power supply cant provide enough power

Hi!

I have the following problem: I have 4 DC motors from CD trays and im using locked antiphase to swing 4 bells with chinese h-bridges and arduinos.

At the beginning I have used 2 AC-DC adaptors of 12V and 800mA/1200mA max currents (I opened 800mA one and found voltage divider and 1000uF cap inside). But the problem was, when bell started to swing alone, it could easily come to full swing, but as soon as two bells that share same power supply start, it can clearly be seen that one of them is having difficulties (interesting, it was always those same two that had problems). Anyways, I added another adaptor and it fixed the problem for one of them. Now im out of adaptors (xD) but need one for the fourth one and Im thinking... what the heck, I would have 5 adaptors there all drawing power and generating heat (one of them is for arduino on 9V). I was thinking of using PC power supply but as soon as I read that it needs to draw at least half an amp to stay stable I stopped thinking about it. Now Im thinking about building a power supply myself, a bit beefier one and I need some help.

If I buy 12V 2A regulator and if I already have diode bridge and transformer (18V 2A), where should I put additional caps? (Im assuming its lack of capacitance thats causing the problem) After or before the regulator? In any case Im afraid that Ill burn out something by adding more caps because they'll draw a lot of power to fill themselves when turned on :S And how much capacitance would be enough :\

Its interesting though... I measured current drawn by bells and its its not exceeding power supply's limit, I guess the power supplies are not as powerful as the label says...

Well, thanks for any help! :smiley:

How much current does a single bell need ?

I think you have the wrong motors and the wrong adapters.
A CD tray needs to spin up very fast (unless you have very old ones that are 1X speed). So the motors might not be very efficient and require high currents.
You should stay away from cheap adapters (wall warts or whatever they are called). A good certified switching requlator can supply the current that it is labeled for.

Hi.

You seem to be missing the fact that a motor draws something like 6 to 10 times the current while starting.
This shouldn't last very long, so you might be able to solve that with a large capacitor.
But not if you are using 4 motors at the same time.
And if you are using PWM to control speed, you'll be switching on and off constantly.
You probably need to beef up the supply a bit more.
800 mA is good enough for a single motor, but not for 2 you told us.
So 800 * 4 is what you need to aim at, and that isn't 2 Amperes.
You also need to know that a 2A regulator can only do that continuously if the generated heat is dealt with.
So in this case i would look for a regulator that can handle 3 A or more to be on the safe side.
Still need to have the caps as close to the load (in this case the H bridge) as possible.
A combination of micro farads to have the buffer function discussed above, and nano farads to filter out spikes that might be generated by the switching of the H bridge transistors and its loads.

@ Koepel:
Isn't a tray motor the motor that drives the mechanism to open and close the tray, instead of the one that spins the disc ?
(i see i read "tray motor", but OP said "motor from a tray" instead.)

Koepel:
How much current does a single bell need ?

I think you have the wrong motors and the wrong adapters.
A CD tray needs to spin up very fast (unless you have very old ones that are 1X speed). So the motors might not be very efficient and require high currents.
You should stay away from cheap adapters (wall warts or whatever they are called). A good certified switching requlator can supply the current that it is labeled for.

Yeah, that may be, but these were easiest and cheapest to get and small enough to fit the design :\ I run them on 12V instead of specified 5V. I tried to measure the current but I dont know if my analog multimeter has fast enough response time :\ Anyway, they are not those slow ones I think but fast ones. I would rather change the adapters than the motors if possible. I grabbed those I had available but it seems they were not suitable for the job (as you said). What about old laptop power supplies? They have high current capablity, only problem is they are 16V >.< If there is a way to lower the voltage nicely...


(scale is to 0.5Amps)

MAS3:
Hi.

You seem to be missing the fact that a motor draws something like 6 to 10 times the current while starting.
This shouldn't last very long, so you might be able to solve that with a large capacitor.
But not if you are using 4 motors at the same time.
And if you are using PWM to control speed, you'll be switching on and off constantly.
You probably need to beef up the supply a bit more.
800 mA is good enough for a single motor, but not for 2 you told us.
So 800 * 4 is what you need to aim at, and that isn't 2 Amperes.
You also need to know that a 2A regulator can only do that continuously if the generated heat is dealt with.
So in this case i would look for a regulator that can handle 3 A or more to be on the safe side.
Still need to have the caps as close to the load (in this case the H bridge) as possible.
A combination of micro farads to have the buffer function discussed above, and nano farads to filter out spikes that might be generated by the switching of the H bridge transistors and its loads.

@ Koepel:
Isn't a tray motor the motor that drives the mechanism to open and close the tray, instead of the one that spins the disc ?
(i see i read "tray motor", but OP said "motor from a tray" instead.)

Possibly but my analog multimeter does not detect that :\ But... I do have them on slow spin-up (they start slowly and then accelerate). On the other hand I read that locked antiphase does have high back surges and it might even get to power supply if there are not capacitors to catch it and release it back in the next cycle (I read that in some article on locked antiphase).

So if I put a cap next to the load, times 4, how do I prevent the caps from overloading the power supply when they are filling themselves up?

The h-bridges do have some caps on them, not sure though if they are for the 5V regulator on them of to catch spikes from the motor.

Terraviper-5:
I was thinking of using PC power supply but as soon as I read that it needs to draw at least half an amp to stay stable I stopped thinking about it.

Not all PC power supplies need half an amp load. I've used one of mine many times without worrying about a minimum load. If it turns out your power supply needs a minimum load, add a fan or something.

As other have suggested, motors can draw a lot of current when starting up. I've had many motor projects which didn't work properly when powered from my bench supply (plugged into the wall), which work fine when powered from a good battery pack. It can be challenge finding a power supply for motor project. My bet is a PC power supply would do the job well.

DuaneDegn:
Not all PC power supplies need half an amp load. I've used one of mine many times without worrying about a minimum load. If it turns out your power supply needs a minimum load, add a fan or something.

As other have suggested, motors can draw a lot of current when starting up. I've had many motor projects which didn't work properly when powered from my bench supply (plugged into the wall), which work fine when powered from a good battery pack. It can be challenge finding a power supply for motor project. My bet is a PC power supply would do the job well.

I have several, but how do I (safely) find out if they do need additional load or not?

Most of the time they won't have almost any (additional) load on them.

[Here's a link to a PCB for using an ATX power supply as a bench supply.](Seeed Studio Bazaar, The IoT Hardware enabler. a link to a PCB for using an ATX power supply as a bench supply.)

The kit includes a heavy duty resistor to use a load if needed but the page says most supplies don't need a load.

There's pin on the connector which needs to be shorted to ground in order for the supply to work. I read how to do this on the ATX Wikipedia page.

DuaneDegn:
[Here's a link to a PCB for using an ATX power supply as a bench supply.](http://Here's a link to a PCB for using an ATX power supply as a bench supply.)

The kit includes a heavy duty resistor to use a load if needed but the page says most supplies don't need a load.

There's pin on the connector which needs to be shorted to ground in order for the supply to work. I read how to do this on the ATX Wikipedia page.

Hm the link doesn't open for me :\

The problem is this will be turned on more than half a day everyday, so it would be really wasteful to have half an amp or more constantly flowing. Before I posted here I searched Google about PC power supplies and there are a lot of mentions of minimum load usually be the cause of problems for those trying to use the psu as standalone power source.

The two power supplies on lower picture are older and still have a switch to be turned on.

Sorry about that. I fixed it.

I suppose you could try the power supplies to see if they will work without a load. I really don't think there's a danger of testing an unloaded ATX supply for the amount of time it would take to check the output.

DuaneDegn:
Sorry about that. I fixed it.

I suppose you could try the power supplies to see if they will work without a load. I really don't think there's a danger of testing an unloaded ATX supply for the amount of time it would take to check the output.

Wow, interesting board, spares you some work. Thanks for the link!

Yea but the problem is, I read some power supplies output erratic voltages in that case, so I would have to hook it up to an oscilloscope to see it. Well Ill just find someone who has it then :smiley:

Thanks!