Does anybody have any experience or any thoughts on hysteretic pulse charging of lead acid batteries (also called current interrupt charging I believe)? I happened to glance at the BQ2031 datasheet, and one of the algorithms is "Pulsed Current: hysteretic on demand."

Basically, the charger operates as a fixed current source up to the point where the battery overcharge voltage is reached. At this point, the charging current is inhibited, and then battery voltage decays towards the float voltage level. When the float level is reached, the charging current is enabled again. The process then continues on and on. The thing that I find attractive about this approach is that when the current is inhibited, the batteries can cool off. Also, there is only a single control loop for controlling current. Any thoughts, advice, experience, and info of any kind is much appreciated.

The first question that I would ask is the voltage and capacity of the battery.

charging the battery at a floating voltage leads to sulfatation of the electrodes. To avoid this kind of failure, we apply a plusation current but I must not be very repetitive and not very high unless we cause the corrosion of the battery. So there is a kind of trade off.

Chris, this will be a 120V EV battery bank composed of either 6V or 12V flooded lead acid batteries (also looking into sealed AGM batteries since they won't spill). I can get much more range out of the 6V batteries, but I would have to series stack 20 of them. The 20hr capacity per battery I am looking at is ballpark at 200 Ah.

For a six volt based system, I am checking out the Trojan T-605 6V (C = 175 Ah @ 5 hr rate, C = 210 Ah @ 20 hr rate) flooded wet cell. A 12V AGM with these ratings would be nice, but probably very expensive.

I would think charging a series string would best be done with a current source type of charger (pulsed constant current). I know AGMs can be fast charged with constant current, but I do not think flooded wet cells can fast charge. Charging current pulse magnitudes for flooded wet cells would probably be around 5 - 10% of the 20hr C rate (ie 10.5 to 21 A).

The reason I was asking the voltage and capacity is that people try this type of technique on small capacity systems and the time saved is so small it does not warrant the complexity. I have been designing SLA chargers for more years than I care to remember but my knowledge of wet cells is very limited. Can you explain the overall system ( are the cells in system when they are being charged) if so will it cope with the maximum charging voltage down to the low voltage cut off. Again I am not an expert on high voltage stacks but most systems have a voltage equalising system and cell stack monitoring to cope with uneven cells and cell failure.
Trusting the above is of some assistance
Best wishes

The overall system will be composed of 10 12-V flooded wet cells in series (AGMs are just too expensive for me at the capacity I want). I found a pretty good 12V flooded cell for what I need from Trojan Battery (J185P-AC). It has a 20hr capacity of 205 Ahr. For the weight, size, and cost, this seems to be the most attractive option (out of the lead-acid family). They will be charged in system. A voltage equalizing system and cell stack monitoring is something I would have to research more. I was thinking one option would be to have a small charger for each battery, to make sure each battery is charged. Just leave the batteries in-system, and hook up a charger to each battery. Probably would never be accepted in industry, but it might be good for now. Anyways, thanks for your helpful advice and comments!

Jay
Is this a one off project, you could charge each battery individually but this is not very cost effective and the monitoring would be a nightmare (individual charger failure). I still do not understand why you would want to use a complicated charging system in the first place. Are you confusing pulse charging with 3 stage charging where the third stage is effectively scrubbing the plates. Is this part of a UPS or standby system

Yes, I agree. Multiple chargers would be a nightmare... It was just a thought I guess. I did some more research, and found that flooded cells can take a bit more abuse during recharge, and a BMS system is not required with flooded lead acids. With flooded cells, you bring undercharged cells up to charge with an equalization charge mode after the normal charge mode. Then top off with water if necessary.

I think I understand three stage charging: constant current, constant voltage, then trickle at constant voltage. I was/am interested in constant current only charging since the control loop is a single control loop (with simple voltage monitoring circuitry). The three stage charger requires closed voltage and current loops, and more complicated sensing circuitry.

To answer your question: this is a standby system only.

Still looking into this stuff and more to come. Thanks for all the comments and suggestions!

now we are making progress, what you need is a single stage constant voltage constant current charger set to the float voltage of the chosen cells used. As you have discovered open wet cells stand more abuse although they need more maintenance. As previously stated most of my work has been on lower voltage SLA systems but I would suspect that you would be able to find this type of product is manufactured for forklift trucks.
Trusting the above is of assistance

Thanks Chris! I appreciate your time and help. I believe you are correct: single stage constant voltage/current charging is probably the best way to go for this application.

Correct me if I am wrong, but I believe this is called a float charger: The charger holds a voltage across the bank, and current is limited to some constant value at the beginning of charge. As the bank nears the end of charge, it draws less current, and the charger comes out of it's current limited mode as charging current tapers off. Charging voltage did not change the whole time.

Your are correct that this type of charger is called a float charger, the way it works is that when you charge a discharged battery the system works in current mode and delivers the maximum current that the charger allows. As the voltage rises it will reach its float voltage and at this point the battery will take current as it wishes which is dependent on the capacity of the battery ( the larger the capacity the larger the current) when the battery is fully charged it will just take a small current which compensates for the efficiency of the battery.

Awesome! I think I am set here on what I need to do. Thanks for offering your excellent advice Chris. I appreciate it

You still have not explained your application, do you need a load shed to protect the batteries.

Hello Chris!

The application is an electric vehicle DC motor drive, so I assume you are referring to shedding the load while the batteries are under load. I have multiple ways to protect the battery bank: I have a fuse and a large circuit breaker.

On somewhat of less related note, filter capacitors (along with cable inductance) ensure smooth DC draw from the bank as well.

You need to be careful with pulsing as the electrodes can be stripped if the current is too high. You would need to have a look at the manufacturers data to determine the maximum current you could use to charge. The regine ought to have a folat facility in order to balance the cells after a lot of punishment.

I have been in the UK for a few days so I did not see your reply, the load shed is to protect the batteries from deep discharge. It might not be so important with you application but I would ask the battery manufacturer how their batteries recover from deep discharge
Best wishes
Chris

Thanks Chris. I will definitely check with the manufacturer. I hope you enjoyed your trip to the UK.

Hi Jay, I looked into this a few years back, have a look at these papers
M Bhatt, W G Hurley, W H Wölfle, “A New Approach to Intermittent Charging of Valve-Regulated Lead-Acid Batteries in Standby Applications”, IEEE Transactions on Industrial Electronics, vol. 52, no. 5, pp. 1337–1342, October 2005.
W G Hurley, Y S Wong, W H Wölfle, “Self Equalization of Cell Voltages to Prolong the Life of VRLA Batteries in Standby Applications”, IEEE Trans. on Industrial Electronics, vol. 56, no. 6, pp. 2115–2120, June 2009.
These papers discuss pulse charging.

I began designing pulsed battery chargers back in the 80's.
Regarding Lead-Acid types, plate-sulfating issues arise and pulsing can provide mechanical assistance in minimizing it.
This is ok for small VLA batteries, but my many years in the Cellular power field shows that it is less practical with the large telco battery stacks as there are hundreds or thousands of amperes involved.
My controllers in this arena utilize temperature controlled floats and careful metering to maintain the 20+ year life spans.
Therefore, application may make the practical choice for you.

Anthony:

Thank you for offering your exprt advice. I was originally looking at using this for large flooded lead acid cells. The cells I have are 12V 200 Ah flooded lead acids. Chris and I discussed a few things, and it seems like an industrial style constant voltage/current limited float charger would be best for my application (120V EV battery bank).

I like the idea of a switch-mode current source for charging flooded batteries though. The majority of chargers are CV,CC, and CV 3-stage smart chargers, which need 2 control loops to implement ACMC.

I still like the idea of a simple switch-mode current source for charging flooded batteries though (one active control loop for current). I know it may sound a little odd, but I've always wanted to build a current source out of a switchmode power supply for an application, other than for charging capacitors (but a battery is kind of like a big capacitor.....). For a 200 AH battery, I would CC charge at 20-30 amps until the float voltage, and then pulse this current on and off. Any thoughts? I guess I could always try it on flooded cells and see what happens!

Flooded lead acid batteries have a long history in telco industries. You won't improve on existing schemes for charging. Consider the 20, 30 year maintenance required at central office installations. If a better technique could be used, it would... this is big money that's involved.

It is a hot and sunny day here in Ireland, I did not reply to you earlier as I had an accident last week whilst cutting trees and ended up in hospital. I noticed that you were interested in adapting a switchmode power supply for a charging application. I have spent the last 30 years designing sealed lead acid chargers. The basic difference between a power supply and a battery charger is that a charger is stable in voltage and current mode and has some form of reverse polarity protection. The systems that I tend to design are smaller than yours and use electronic protection where a system of your size would use relays/contactors electronically controlled for protection. 24V and 48V power supplies are quite common and cost effective and if you chose a power supply that is capable of load sharing this means that it will be stable in current mode. Another possibility is to look at chargers made for forklift trucks which tend to run at higher voltages. If you do go down the road of adapting a switchmode power supply be careful of the output capacitors as applying the battery when they are discharged will cause a large splash and is dangerous.
The normal way over this problem is to monitor the battery and the output of the charger and only close the relay when both are up. The other problem you might have is when the battery is deep discharged, the way over this is to apply a small current through a resistor which charges the battery slowly and when it reaches a suitable level you close the relay and charge as normal.

Chris: I assume (and hope) that you are OK. Cutting trees can be pretty dangerous. I cut mine in my yard a few weeks ago with an extension ladder and a hand-held saw. It was a dauting task at times. Anyways, I hope you are well.

Wow 30 years of design...I hope I will get to that point one day. Thank you for all of your practical tips, they probably saved me a few blown caps!

I plan on checking with the forklift folks as a last resort, since I am pretty determined to give this a shot with a switch-mode charger.

Right now I am looking at a PFC front end, with some type of isolated buck converter (2-SW forward, full bridge, etc). The other thing I was thinking about is something that provides PFC capability and isolation in one topology. A 2-SW flyback and a UC3854 PFC control IC could possibly do the job, but at my power level, this might not be practical.

As you mentioned, when the batteries are deeply discharged, there will be a serious current draw. A resistor would work, but I would think a current limit in the SMPS would do the job too? But then again, a resistor/relay combo might be better.

Anyways, thanks again and hope to talk to you soon.

Power supplies of the size that you are looking at will automatically have PFC and probably soft start, as previously stated it is important to have a contactor / relay between the power supply and the battery this can be used for basic reverse polarity protection and should only be energised when the positive of the battery and the positive of the power supply are detected. By the sound of things you do not quite understand deep discharge, when a battery becomes deep discharged it is almost transparent (does not work like a battery) and to overcome this condition you will feed a small amount of current into the battery and eventually it will start to charge (could take several hours). When the battery reaches the the low level it is safe to start to charge the battery as normal.
The battery should not be allowed to get into this condition but if it can happen it is worth fitting the extra circuit as it will save an Engineer having to manually bring the batteries out of deep discharge.

Hello Chris!

I see it now, and yes I did not have my head completely wrapped around it yet.

I believe some battery charger ICs perform a test to see if the batteries are deeply discharged (or even if internal cells are shorted) first before applying full charging power. A small charging current is applied as you said and voltage rise on the battery is detected. Once it comes up to voltage, full power is applied. If it does not come up, then the charger is shut off. Polarity protection also is a must.

Let me know if I have the correct idea. Thanks again!

Jay,

I just finished a project for a lead acid battery float charger on a hospital bed. Look into the UC3909. That actually provides all the functionality you are looking for in a lead acid battery charger. It may take a bit of technical wizardry, and a bit of math, but that chip can be adapted for your application by using a high-side FET driver and a high-side current sense chip, like I did. You can add a PFC front end to the system, if so desire or required.

Good luck.

Hello John!

Thank you for pointing me towards the UC3909. It looks like a pretty neat IC... Float charging is what I'm looking at right now too. Just out of curiosity, were your batteries flooded or sealed?

Anyways, thanks for taking the time to post. I appreciate it!

The batteries I used were sealed, VRL type batteries, 12V, 33Ah, BP33-12S from BB Battery, two in series. My max charge rate was only 2.2A, so rather slow. My charge current was limited by my power source's capability. The UC3909 controls external devices, so you can easily use whatever is required for whatever charge current and voltage you need. Using a high-side N-FET driver chip, FAN73711, and a high-side output current sense amp, LT6106, solves the high-voltage problem. I took the formulaes from the U-155 app note and put them into a spreadsheet so I could dynamically vary elements, output voltage, current, and other factors to tailor the resulting device values to get exactly the response I wanted. It was not a trivial task, but well worth it. I can't link files here or I'd send you a copy of my spreadsheet. Anyway, the result is a very reliable and safe system for this medical application. Efficiency in the design is only as good as you make it, and greatly keyed to the inductor and pass-FET chosen.

Good luck.