When you use the full wave bridge and transfer the earthed scope to the bridge negative you are shorting two diodes of the bridge.

The capacitor is not necessary for battery charging.

Ideally you need an isolating transformer before or after the variac but if the battery is isolated and you keep the scope out of the way you can use the variac direct, this is not particulary safe but if you take care it can work. Just put an ammeter in the battery circuit to see the charging current and forget the scope it tells you nothing.

Flux

[OperaHouse]

The variac hums because it doesn't know the words.  Technically only one diode is shorted.

[jack11]

thank you all, I have some time now and wanted to catch up with some of my projects, the GTI I posted about, this battery charger, etc. Hopefully others can benefit when they deal with similar issues.

You are absolutely right, I don't know why this didn't dawn on me. After I drew a schematic of what I was doing, it shows one of the bridge's AC terminals shorted to ground through the variac and the neutral (white wire grounded at the house), and the DC(-) terminal also grounded through the scope probe shield (I haven't checked yet, but I assume this shield is connected to neutral/ground at the house). So, the diode between DC(-) and AC (neutral) was shorted. I'll check it out tonight on the actual hardware.
And, my variac does not have an isolation transformer built in.

BTW, that large capacitor very nicely smooths out the ripple resulting from half-wave rectification (makes almost constant DC out of the half-wave ripple), and I assume it'll do the same for the full-wave bridge.

Out of curiosity, what about this more general question, would a generator be affected long-term in any way if the rectifier was half-wave, and the energy drawn out of the generator was only happening on half-cycles?
Being a piece of rotating and commutating machinery, it seems like there would be some negative long-term effects (unbalanced operation, etc).
I am talking about an average cheap generator, mine is a 3.5kVA Champion.

[Flux]

The capacitor makes the no load waveform look good on the scope but the ripple will come back when you load it. The battery will only charge when the peak of the ac voltage is above battery volts, in effect the battery behaves as a very large capacitor and adding a few uF in parallel will do nothing.

Yes the capacitor will also smooth the waveform on open circuit with the full wave rectifier ( it in fact works better).

You are right that the generator will object to significant half wave rectification, in your case the charging current probably is not high enough to cause it much problem but loads that are unidirectional will tend to cause the core to saturate.

In fact although full wave rectification will solve this you will still find that a small generator will perrform very badly with a simple full wave battery charger. As the current only flows at the peaks of the waveform you will be very limited in the current you can take, as the reactance of the generator will reduce these peak currents. With a generator as small as 3.5kVA I doubt that you will get much more than 500W into your battery with a simple charger.

The only round this is with a very large generator with low reactance or with a clever battery charger with power factor control that takes the current over virtually all the cycle.

Flux

[jack11]

Thanks Flux, great info.

I've already seen some of the ripple come back when I tested the charger with the half-wave rectifier under load.

I'll be watching the efficiency of this whole setup next time I start the generator. This charger is for backup only, so the efficiency is not a big deal (the place where we do not have grid access, and where I use the generator, is in the southern mountains in NM, and the sun is almost always shining there).

[OperaHouse]

You should remember that the sine wave is the po boys PWM when it comes to charging batteries.  This allows longer charge currents when the battery is 12.5V and allows the current to drop down when it reaches 14.5V.  A very large capacitor would defeat that.

I remember a long time ago our R&D newsletter has a RAPID 50A battery charger that was only a 6 inch cube.  That was impressive enough for around 1980 but it had an interesting feature that gave a high discharge every few cycles that was supposed to eliminate gas buildup on the plates that compromised very rapid charging.  Can't say that I have heard much of that technology today.

[boB]

thank you all, I have some time now and wanted to catch up with some of my projects, the GTI I posted about, this battery charger, etc.

In the GTI posting, I thought you said you did not want to use batteries or maybe a battery based inverter.

Have you changed your mind ?  If so, you can use a battery based inverter as a charger or AC coupled from your GTI and not have to worry about this diode connected to your generator AC source.

[jack11]

This full-wave charger is working well now that I have removed the ground from one of the bridge's points DC(-).
Pretty easy to set the charging current with the Variac, and minding the battery voltage at the same time, of course doing it all manually.
And, universal, should be able to charge any battery bank up to about 72V, maybe higher.
The bridge I use it a commodity frequently used in AC/DC welders, 40A at 800V, seems to be working well in this application, and cheap around $10, in case someone wanted to build a similar charger.

Good point OH, I haven't thought of this self-regulation feature when I put the capacitor in (I think it's around 250uF and it came from the same Miller welder as the bridge, and I've simply replicated the welder's circuit for the charger - I junked that welder some time ago).
Now that I know, I'll be taking some measurements to make sure I am not defeating this self-regulation with the cap.

Bob, we have two places where we use solar. One is in TX where we have grid access, and the other one in NM where we don't.
The charger is for NM, where I need to crank up the generator sometimes, if the sun is not adequate (or just to warm the generator up occasionally). I have a battery bank in NM.
The GTI is for TX, where I just want to use the grid for the reference signal to keep the GTI going, but get my usable power from the GTI and not the grid. I do not have batteries in TX, and don't plan to put in any for this GTI system.
One exception, I have here another separate battery-based solar system that's entirely stand-alone and feeds some of the house, and I don't want to mix it with the GTI or anything else, for some valid reasons.
Also, I don't want to feed the grid in TX because the return is lousy in comparison with the headaches.

So, these two projects are for two different places, and two different sets of conditions.