Can i use 2 of these PSU to make +- 24V?

[neutron7]

I found these 24v 6A SMPS for cheap

http://www.mpja.com/email/04-14-09.a...851&p=17808+bx
about half way down

"24V @ 6.5A POTRANS POWER SUPPLY"

is it safe to connect the positive side of one to the "ground" of the other to get -24 0 +24 for some chipamps?

[AndrewT]

I see three problems.
1.) the 6.5A limit is too low for many 8ohm speakers if you want the full 60W that a 3886 can supply. 4ohm and 6ohm speakers require even more current.
2.)the units may or may not need to be earthed.
3.)the units need to be cascaded in series to generate the 0-24, 24-48Vdc. The centre 24Vdc then becomes the PSU zero volts giving you the -24-0+24Vdc dual polarity supply. If 2.) requires earthing then this cascaded arrangement is impossible.

[neutron7]

I dont really mind about not getting the full wattage, but the other problems make it look like something i better not mess about with.

[infinia]

I think it should work OK for a couple of chip-amps, with both PS outputs tied in series. The PS data sheet (see link) shows both inputs and outputs are floating from earth ie FG - Frame Gnd. The chip-amp inputs and outputs should be designed as floating from earth ground as well. Current limit is sensed at primary and might be adjustable, but should be tested first. Input/output power can vary from 110-200% so mostly factory targeted at 130%. Not sure what would happen to the amps if one of the PS goes into overload limit, but not to be pretty, since the PS goes into hiccup mode ie periodic cycling on and off. I would carefully test this tho at fullpower before running good speakers.

http://www.circuitspecialists.com/pr...I-15024-1M.pdf

[infinia]

Originally posted by AndrewT
I see three problems.
1.) the 6.5A limit is too low for many 8ohm speakers if you want the full 60W that a 3886 can supply. 4ohm and 6ohm speakers require even more current.

The output of chip amp using +/- 24 Vdc can provide around 21 Vp max without clipping. It's reasonable to have 8 ohm speakers with say a low imp. of 6.4-6.0 ohms (Re). Then calculate Ip = Vp/Re or close to 7.0 amps peak for 2 channels. According to the PS spec Io = 6.50A nominal or 7.15 A max at the minimum current limit set point. Most likely the current limit is a bit higher than 110%, which a quick test would verify.

So I think it would be OK to use 2 of these PS for a stereo 60W chip-amp at maxpower using 8 ohms speakers. As AndrewT pointed out, it would not be recommended or wise to use 4 ohm speakers.

[AndrewT]

Infinia,
your current advice is misleading.
You have shown the maximum peak current into a resistive load assumed to be equal to the Re of the speaker/driver.

A reactive speaker load does not work like this.

The peak current can approach and some severe reactance speakers can exceed, max current ~ Vpk / Nominal impedance / 0.35.
For a 30Vpk (56W into 8r0) the resistive current is 3.75Apk but the reactive load transient current can be 10.7Apk. A cold lm3886 struggles to meet this demand.
A fixed output 6.5A SMPS cannot meet this demand for one channel.
A simple transformer+rectifier+2smoothing caps PSU can easily be designed to meet double this load current demand for a two channel amplifier.

A +-24Vdc PSU supplying two channels of lm3886 (28W+28W into 8r0) must be able to meet a peak transient demand of at least 15Apk for a pair of 8ohm speakers.

[infinia]

Originally posted by AndrewT
Infinia,
your current advice is misleading.
You have shown the maximum peak current into a resistive load assumed to be equal to the Re of the speaker/driver.

A reactive speaker load does not work like this.

I don't know how you get 15A ? Is this short term and due to what since the amp holds V at a constant.
dont know but all of my Unibox models show max peak current occuring at Re either ported or closed?
OK I can post some real plots for data using a 10" subwoofer with highly reactive motor with a ported enclosure with Po=40Watts.

[infinia]

In addition> Here the in box impedance for the Ipk plot shown above.

[AndrewT]

Originally posted by infinia
I don't know how you get 15A ? Is this short term

it is very much short term.
It is a transient peak on fast starting and/or fast stopping transients.

I don't get and I am incapable of measuring these transient output current demands.

I read technical reports and I am usually able to sift the wheat from the chaff. Subjective reporting simply tells us if they can hear what the technical reporting has said may be possible.

What I do know is that using a resistive load to model the behaviour of an amplifier is almost useless.

BTW,
what is Re of that peerless you posted data for?

[infinia]

Originally posted by AndrewT
it is very much short term.
It is a transient peak on fast starting and/or fast stopping transients.

I don't get and I am incapable of measuring these transient output current demands.

I read technical reports and I am usually able to sift the wheat from the chaff. Subjective reporting simply tells us if they can hear what the technical reporting has said may be possible.

What I do know is that using a resistive load to model the behaviour of an amplifier is almost useless.

Andrew
Using Re works better than 8R0 for a first order approx.
If it's short term then the SMPS current limit loops are rather slow to react and would have to be supplied by the raw output C.
Ipeak that large implies a rather largish capacitive load any reports that you could share? Would help TIA