Couple years back i picked up about 20 dead dewalt cordless tool battery packs from a guy at the flea market. I took them apart and zapped all the 1.3v cells individually with 48V from a welder. Tcmtech says 1 i. 10 can be saved, my results were a little worse.

I'm glad i didn't build a desulfator, would have been a waste of time. Maybe lead acid takes a little better to it, i don't know.

 

Lead acid are at best 1 in 10 that will be recoverable Nicad and NIMH are as you found far less than that.

I too now have a literal 5 gallon buckets worth, 16 or so now, of the Craftsman 19.2 volt nicd packs that are junk and so far I have found zero ways of getting any of them to work long enough to be of any use under a load.

 

I can understand why folks have problems with the OTC and DIY pulser circuits that abound.
Many of them claim amazing results with no respect to conservation of energy or battery electro chemistry. What's worse is all the misinformation that goes with it, such as magic elixirs to add to the battery such as battery vitamins.

Plus there's a statistical element in play. On occasion I have recovered 3 batteries in a row, and i have also seen 0 out of 10. Thus, unless you have the numbers of batteries to process you may end up with the short end of the stick.

 

A guy I work with sometimes claims to bring NiCds back on a regular basis, but I too have batted an ohfer. I'm getting to where I just solder in replacements and dispose of the old ones.

I am also wondering about the outrageous claims of those DIY circuits I have seen. Some makes it look like I can just put a transformer in series with a charger. If I haven't asked for it before, I'm asking for it now. I was hoping to see some of the science behind what these desulfator circuits are actually doing, and how to improve my odds of finding and correcting the actual problem after reading cell voltages and checking water levels. I bet 8 of the 10 lead acids just got low on water.

 

OK, simple lead acid battery knowledge says when a battery is discharged, sulfates form on the plates. When a LA battery is charged, those sulfates reform back into sulfuric acid. If a battery is left discharged for a long period of time, the sulfates on the plates will harden and normal charging current will no longer cause them to reform back into sulfuric acid. More robust batteries can survive a process called equalization to reduce sulfation. This is when the sulfated battery is severely over charged in an attempt to desulfate the plates. Equalization creates high heat and is very rough on a battery. Sometimes equalization works, sometimes it doesn't, and sometimes it destroys the battery completely. Less robust batteries normally can not survive an equalization charge at all. Now, everything I have said so far is text book, lead acid, care and maintenance.

Enter the pulse desulfator
A pulse desulfator does the same function as equalization, but is spread over a longer time period, and thus produces lower heating and is much less stressful. Also pulse desulfating can incorporate much higher currents than traditional equalization. Because pulse desulfators do not generate the heat that traditional equalization generates, pulse desulfators can be used on most any battery.

My first experience in pulse desulfating was with a motorcycle battery. It would have never survived a equalization, but pulsing worked incredibly well.

 

Lectra:
I have not done any NiCad R&D. I have done perhaps two thousand hours of lead acid R&D science since 2012.
Statistically the ratio of recoverable junker lead acid batteries is around 1 in 3
Recoverable is defined as returnable to general service either as a storage or SLI battery with > 40% Ah capacity.
I am in tropical territory so cold weather over capacity requirement is of no consequence.
I mostly work with Japanese 35 to 70 Ah batteries that have failed in automotive service.
Sometimes a recovered battery is as good as new; probably got sulfated due to poor charging procedures.
I have photographic evidence of desulfation, thermal imaging of dendrite removal and data logs of all battery recoveries tracking voltage, temperature, avg current, pulse current, spec. gravity etc.

I have the science data, I am almost finished with the final build of a practical rejuvenator that incorporates my empirical research. It is merrily humming along as I type this , processing a Jap. 55B24, 42/45 AH SLI battery. It averages around 15- 20 battery recoveries per month (35-70Ah) out of a 50-60 junker batt. feedstock.

The key to battery recovery lies in the thermal management at a cell level. Pulsers that don't do this are snake oil. i will point out here though, that people who manually track the battery temps & spec. gravity CAN get results with basic pulsing and their intelligent interventions & adjustments.

 

Sounds good Ancel, but I'm looking for how the device "knocks the cured sulfate" back into solution. This is what is quoted as the process. As mentioned, sulfate forms as the battery is pulled down, and should recombine when the battery is charged. The sulfate sometimes cures when the battery is left down for too long. Equalization supposedly misses this cured sulfate, though it can prevent new from forming by fully saturating the cells with what is available for reaction. It is used at 1/5C to correct when the cells get out of balance from each other. Supposedly I can up the voltage and current (to what?) and knock the harder cured sulfate loose. Does the 1/5C RMS rule apply? ...or is it something else? ...and what is different than other charging methods?

 

The current flow into the battery and the resulting elevated voltage across the cells equalizes the battery similar to DC equalization and reverses the large crystalline formations of the 'hardened' sulfate back into solution. It is simply a different and often slower and more gentle means of equalizing the battery. I have found no 'magic' resonance frequency that is best, but due to cable inductance a high frequency (Mhz) achieves very little 'pulsing' as inductance resists current change and averages the voltage out to little more than ripple DC.

Note that sulfate does not 'cure', in time, it simply grows into larger crystals which can eventually disrupt the paste & grid as well as insulate the plates. DC equalization does not 'miss' this. It occurs because the battery is not being fully charged on time. Timely DC equalization reverses this just fine.

A high degree of sulfation requires a higher electropotential across the cells to denature the larger sulfate crystals. This causes heat generation which coupled with large sulfate crystals can propagate cracking and destroy the plates via thermal stresses. 'Soft shorts' causing higher self discharge can also develop if the large 'sharp' crystals puncture the cell plate separators.

You do not want to 'knock' sulfate loose nor do you want do 'convert' it to something else (via adding EDTA etc.) as that removes some sulfate ion from the electrochemical process and this reduces the conductivity of the electrolyte and storage capacity of the battery. You need to reform the electrolyte evenly in all the cells.

The approach that is required depends on your goals. If you're trying to recover a known good battery that has sat a couple months too long between charges, DC equalization as detailed by MikeMl is all you need to do. If you're handling batteries with a couple years of neglect and unknown failure quantification and >1 Ω internal resistance then you need a decent pulse desulfator capable of > C/40 A avg current delivery, a non contact temperature sensor, a spec. grav. tester and a carbon pile load tester. Also a fair amount of time to monitor the battery and decide if it's improving or not. As I noted before only 1/3rd on average are recoverable as there are other battery failure mechanisms that could be active here.

Once you get the S.G> up to reasonable values (1250+) a 20% discharge/charge cycle or two is advisable as this seems to reform the even distribution of active material on the lead plate grid and provide better AH capacity.Over vigorous out gassing during the EQ process can break off any loosened lead paste that developed due to sulfate crystal expansion, which leads to small pores in the paste exposing virgin paste or lead grid. Cycling the battery helps compensate for some of this.

All of this takes time, as overheating the battery can destroy it. Further AGM 'starved' electrolyte batteries and other recombinant SLA batteries need a special regimen to be safely recovered due to thermal considerations. I have decided not to pursue recovering GEL batteries as they don't usually sulfate (no stratification) and out gassing can cause voids in the gel. Wet cell storage batteries are probably the best candidates for recovery as their thicker plates can tolerate more abuse.

As I deal with junker batteries mainly, I use a battery (pun) of approaches which combine to achieve statistically viable results. This includes reflex pulsing sometimes called negative pulsing during the out gassing process, to limit heat generation and efficiency loss due to too many gas bubbles ,caused by a very polarized surface charge, degrading electrolyte conductivity.

Here are a couple pics of the same busbar in a 12V SLI battery (b4 & after recovery)

 

That was the most imformative anything I have run across, and contains information I have been looking for, but not finding.

...anyway, since 1/40C is a good RMS current, what is a good peak current? How much difference does it make?

Also, I have been using the "too hot to hold" rule of thumb for thermal management. If it's very warm to the touch, it's too hot. What is the maximum temperature rise you consider safe?

"Gel cells" I also process, but those too I open to check the electrolite before I go too far. They rarely have gel.

 

That was the most imformative anything I have run across, and contains information I have been looking for, but not finding.

...anyway, since 1/40C is a good RMS current, what is a good peak current? How much difference does it make?

Also, I have been using the "too hot to hold" rule of thumb for thermal management. If it's very warm to the touch, it's too hot. What is the maximum temperature rise you consider safe?

"Gel cells" I also process, but those too I open to check the electrolite before I go too far. They rarely have gel.

 

That statement doesn't tell the whole story.
While it is true that pulsers are just slow equalizers in MOST cases there are certain pulsers that do MORE than a lab supply can.

One task that a low impedance pulser can execute that a lab supply cannot is: dedendriting AND stress testing of the battery interconnects & grid for conductive integrity.

Various projects in magazines for pulsers usually involve energising an inductor and producing a back emf that goes back to the battery as a narrow spike of higher than the normal terminal voltage.

A sulphate layer on the plates is basically an insulator, so you need higher voltage to stand any chance of breaking it down. In the past I've used a current limiting transformerless method that isn't allowed on this forum - but if a battery is that bad, it pretty much isn't worth reviving.

All the magazine project pulsers I've seen rely entirely on the energy from the battery itself - inevitably they will eventually run the battery flat and cause what they're supposed to cure. Its the sort of thing I'd set going and only remember to check it by the time the battery is ruined.

 

I've seen that too Ian Field. My build will certainly pull from an external source, even if it's my Schumaker (charger). Those projects that pull from the battery to fix it looks ineffective. Most OTS solutions go for what several batteries cost me, and they all have the same snake oil salesman claims--CURES ALL BATTERY PROBLEMS. Show me some science, guys.

 

That was the most imformative anything I have run across, and contains information I have been looking for, but not finding.

...anyway, since 1/40C is a good RMS current, what is a good peak current? How much difference does it make?

Also, I have been using the "too hot to hold" rule of thumb for thermal management. If it's very warm to the touch, it's too hot. What is the maximum temperature rise you consider safe?

"Gel cells" I also process, but those too I open to check the electrolite before I go too far. They rarely have gel.

I set the case temp control (triggering current limiting) for Wet cells at 43°C and 37°C for AGM/SLA respectively. Be aware that internal core temperatures are about 5° - 7°C higher. Passive Infra Red imaging works well on the thin plastic cases of modern batteries. I use a FLIR E8 cam to track the internal cell structure thermal profiles & hotspots.

On the matter of pulse current amplitude: It serves two purposes:
1) Forced electron and hence ion mobility at the cost of I*I*R power dissipation outcomes. pulse duty cycle impacts this as well.
2) A high enough pulse current can decompose dendritic growth 'mossing' which occurs within batteries being slowly desulfated.

My observations show that more is better ONCE you apply proper thermal runaway limiting on the battery. Desulfation occurs more rapidly the hotter the battery becomes. Ion kinetic energy is directly related to thermal energy. This can be observed directly by the reduction in electrolyte resistivity as it heats. This effect is beneficial to the current flow and helps the electro chemical recovery process. However, due to battery construction , the use of dissimilar materials , alloys and the composite nature of battery plate paste, too much heating or hotspots can cause physical deformations which are not reversible. Once more, some heat is good, too much or too fast and it's bad. Thus it is advisable to also ramp up the pulse current over a few hours to provide for battery electrochemical latency if your pulser can deliver high amplitudes upwards of 2C.

Note that current monitor transformers are the best way to accurately track pulse currents IMHO, thanks to @Lestraveled for that advice. (Note my Icon)
I have developed a reasonably accurate (+/- 5%) means of deriving pulse current from average current and pulse widths for my application. The calculation is an empirical one and depends on the known efficiency of the individual pulser, cabling, connectors etc.
Here's a pic shown an example of my pulse train at work.

 

I've seen that too Ian Field. My build will certainly pull from an external source, even if it's my Schumaker (charger). Those projects that pull from the battery to fix it looks ineffective. Most OTS solutions go for what several batteries cost me, and they all have the same snake oil salesman claims--CURES ALL BATTERY PROBLEMS. Show me some science, guys.

What I usually use is an Optimate charger/conditioner, but it has polarity detection and a dead flat battery with no voltage at all won't activate that and it won't start charging.

In most cases, anything the Optimate can't recover is suspect at best, but to get by the no terminal voltage problem, I use a heavily modified old iron-cored transformer charger. First of all, I added a pair of non-electrolytic 2u2 capacitors to the bridge rectifier to convert it to a voltage doubler - the off load voltage reaches over 40V! That will start charging on a mildly sulphated battery and give it enough terminal voltage to activate a smart charger with polarity detection. Later on I added a hefty electrolytic so I can use that 40V or so to zap whiskered nickel cells.

 

Hi Ian:

Since my design also uses polarity detection to prevent accidental pulsing into a short or rev. connected batt I have tweaked it to bring the batt. detection down to under 2V. In the >100 'flat' 12V batts I have processed the lowest resting V has been around 3V.