Batteries

V2.07  5-Jul-2004

1. Introduction

A battery converts chemical energy into electric energy through an electro-chemical process. The basic unit is called a "cell" and can be manufactured in a wide variety of shapes and sizes. Batteries are made up of one or more cells in series or parallel combinations to create the desired voltage and output capacity.

The electro-chemical cells consistsof two terminal suspended in an electrolyte. The terminals are called the anode (-) and the cathode (+). An electrical current is essentially a flow of electrons, and the battery can be regarded as an electron pump. The chemical reaction between the anode and the electrolyte forces electrons out of the electrolyte and into the anode metal, through the circuit, then back to the cathode. From the cathode metal, the electrons re-enter the electrolyte. This direction may seems strange, from negative to positive. We regard ‘current’ as flowing from positive down to negative, but in fact, this current is a flow of electrons in the opposite direction! The anode and cathode both get converted during this reaction, one is ‘eaten away’, and the other has a build-up of material on it. When a rechargeable battery is recharged, this chemical reaction is reversed, and the terminals are restored.

2. Primary and Secondary cells

Batteries can be divided into two classes: primary, and secondary. Primary batteries are designed for a single discharge cycle only, i.e. they’re non-rechargeable. Secondary cells are designed to be recharged, typically, from 200 to 1000 times. For use in robot wars, primary cells would be far too expensive. Among secondary cells, the options open to us include NiCd, NiMH, rechargeable Lithium, or lead-acid.

3. Ampere hour ratings.

The energy stored in a battery is measured in the units Ampere-hours (Ah). The units of energy are actually Volts × Amps × Seconds, but since the voltage of the cell is constant, it is only the product of current and time which determines the amount of energy in the battery.

Rechargeable cells generally have a lower energy density (that is the total amount of energy they can hold, Volts × current × time, divided by the physical size of the cell) than primary cells. A nicad D cell provides 5Ah at 1.2V, a lead acid D cell provides 2.5Ah at 2V, and the comparable alkaline cell provides 10Ah at 1.5V.

4. Comparison of secondary technologies.

The table below shows a comparison between the main three types of rechargeable battery.

Chemistry
Volts per cell (V)
Temperature range (ºC)
NiCd
1.2
-35 to +70
NiMH
1.2
-5 to +70
Lead-acid
2
-65 to +65

To provide a reasonable energy for a heavy-weight robot wars bout requires a minimum of 5 Ah realistically, It would require many Nicad or NiMH cells to achieve this, so the rest of this page will concentrate on lead-acid batteries, which most people are using.

5. Lead acid batteries

In lead-acid batteries, the positive electrode is lead dioxide, while the negative electrode is metallic lead. The electrolyte is sulphuric acid. As the cell discharges, the acid electrolyte is consumed producing water and both electrodes change into lead sulphate. When the cell is recharged, the chemical reaction reverses. As in most battery types, it is good to store them at cool temperatures and operate them at warm temperatures. Most lead-acid batteries have the highest energy density at about 30-40ºC. If the cell undergoes overcharging, hydrogen and oxygen gassing will occur with the loss of water.

The sealed type of lead-acid battery (SLA battery) is similar to the non-sealed type except a few changes in the electrolyte. This commonly takes the form of a gel or is absorbed by the separator. Methods to recombine oxygen formed at overcharging are employed. Although specifically designed for the reduction of excess gassing, pressure valves usually are installed to vent excess pressures to the atmosphere. The sealed construction allows a maintenance-free life which is not restricted to a horizontal orientation.

6. Safety

Non-sealed lead-acid batteries contain a liquid sulphuric acid electrolyte. If the battery casing is split, the acid can leak out. In batteries is it quite concentrated and will easily burn the skin and clothing. Therefore, the RW rules state that liquid-electrolyte batteries of this sort are not allowed. Appropriate SLA batteries are the only type allowed (apart from the much more expensive NiCd or NiMH types).

All lead-acid batteries produce hydrogen and oxygen during the recharging process. This production is increased if overcharging occurs. Sealed designs cause recombination of the oxygen at the same rate is produced, therefore eliminating the explosive mixture. Hydrogen which is produced will permeate plastic containers and as long as the sealed battery is not in a sealed small area, the hydrogen will harmlessly dissipate into the atmosphere. It is always good practice to allow for ventilation even with sealed batteries because of the possibility of a charger failure causing abnormal charge rates. If this happens, the battery will vent to prevent pressure build up. Another precaution is to prevent short circuits of the battery terminals. This can cause high heat and potential fire hazard.

7. Capacity

7.1. The ‘Cn’ rating.

A common way to specify the charge and discharge current of a battery is the C rate. It denotes the rate of total battery capacity. It is calculated by dividing the charge or discharge current by the total numerical battery capacity (Ampere-hours, Ah). For example a 30Ah battery is discharged with a current of 6A. The C rate is equal to 6A / 30Ahr=0.2C rate, also expressed fractionally as a C/5 rate, or C5.

7.2. Effect of current on Ah rating

Although a battery is quoted as having a capacity of, say, 26 Ampere-hours, this is generally its best possible capacity. The actual capacity depends upon a number of factors, including operating temperature, discharge current, and battery age. The Ampere-hour rating is normally for a discharge current of C/10 (2.6 Amps for a 26Ah battery). To see how this changes when the discharge current is different, look at the graph below:

Have a look at the top line in the graph, this is for the 70 Ah battery in the Hawker Genesis range. At the C/10 current, 7 Amps, looking along the graph we see that the battery will last 10 hours at this current, that gives 7A × 10hr = 70Ah. At 70 Amps, we may expect the battery to last 1 hour, but if you look closely, it will only last for 0.8 hours (48 minutes), that’s 56Ah. At 280 Amps drawn (a current of 4C), it will only last about 0.11 hours (6.6 minutes), that’s about 31Ah.

Taking the 26Ah line and plotting the percentage of nominal capacity against the current drawn, we have the following graph:

The I20 value is the current I = C ÷ 20, which for a 26Ah battery is 1.3 Amps.

Thus when we come to work out how large a battery we need, we are going to have to work out an approximate load-profile during the course of a battle. This is done later in this page.

7.3. Effect of temperature on Ah rating

Not only does load current reduce the rating, the temperature does also. Batteries are normally quoted for capacity at 25ºC. Below this temperature, and also far above it, they will operate less efficiently, The graph below shows how temperature affects the Hawker Genesis range of lead-acid batteries.

Also, the effective capacity of the battery improves as the temperature increases:

However, this doesn’t mean "the hotter the better"! The grid-corrosion rate in a battery is intimately linked to the battery’s ambient temperature. The higher the temperature, the greater the corrosion rate—and the sooner the failure of the battery. This accelerated corrosion at higher temperatures occurs regardless of the charge current flowing into the battery. However, since higher temperatures give rise to increased currents at a given voltage setting, the net result of an elevated battery ambient temperature is to intensify the negative effects on the battery.

8. Discharge

In an ideal battery, the terminal voltage would be constant over the whole discharge time, until when the battery is finally fully discharged, the voltage would drop right down (although this could make battery level detection quite hard). However, in real batteries, the terminal voltage decreases as the charge reduces. The flatter this drop the better. Let’s look at some typical discharge curves of various battery technologies:

It can be seen that lead-acid batteries have a reasonably flat discharge curve until they are nearly exhausted, although they are nowhere near as good as the NiCd secondary, and silver, and mercury primary cells.

8.1. Depth of discharge

The capacity taken from the battery is termed as the depth of discharge. For example, if 80% of the nominal capacity is used from the battery then the depth of discharge is also 80%. The maximum permitted depth of discharge depends on the level of the discharge current and the temperature. To maintain a healthy lead-acid battery, it is important not to discharge the battery too much. 80% is generally the most that it should be discharged to. The number of cycles obtained from a battery can be significantly increased simply by oversizing the battery, if you can tolerate increased size and weight, thereby lowering the depth of discharge in each cycle.

8.2. Deep discharge

The final discharge voltage is specified as the value below which the voltage must not fall when discharge occurs at the respective current. If the battery is discharged further, (i.e. the battery voltage is lower than the final discharge voltage), the area of deep discharge is entered. Discharging levels are key to long battery life as a deep discharge will shorten the life, and partial discharges will extend life.

8.3. Calculating charge level based on terminal voltage.

Since the voltage does gradually decrease as the battery is discharged, an approximate determination of the state of discharge van be made by measuring the open-circuit voltage to within a tolerance of about 15%:

8.4. Calculating charge level based on current accumulation

The only way to determine the amount of charge left in the battery with any degree of accuracy is to measure how much current goes into the battery during charging, and out during discharging. This sounds frighteningly complicated, but luckily there are chips that'll do the job for us. One I have come across is the Texas Instruments bq2013H. This requires a sense resistor to be fitted between the battery negative terminal and the ground line of the speed controller circuit. The IC takes into account time, battery temperature, rate of charge, self discharge, and initial battery capacity to determine the up-to-date battery capacity. It can drive basic displays to indicate this, or can deliver the information to an on-board micro via a serial link. The sense resistor can double up as a means of providing short circuit protection on the speed controller bridge circuits too.

9. Battery preservation and maintenance.

There are a variety of parameters which must be kept within certain thresholds to maintain the life of your battery. The most important the depth of discharge (DOD). At a DOD of 80%, the Hawker Genesis battery will deliver about 500 cycles; at 100% DOD, that number decreases to about 400 cycles.

Next in importance to DOD in determining the battery’s cycle life is cycle time. The time allowed for a recharge between discharges is critical to the battery’s life expectancy. Generally speaking, the longer the time allowed for a recharge, the longer the battery’s life expectancy. The preceding paragraph may be better understood if one remembers that a battery is fully recharged when between 105 to 110% of the ampere-hours discharged are put back in the recharge. If the time allowed for a recharge is less, then the current magnitude for a given ampere-hour must be increased. An increase in the charge current can be accomplished only if the charge voltage is also increased. This, in turn, leads to a higher level of overcharge, which speeds up the battery’s ageing process.

Thus, the longer the recharge time, the lower the overcharge rate, and the better it is for the battery. Conversely, the shorter the recharge time, the higher the overcharge and the harsher the condition for the battery.

10. Storage and self-discharge

Any battery, primary or secondary, loses charge over time when it is kept on open circuit. This phenomenon is termed self-discharge. If the capacity loss due to self discharge is not compensated by recharging in a timely fashion, the battery capacity may become irrecoverable due to irreversible sulphation, where the active materials (PbO2 , lead dioxide, at the positive plates and sponge lead at the negative plates) are gradually converted into an electroinactive from of lead sulphate, PbSO4.

Therefore, if the capacity loss associated with self-discharge is not replenished, the battery ultimately fails because storage is equivalent to a very low rate of discharge. The key factor influencing the self-discharge rate is the storage temperature. This is because the ambient temperature plays a major role in determining the speed at which the internal chemical reaction proceeds. The higher the temperature, the faster the speed of chemical reactions.

11. Charging

11.1 Charging schemes

There are many methods of recharging a lead-acid battery. Cheap car battery chargers (often termed ‘taper chargers’ because of the shape of the current waveform) generally have a transformer to reduce the mains voltage to around 12V RMS, then rectify this to produce a very dirty DC signal. In a normal DC power supply using this scheme, the output is capacitively smoothed, and often regulated. However, the battery itself provides the smoothing in this scheme, and regulation is unnecessary. The battery is then charged with a constant voltage, although it has a large AC content (ripple) on top of it. This ripple is a contributing factor to the shortening of the battery’s life, and ideally should be removed. The maximum ripple quoted for the Hawker Genesis battery is ±2%.

11.1.1 Constant voltage

In a constant voltage scheme, the current is dictated by either the battery itself, or by a current limit in thecharger, and the voltage is maintained constant. For example, a constant 2.25 to 2.6 volts per cell may be applied to the terminals. When the battery is initially discharged, its terminal voltage might be quite low, say 10.5V, and the equivalent resistance may be as low as 0.008 Ohms. This would cause a very large inrush current at the start of charging. The Hawker range of batteries allow this current, but most do not. The Sonnenschein range must be current limited to a value 0.2C (remember C is the Ah rating), so a 25Ah battery would be limited to 5 Amps. The Yuasa NP range is limited to 0.25C.

As the battery charges up, its terminal voltage increases, and the internal resistance decreases. Eventually, when it is fully charged, it will be taking a trickle current from the charge which maintains its fully charged state. The graph below shows a constant voltage charge cycle with no current limiting.

11.1.2 Constant current

In the constant current charging scheme, the current is controlled, and the voltage is allowed to go to whatever value is required to maintain the required current. This is the normal charging scheme for NiCd and NiMH batteries, but is not very suitable for lead-acid batteries.

11.1.3. IU scheme

In the IU scheme (I stands for constant current, U for constant voltage), the battery is first charged with a constant current (effectively the current limit mentioned above), and then with a fixed voltage. Effectively then, the graph is the same as the one above, but the current is limited to, say, 10 Amps:

Note that the voltage is dipped at the beginning. This is because during a constant current charge, the charger will supply the correct voltage to satisfy the Ohm’s law equation V = IchRbatt. The internal resistance of an uncharged battery is higher than that of a charged battery, so as the battery starts to charge up, its internal resistance decreases, and so the voltage consequently increases.

Obviously, by limiting the initial current, this increases the amount of time it takes to charge the battery.

11.1.4. IUI scheme

This is the same as the IU scheme above, but near the end of the charge, a constant current is supplied to finish charging the battery. This can result in a faster charge time for batteries that support this charge profile. It is recommended for the Sonnenschein Dryfit range. Sonnenschein recommend a final charge current of 0.24C, but this must not last longer than 4 hours until being turned off.

This method must only be used for batteries that are specified to use it, and only under careful control of the final constant current. If the final current is too high, the voltage will exceed the ‘gassing voltage’, which means the recharging chemical reaction produces gases faster than the battery can recombine them, and the pressure seal may blow, damaging the battery.

11.1.5. Pulse charging

Pulse charging applies a current to the battery for a fixed length of time, disconnects from the circuit and measures the open-circuit voltage to determine the state of charge, then resumes as required to complete the charge process. This is quite a complicated scheme, and requires some sort of microprocessor control. As we have seen, the terminal voltage gives about a 15% error in the indication of charge level. Also, a battery will have a highish voltage immediately after a charge session, and may take a few tens of seconds to settle to a stable value as the chemical reactions inside stabilise and gases are reabsorbed.

11.1.6. Temperature compensation

In all the charging schemes, it is preferable to have adaptive charging voltages based on the temperature of the battery electrolyte. It is only by measuring this temperature that you can tell exactly how the battery is responding to the charge. At higher temperatures, the chemical reactions proceed more quickly, and so less voltage is required.

Since the electrolyte in a SLA battery is inaccessible, the ambient temperature around the battery must be used. The charge voltage is then adjusted according to a graph like the one below. The adjustment may be different for each battery type or manufacturer, so you must refer to the battery datasheet for details.

11.1.7 Comparison of the schemes

Which charging scheme you want to use will come down to cost, and what you have available. I would recommend that you do not use a car battery charger on these expensive types of batteries, unless you know that it provides a high quality output.

The constant voltage scheme is the simplest to implement, but you must make sure your charger can handle the high initial currents, or is suitably current limited. The IUI scheme requires more complicated control and may need a special charger chip or microprocessor.

11.2 Battery charger circuits

I’m not going to present a whole host of battery charger circuits here because there are already plenty on the web, and links are given to them in the links section below. However, a description of viable methods is worth describing.

11.2.1 Conversion of a car battery charger

As we discussed earlier, car battery chargers generally do not provide a clean enough signal to charge these expensive SLA batteries. However, it may be possible to modify them so that they do.

The main problem with them is the amount of ripple current that they generate. First, there will be less ripple if a full-wave rectifier circuit is used rather than a half-wave rectifier. The circuits below show how the half-wave rectifier can be modified to full-wave, using three more diodes of the equivalent type to that used in the half-wave rectifier.

The second alteration we can make is to filter the signal. This requires as large a capacitor as possible tobe connected to the output of the charger, in parallel with the connected battery. The charger itself should limit the current that will be dumped into this capacitor on power-up. Some care may be needed here however, since car battery chargers often only have current limiting based on the resistance of the wires in the secondary of the transformer!

11.2.2. Using a battery charger control IC

Unitrode (now part of Texas Instruments) make a variety of complex battery charger control chips that can generate the appropriate signals to control various charging profiles. The UC3906 is their standard lead-acid charger part.

They also produce the BQ2031 which is a switch-mode charger control IC. This can charge using a dual-voltage scheme, a dual-current scheme, or the pulse scheme described above.

Both these ICs have temperature compensation capabilities built in, and various safety and overcharge mechanisms to prevent damage to the battery.

12. Series and parallel connection

12.1 Series

Batteries may be connected in series to generate higher voltages. The total voltage is the sum of all the individual voltages. However, some care must be taken depending on exactly what load is attached to each battery. Take the following connections:

In connection A, two batteries are connected in series to supply a single load. The batteries should be of the same type and capacity, and should be at the same state of charge,i.e. when you connect them up, make sure both are charged up.

In connection B, it doesn’t matter if the two batteries are of different type. The upper 12v battery may be just a small one to supply high voltages to turn MOSFETs in the upper arms of an H-bridge for example. The low one will be the high power battery.

12.2 Parallel

If you connect batteries in parallel you must be very careful that they are exactly the same type, same state of charge, and the same age, much more so than in the serial connection above. Batteries have an internal resistance, which goes up as they discharge. Imagine the following connection where one battery is fully charged and the other is discharged:

The discharged battery on the right has a terminal voltage of 10.5V and an internal resistance of 8 milliohms. The fully charged battery on the left has a terminal voltage of 13.2V and an internal resistance of 4 milliohms. The internal resistances are shown in the circuit for clarity. When the two batteries are connected together in parallel, the fully charged one will start charging the discharged one. A current will flow around the loop between the two batteries of

This current is very high and may damage some batteries.

If the batteries are at an identical state of charge, their terminal voltages will be the same, and so no appreciable current will flow when they are connected in parallel. Any imparity will be evened up as the one with greater charge and hence terminal voltage charges up the other.

It may be easier in the long run if you want to use two batteries with two motors to connect them separately. The negative terminals can be connected together, but the positives should then go to the positive connector of two separate speed controllers.

Batteries can be connected up in series-parallel configurations to get the required voltage and current capability, although it is recommended that no more than 4 in series should then be connected to a second chain of four:

When batteries are connected in parallel, they will not share current exactly. If one supplies a slightly higher current, then it will get warmer. As it gets warmer, it will become more efficient, and it will start to supply even more of the current. Hence there is the possibility that it will end up supplying most of the current while the other sits there getting charged up! A solution to this is to insert small value resistors in series with each battery before paralleling them up.

Remember, however, that Robot Wars has a voltage limit of 36 Volts DC!

13. Maximum current capability


All batteries are quoted for a maximum current source. Sometimes there is a continuous rating, and a pulse rating. The pulse times are often very short, so are not much use. The continuous rating is also normally quoted at 25C, and may have to be derated at higher temperatures. Remember, this current is the maximum that the wires, plates, and terminals in the battery can take before getting too hot. If the battery is already hot, then the power they can take will be less because of that.

If you have batteries in series, the maximum current you can take is equal to the maximum current of the least capable battery in the series chain (although as stated elsewhere in this page, all your batteries in series should be the same type anyway). If you have batteries in parallel, you can’t just double the maximum current capability, as explained in the series and parallel connection section, they will not current share very well.

14. Estimating Ah required for RW.

As we saw in section 7.2, the capacity of a battery depends very much on the ambient temperature and the actual load current being taken. At a fixed temperature and a fixed load current, we could quite accurately determine how large a battery we need. However, during the five minute robot wars bout, the load current is going to be anything but constant.

What we need to do is determine a typical load profile during the bout. This is going to need examination of several typical bouts, and timing of the periods spent stationary (but ignoring dead robots!), accelerating, pushing and shoving, and at constant speed. All these phases of the bout represent a different load on the battery. Some estimates are shown in the table. (NOTE: I would appreciate it if someone who has several videos of robot wars could give more accurate average figures for this table). The current values are based on using two 12 Volt, 750 Watt motors, which seems to be a fairly typical setup.

Phase in bout
Time (s)
Typical current draw (A)
Stationary
30
0
Accelerating
20
100
Stalled
10
300
Shoving
20
250
Constant speed
220
40

The graph below shows the effective derating of battery capacity for the Sonnenschein Dryfit A500 battery based on the load current.

Using this graph, we can work out how much capacity is taken out of the battery for each phase of the bout. The capacity used up is the actual ampere-hours, divided by the derating factor. For example, when accelerating, we assume a derating of 50%, and this goes on for 20 seconds. The actual ampere-hours taken in this time is 100 Amps x 0.0056 hours = 0.56Ah, but the battery is only operating at an efficiency of 50%, so this takes 0.56 ÷ 50% = 1.12 Ah off the nominal ampere-hour rating. The Capacity used up column is therefore given by the equation:

where the 3600 is to convert seconds into hours. Note that the percentage value is converted to a fractional value before being used in the equation!

Phase in bout
Time (s)
Current as a proportion of I20
Derating (%)
Capacity used up (Ah)
Stationary
30
0
100%
0
Accelerating
20
77
50%
1.12
Stalled
10
231
18%
4.63
Shoving
20
192
22%
6.31
Constant speed
220
31
70%
3.49
Total capacity used
15.55

Therefore, for this robot, we require a battery of at least 16 ampere-hours nominal capacity.

What assumptions have we made in this calculation which affect the result?

15. Which battery is best?

So here's the big question. There has been much debate over the years on the Robot Wars forum about which battery is the best to use. It has been fairly universally agreed that buying second batteries is a bad idea. Many people have done this and have ended up with poorly performing batteries. SLAs are easy to damage, and it appears that many second hand batteries have been treated badly.

As to which new battery to buy, there are pros and cons for all of them. The various opinions that are based on experience that I have seen on the forum are summarised in the table below:

Manufacturer

Model

Pros

Cons

Hawker Genesis Very high charging current allowed.
Good efficiency at high currents
Discontinued
Hawker SBS Expensive
Hawker Odyssey Expensive
Hawker Cyclone Expensive.
Small terminations
Kiel   Cheap Low power to weight ratio
High internal resistance
Steatite   Cheap  
Sonnenschein Dryfit Not enough information Not enough information
Gaston  
Panasonic  
Yuasa  

Andrew Marchant from Team Tornado has performed battery discharge tests on a variety of the common SLAs used by roboteers. The results can be found here.


15. Links

General

Lots of battery information
http://www.battery-index.com/

Battery FAQs
http://www.mywebplace.com/batterywholesale/batinfo.html
http://www.mywebplace.com/batterywholesale/batfaq.html

Basic stuff
http://www.jmclarke.co.uk/page47.html

sci.electronics battery FAQ
http://www.repairfaq.org/ELE/F_Battery_info.html
 

Ibex make battery chargers and gave lots of useful information on their site.
http://www.ibexmfg.com/index.htm

Steatite

Site
http://www.steatite.co.uk/defaultflash.html

24Ah SLA datasheet:
http://www.steatite.co.uk/batt/pdfs/rg12v-24.pdf
 

Sonnenschein

Dryfit datasheet
http://www.mhpower.com.au/Batteries/leadbatt.html

Panasonic

SLA batteries
http://www.panasonic.com/industrial/battery/
http://www.panasonic.com/industrial/battery/industrial/special/sealed_lead_acid/sealed_lead_acid.htm
 

Hawker

Genesis battery
http://www.hepi.com/genesis.htm

Cyclone battery
http://www.hepi.com/cyclone.htm

Literature list
http://www.hepi.com/lit.htm
 

A Supplier of Hawker batteries:
http://www.dmstech.co.uk

Yuasa

datasheet list
http://www.yuasa-exide.com/catalog.asp#5

NP24 datasheet
http://www.yuasa-exide.com/pdfs/NP24-12B.pdf
 

Unitrode (Nowpart of Texas Instruments)

Battery management ICs
http://www.ti.com/sc/docs/apps/analog/battery_management.html

Battery charging documents:
http://www.ti.com/sc/docs/apps/analog/chargers.html

Improved Charging Methods for Lead-Acid Batteries Using the UC3906:
http://www.ti.com/sc/docs/psheets/abstract/apps/slua115.htm

Battery chargers on the web

A 3 Amp SLA battery charger project, very well laid out and explained.
http://www.geocities.com/vk3em/sla-charger/sla-charger.html

High efficiency battery charger using power components.
http://vicr.com/support/apps-info/pdf/pb_battery-charger.pdf

A commercial battery charger manufacturer
http://www.statpower.com/prdlst3.htm

A commercial battery charger manufacturer. Not sure whether any of these are suitable for SLA batteries though, even though they say they are suitable for all lead acid batteries.
http://www.gunson.co.uk/acatalog/index.html

Linear Technology battery charger page, and a SLA charger circuit.
http://www.linear-tech.com/ltc/powersol/chargers.html
http://www.linear-tech.com/ltc/powersol/p09.pdf

SmartKit of Greece
http://www.smartkit.gr/kits/sk1095.htm

A dual rate charger (amongst other things)
http://www.geocities.com/CapeCanaveral/5322/

WICEN (Wireless Institute Civil Emergency Network) SLA battery charger circuit:
http://www.users.bigpond.com/lhoskin/power_&_battries.htm

Specialist battery suppliers

[C1] US supplier of many types of batteries. Limited technical information on the site.
http://batterymart.com