Rugged transformer based 2A charger for 12V lead acid battery

[Vindhyachal.takniki]

1. We are facing a issue, our SMPS based 2A charger for lead acid charger gets burnt sometimes when they run on old type of generators. However on ac mains, we have tested upto 310Vac they work fine.

2. So we want to make it transformer based. Although there are lots of circuits available on internet for 2Amps for 12V lead acid charger like this
https://dmohankumar.wordpress.com/2011/08/03/4653/


Want to know if someone has experience in designing transformer based chargers for lead acid chargers, for suggestions or references, which we can keep in mind before designing or there is dedicated IC for this purpose

[Zero999]

1. We are facing a issue, our SMPS based 2A charger for lead acid charger gets burnt sometimes when they run on old type of generators. However on ac mains, we have tested upto 310Vac they work fine.

2. So we want to make it transformer based. Although there are lots of circuits available on internet for 2Amps for 12V lead acid charger like this
https://dmohankumar.wordpress.com/2011/08/03/4653/


Want to know if someone has experience in designing transformer based chargers for lead acid chargers, for suggestions or references, which we can keep in mind before designing or there is dedicated IC for this purpose

Why do you think that's going to be any more robust than a switched mode charger?

Did you find out why the SMPS charger failed?

[David Hess]

I assume you are asking about something more sophisticated than a charger for a automotive starting battery.  The inexpensive ones use a custom current limiting transformer making them very robust.

There is nothing to preclude using line side transformer isolation with a switching regulator in lieu of a linear regulator to keep power dissipation down.

There used to be some pretty nice lead acid battery charger ASICs but the selection is smaller now.  TI and Linear have some.  There are a lot of good discrete designs published.

AC generators can be hard on off-line switching power supplies which were designed for relatively placid power lines.  Derating the design for line side problems in excess of what would be expected from a power line would solve this.  There are some safety features which would improve reliability.

[Damianos]

The problem is most probably the unstable output from the generators. This stresses the input circuits of switching power supplies/chargers, especially the bridge rectifier and "bulk" capacitors. When you test the charger with higher, but stable, voltage they behave well because of the smoother operation...
The above will affect also a linear circuit, with less problems.

If you study the datasheet of your battery (or an equivalent one), then you may find that the article you linked is not very relevant with the subject.

Taking information from manufacturer(s) you can decide of what charging mode you need/want.
Tip: the simpler to implement is a float charger. Next is the cycle use charger and the more complicated is the "topping-up" with load connected.
Bring the information and the specifications here to discuss it more specifically.

[Zero999]

The problem is most probably the unstable output from the generators. This stresses the input circuits of switching power supplies/chargers, especially the bridge rectifier and "bulk" capacitors. When you test the charger with higher, but stable, voltage they behave well because of the smoother operation...
The above will affect also a linear circuit, with less problems.

I don't think it will be any easier on a linear PSU. The bulk capacitors still have a finite voltage rating and the transformer might blow, especially if the generator decides to output a low frequency at a high voltage.

[james_s]

My own experience suggests that linear power supplies are considerably more robust than the average run of the mill switchmode power supply. I don't personally ever recall seeing a transformer fail on a linear supply, but there are some less than perfectly engineered SMPS's out there that will blow up if you look at them wrong. It's a lot harder to screw up the design of a linear PSU.

[Damianos]

The problem is most probably the unstable output from the generators. This stresses the input circuits of switching power supplies/chargers, especially the bridge rectifier and "bulk" capacitors. When you test the charger with higher, but stable, voltage they behave well because of the smoother operation...
The above will affect also a linear circuit, with less problems.

I don't think it will be any easier on a linear PSU. The bulk capacitors still have a finite voltage rating and the transformer might blow, especially if the generator decides to output a low frequency at a high voltage.

I don't refer to voltages but currents, think about the difference having a transformer at the input, especially if it is an inductive protected type.
If the case of low frequency with high voltage is probable, then think again the currents... On the other hand, a transformer will protect the rest of the circuit.
A switching circuit, as we know, will have much higher efficiency, but needs more complex protection, if it is to be used with a "bad" power source.

[David Hess]

A switching circuit, as we know, will have much higher efficiency, but needs more complex protection, if it is to be used with a "bad" power source.

There is nothing to preclude using a switching circuit with an input transformer for high efficiency.  Even power factor correction can be used to get the most out of the transformer.

2 amps is pretty low making the added weight from an input transformer reasonable.  For a pure switching design, I might use a robust power factor correction input stage so it can handle very dirty power.  There are a lot of parts intended for use with a 480 volt AC input which could be used for higher voltage derating.

[oldway]

For this voltage (12V) and this current (2A), the best solution would be to use a special transformer with high short circuit impedance (15 to 25%) with low induction (8000 Gauss, 0.8T for example, to allow a large overvoltage without saturating) with center point secondary and two thyristors to phase control the voltage and the charging current.

Low cost scr's like TIC116 can be used.

[Damianos]

A switching circuit, as we know, will have much higher efficiency, but needs more complex protection, if it is to be used with a "bad" power source.

There is nothing to preclude using a switching circuit with an input transformer for high efficiency.  Even power factor correction can be used to get the most out of the transformer.

2 amps is pretty low making the added weight from an input transformer reasonable.  For a pure switching design, I might use a robust power factor correction input stage so it can handle very dirty power.  There are a lot of parts intended for use with a 480 volt AC input which could be used for higher voltage derating.

Am I saying something else? Maybe I have misunderstood your statement.

[David Hess]

A switching circuit, as we know, will have much higher efficiency, but needs more complex protection, if it is to be used with a "bad" power source.

There is nothing to preclude using a switching circuit with an input transformer for high efficiency.  Even power factor correction can be used to get the most out of the transformer.

Am I saying something else? Maybe I have misunderstood your statement.

The implication was that using the input transformer associated with a linear regulator results in low efficiency.  There is nothing to preclude using an input transformer with a switching regulator for high efficiency.

[oldway]

For fun, here is a prototype 12V 25A automatic battery charger with thyristors I made more than 25 years ago ....

He suffered the worst possible abuse during 25 years, he even remained under muddy water during a flood.

Still perfect working ..... but need a serious restauration     

It is protected against short-circuits and reverse polarity.
The charging current is adjustable.
Display of voltage and current by led's bargraph.

[Zero999]

The problem is most probably the unstable output from the generators. This stresses the input circuits of switching power supplies/chargers, especially the bridge rectifier and "bulk" capacitors. When you test the charger with higher, but stable, voltage they behave well because of the smoother operation...
The above will affect also a linear circuit, with less problems.

I don't think it will be any easier on a linear PSU. The bulk capacitors still have a finite voltage rating and the transformer might blow, especially if the generator decides to output a low frequency at a high voltage.

I don't refer to voltages but currents, think about the difference having a transformer at the input, especially if it is an inductive protected type.
If the case of low frequency with high voltage is probable, then think again the currents... On the other hand, a transformer will protect the rest of the circuit.
A switching circuit, as we know, will have much higher efficiency, but needs more complex protection, if it is to be used with a "bad" power source.

Sounds crazy but how about using the existing SMPS, with something to limit the current, such as a large incandescent lamp or an inductor, such as a fluorescent tube ballast?

How about just using an isolation transformer to run the SMPS?

[james_s]

Assuming the output of the SMPS is DC an inductor isn't going to do much. Certainly you don't want to limit the input current, that's a good way to blow up some types of SMPS.

[David Hess]

For fun, here is a prototype 12V 25A automatic battery charger with thyristors I made more than 25 years ago ....

For robustness and reliability, a thyristor based charger is a great way to go if the extra weight of the transformer and inductor is acceptable.  The only thing tougher would be a saturable reactor design.

I suspect though that a voltage derated high frequency off-line switcher could be made tough enough with the right set of protection features like cycle-by-cycle current limiting and over-voltage shutdown.  The right kind of active power factor correction can get rid of the surge current into the input capacitor providing the same benefit that Hero999 suggests with an input ballast of some kind.

[Zero999]

Assuming the output of the SMPS is DC an inductor isn't going to do much. Certainly you don't want to limit the input current, that's a good way to blow up some types of SMPS.

I know that! I wasn't talking about adding an inductor to the DC, but the AC side. The battery charger's internal circuitry should limit the DC current anyway and interfering with that would be a bad idea..

[Damianos]

A switching circuit, as we know, will have much higher efficiency, but needs more complex protection, if it is to be used with a "bad" power source.

There is nothing to preclude using a switching circuit with an input transformer for high efficiency.  Even power factor correction can be used to get the most out of the transformer.

Am I saying something else? Maybe I have misunderstood your statement.

The implication was that using the input transformer associated with a linear regulator results in low efficiency.  There is nothing to preclude using an input transformer with a switching regulator for high efficiency.

Are you answering to yourself? Where I said that?

[Damianos]

The problem is most probably the unstable output from the generators. This stresses the input circuits of switching power supplies/chargers, especially the bridge rectifier and "bulk" capacitors. When you test the charger with higher, but stable, voltage they behave well because of the smoother operation...
The above will affect also a linear circuit, with less problems.

I don't think it will be any easier on a linear PSU. The bulk capacitors still have a finite voltage rating and the transformer might blow, especially if the generator decides to output a low frequency at a high voltage.

I don't refer to voltages but currents, think about the difference having a transformer at the input, especially if it is an inductive protected type.
If the case of low frequency with high voltage is probable, then think again the currents... On the other hand, a transformer will protect the rest of the circuit.
A switching circuit, as we know, will have much higher efficiency, but needs more complex protection, if it is to be used with a "bad" power source.

Sounds crazy but how about using the existing SMPS, with something to limit the current, such as a large incandescent lamp or an inductor, such as a fluorescent tube ballast?

How about just using an isolation transformer to run the SMPS?

For an SMPS it is not so simple. When the input voltage lowers, it will try to overcome this by changing the drive of the transistor(s) and, as a consequence, discharges the capacitors deeper. When the voltage increases again, the charging current... Also most capacitors are not made for this type of operation, there are special types for often discharge and recharge. On the other hand a linear regulator, when the input voltage is low, simply gives less power to the load...
All these according to my assumption that the problem derives from the variation of the generator voltage.

[Zero999]

The problem is most probably the unstable output from the generators. This stresses the input circuits of switching power supplies/chargers, especially the bridge rectifier and "bulk" capacitors. When you test the charger with higher, but stable, voltage they behave well because of the smoother operation...
The above will affect also a linear circuit, with less problems.

I don't think it will be any easier on a linear PSU. The bulk capacitors still have a finite voltage rating and the transformer might blow, especially if the generator decides to output a low frequency at a high voltage.

I don't refer to voltages but currents, think about the difference having a transformer at the input, especially if it is an inductive protected type.
If the case of low frequency with high voltage is probable, then think again the currents... On the other hand, a transformer will protect the rest of the circuit.
A switching circuit, as we know, will have much higher efficiency, but needs more complex protection, if it is to be used with a "bad" power source.

Sounds crazy but how about using the existing SMPS, with something to limit the current, such as a large incandescent lamp or an inductor, such as a fluorescent tube ballast?

How about just using an isolation transformer to run the SMPS?

For an SMPS it is not so simple. When the input voltage lowers, it will try to overcome this by changing the drive of the transistor(s) and, as a consequence, discharges the capacitors deeper. When the voltage increases again, the charging current... Also most capacitors are not made for this type of operation, there are special types for often discharge and recharge. On the other hand a linear regulator, when the input voltage is low, simply gives less power to the load...
All these according to my assumption that the problem derives from the variation of the generator voltage.

Yes, I am aware of the fact that in an SMPS, the current draw will increase, if the voltage drops due to the source impedance being higher.

The reason for suggesting a higher impedance supply, was to limit the ripple current in the DC smoothing capacitors and surge currents in the rectifier. Whether it will have any effect or not, depends on if it has active power factor correction, in which case it will make no difference. Another factor is the SMPS's minimum working voltage. If the increased load impedance, causes the voltage to drop below the SMPS's minimum specified value, the transistors may be more stressed because they will have to turn on harder.