Lead Acid (SLA) battery desulphation.

Thread Starter

recklessrog

Joined May 23, 2013
985
At Christmas I was very ill with the flu and have been recovering for the last two weeks. This has given me time to get into my workshop for some play time.
A while ago, another member posted a question about the Dave Barker Lead Acid battery de-sulphator, and as my little club had been given a collection of used alarm back-up SLA batteries in various states of health ranging from "as new" to totally dead, it aroused my interest as to whether or not the process worked. I read many articles on the subject, and concluded that most had been written by "armchair experts" who probably were very knowledgeable on the theory of the chemistry, but had not actually done any practical testing. (Throughout my life, many time I've been told "that won't work" only to find in practice it does)
So, as the consensus of opinion seemed fairly evenly split between the advocates, and the dismissive, I decided to build one and see for myself.
First of all, I attempted to charge a small selection with my commercial maintenance chargers. Those that accepted a charge were then given a discharge/time test carefully noting the current and voltage over time. Trying this a couple of times showed pretty much the same results, they were all well below original new capacity, i.e a pair of 7A/h sla's only had about a quarter of that rating.
I built the Dave Barker design exactly as his schematic showed, and by using differential 'scope measurement across a 1 Ohm 10W resistor in series with
the battery, observed initially very short 15-20 microsecond 20Amp H.F ringing pulses increasing to over 30Amps into the battery at a repetition rate of around 2 KHz. After 24 hours, the discharge test was repeated, and to my surprise, now showed some recovery, up to about a third of full capacity. I reconnected the desulphator and left it for another 24 hours. Now the off load voltage (after allowing time for settling) had increased by 0.75 volts from the previous charge.
The discharge cycle was repeated and now was up to 2 thirds full capacity. The process seemed to be successful. At this point, I connected a completely dead 2.1Ah battery that had failed to accept a charge from the maintenance charger, and left it on for 24 hours.
( I should point out that I was powering the de-sulphator from a bench psu that had current limit set to a little over the maximum charge rate for the battery under test, but voltage was only 12 Volts. The final battery voltages were above this so it was not just "pass through" charging.)
Back to the "dead" battery", after 24 hours, it showed 11.5 volts, and a discharge capacity of about 0.25 Ah. I gave it a further 24 hours, let it rest for 24 hours while I charged the first one again, then gave it 36 hours. After that, I ran the discharge cycle and it was up to 1.8 Ah and the voltage was 13.7V!!!!! Wow, something good is going on.
I will continue charge/discharge test as time permits and see if the is any further improvement.
I have made a lash up of a somewhat more powerful version which is able to provide 90Amp pulses into a 17 Ah battery that is also recovering well.
My conclusion so far, is, They were previously useless batteries, and now they are working and useful, so from the above, feel free to comment and draw your own conclusions. :)

Once my camera battery is charged, I will post some pictures
 

Thread Starter

recklessrog

Joined May 23, 2013
985
Maybe the battery needs to be desulfated. I know a guy that can help wi...
Never mind.

PS, good work. I would love to see some graphs of what your results are on discharge before and after the desulfurization process.
Ha ha! yes, I will try and put my scribbles down into graph form once I've done some more testing.
Trouble is, now I'm getting better, I am expected to do all the other things not done while I was dying! lol
 

GopherT

Joined Nov 23, 2012
8,009
Ha ha! yes, I will try and put my scribbles down into graph form once I've done some more testing.
Trouble is, now I'm getting better, I am expected to do all the other things not done while I was dying! lol
Well, make a note to remind yourself to make the graph next time you are dying!
 

SgtWookie

Joined Jul 17, 2007
22,230
Years ago, I did some experimentation with desulphation circuits, and did get some success with recovering a few very heavily sulfated cells. Several had a specific gravity of zero. It took a month of desulphation on one battery, but I finally got normal readings on specific gravity on all cells.

If the problem is merely sulphation, then a desulphator is what's called for. If the cell has structurally degraded, then there isn't much you can do to save it, besides recycling.
 
Last edited:

Thread Starter

recklessrog

Joined May 23, 2013
985
Here are some pictures of the set up used. The large 200Watt W/W rheostat is my discharge load. Sorry the scope pics are a little fuzzy, taken hand held. (note to self, do not lend camera tripod then forget who I lent it to)
P1010012.JPG P1010011.JPG P1010015.JPG P1010030.JPG
 
Last edited:

Thread Starter

recklessrog

Joined May 23, 2013
985
I've made some improvements to the lash up version and can now pulse 200 amps max, so will try this on an old diesel car battery that gave up the ghost a couple of years ago. It will probably take a long time but I will post the result.
What is interesting is that the choice of power fet is fairly tolerant of different types, the inductors too are not critical. I have a collection of transformers from scrap Plasma T.V power supplies that seem to be ideal, I just look for ones with a thick copper wire windings, and measure the inductance. Anywhere between 230 to 500 micro Henrys is easily accommodated by adjusting the drive "ON" time to the power fet.
No need for the Hex inverter buffer 4049 I.C used in the original Dave Barker design, just directly drive from a good old 555. It is important to include a 15-30 Ohm resistor in series with the gate to prevent high frequency parasitic oscillations that can occur due to rapid switching of the capacitive fet gate. this resistor needs to be mounted very close to the gate terminal. Without this resistor, I fried one fet.
If I make any substantial changes, I will post an update.
Oh, nearly forgot to mention that it is important to keep the output leads short, less than 10 inches (approx 25 C.M) as it is a powerful 3-5MHz transmitter with lots of harmonics!!!!

Here is a picture of the inside of the original Dave Barker design I made. I included a small fan to draw air through the box to keep it cool during long periods of operation.

P1010033.JPG
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
It's great that you included photos, but I don't see any schematics. It would be great if you would upload (a) schematic(s) in .png or .pdf format so that others may replicate what you have done.
 

Thread Starter

recklessrog

Joined May 23, 2013
985
It's great that you included photos, but I don't see any schematics. It would be great if you would upload (a) schematic(s) in .png or .pdf format so that others may replicate what you have done.
Hi, If you google "pulse De-sulphator for Lead-Acid battery resurrection" by Dave Barker, the schematic for the one I built first is there .www.barkeraircraft.com/files/Pulse3_web_layout_.pdf
Once I have more proof that my new design is working as required, I will post a schematic :)
P.S. Anyone building the Dave Barker design should note that the unused input pins of the 4049 (pins 9-11-14,) MUST be connected to the negative 0V rail or very odd things will happen. All INPUTS must go somewhere and not be left floating. This goes for all Cmos logic chips.
 
Last edited:

Janis59

Joined Aug 21, 2017
1,834
Maybe Your American batteries are something more likely a russian batteries, but here in Europe them are long ago forgotten. My youth the battery weighted some 40 kg, but now maybe 5, so I even doubt how much the lead they contain. Those old batteries I remember my first cars from 70~ies was working some two years and then some 5 years was dying step by step. At least, if engine may make a one turn, You have a hope that next summer something like a half-turn will be still in it. But nowaday batteries act completely other habit. When You see that while starting the radio change the sound or mp3 device loss the memory, be in true horror. Because probably You shall start the engine as normally, shall drive a so small as one kilometer to the belowed (I ironize) Systembolaget shop, but there may realize that accumulator is unable to give even so small as the sound to radio. Click, and it nonexisted. How that may happen so sharp I dont know, but now its the second acting like this. Dead in the 1 kilometer trip. Yet there are a good news too. The first 5 years they are working faultless.
 

Thread Starter

recklessrog

Joined May 23, 2013
985
Just a quick update, I made a few changes to the new de-sulphator I designed and am pleased to report that so far it appears work better than the original design. One 17 AH battery was unable to take any charge from a conventional maintainance charger no matter how long I left it on. I connected my new design to it, and after a couple of hours it was charging so I disconnected the de-sulphator and left it overnight on the maintainance charger.
The following day I checked the voltage off load at 10.1 Volts, performed a discharge cycle at 1 amp and it was down to 9 Volts in about 15 minutes. I re-connected the new design de-sulphator and limited the current pulses to 75 Amps, left it running for 24 hours. After allowing the battery to settle for a few hours, I checked the voltage as 12.9 Volts off load.
Another discharge cycle was started at 2 Amps and the voltage stabilised at 12 Volts for over 90 minutes then fell to 10 volts in the next 10 mins.
From this I can safely say the battery is "recovering" some usefulness. It is currently back on the de-sulphator at a pulse limit of 50 amps and I shall leave it runnining for a couple of days before testing again.

I will scan a drawing of the schematic and post it here for anyone who may want to build one. Photo's to follow once I finish the case.
 

Thread Starter

recklessrog

Joined May 23, 2013
985
Here is the schematic of my MK11 De-sulphator, I have not yet included the power supply as it's still under test but I will post it soon along with details of the inductor and the semiconductors. Please bear with me as I don't have a lot of free time now I've recovered from the flu.
MK11 Desulphator.jpg





Important note! I was making some changes to the 555 circuit and discovered that the rise/fall time of a NE555 is not as fast as an LM555. This may be why the Dave Barker design uses a buffer to improve the drive switching of the Mosfet. Whether this is something particular to the ones I have or is a function of the make I don't know.
My batches of both came from R.S supplies so are not Chinese rejects or copies.
I will order some more of each type and the ua555 version and see what results I get.
At the moment, I would recommend only using the LM555.
 
Last edited:

Thread Starter

recklessrog

Joined May 23, 2013
985
img126.jpg I think I have conclusively proved that the "de-suphator" is able to restore otherwise useless SLA batteries with varying degrees of success. Whatever is occurring, whether it is "de-sulphation" or some other process, I cannot say. But, out of 10 otherwise scrap batteries, we now have 7 that happily accept a normal charge and are at least up to half of the rated Ah capacity, 3 are actually nearly "good as new" and all seem to be improving with each charge/ discharge cycle.
The remaining 3 had been unable to accept any normal charge to start with and after several "de-sulphator" charges, regained some use so far but only time will tell if repeated cycling will give any further improvement.
I have also done enough testing to satisfy myself as to the better power supply design and will attach a schematic.
The first one I built used a 20 VA 12 + 12 Volt secondary that can be wired in parallel or series and so become 12-0-12 Volt.
In the first instance, I wired the secondaries in parallel and used an encapsulated bridge rectifier type KBPC 35-01 rated at 35 amps 100 volts.
The forward voltage drop of this was 1.3v per diode, so I was losing 2.6 volts in total, giving an output of only 10 volts at full load, it also ran very hot when powering the de-sulphator.
Looking through my accumulation of used parts mainly from scrapped plasma t.v's, I found a nice double diode with only 0.41 forward drop so I rewired the secondaries in series, as per the attached drawing and the diode stayed very much cooler, plus the output to the de-sulphator was maintained at 12 volts at full load.

As for the diodes and FET, pretty much any 30 volt 8 plus amp or more (FMCG28,FSU05A40,F10UP60S, BYC10X600, RHRP1560 etc) Shottky diodes work well providing they are mounted to a heatsink.
The Fet's I've tried are mostly enhancment mode 250 volt or greater, even IGBT's RJH3047, work well.( some type no.s of ones I have tried, FQA19N60, 88N30W, IRFPC60, K3911, 20N60C3 or anything similar to these.) Nearly all mine came from scrap old Plasma tv power supplies as did the inductor.
I will follow up on how to choose one soon.
P.S, In the schematic, I have not shown an on/off swich or fuse, but it is advisable to add these if you build one. Also, if you fuse the output of the power supply, use a slowblow anti surge type of around 5 amps.
 
Last edited:

dendad

Joined Feb 20, 2016
4,452
This has been a very interesting entry, thanks. I must dig my hacked together desulphator out and have a go on the 48V worth of SLA batteries I took out of a rack mounted UPS. I was given it and all batteries were dead as it had not been used from new but just stored for some years. It will be a good test!
 

Thread Starter

recklessrog

Joined May 23, 2013
985
This has been a very interesting entry, thanks. I must dig my hacked together desulphator out and have a go on the 48V worth of SLA batteries I took out of a rack mounted UPS. I was given it and all batteries were dead as it had not been used from new but just stored for some years. It will be a good test!
That is an excellent idea, there must be hundreds of dead or failing UPS batteries kicking around, in fact it would be a good idea for anyone who relies on their UPS to protect their computer data, to put a load on the UPS, like a lamp(s) and do a couple of charge discharge cycles and see if the batteries hold up. If not, build a de-sulphator and give them the treatment. It may save your important data should you get a power cut.
Same goes for mobility scooter and golf cart SLA's.
 

Thread Starter

recklessrog

Joined May 23, 2013
985
I am sorting through a selection of inductors that have been removed from scrap power supplies. Many have windings that seem to be suitable.
I designed a saturation/inductance curve tracer some time ago, and it is invaluable for measuring unknown inductors of all sorts. not only can you get a very good approximation of the inductance, but also see the saturation profile. Although I have peak electronics atlas LCR ( https://www.peakelec.co.uk/downloads/dca_and_lcr_review_pw_mar05.pdf ). it does not show the saturation characteristics, so the two in combination, enable me to easily select a suitable inductor for a given purpose.
I will follow up with some more pictures and waveforms, also a slightly redesigned 555 driver circuit that is independently adjustable for pulse period and frequency which will be useful to optimise performance to accommodate various combinations of coils and fet's etc.
My thought's are, don't buy expensive components when a trip to the local rubbish tip/reclamation yard can often yield scrap electronics that would otherwise be crushed.
( Be polite, and ask the guys in charge if they mind if you can take some scrap tv's etc. Old plasma psu's have a wealth of power Mosfets, shottky diodes, and inductors, don't expect them to remove parts for you, and always offer a small tip)
 
Last edited:

Thread Starter

recklessrog

Joined May 23, 2013
985
Here are a couple of example curves that clearly show the point of saturation. The voltage in millivolts is multiplied by 10 and become the current in the inductor.


P1010058.JPG

P1010050.JPG
 
Last edited by a moderator:

Thread Starter

recklessrog

Joined May 23, 2013
985
I'm repeating the note I added to the schematic of the MK11 version.

Important note! I was making some changes to the 555 circuit and discovered that the rise/fall time of a NE555 is not as fast as an LM555. This may be why the Dave Barker design uses a buffer to improve the drive switching of the Mosfet. Whether this is something particular to the ones I have or is a function of the make I don't know.
My batches of both came from R.S supplies so are not Chinese rejects or copies.
I will order some more of each type and the ua555 version and see what results I get.
At the moment, I would recommend only using the LM555.
It is feasible to use 4050 non inverting buffer (same pinout, just doesn't invert the signal) wired as in the Dave Barker design to speed up and shape the pulse if you only have a NE555
 
Top