My new pulser/pulse-charger.

[mael]

[SeaMonkey]

[mael]

[SeaMonkey]

Mael,

Private Messaging is disabled but you may email me by clicking the [EMAIL] icon at the bottom of my postings.

Please send email to me.

[mael]

[SeaMonkey]

[mael]

Living on this island a thousand kilometres away from the nearest shop which has parts on the shelf has presented some problems.

Relying on discarded components has been interesting, but for a few reasons it is becoming tiresome. The most pressing is that our local garbage collectors are becoming difficult (for want of a better word) about picking up anything other than metal drink cans in the rubbish bags. - Taking the bags to distant locations and leaving them in different villages makes it seem more trouble than it's worth.

A week ago I made the brave step of trying to order parts in Japanese, and surprisingly I was successful. It was only non-resettable thermofuses but it's a start.

Actually the 40 A MOSFET was one of the ones which didn't work from the git-go. That was disappointing. So wearily I tried four or five others just to see if one of them would do something resembling 'work,' and one of them didn't die. His friends of course are deceased - martyrs for the cause y'know... . The one that works is a freak. It's a normal bi-polar power transistor. My driver circuit fortunately has a lot of spare oomph, and it drives the transistor very nicely. I reckon the poor bugger is shifting going on 30 Amps if not a shade more. I am afraid to look at the datasheet on that one again as I'm sure it will worry me.

It's actually rigged-up to a sulphated battery and ticking away.

I noticed the coils on my 46V 1A transformer are too hot to hold, and that's with a fan on it. The iron part of the transformer is much cooler.

The freak transistor is on a large heatsink with a fan, and gets hand-hot. But somehow it is holding together.

As I'm putting at least 4 A across the relay contacts I'm getting some arcing, and so I can safely assume it isn't going to last very long.

So it is working through pure applied pig-headed stubbornness.

I think this pulser is going to be my 'Edsel' model.

Ho hum! So I'm going to order some semiconductors from a shop! Brand-new!!! I'm working up to it.

... But which one?

I would appreciate your recommendation for a MOSFET which will do the job for me.

The specifications I would prefer are:-

1. (Let's say) 100 A capability on a pulse, but not necessarily a short one. Perhaps with all the variables with my systems I'd expect it to function on at least a 50% duty cycle at 100 A.

2. 100 V would be fine.

3. 'tolerant' GATE requirements.

4. And finally it must be tough. Y'know what I mean... .... .

* I'll send an email in a day or so.

ZZZzzzZZZzzz

[mael]

Of course it is too soon for me to be able to provide reliable data on the efficacy of the high(er) power desulphator which is now limping-on with a hot(tish) transistor.

Last night (actually 08:30 this morning) I had rigged it to a 19 A/h battery and when I was reasonably sure it was OK, I left it to be pulsed. (So I could get some sleep).

The next time I tested it it was done. I finally did a SG test after it had been pulsing for about twelve hours. So I was unable to watch the progress of the desulphating because the sulphates appeared to be gone.

That battery must have been a wierd one. From the start it was overfilled. All of the cells wwere over maximum and three of them were approaching the top of the battery case.

Usually when a battery is overfilled like that it will have a low SG reading. But All the cells were at least in the green (over 1250) and the ones which were only overfilled were well into the green. Strange.

I would guess the pulse rate was in the region of 2 pp/s at a current of over 20 A. I've got around 16,000 mFd caps charged to about 42 V being pulsed in 100millisecond pulses.

Unfortunately I have had to rebuild some parts due to component failure (murder), but thanks to having pretty-coloured wires with croc clips on each end I have it running.

I just wanted to provide my preliminary result from my experiment with pulsing at a high amperage in a fat pulse (a thunk is probably not the scientific term for it).

The transformer I'm using is rated at around 50 Watts, and it is certainly delivering well over that. It remains viable due to a (computer) fan blowing on it. The transistor I'm using is mounted on a 'U' shaped heatsink with the flat side up, and the fan is mounted underneath it with a hole the diameter of the blades cut into the wooden base. The air hits the top of the heatsink - under the transistor and is then blown out of each side. On one side is the main transformer which charges the caps, and the other side is a smaller transformer which I am using to power the gate/base driver.

I'm in a sort-of hurry to find some semiconductors I can use without worrying. I'm using a transistor due to it being the only thing I can find which will survive the amperage, and I'd much rather use a MOSFET.

The base of the transistor is drawing up to 4 A at max, and this will take its toll on the (somewhat large and noisy) relay I have settled on. - A small relay rated at 3 A was toast in half an hour. - I mean the contacts welded themselves shut. I am wondering whether they'll last much longer either in a relative vacuum or running in a mineral oil bath? I Might try this potentially awkward and messy experiment at some time.

* These days I am fairly selective with choosing the batteries I deem suitable fo an attempt at desulphating. I'll reject the ones with seriously corroded plates. I think those with disintegrated plates may work well. But I doubt they'd last long, and I would imagine the capacity of a battery with crumbling plates would be appreciably less than one in which the plates appeared sound.

Slow & high-A pulsing???? Hmmm! So far so good.

[AlaskaStar]

[mael]

[SeaMonkey]

It's good that your transistorized desulfator is pulsing along nicely and getting the job done! In truth, Power Transistors are not a bad choice at all for the kind of pulser/desulfator that you're using. The transistor has the ability to turn 'on' very rapidly, thus producing a sharp leading edge on the current pulse; but it turns 'off' much more slowly which makes it less than ideal for driving inductive loads and transformers. For use as a Current Switched Pulse Charger/Desulfator their characteristics are just about ideal.

Most power transistors have a relatively low current gain (beta) so the base drive current is often measured in Amperes. This means that a Power Transistor with a beta of 15 would require a Base current of 2 Amperes in order to switch a Collector current of 30 Amperes. Anything beyond the necessary 2 Amperes of Base Current would be essentially 'wasted.'

When transistors are connected as 'Darlington Pairs' the total current gain of the 'pair' is equal to the product of the individual transistors current gains. Say a small power transistor with a Collector current rating of 8 Amperes, and a 'beta' of 30 is used in conjunction with a larger power transistor which has a Collector current rating of 20 Amperes, and a beta of 15. Together the 'pair' will operate as a 'single' transistor with a Collector current rating of 20 Amperes and a current gain of (30 X 15 = 450.) So the Base current requirement for 20 Amperes would be approximately 44 milliAmperes. To drive it into 'hard saturation' for a large current pulse of about 80 amperes the Base drive current pulse might be nearer 80 milliAmperes.

Since Power Transistors run 'hotter' than MosFets, they always need a good 'heat sink' to prevent damage. But apart from that, they're perfectly suitable.

The advantage of Power MosFets is their very low Rds (on) and consequent higher current ratings and cooler operation.

To find a MosFet type no. that corresponds to your application needs, use one of the following Selection Guides: or

As a general rule with desulfators/pulse chargers, the higher the voltage source the narrower the pulse. Pulsing at 60 Volts could easily result in a narrow pulse magnitude of 100+ Amperes. In order to avoid damage to the battery being rejuvenated it would be necessary to limit the pulse width (time) to a few (not more than 20) microSeconds. This kind of timing can only be achieved with a semiconductor circuit (multivibrator) which is quite easily constructed.

It would be best to drive a Power MosFet to do the high current pulse switching with a MosFet Driver Chip which itself can be driven by a simple CMOS Multivibrator which would have adjustable Pulse Width and Pulse Frequency. Then you'd be able to 'tailor' your pulses to virtually any size battery from small to very large.

You're presently doing amazingly well in your rejuvenation projects with the circuits you've designed so far! With just a little more work you can put together an 'all purpose' desulfator to handle any job!

Yes indeed, in no time at all you could have a very lucrative 'rejuvenation' business with a host of very satisfied customers!

[SeaMonkey]

By the way, the IGBTs that AlaskaStar mentioned are a very good alternative as well. (Insulated Gate Bipolar Transistors) They come in a variety of voltage and current ratings and are very effective. Their 'input' is a MosFet and their 'output' is a BiPolar Power Transistor so their drive requirements are the same as a Power MosFet.


To make a really good transformer for battery charging and high current desulfators, get a transformer from a discarded Microwave Oven. Invariably in those the Secondary Winding is the one that fails and by carefully removing it, you can replace it with a heavy wire low voltage secondary which can easily handle 40 Amperes continuously at about 15 Volts AC. When rectified this would peak at about 21 Volts DC. The Secondary winding turns can be increased to produce a higher output voltage (50 Volts or so) but the output current capability would decrease to about 10 Amperes continuous.

With the kind of capacitor bank that you envision, this could easily translate into a pulse current of 100+ Amperes.

The Primary winding of the Microwave Oven Transformers are very rugged and can handle a power input of up to 1500 Watts.

Happy hunting!

[mael]

I am not unfamiliar with Darlington pairs. It's just I am not very experienced in making them.

I have used a simple flasher circuit using two transistors, a resistor and a capacitor as the timing for my pulses through a relay. But I'm tired of this circuit and its limitations. So I'll build another astable multivibrator and take the positive side of one state and use it to drive a power semiconductor through a Darlington pair.

I'm just a bit worried I'll blow too many parts before I get it right, but I am becoming more confident.

... So I'd put a variable resistor on the base of the first leg of the Darlington to determine the output to the power transistor/MOSFET? Seems this would be the way to go.

... And I've got at least three 555 timer chips knocking about the place somewhere - at least two are OK. It isn't a daunting task for me to wire those up - I mean, the pulsing circuits only require a half-dozen of components and they are all as common as weeds.

Yes! Time to move up!

I have another large and high speed pulser running through a rattly relay driven by an astable multivibrator. It seems it will be needing some attention soon, so I'll try to get that working without relays shortly after I've modified the pulser I've just got running.

... Then there's a mountain of stuff I have been neglecting since last week when I began making that pulser.

[mael]

Oh! I forgot to add:-

Can I use a PNP on the output of an NPN astable multivibrator to chop the pulse?

Would I couple the base of the PNP via a cap to the collector of the chosen NPN?

[mael]

The nesest pulser is now back on the bench.

I removed the 'flasher' pulsing circuit and relay. I wanted to make it solid-state.

I made a 'relaxation oscillator,' and have been trying to use it to drive the power transistor.

I got a 'saw-tooth' oscillator running, and I can hear it through a piezo speaker I use as a signal locator. But I can't seem to get it right.

It is driving the output transistor, but it doesn't have enough power to deliver enough current.

I can only tell it's oscillating because it whines or buzzes. But I have no idea of the actual wave-form, and cannot tell if it has a sharp enough leading-edge.

makes me pine for relays where at least I can see what is going on and know for sure the pulses getting to the output stage are full-on and full-off.

Hmmm! Shall I do this pulser suing methjods I am familiar with and postpone the solid-state version to the next time? It is tempting.

... And I'm whining as well! I think I feel better now.