Battery-powered supplies

[protos]

Originally posted by Verbal
Would need a symmetric supply...Perhaps like this?

This one from page 4 of this thread.

[ALW]

Been meaning to do this for a while, and the oppportunity arose recently whilst doing some other stuff.

The plot below is the noise of a sealed lead-acid gel-cell (Yuasa NP series) under no-load and resistive load conditions. For comparison there's a discrete regulated supply. The resistive load represent about a 35mA current drain - it gets worse with increasing load current. Even more inportant the noise doesn't stop when a heavy load is removed, it takes seconds to settle, presumably due to the chemical reaction going on inside the cell.

Bear in mind these are resistive load conditions i.e. easy, when placed under a dynamic load (plots to come at a later date) things look much worse.

Particularly bad is that the voltage output bears no relation to A.C. load, and produces other noise products that are not remotely related to the load it sees.

Adding a 22,000u cap (not shown) gets the noise above the audio band down to the regulator levels.

As you can see the battery is only quieter under no-load conditions - not a great deal of use!

I'll do plots to compare to 3-term reg's later, which at low currents look cleaner, but noisier, the noise is inocuous though as they display far better linearity.

[fdegrove]

Hi,

Which is why I always say tht it's good practice to decouple the batteries with a fair amount of caps.

If not you'll get "Assault and Battery".

Guess, it helps when people see what they listen to?

[Steve Eddy]

Originally posted by ALW
Been meaning to do this for a while, and the oppportunity arose recently whilst doing some other stuff.

Well, it's about damn time!

Thanks.

Before commenting further, a couple of questions.

What's the reference level for the relative amplitude?

Are you applying any averaging? If so, how much?

se

[RichardJones]

Hi,

I am in the process of building a 36v (3x12V SLA) power supply bypassed by 68Kuf to 136Kuf of capacitance per channel.

Going to make a float / trickle charger at 40v that I just plug in and charge... the amp will then be running of the charger and batteries at 40v not 36v... which is NOT a problem.

Can I make a simple circuit with a tranny and bridge rectifier?? Or do I need a couple of voltage regulators in there?

If so, what voltage should my tranny be? What current rating should the recitifier be?

Thanks!!!

p.s. how long should I run my amp at 1A from 7Ah batteries before I recharge them?

[mikelm]

>I am in the process of building a 36v (3x12V SLA) power supply bypassed by 68Kuf to 136Kuf of capacitance per channel.

Well get ready for some major arcing & sparking when you connect the batteries to caps of that size...

my caps are 10,000uf and that is quite wild enough. unless you use a soft start resistor you might end up arc welding the wires to the battery or cap terminals ! The initial current surge could well be 100 amps

If you are not in balanced working how are you planning to organise your speaker return if you are using three batteries ?
For non balanced working 4 batteries would be simpler as it gives you a zero volts centre point. In this case I suspect that 2x good quality 1000uF caps will be more ideal than the bigger ones. Otherwise you will need to organise a low impedance earth return in which case the big caps may be necessary.

>Going to make a float / trickle charger at 40v that I just plug in and charge... the amp will then be running of the charger and batteries at 40v not 36v... which is NOT a problem.

ideally each 12V cell should float at about 13.5 or a little more so 40 is about right. obviously in theory the amp should sound better when the battery is off charge but to be honest I can hardly tell the difference.

>Can I make a simple circuit with a tranny and bridge rectifier?? Or do I need a couple of voltage regulators in there?

I am quite happy with tranny & bridge, seems to work well

>If so, what voltage should my tranny be?

9 volts AC per 12V battery. ( This multipies up by 1.414 to DC and another 10% for typical tranny regulation minus diode drops = about 13.5V )

>What current rating should the recitifier be?

15 amps should be OK

>p.s. how long should I run my amp at 1A from 7Ah batteries before I recharge them

Well 7 hours I guess... ...but for best battery life charge them all the time you are not listening.


hope some of this is useful

mike

[Alex S]

ALW,
the plots don't look right... how come the 50Hz spikes on the battery plots are higher than the (regulated) AC supply? Are you sure the BUT (battery under test ) is not still connected to it's AC charger and that you were measuring the noise of the charger?

[Verbal]

Originally posted by Alex S
ALW,
the plots don't look right... how come the 50Hz spikes on the battery plots are higher than the (regulated) AC supply? Are you sure the BUT (battery under test ) is not still connected to it's AC charger and that you were measuring the noise of the charger?

In any case a battery won't produce a 50Hz hum...(nor all those harmonics)

[ppl]

AWL could you do similer comparrisons but using NI-MH Cells. This would be somthing to consider in low current line level stages as apposed to gain card Clones. I wish reasonable NI-MH Power was available.

[peranders]

ALW is maybe proving something but in the same time the complete circuit isn't defined so you aren't really sure what you are seeing. This hum pickup is a bit disturbing and the noise from the battery may come from switched PS or other sources, so take the results with a grain a salt.