Dirty but cheap DIY charger?

Hi guys,

In my spare time, I'm doing a Youtube course on electronics. I'm not taking it too seriously at the moment as it's mainly for interest, rather than educating myself to design or even diagnose faulty equipment. (Though that may change).

With the basics I've learned so far, I think in theory, a high current charger should be very cheap and easy to build. From what I know (and this could be wrong):

1. Get a transformer with the right number of turns to reduce 240vac (or your local supply), to say, 48vac (or a little bit higher to account for losses in the next steps).
2. Insert a full wave rectifier, to turn it into pulsing 48v DC.
3. Put a large electrolytic capacitor in parallel with the rectifier to smooth the pulsing DC into smoother DC power.
4. Put a resistor in series to lower the current to the desired level.
5. Use some diodes in series to lower the voltage to exactly the desired output.

So, my question is, would this work, and if it does, why don't they exist? Surely there is a demand for cheap but low quality power supplies. My thoughts are:
[*]The voltage would not be regulated - but charging a battery, as long as the variation wasn't big, that shouldn't be an issue.
[*]I don't know if the DC would be "dirty" (i.e. still get noise from the AC side, or nearby other electromagnetic fields), but again, as a pure battery charger, shouldn't matter.

Have I misunderstood the lessons? Any input appreciated.

Sunder said:

Hi guys,

In my spare time, I'm doing a Youtube course on electronics. I'm not taking it too seriously at the moment as it's mainly for interest, rather than educating myself to design or even diagnose faulty equipment. (Though that may change).

With the basics I've learned so far, I think in theory, a high current charger should be very cheap and easy to build. From what I know (and this could be wrong):

1. Get a transformer with the right number of turns to reduce 240vac (or your local supply), to say, 48vac (or a little bit higher to account for losses in the next steps).
2. Insert a full wave rectifier, to turn it into pulsing 48v DC.
3. Put a large electrolytic capacitor in parallel with the rectifier to smooth the pulsing DC into smoother DC power.
4. Put a resistor in series to lower the current to the desired level.
5. Use some diodes in series to lower the voltage to exactly the desired output.

So, my question is, would this work, and if it does, why don't they exist? Surely there is a demand for cheap but low quality power supplies. My thoughts are:
[*]The voltage would not be regulated - but charging a battery, as long as the variation wasn't big, that shouldn't be an issue.
[*]I don't know if the DC would be "dirty" (i.e. still get noise from the AC side, or nearby other electromagnetic fields), but again, as a pure battery charger, shouldn't matter.

Have I misunderstood the lessons? Any input appreciated.

Click to expand...

1. yes it would work
2. they do exist

now for the real world application
power supplies used to be made like this but there are a few issues that prevent a product being made and sold like that any longer.
when you design a mains powered device you need to comply with many rules/standards, off the top a few of the relevant ones here would be.
1. power factor
check out https://en.wikipedia.org/wiki/Power_factor the supply you suggest would have a pretty bad PF and would get worse the more capacitance you added to the output. so the better you try make your output waveform the worse your input would be, so the only way to get around this one is with minimal capacitance and a really bad output. Or just use a SMPS with power factor correction
1b. you might find some low power transformers still sold like this as this is only enforced above the 10's of watt range. (sorry cant remember exactly)
2. conducted emissions
the large current flowing into the caps on the output end up making a lot of conducted(and radiated) emissions back down the mains, so you'll end up using common mode chokes, caps,.. etc on the mains side just to clean this up. these parts are generally physically big due to their low frequency operation and hence become expensive.



regards

There were some refurbished on ebay for $75.
By the way, you don't need caps. You are studying power supplies.
http://www.ebay.com/itm/Schumacher-5-10A-Multiple-Battery-Charger-New-Free-Shipping-/182679890614?epid=671821702&hash=item2a8891eab6:g:VboAAOSwGhxZcrxA

I got one and it works great as a manual multi voltage charger. Aluminum TX though.

Thanks guys. Some very knowledgeable people here.

Since dirty unregulated power doesn't bother batteries, I wonder if I could build a "guerrilla" charger (except that I would use it only in approved situations.) You know, when people build chargers basically to charge their bikes as fast as their batteries will take, even though it might shorten their life, because they just want to be at a cafe or service station for 15-30 minutes, grab a very quick top up, and go.

My local chargers do 240v @ 30A (J1772 Level 2). I *think* my bike could soak that, as it's got a 5kwh battery (motorcycle), so that's only just over 1C, but I suspect a 5kw capable transformer is going to be huge...

I think for now it's going to stay in the "one day" basket, and then will probably experiment with much, much smaller models before deciding whether to build a real usable one later.

You might want to look up "bad boy" chargers; there've been various threads and posts on them here on ES and over on DIY Electric Car forums, that might help you out.

Sometimes called "freddy" charger, "poor man" charger, etc.

One example
https://endless-sphere.com/forums/viewtopic.php?f=14&t=4047

another
https://endless-sphere.com/forums/viewtopic.php?f=14&t=36646&hilit=+bad+boy+charg*


The catch with a portable charger that does a lot of power is that if you don't use SMPS tech to do it, the transformers are going to be huge and heavy.

If it doesn't have to be portable that's no problem. But if you need a lot of power *and* portability, it gets more complex and expensive.

It's probably cheaper and certainly easier to use existing PSUs, like the Meanwells (either openframe if you can keep them from vibrating apart while riding, and don't need weather resistance, and can add sufficient forced cooling, or potted like the HLG/etc series if you need bulletproof, weatherproof, and fanless).




Notes:

--resistor in series to lower current only limits to a maximum, it doesn't regulate the current at all, and it has to be able to dissipate the wattage at the lowest voltage the pack would be at, at the highest current the resistor would see. Let's say you had a big pack that has 20v difference between full and empty, and you want to charge it at a max of 20A, that's 20x20=400W of power the resistor has to dissipate continously. That's a really big resistor making enough heat to cook your dinner on; you'll probably have to have quite a lot of forced air cooling on the resistor and it's large heatsink just so it's heat doesn't damage stuff around it.


Diodes to drop the voltage: they aren't constant voltage drop; they generally drop more voltage at higher currents. So you'd need to make sure they are dropping enough voltage at the lower currents near full charge.


There are plenty of existing old power supplies designed this way, but they're big, heavy, very inefficient, hot, and noisy (or even bigger due to heatsink size if they dont' have fans).


I have a 5v 100A adjustable voltage PSU that works basically like that (no diodes though). that's only 500w. But it's as big as at least two FLA car batteries, and weighs as much as one or two of those. :/

You can see it in these videos:
https://www.youtube.com/watch?v=kh42DL20l2E
https://www.youtube.com/watch?v=UjtGYUt_XgQ

Thanks for that Amberwolf.

It's probably best I didn't play around with this kind of stuff then. I just looked at the price of the transformers I would need. It's not that much of a savings, and would be very heavy indeed.

Don't forget that diodes have a voltage drop that varies with current (not fixed).

There are many problems (and risks) with these simple approaches. Line voltage is not constant, for one. One day it works, another day batteries burn.

During the early charging phase with very low batteries the current may exceed safe levels.

A lot of heat is produced by the simple regulation.

Not easy to adjust either voltage or current.

Best to study but not build.

The amount of material that goes into an old-school line frequency transformer is about 10x what it takes for a modern switching transformer of the same capacity. They are heavy beasts. You could probably rewind a microwave oven transformer to make a crude charger. At least the transformer would be cheap.

Simple transformers do have some advantages, like they're simple and don't have too many things that could fail.

These days you can find surplus server power supplies really cheap and hack them into a pretty nice charger.

How do you adjust voltages on server power supplies? What kind of technology do they use? Presuming not transformers, as they are very small and light.

SMPS (switched mode power supply), which includes transformers, but at much higher frequencies so they are much smaller.

https://en.wikipedia.org/wiki/Switched-mode_power_supply
https://www.google.com/search?q=SMPS+%28switch+mode+power+supply%29&num=100&newwindow=1&source=lnms&tbm=isch&sa=X&ved=0ahUKEwista-qieDVAhUVwWMKHfn9B4sQ_AUICigB&biw=1188&bih=918


Server PSUs are not always adjustable, but lab PSUs are, and some others like the Meanwell HLG LED PSUs (with A suffix) and someof the other Meanwell open-frame types are.

But one fo the important things about any SMPS PSU you have to be sure of before using it as a charger is if it's current-limiting is "hiccup" mode (where it shuts down completely at any overcurrent, then attempts to restart, and shutdown again immediately because the load is sitll there, etc.), or if it is a voltage-drop type, like a charger or adjustable-current lab-PSU.