Need a stable 5V 7A Power Supply

[cdxer]

I am trying to power 230x Laser diodes at 30mA each.
Applying 5V to each would accomplish this.
However, I've converted two PC Power Supplies to
general purpose power supplies and neither of them
were able to support the required current.
They would drop from 5V to maybe 3.5V at around 1Amp.

I've even ordered a power supply from eBay that was rated at 5V 40A.
It had an adjustable voltage of -0.5V to +0.5V.
At 5.5V, it would drop to 4.5V and only 1A trying to power the load.

Is this an expected behavior even from good power supplies?
Or am I getting ripped off constantly?
Is there a better way to power this load?

 

[Jony130]

Do you use any current limiting resistor in series with the laser diode ?

 

[cdxer]

No, the current is coming straight from the power supplies.
The problem is not current protection but lack or current because every power supply I use have a massive voltage drop on a heavy load.

 

[pantaz]

Easier to diagnose if we could see the driver circuit you're powering.

 

[berkeman]

I am trying to power 230x Laser diodes at 30mA each.
Applying 5V to each would accomplish this.
However, I've converted two PC Power Supplies to
general purpose power supplies and neither of them
were able to support the required current.
They would drop from 5V to maybe 3.5V at around 1Amp.

I've even ordered a power supply from eBay that was rated at 5V 40A.
It had an adjustable voltage of -0.5V to +0.5V.
At 5.5V, it would drop to 4.5V and only 1A trying to power the load.

Is this an expected behavior even from good power supplies?
Or am I getting ripped off constantly?
Is there a better way to power this load?

No, the current is coming straight from the power supplies.
The problem is not current protection but lack or current because every power supply I use have a massive voltage drop on a heavy load.

Sounds like the supplies may be in current limit. They should be able to put out their rated current at their rated output voltage. But if you put too heavy a load "R" on the supply (so that I = V/R would exceed the rated output current), the power supply will go into current limit and drop the output voltage and current.

Try putting a resistive load on the output of the supply that matches the rated output current and voltage (this will require a large power resistor, or lots of smaller rated resistors). Do the supplies behave as expected at their maximum rated output?

 

[cdxer]

Sounds like the supplies may be in current limit. They should be able to put out their rated current at their rated output voltage. But if you put too heavy a load "R" on the supply (so that I = V/R would exceed the rated output current), the power supply will go into current limit and drop the output voltage and current.

Try putting a resistive load on the output of the supply that matches the rated output current and voltage (this will require a large power resistor, or lots of smaller rated resistors). Do the supplies behave as expected at their maximum rated output?

The eBay power supply is described as 5V 70A 200W, adjustable from 4.5V to 5.5V.
With a 0.33Ohm resistor, the voltage drops to about 4V from 5.5V adjustment.
Simple case of false advertisement?

 

[berkeman]

The eBay power supply is described as 5V 70A 200W, adjustable from 4.5V to 5.5V.
With a 0.33Ohm resistor, the voltage drops to about 4V from 5.5V adjustment.
Simple case of false advertisement?

Hmm. 5V * 70A = 350W, so it is weird that they only quote 200W. Does sound fishy...

 

[vk6kro]

The eBay power supply is described as 5V 70A 200W, adjustable from 4.5V to 5.5V.
With a 0.33Ohm resistor, the voltage drops to about 4V from 5.5V adjustment.
Simple case of false advertisement?

Assuming you want to use this power supply, you can work out how to do it from the measurements you gave.

The power supply gives 5.5 volts off load and 4 volts at 12.12 amps. (4 volts / 0.33 ohms =12.12 amps)

So, it has an internal resistance of (5.5 volts - 4 volts ) / 12.12 amps or 0.123 ohms.

So, if you did take 7 amps from it, it would give 5.5 volts - ( 7amps * 0.123 ohms) or about 4.6 volts.

So, each diode would require a resistor of (4.6 volts minus the rated voltage of the diodes) / 0.03 amps in series with it.

If the diode voltage was supposed to be 3.5 volts, the resistor value would be about 37 ohms. You can buy 39 ohm resistors.
So, you would need 230 of these 39 ohm resistors.

Note that you would have to substitute the actual rated value of the diode voltage in the above calculation.

Unless this rated voltage is 5 volts, you should not put the diodes across the supply without a suitable series resistor.

Are these really LASER diodes, though? There are safety problems with laser diodes, but not with LEDs.
Do you know about the safety aspects of laser diodes?

 

[cdxer]

Sorry, the supply is actually 4.5V to 5.5V at 40A, 200W. That's my mistake.
However, it doesn't come close to 40A.

These laser diodes are rated at 5V 30mA.
The reason I am not using a 12V is to avoid having 230 resistors.
The supply is barely able to power even 30x diodes at a time without significant voltage drop.
Yes I do know about thermal runaway conditions on laser diodes: heat -> more current -> heat -> more current.
I decided to skip on the driver circuit(s) and I understand the risks.

Is there a better way to do this?
At the same time I can't have 230x resistors or driver circuits.

 

[vk6kro]

Sorry, the supply is actually 4.5V to 5.5V at 40A, 200W. That's my mistake.
However, it doesn't come close to 40A.

These laser diodes are rated at 5V 30mA.
The reason I am not using a 12V is to avoid having 230 resistors.
The supply is barely able to power even 30x diodes at a time without significant voltage drop.
Yes I do know about thermal runaway conditions on laser diodes: heat -> more current -> heat -> more current.
I decided to skip on the driver circuit(s) and I understand the risks.

Is there a better way to do this?
At the same time I can't have 230x resistors or driver circuits.

The risk is that laser diodes can damage eyesight. This is a serious problem. With so many of them, you would need to take precautions against this.

If your diodes are able to stand 5 volts without any resistors, then they may work OK at 4.6 volts which is about what you would get with 7 Amps load on your power supply. Try one on 4.6 volts and check.

You wouldn't need any resistors if you can run the diodes directly from the power supply.

If you did use 12 volts, you could put two diodes in series with a resistor, so you would only need 115 resistors. Still a lot, but resistors are cheap.

Incidentally, the power supply on my computer has a label that says it can deliver 30 amps at 5 volts.
I know I have to join the green wire to one of the black wires to make it work, but this would be a good alternative. I don't know why your computer power supply wasn't able to do this.

 

[Carl Pugh]

Are you measuring the power supply voltage at the power supply or at the load?

At high current there can be considerable volt drop in the leads.

 

[cdxer]

Are you measuring the power supply voltage at the power supply or at the load?

At high current there can be considerable volt drop in the leads.

That won't make a difference since they are the same nodes.
I've come to believe that it's simply an advertising mechanic.
It is capable or 5V, or 40A, but not at the same time.
Just like how disk space is sold by gigs division by 1000 instead of 1024.

I think my solution would be to find a 7V supply and expect it to drop into range,
but that is not a typical number for power supplies.
]

 

[vk6kro]

Legitimate power supplies would not behave like that.

If the supply is old, it may have dried-up electrolytic capacitors in it. These would be unable to hold a voltage and the output would have a lot of ripple on it, but with a meter you would just see a drop in voltage.

If it was rated at 40 amps, it probably did deliver 40 amps at one stage. This is always AT the stated voltage.

Ebay sellers are very nervous about their star rating, especially if they are regular sellers. So, all you have to do is complain to the seller and, if that doesn't work, complain to Ebay.
On rare occasions, I have complained about something that didn't work and I always got a refund or a replacement.

You can check the ripple on the output by putting a capacitor in series with an AC voltmeter and reading the voltage. It should be less than a volt under load.
A better way is to observe the output with a oscilloscope.
You could easily replace the electrolytics if the ripple test showed they were faulty.

What are you trying to do with 230 laser diodes? What color are they?

 

[jim hardy]

i've used secondhand PC power supplies for hefty 5v loads.

200 watts at 5 volts is forty amps.
i got twenty , all i needed, with no problem.

But they do have a minimum load , ie won't maintain 5v into an open or nearly open circuit .
So provide a lamp or fan or something to provide that minimum load.

Be aware some PC's operate on 3.3 volt logic and those supplies would behave like you described in very first post. Be sure you get a supply for 5 volt computer .

 

[cdxer]

Legitimate power supplies would not behave like that.

If the supply is old, it may have dried-up electrolytic capacitors in it. These would be unable to hold a voltage and the output would have a lot of ripple on it, but with a meter you would just see a drop in voltage.

If it was rated at 40 amps, it probably did deliver 40 amps at one stage. This is always AT the stated voltage.

Ebay sellers are very nervous about their star rating, especially if they are regular sellers. So, all you have to do is complain to the seller and, if that doesn't work, complain to Ebay.
On rare occasions, I have complained about something that didn't work and I always got a refund or a replacement.

You can check the ripple on the output by putting a capacitor in series with an AC voltmeter and reading the voltage. It should be less than a volt under load.
A better way is to observe the output with a oscilloscope.
You could easily replace the electrolytics if the ripple test showed they were faulty.

What are you trying to do with 230 laser diodes? What color are they?

I have tried 3 PC power supplies.
a 250W 5V rail rated at 15A is 5.2V open, and loses 2V on 0.33Ohm load
a 300W 5V rail rated at 20A is 5.2V open, and loses 1.5V on 0.33Ohm load
a 500W 5V rail rated at 30A is 5.2V open, and loses 0.5V on 0.33Ohm load

I have tried 2 LED/CCTV power supplies from eBay:
"5V 40A 200W Regulated Switching Power Supply"
5V drops to 3.85V at 0.33 Ohm

"Cosel PAA/PAA75F-3 Power Supply Unit PSU 75W/3V/15A"
3.5V drops to 2.4V at 0.33 Ohm

This is why I've come to believe this is a normal behavior.
Yes, I will definitely complain to the seller.
A 200W supply SHOULD be able to handle a 15W loaded without V drop.

Can you give a little more detail on measure ripple with a capacitor?
My multimeter may not be able to measure AC in series.

I can't say much what the lasers are used for.
They are rated at 6mW consumption and 5mW light output,
which is the same as a typical laser pointer power.
The lights are not focused with a lense, so eye safety is not an issue.

 

[vk6kro]

PC power supplies are rated on the total current they can deliver, but anyone of the output leads, such as those that go to the hard drives or DVDs etc, probably can't deliver more than 5 or 7 .5 amps just because of the thin wire that is used.

So, you would need to open the power supply and go to the place where all the 5 volt wires come from and make a connection there and at a good ground point where all the black wires come from..

A 40 amp wire would be more like something you would find under the hood of a car. Probably with many strands of thick copper.

Older style (moving coil) multimeters had a capacitor connection for measuring AC ripple in the presence of a DC voltage.

For a modern digital meter, (depending on the rectifier circuit used) you may have to make up a half wave voltage doubler like this, but without the transformer:

The output of the voltage doubler circuit would go to a DC voltage range on the meter.