Regulating battery to 9V

[Nicho]

Hi,

I'm looking to make a rechargable battery pack that can power a few 9V effects pedals. I was thinking of using four of these LiFePo4 batteries to get a nominal 13.2V and regulating it down to 9V. Can anyone advise me of the best way to do this in order to achieve a nice smooth output voltage, while minimising additional current draw?

 

[JMFahey]

If voltage comes from batteries, it is already nice and smooth.
Use 3 packs instead of 4 for nice nominal 9.9V.
Your pedals will be happy.
If for any reason you *insist* on 9V (your pedals will not complain, by the way), you can add a series diode to lower it to nominal 9.2V , or even 9V at higher current consumption.
Zero extra current consumption , which will not happen with any other regulating system.

 

[Nicho]

If voltage comes from batteries, it is already nice and smooth.

I wanted to keep it that way. I suppose any regulator will do this will it?

Use 3 packs instead of 4 for nice nominal 9.9V. Your pedals will be happy.

This was my first thought. Presumably the sound will be altered slightly? But I thought if they sounded better with 9.9V, great, if not I could adjust appropriate parameters to repair the damage. But while the batteries spend most of the time around 3.3V, I would probably operate them from about 3.6V fully charged to 3V cutoff, giving a 1.8V charge dependent voltage variation, which I wasn't sure would be ideal - or wouldn't it be worth bothering about?

If for any reason you *insist* on 9V (your pedals will not complain, by the way), you can add a series diode to lower it to nominal 9.2V , or even 9V at higher current consumption.
Zero extra current consumption , which will not happen with any other regulating system.

Since my main (unspecified) goal was to keep the voltage constant, I don't think just a diode would work for me. Another idea I had was to use an (as low a current as possible) LED in series with a zener as part of the regulation, which would be on until the battery voltage dropped below a certain level, informing me when the battery needed charging.

 

[Nicho]

Would this work?

 

[JMFahey]

Conceptually, yes.
The idea of using a MosFet is very good, because gate current is 0.
Practical problem is that most MosFets readily available need *around* 3.5VGS just to be turned on, "wasting" about what 1 cell supplies.

How much does a TL78xx need for itself?

Remember than the Zener itself needs a minimum current through it to "zene".

But keep throwing ideas, one of them might be the ticket.
Good luck.

 

[sofaspud]

If you're going to go in that direction, why not budget that current for a regular voltage regulator? Using an LM317 with ~1k for the reference voltage, the quiescent current through the load then need be only a few milliamps to meet rated specs.

 

[Nicho]

Thanks for the responses JMFahey and sofaspud. I could drop 3.5V accross the MosFet, and use a 50uA Zener, and LED to drop about 12.5 volts accross, or parallel 50uA Zeners if I need more current for the LED. I don't really mind wasting one cell of voltage - it means I need the fourth battery, but that would seem much better than a large current drain.

An adjustable regulator chip would also be a possibility - I will have to research them a bit. Is there any advantage to using these other than the ability to adjust the voltage (which might be nice)? Presumably the greater ripple rejection is not particularly important in the battery context.

My main concern with either form of regulation may or may not be related in some way to the term "transient response". Will there be a negative effect in this or in some related area.

 

[Nicho]

I don't know if that last paragraph makes sense. I assume that the regulator must respond ex post to changes in current demand, and imagine (without much knowledge) the possibility of the lagging behind slightly, cutting corners, overshooting, etc., while the unregulated battery simply has lots of little electrons jumping up and down just waiting to go. Is there any truth to this or not?

 

[counter culture]

You might be better to use a battery like these:- Buy UltraFire Protected 18650 3.7V 2400mAh Lithium Batteries (2-Pack Grey)

These are 'protected' cells. They have a little built-in circuit that means you can't overcharge or over discharge them. Quite a lot cheaper, although the capacity tends to be a bit exaggerated with these Chinese cells. You'll get 6 for the price of 2 from the German site though. Delivery takes ~14 days however.

They do run ~3.6 to 4.2V fully charged, so an LDO regulator (MIC2941 is a good one) might be advisable, because 3 fully charged will be 12.6, which might be a bit too much for your pedals. It'd probably be OK, but nobody wants to say that for sure and have you blow up your pedals. Dropout voltage on that reg. is only 0.6V @ 1.2A out.

You won't find any practical difference using a regulated supply, the reg responds so fast.

 

[KMossman]

Not to be rude, but you seem intent on making something quite easy, difficult.

JMFahey gave the best advice.

Most batteries do NOT have exhibit the same voltage through-out their discharge cycle. ONLY as an example, a '12'V lead-acid battery will vary from 14+V to 10+V.

Your stuff will be fine with 9V9. No worries at all.

The only wrinkle is adopting a battery charger to charge them.

 

[KMossman]

Since my main (unspecified) goal was to

a) keep the voltage constant, I don't think just a diode would work for me.

b) Another idea I had was to use an (as low a current as possible) LED in series with a zener as part of the regulation, which would be on until the battery voltage dropped below a certain level, informing me when the battery needed charging.

a) I do not understand your comment. The variance of Vd across the diode is trivial relative to the Vd of the batteries.

b) Perhaps you mean a LED in series with a resistor and zener across the battery? You will need at least 5mA thru it. I never go lower than 8mA to ensure the zener will take effect - some-one correct me if I am wrong! If I am right, then remove the zener and run 3 or 4 mA thru the LED - you will see it dim as the voltage drops: not precise but likely good enough.


You might to look at http://www.yesa.com.hk/pages.asp?id=19
for the voltages over the discharge cycle.

 

[counter culture]

3 of the German-sourced batteries will cost U$43 for a total capacity of 2500mAh.

6 of the dealextreme batteries will cost U$22 for a total capacity of 4800mAh, and there will be a spare set, so one set can be charged while the other is in use.

Seems to me you're oversimplifying a situation that you never took the time to understand fully.

Anybody looking at dealextreme will quickly discover many cheap chargers.

Funny how often people who intend to be rude prefix their remarks with something like 'not to be rude...'

 

[KMossman]

3 of the German-sourced batteries will cost U$43 for a total capacity of 2500mAh.

6 of the dealextreme batteries will cost U$22 for a total capacity of 4800mAh, and there will be a spare set, so one set can be charged while the other is in use.

Seems to me you're oversimplifying a situation that you never took the time to understand fully.

Anybody looking at dealextreme will quickly discover many cheap chargers.

Funny how often people who intend to be rude prefix their remarks with something like 'not to be rude...'

QUOTE=Nicho;3447507]Hi,

I'm looking to make a rechargable battery pack that can power a few 9V effects pedals. LiFePo4 batteries to get a nominal 13.2V and regulating it down to 9V. Can anyone advise me of the best way to do this in order to achieve a nice smooth output voltage, while minimising additional current draw?[/QUOTE]

Seems clear to me. Although the OP did later add requirments. And no, I did NOT intend to be rude. Just direct. He did say "nice smooth"; not stable, nor precisely X voltage. Whatever 'nice' means. Batteries do tend to give "nice" smooth power. Introducing components introduces noise and potential instability.

I did not see the OP about asking where to buy. Ask your doctor about Valium.

 

[Lumbermatt]

From my little experiments with LiFePO4 : the fully charged voltage for one cell is about 3.6V, you should always keep about 30% of charge inside not to damage them, which brings you around 3.2V at 70% of discharge. (those are generic numbers for this technology, i'd advise you to exactly check the datasheet of the cells you plan to buy).

With 3 cells in serial, you'll have from 10.8V to 9.6V as supply voltage for your pedal. Even if the voltage will drop rapidly after the first 5% of discharge, you will have an initial (full charged) situation with 10.8V. Can you device handle this (+20%) ?

I would advise you to put a VLDO regulator with a maximal 600mV dropout indeed.

 

[Nicho]

Not to be rude, but you seem intent on making something quite easy, difficult.

This is true.

Most batteries do NOT have exhibit the same voltage through-out their discharge cycle. ONLY as an example, a '12'V lead-acid battery will vary from 14+V to 10+V.

This is why I don't want to use LED acid. 14V would be over some component tolerances.

Your stuff will be fine with 9V9. No worries at all.

I'm not worried about blowing them up, but I thought when the batteries were fully charged the 10.8V might alter the "headroom" of the pedals more than I wanted. Not a big deal at all, but adding another battery and regulating didn't seem too distressing a thought either. I suppose I should be balencing the batteries as well, so maybe I should work out how to do that first.

The only wrinkle is adopting a battery charger to charge them.

I will be working on that next. I have the recommended charge parameters, so it shouldn't be too bad. I'll be posting on here if I get stuck.


Any thoughts on my last question about the effects of regulation would be greatly appreciated.

 

[Nicho]

a) I do not understand your comment. The variance of Vd across the diode is trivial relative to the Vd of the batteries.

A diode would drop voltage but not regulate it - good if I want to drop a constant amount of voltage, but not if I want to hold voltage to my pedals at a fixed level.

b) Perhaps you mean a LED in series with a resistor and zener across the battery? You will need at least 5mA thru it. I never go lower than 8mA to ensure the zener will take effect - some-one correct me if I am wrong! If I am right, then remove the zener and run 3 or 4 mA thru the LED - you will see it dim as the voltage drops: not precise but likely good enough.

Did you see the schematic I posted? You can get very low current zeners now (50uA). I was thinking of a zener LED combination summing to the required voltage. When the battery was over this voltage, the zener should conduct and the LED will be off. When the batteries voltage drops below the voltage the LED will go off -> recharge time. There would be a restistor between the zener and the battery to "set" the current.

You might to look at http://www.yesa.com.hk/pages.asp?id=19
for the voltages over the discharge cycle.

I have the datasheet for the products, which was what I was going by

 

[sofaspud]

In general, transient response is best dealt with at the load. When using a battery, the main advantages of regulation is likely to be reducing the supply voltage to a stable and usable level, at a cost of some power "wasted." If using a battery only, you can still include a capacitor across the + and - to improve the response. A capacitor may well have a lower impedance than a battery and thus be able to provide energy quickly when the load demands it.

edit: I don't quite understand your circuit. What is the purpose of the LED? If you plan on 50uA through the zener that isn't enough to light an LED.

 

[Nicho]

You might be better to use a battery like these:- Buy UltraFire Protected 18650 3.7V 2400mAh Lithium Batteries (2-Pack Grey)

These are 'protected' cells. They have a little built-in circuit that means you can't overcharge or over discharge them. Quite a lot cheaper, although the capacity tends to be a bit exaggerated with these Chinese cells. You'll get 6 for the price of 2 from the German site though.

Thanks for the info. I was wanting to have a go at making a battery pack and charger from parts. I'm expect it is something I will only want to do once. But I am pretty new to electronics, and it seemed like a good learning experience. I'm still liking the cells from Germany (made in China also I think). The solder tags are nice and I have the datasheets and design manual for them. I'm also expecting them to be pretty consistant and reliable.

They do run ~3.6 to 4.2V fully charged, so an LDO regulator (MIC2941 is a good one) might be advisable, because 3 fully charged will be 12.6, which might be a bit too much for your pedals. It'd probably be OK, but nobody wants to say that for sure and have you blow up your pedals. Dropout voltage on that reg. is only 0.6V @ 1.2A out.

12.6V is probably just OK, but from what I have read, if they are LiFePo4 they would probably be happy only being charged to 3.6V and not lose much capacity.

 

[Nicho]

In general, transient response is best dealt with at the load. When using a battery, the main advantages of regulation is likely to be reducing the supply voltage to a stable and usable level, at a cost of some power "wasted." If using a battery only, you can still include a capacitor across the + and - to improve the response. A capacitor may well have a lower impedance than a battery and thus be able to provide energy quickly when the load demands it.

LiFePo4 batteries are supposed to have very low internal impedance. The data sheet for these ones says 6 milliohms at 1kHz.

edit: I don't quite understand your circuit. What is the purpose of the LED? If you plan on 50uA through the zener that isn't enough to light an LED.

The LED indicates when battery voltage is dropping and recharging is required. How much is required to light an LED? I saw somewhere that 250uA would work. If it was not too much I could put a few of the zeners in parallel.

 

[Nicho]

I would advise you to put a VLDO regulator with a maximal 600mV dropout indeed.

Hi Lumbermatt. Thanks for the suggestion. Would this one be suitable. It has a max 18V input, 300mV dropout, and an adjustable or fixed output. That would enable me to regulate while using only 3 cells. I would just need to work out another way of giving myself an indicator of when it was time to recharge.