Will this work?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello all.

I currently run my bench off of two 12V 7Ah lead-acid batteries, but if I'm to use my latest amplifier designs, I might need help keeping them charged. I use the batteries because I don't have a transformer that will supply the required current.

So I thought of this.

The batteries are kept charged constantly, and the trafo is protected by the current limiters. Large transients will be pulled from the batteries, with a limited amount of help from the input transformer.

How will this do?

An externally hosted image should be here but it was not working when we last tested it.


Thank you,
- keantoken
 
I don't see how this could work. The MOS transistors will be reversed biased by the charger, and they will be equivalent to a 12ohm resistor in series with a diode. Is this your intention?

And if the reversal is just a mistake, it won't work either: such current sources only work with depletion devices, such as jFETs.
 
Elvee said:
I don't see how this could work. The MOS transistors will be reversed biased by the charger, and they will be equivalent to a 12ohm resistor in series with a diode. Is this your intention?

And if the reversal is just a mistake, it won't work either: such current sources only work with depletion devices, such as jFETs.

I realized my mistake after posting. At any rate, the base concept is to use a CCS as a current limiter so as not to ruin the trafo.

AndrewT said:
a 317 can give you a 1A CCS.

Are the peaks relatively short term?
This will determine the size of the heatsink on the 317s.


I have an LM7915C and an LM1815C, but no 12V regulators.

I could probably make a two-transistor CCS using 2N3055/2955, but the JFET current source is more efficient and concise (I will probably have to use the BJTs since I don't have any JFets).

So I'm assuming that since no one has pointed something out, it is okay for me to have the batteries constantly connected to power, so they'll be forever charged when I need to use them?

- keantoken
 
keantoken said:


I realized my mistake after posting. At any rate, the base concept is to use a CCS as a current limiter so as not to ruin the trafo.




I have an LM7915C and an LM1815C, but no 12V regulators.

I could probably make a two-transistor CCS using 2N3055/2955, but the JFET current source is more efficient and concise (I will probably have to use the BJTs since I don't have any JFets).

So I'm assuming that since no one has pointed something out, it is okay for me to have the batteries constantly connected to power, so they'll be forever charged when I need to use them?

- keantoken
A pair of 12V 5W bulbs would probably be sufficient as current regulators. They are simpler, cheaper, and more robust than any semiconductor regulator.
You have to make sure the open circuit voltage of your transformer + rectifier cannot rise above 14V, otherwise you risk damaging your batteries in the long term.
 
What about the impedance of the batteries? My reason for preferring the CCS variation is so that the trafo will be allowed to improve the line regulation, without high series impedance. Is this reasonable or do the batteries already have a fairly low series resistance?

Thank you,
- keantoken
 
Elvee said:

A pair of 12V 5W bulbs would probably be sufficient as current regulators. They are simpler, cheaper, and more robust than any semiconductor regulator.
You have to make sure the open circuit voltage of your transformer + rectifier cannot rise above 14V, otherwise you risk damaging your batteries in the long term.

Okay, next questions...

Will it damage my batteries if the voltage goes above 14V while charging, or does this limit just apply to float charge?

I think there is something wrong with my setup. At first while I was charging one battery, I found that the other battery was getting warm instead! Following all the wire, there was no logical way this could happen. Anyways, this resulted in the battery getting charged the wrong direction... I don't think it got completely discharged, but now the voltage across it goes to 15V when charging, and the lightbulb does not light up.

Thanks for your help,
- keantoken
 
If your charging input current is relatively low (ie under 5A) you can use a smiple Switchers like the LM22678. to provide voltage regulation. Instead of try to use a 13.8/-13.8 regulation you can use a 27.6V without using an common ground in the middle. There is a slight issue of charging imbalance if on of the cells has a abnormally high internal resistance, but that probably unlikely unless your batteries are aging or have sulfation buildup on the plates. Just keep your batteries fully hydrated by adding deionized\distilled water periodically, and don't excessfully draw down the batteries. If you want to maintain long battery life add a electronic low battery cut-off switch if the voltage drops too low.

http://electronicdesign.com/Articles/ArticleID/4423/4423.html

Using a simple switcher is far more efficient then using a linear regulator. The batteries won't be upset by the small amount of noise a simple switcher will create. If your concerned about output noise then add a Pi filter between the switcher output and the batteries. Simple switchers also have built-in current limiting.

Best of Luck to you!
 
I'm using sealed batteries, so adding water I don't think is an option.

I remember that you can desulfate batteries with high-current pulses. I imagine I would need more than a 160uF photoflash charged to 200V, if I ever needed to?

Thank you all for the information.
- keantoken
 
This IC might be of interest for your project:

http://cds.linear.com/docs/Datasheet/1479f.pdf

PowerPath Controller for Dual Battery Systems:

Complete Power Path Management for Two Batteries, DC Power Source, Charger and Backup

Compatible with Li-Ion, NiCd, NiMH and Lead-Acid Battery Chemistries

“3-Diode” Mode Ensures Powers is Available under “Cold Start” Conditions

All N-Channel Switching Reduces Power Losses

Capacitor and Battery Inrush Current Limited

“Seamless” Switching Between Power Sources n Independent Charging and Monitoring of Two Battery Packs

New, Small Footprint, 36-Lead SSOP Package
 
That is a good solution, methinks, but I had something different in mind.

What I want to do is have my batteries constantly on float charge, and never have to worry about taking apart my bench supply to charge them. I was thinking that if I couldn't buy a transformer with high enough current, I would still be able to test hungry circuits because the batteries would take care of the transients.

I don't like the idea of switching, because it's more complex, more expensive, and requires PCBs, which I can't make yet.

If the trafo puts out too much voltage at floating output, I can just make a shunt thing to pull down the voltage by drawing some current at excess voltages.

To limit current, I can simply use lightbulbs as suggested.

This is the transformer I am currently looking at:

http://www.radioshack.com/product/index.jsp?productId=2102703

So here is the schematic.
 

Attachments

  • benchsupply2.gif
    benchsupply2.gif
    13 KB · Views: 307
keantoken,

You need to do some research on charging SLA batteries. For descent battery life you will need at least a dual mode charger. It will also need pretty accurate voltage settings. Basically, what I am saying is that you need to specify the chargers requirements and then design the charger. The circuit in your last post won't meet the requirements you come up with.

Rick
 
Okay, out of curiosity, I though I would see how little components I could use to try and make a discrete charger like the one in the post above.

Charging current will lower with temperature, so I would suggest to use a fan and put Q2, Q3, and J1 somewhat away from power components and "upwind".

The only reason this shouldn't work is if I don't understand FETs properly. I'm assuming the FET turn-on voltage is around 2.5V, so the FET begins to turn off when charging current goes below somewhere near 30mA. This lowers the voltage.

According to something I read, SLA batteries have a 5% discharge per month. Since the float mode is triggered with current sensing, would we have to match the current exactly with the self-discharge current or would the battery self-compensate? I heard that FETs variate a lot between same devices, but I don't know whether this tolerance would have enough of an affect to cause problems.

I am buying the Antek 0515 transformer (15V, 50VA), and with my 8800uF caps, it should be enough for what I want I think, but I would still prefer to have the batteries to stiffen the supply.

If I were to simply limit voltage to 13.8V, would the batteries be fine?

- keatoken
 

Attachments

  • batterycharger.png
    batterycharger.png
    12.8 KB · Views: 227
keantoken,

Here's some feedback.

1. The article you posted pretty well tells you what you need to do to charge SLAs. The battery will never reach full charge with only 13.8 volts applied, and you still need to limit the current; especially if the battery voltage is very low when the charger is turned on.

2. I'm not an expert on batteries, but for batteries in the 5 - 10 A Hr range, NiCad may be the next best choice. I believe NiCads require 2 mode chargers for best performance (life).

3. If you don't want to use an automatic dual mode charger, you might consider a manual one, where you manually switch in and out a high current charge circuit that will be in parallel with the 13.8 volt supply.

4. You need to realize that batteries can put out very large currents under overload conditions and that could cause safety problems. Fuse the battery for safety.

5. Almost all JFETs (not MOSFETs) are depletion mode devices. This means that at Vgs equal zero the device current will be at the maximum (Idss). As you reverse bias the gate (wrt the source) the drain current will decrease (like a vacuum tube). (Thus the FET regulator you posted won't do what you want.)

6. All semiconductor devices can have wide parameter variations from device to device. (Check the data sheet on BJT current gain, or a JFET on Idss, or a MOSFET on Vt.) Any "open loop" circuit where it is desired to operate transistors in their active region will give disappointing performance. Therefore, feedback is almost always used. Note that a voltage follower circuit uses negative feedback. A common emitter circuit with a (bypassed) emitter resistor has negative feedback.

7. Precision analog circuits require feedback. Period

Rick
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.