(Lead-Acid) Battery desulfator

[kathmandu]

Hello,

I need to recover two of my lead-acid batteries (from a large battery bank). They seem to be affected by sulfation (some spongy white deposits on top of the plates).

Usually, this "desease" is caused by insufficient charge or deep discharge but it wasn't the case anyway (the rest of the batteries are OK).

Nevermind, I've read about some desulfator circuits that might help. It's all about applying some short high current pulses during a long period (24-48h) to break down those sulfate crystals.

The problem is I didn't find any informations about those pulse patterns; I've read about using fixed or variable frequency (in 10khz - 100khz range), 50% pwm ratio or much smaller and so on.

Does anyone have any experience with that? The batteries are just 1.5 years old and I don't want to loose them (though I can't explain what really happened).

The two batteries were connected in series (I have a 24V battery bank) and I've noticed few days ago that they were bubbling hard (even at very low charge currents).

Btw, they are rated at 225A and the maximum charge current is 15A.

 

[BradtheRad]

A chief cause of sulfation is sitting partially discharged for an extended time. The battery chemistry likes to be 'exercised'.

The topic of desulfators has a few threads here. You can find them via forum search on:
sulfate
sulfation
desulfate
desulfator
desulfation
etc.

I read about a commercial unit which delivers pulses to the battery as you describe. Once it is hooked up, it discovers a specific frequency which is most effective. Perhaps it detects a resonant frequency based on physical characteristics of the plates. Perhaps the secret is to create an ultrasonic cleaner action in the battery, so that fresh acid continuously scrubs the sulfated surfaces.

 

[kathmandu]

Your explanations sound reasonable; maybe that's why some "cheap" desulfators use variable frequency - to be sure they're passing through resonant frequency, whatever it is.

Have you read any "official" papers on this issue? I've searched the board (and googled a lot) with no success. I've seen some schematics but they seem "jack of all trades, master of none" (too generic and too empirical to attempt for any results).

Btw, the ultrasonic cleaner might be an option, but I think it's better to get rid of those crystals by chemical means (reversed reaction). If one just clean the plates and those sulfate deposits go to the bottom, I'm afraid the whole thing is getting worse.

PS: I've read about the famous sulfate salt, too. Well, at this point, I would not take the risk. Not before reading some scientific facts.

 

[Warpspeed]

There seems to be a lot of mystery about how these desulphators work, plenty of theories out there though.....

I doubt if its a resonance, but what is needed is a very fast steeply rising pulse edge, with sufficient voltage and current. My own theory is that the sulphate crystals make excellent electrical insulators, but when subjected to a very fast rising pulse are subject to high dielectric stress which very slowly breaks these insulating crystals down.

The usual way is to use the flyback energy from an inductor. The internet is full of circuits that do this, but most are flea power devices that can take weeks or several months to make any real impression on a badly sulphated large battery.

High power desulphators that also charge the battery at the same time are a much better idea, but if it is too powerful it could potentially damage small sealed gel cells, so you need to use some caution with how much pulse power is applied.

As you say "large battery bank" an externally powered high power desulphator might be your best bet.

I also have some information here on chemically cleaning away the sulphate, but it is a fairly radical process. Often single cell failure is due to collapsed plates or shedding of active material, or other damage which a desulphator obviously cannot fix.

 

[kathmandu]

I'm just reading an interesting thread on another forum and I came to understand why the battery case is pitch black.

I wish it was transparent so I could play with those desulfators.. but the manufacturers have been chosen this "color" on purpose. There is a secret hidden that could ruin their business..

If only I could feel those crystals (or remove them by hand, if that hole were a little bit larger).

Anyway.. no one talks about that sulfate salt. What if it reacts like the table salt on ice?! What about dropping some magic powder and watch those crystals disappear?

Once again, I guess these secrets are hidden on purpose.. who need an everlasting battery? Not the manufacturers, not the retailers, not the car services, not the "desulfator" sellers(!).. who's left, just few of us?!

 

[kathmandu]

I'm thinking of the following circuit (trying to mix both worlds: variable frequency and high current pulses):

I'm going to use a full-wave rectified voltage (100Hz, 25V peak or higher) chopped by a variable frequency in 10kHz - 100kHz range.

This way, I could generate various pulses with pseudo-random period and current intensity (to not let the crystals accomodate with the new "threat").

Moreover, I could generate a complementary discharging pulse (by chopping a load across the battery). This way, I could force those crystals to bounce back and forth "till death break them apart".

Do you think it's a feasible solution?

 

[Adri67]

My experience is that sulphation occurs in stages.
If you catch it early (i.e. normal use and never "fully charged" such as the classic 13.8V auto alternator-regulated system and most PV- or alternative charging systems, because "13.80V" does NOT mean fully charged), you can easily over-charge the battery comfortably up to 15.0V or even higher once a month, provided that
(a) you monitor the electrolyte level and keep it topped up
(b) you control the charge current to prevent the battery over-heating (say, more than 30 degC above ambient).
The problem with desulphation of the severe (late) stage is that the plates are permanently damaged and then no chemicals or fancy charging methods will fix the destruction.
Worse still is that the junk accumulates at the bottom of the cells and causes shorts between the plates.
The worst of it all is that when one of the 6 cells (of 12V battery) expires due to sulphation (which happens 99% of the time) the whole battery is lost, and your battery salesman is laughing his head off! (I.e. we cannot simply replace one cell.)

 

[kathmandu]

The worst of it all is that when one of the 6 cells (of 12V battery) expires due to sulphation (which happens 99% of the time) the whole battery is lost, and your battery salesman is laughing his head off! (I.e. we cannot simply replace one cell.)

Actually, it happens to me almost one year ago. One of my 6 months old battery has (suddenly) lost one cell; the battery was in series with another one (identical, bought at the same time) which had absolutely no problem.

I went to the shop and after an one-hour test, the rest of the cells were bubbling but in the dead one the electrolyte was absolute still. Btw, the electrolyte was perfectly clear in all the cells (no sulphation whatsoever).

I left the shop with a brand-new (identical) battery. But yeah, it supposed to be one lucky day. I guess it had a manufacturing problem (broken/shortcircuited plates or such).

My experience is that sulphation occurs in stages.
If you catch it early (i.e. normal use and never "fully charged" such as the classic 13.8V auto alternator-regulated system and most PV- or alternative charging systems, because "13.80V" does NOT mean fully charged), you can easily over-charge the battery comfortably up to 15.0V or even higher once a month, provided that
(a) you monitor the electrolyte level and keep it topped up
(b) you control the charge current to prevent the battery over-heating (say, more than 30 degC above ambient).

I guess I'm in this first stage, but I wonder why the rest of the batteries (six of them) have no trace of sulphation. Only these two batteries have an identical aspect (trace of sulphation on every cell). They were also bubbling hard and get moderated warm.

 

[dick_freebird]

The sulfation web pages I have seen all look like very
unscientific mumbo-jumbo.

I've been thinking about the possibility of making my own
clear plastic per-cell cases and pulling cells out of the
batteries I have hoarded, so they can be cleaned and
reconditioned individually. But that's liable to be messy
and one of those projects I never quite get around to.
Pretty likely that no optically clear material (other than
perhaps glass) will stay that way over time and acid.
Although I do have batteries in both black and white
poly cases, the white is only translucent.

I suppose with enough care you might be able to remove
the top and leave the bottom wells intact, and only have
to recreate (or reattach) the top cover. If you don't have
to move the battery or worry about spillage then maybe
the (re-)seal doesn't have to be perfect.

 

[Warpspeed]

Part of the problem with these black multi cell lead acid batteries is that its usually not easy to monitor individual cell voltages.
With more modern cell chemistries, its much more common to buy individual cells and hook them up to some pretty sophisticated charging and monitoring circuitry.

If you know that one particular cell is diverging from the others, it is much more likely that something effective can be done about it at an early stage.
If only one cell is visibly sulphating, that is probably not the real underlying cause of the problem.

Some batteries respond to mild desulphation quite well, but others fail to show any improvement, so its not a universal cure.

Don't laugh, but I have fixed some battery problems by dropping the battery a few inches onto solid concrete. The sudden shock can sometimes dislodge an internal short.
If the battery is unusable, it probably cannot be harmed any further with some pretty rough treatment.

Other solutions such as draining, flushing, and chemical cleaning can be a waste of time too if the plates have shed their active material, are warped, or the separators are disintegrating.

 

[kathmandu]

My batteries have a strange behaviour. They were left open circuit, and after 12 hours, I've heard them still bubbling. Now, after 24h from the disconnection, they finally felt asleep. During this period, the open circuit voltage was steady (around 12.5V).

I guess I'll turn their sleep into a nightmare. I'll take a shot (in the dark?) with a.. umm.. "high current pulse circuit" (I won't call it "desulfator" for now) to see what's happening. And I'll further choose the hard way, by alternating strong high frequency charging pulse with strong discharging one.

I'll eventyally open the refilling caps so I can watch de "desulfating" process (though I might getting old in the mean time).

Wish them luck!

@dick_freebird: you're right, all discussions about desulfators seem to be fun or flame.

@warpspeed: if electronics fail, I'll try a mechanical (hammering) resurection, too. For now, I'll try to stick a wooden rod inside one cell, to check for the crystals consistency. It's like getting to know your enemy, you know..

 

[Warpspeed]

Here is the business end of my externally powered desulphator:

The two mosfets turn on simultaneously ramping up the current through the inductor.
When both mosfets turn off simultaneously, the flyback energy flows through D3 and D1 into the battery.
Initial rise time of flyback is very fast, a few tens of nanoseconds, and the peak discharge current will be whatever the original ramp up current was set to.
Diode D2 provides an alternative inductor discharge path back to the main dc supply, if the battery impedance is very high, or when there is no battery connected.

The repetition rate is around 1Khz, and it causes a current mode control chip to set the desired peak current (up to 30 amps peak max)
Average current into the battery is quite low up to around perhaps one amp average maximum, because of the very short duty cycle.

With a pristine battery, the flyback voltage right at the battery may only be a volt or two above nominal battery voltage. with a heavily sulphated battery it might go right up to the full twenty volts for a while, and then gradually decrease. Monitoring this voltage with an oscilloscope is a good indication of battery internal resistance and the desulpating process.

 

[kathmandu]

Your schematic is pure genius. I've just built a similar circuit (two-transistor flyback converter) but using D1 to discharge the inductor is the key. Awesome!

What do you think about getting rude and firing up a complementary discharging pulse? Could it help to push back an forth those crystals?

 

[Warpspeed]

I don't think you really need a discharge pulse.
This thing hits pretty hard, but only for a very brief instant.
The total charging energy is very small.

The above circuit is only part of what I built.
The rest of it measures the battery voltage with a sample and hold circuit just before the next pulse is due.
The sample and hold drives a commercial panel voltmeter that has programmable alarm settings. Its arranged so that it will keep desulphating until the battery voltage reaches 14.2 volts, where it stops desulphating.

It starts up again when the battery voltage falls below 12.6v.
So it cycles on and off, and you can just leave it switched on for very long periods without fear of overcharging the battery.

 

[kathmandu]

I thought about alternating discharging pulses as to push those crystals from both sides.

Nevermind, I'm going to build the base version of your circuit (I don't need timers for now as I'll keep an eye on it anyway).

Great job, thanks for sharing!

 

[Warpspeed]

Its pretty basic, but here is what is on the printed circuit board.

 

[kathmandu]

Thanks for sharing the full schematic. I just can't wait to test it so I'll get rid of most of the parts (PWM driver, optocoupler, CT and so on). I'll use a simple asymmetric pulse width oscillator (555) and the power stage (Mosfets + driver).

Have you got the best results with the 1khz chopping frequency?

 

[Warpspeed]

I never tried any different frequencies.
From memory the inductor took a bit less than 100uS to reach 30 amps.
Then *ZAP* into the battery.
The current run down phase varies depending on battery impedance, but is always less than about 500uS in practice.

Your operating frequency really depends on the inductance, dc input voltage and battery impedance. Its not difficult to fiddle the numbers once you get it all up and running. Just make sure your inductor does not saturate and it should all be sweet.

 

[kathmandu]

The current run down phase varies depending on battery impedance, but is always less than about 500uS in practice.

So that's where the 1khz comes from (1000 us pulse period). OK, thank you once again for your kind support, I'll post the results asap.

 

[Warpspeed]

The original idea was to keep the inductor charge/discharge period within half the repetition rate.
That keeps the second half clear for the sample and hold, and battery voltage monitoring.

I doubt if the repetition rate is that important as far as desulpating is concerned.
Its just an overall design parameter for the flyback supply.

Flat out, the charging ramp is a bit less than 100uS to 30 amps.
The off time is then at least 900uS.
The average current into the battery will be less than 1.5 amps, actually it is around 1 amp (measured) flat out.

Its not a humongous amount of charging power for a really large battery.
On switched ranges less than 30 amps, the average charging power is even less.