/diy/ - Do It Yourself

>>163440
Use a start capacitor?

>>163445
How?

I though about it, but if I add a capacitor to the circuit, wouldn't the PSU need to charge it first, thereby increasing the initial current further?

>>163447
A capacitor will need to be charged first, yes, but its takes no time at all for one as small as you need. It also smooths things out so it shouldn't make the PSU turn off. The charging builds up slowly then releases it suddenly like a battery and starts the motor. This does not draw instant large power like the motor by itself would.

www.google.com/search?q=start+capacitor+circuit+diagram&tbm=isch

>>163447
that's why you put switch after the capacitor, not before it.
And you may want to limit its charging current with a resistor so it will not be damaged.

>>163453
>>163456

not op, but out of curiosity, what capacitance are we talking here?

>>163456
A relay can be added so that when the motor gets the start charge from the capacitor the relay will switch the current from going through the capacitor to normal wiring.

>>163458
That depends on what the motor needs really. But, it should only need a brief instance to kick it up to speed before a relay switches it to normal current. So, it can have a low capacitance. For instance my 1/4HP sewage sump pump motor only needs a 20uF capacitor (220VAC).

>>163460
would this do?

>>163468
I was thinking of something like this. It's painfully obvious I'm not an EE so I'm not sure about this.

>>163463

AC, start capacitor....

>I don't think you understand why AC motor use a start capacitor...

>>163492
I don't think that particular circuit would work. I mean, you're bypassing the relay.

Since everyone agreed I needed a start capacitor I tried a circuit like >>163468 posted, with varying capacitances(between 470u and 4700u), but no dice:(
I know the caps charge and discharge, but it has a negligible effect on the RPM the motor attains before the PSU shuts down.

Thoughts?

>>163493
OP's motor is DC, not AC.

If it helps, the DC motor is from a cordless drill, so I think originally the battery could simply handle the peak starting current of the motor.

>>163468
Yep, do that exactly.
One thing I should say is that this is NOT the same as an AC start-up capacitor. What you're going for is a cap to provide peak current demands. I'd recommend a nice big one, like, over 1000uf. Make sure it's a nice beefy one so its leads can handle the current.

>>163500
First of all, that schematic wasn't a suggestion, I was asking if it'd work.

Second, I think you need to add a diode between the PSU and the capacitor, and I'm no EE either, but I think a 470µF cap takes a while to charge up. You should make sure the thing is charged before pressing the switch (clamp a multimeter on there, and remove the input while measuring, just to get a feel for how long it'd take).

>>163506
OK, i didn't see your post before I posted...
It seems that the capacitors you're using aren't supplying power for long enough, the discharge, and then the PSU has to supply all the current. Go bigger.

>>163499
>omit this if you're a wimp

>>163509
Got a good, cheap way of measuring the charge of a capacitor while it's charging? Because clamping a multimeter on it would just show the voltage of the psu, right?

>>163506
> like, over 1000uf

That was kinda vague, since 1000uF @ 12V takes just 2ms to fully discharge at 6A load.

It's likely that you'll need to limit the capacitor's charging current, or it alone triggers the PSU's overload protection circuit.

>>163515
I was thinking of measuring the current, if it's negligible, the caps are full, right?

I'm going to try 2x 4700u in parallel now.

>>163440
>The motor requires 3-6A at 12V, which should be fine, as the +12V rail on the PSU is rated at 14A.

Try it on the big amp 5v connection. I've always had bad luck with the 12v connections shutting the PSU down, but the 5V 30A connection goes strong. All you need is a transformer to change the 5V 30A into 12V whatever Amp is left over. It should work well actually.

>>163517
>It's likely that you'll need to limit the capacitor's charging current, or it alone triggers the PSU's overload protection circuit.
It has not, as of yet, done that. If it does I suppose I'll just add a resistor in series?

Basically we need values for R, V and F

I suggest:
R = 1kΩ
V = 15v
F = 2 x 4700µF (2 caps in parallel)

>>163518
Won't they just keep on charging until they pop?

>>163440
Add a heatsinked mosfet that shorts a connection over a resistor when switched on. Run it over resistor for start up current, then switch it to maximum power.

Alright, I tried two times 4700u parallel, and the RPM gets higher before it shuts down, but it does still shut down.
Those are the biggest caps I have access to right now, and I would hate to spend money on ridiculously large caps just to solve this.
How about temporarily disabling the overload protection circuity?

>>163526
Samefag with a question:
Couldn't you make a simple switch with a zener and an NPN?

>>163521

Yeah, the PSU alone has quite large capacitors at output. Adding 1000uF shouldn't be a problem. However, it wouldn't be a surprise if you needed a truly huge capacitor, for example 100mF, to jerk the motor.

Resistor is enough to limit the charging current. You need some kind of bypass switch for it, of course.

Btw, does this shit have something to do with R/C cars?

>>163519
this

make a simple transformer with equal numbers of windings on the coils. test it to see how much it is outputting. then increase the number of turns on the secondary coil, testing it again every so many turns until you reach 12V.

>>163528
Are you sure the diodes are fully charged? I would disconnect them from the PSU, measure the voltage (to make sure they're full), then put the PSU back online, and turn the switch on.

If the problem persists, then you probably need bigger caps..

>>163532
But the PSU delivers DC.. You'd need AC for a transformer to do anything other than heating itself up.

>>163536
Yeah, you make an inverter. It's fairly simple. The transformer is at the time of this diagram.

>>163538
>The transformer is at the TOP of this diagram.

fixed

Where the hell did "time" come from?

>>163538
And then what, pipe the AC to a rectifier bridge? This is starting to sound awfully complicated, since a bigger cap would do.

>>163543
The motor will run on AC, fyi.

>>163526
I don't think caps can kill themselves as long as you stay under their rated voltage.

>>163530
I agree on the large cap count, caps are really a shitty storage medium imo.
Not RC cars, it's powering the cutting tool of my DIY CNC mill:)

>>163532
But I have a DC output?
And I got the PSU for free, so I'd hate to spend money on something like a DC/DC converter while this might be solvable by using some extra circuitry;)

>>163534
Yeah, they're fully charged, and they're the biggest caps I have access to:(

>>163544
Oh, I missed that. Wait, didn't you say it was from a cordless drill or something? Why would they make it operate on both AC and DC, when it's likely that nobody will never use it with AC?
I'm pretty sure that if you stick AC in it, it'll either shake slightly or stay perfectly still, unless you modified it with a rectifier bridge or a diode already.

THIIIIIIIIIIIIIISSSSSSSSSSS THREAAAAAAAAAAAAAAD

Alright, does anyone perhaps know of a way to (temporarily) disable the overload protection?
Or else maybe a timing circuit that pulses the PS_ON a few times? 'cause I know I can get it to work if I just let the PSU try handling the current a few times while the motor has not stopped turning from the previous attempts.

http://www.ebay.com/itm/Aerovox-M-CGS503U025V5L3PH-50000uF-25V-capacitor-/390026356028?pt=LH_Default
Domain_0&hash=item5acf61e53c#ht_500wt_1413

I'm just going to leave this here.

>>163492
>>163499
That's very close, though. It should be this: (pic related).

What would happen is you'd supply current to the circuit long enough for the cap to fully charge, then trigger the relay. The cap can charge without the motor trying to start because its circuit is complete, whereas the motor circuit requires the relay to close before it can be complete. The diode is to protect the PSU from the current induced as the motor spins up or down.

Use caps with a rating of at least 13v working if they are electrolytic (don't go too high on the voltage rating, though - full charge can only be achieved by a voltage very near the rating), and try to track down something large (greater than 2200uf). Remember you can put them in parallel if you need more.

>>163555
No, sorry. I guess you could use the battery pack from the drill as some sort of booster, but be careful so it doesn't blow up.
That means disconnect it when the motor is up to speed, and immediately after put the PSU online.
This can be done with a few resistors..

Also, it sounds like you could probably use a flywheel somewhere, and a diode to stop voltage from hopping back into the PSU. (Because when you crank a motor, it turns into a generator, but you probably knew that).

>>163563
>This can be done with a few resistors..

I meant transistors, damn it. It's getting late on this side of the planet.

>>163563
The OP could attach a hand crank to it with a fly wheel. Then crank it up to speed then engage the power. The flywheel will maintain the starting needed when the power comes in so it won't trip the circuit breaker thing. Actually, using a bicycle would be better.

OK, we're overcomplicating this. Use the capacitor in >>163559, or similar. You could just parallel a bunch of salvaged caps. Try to exceed 20,000uf.
I drew up this schematic for you. Excuse the poor quality, my scanner currently doesn't exist.
The two switches there are just two poles of a DPDT switch. That way, in the OFF position, the capacitor charges through a current-limiting resistor. Flip it on, and it bypasses the resistor and connects the motor. I'm not entirely convinced the diode is needed, but what the hell, why not.

>>163571
Hahahaha.. Reminds me of that game I used to play when I was little... The Incredible Machine, I believe.

>>163440
OP, I made this diagram for you. It illustrates the proper way to do this. The part where the screen reads "NO" is where you need to make the circuit from this diagram, >>163511 which will turn it to "YES" and it will run the motor.

>>163578
You silly goose, everyone knows this is plumbing, not electricity

>>163579
You missed the Ad-hoc LAN connection near the middle, lower left, didn't you?

>>163580
Have you never heard of Ad-h2oc?

>>163555
>Alright, does anyone perhaps know of a way to (temporarily) disable the overload protection?
VERY NO

>Or else maybe a timing circuit that pulses the PS_ON a few times?
Use a transistor, specifically a MOSFET (since they can handle the current load) and build a smaller capacitor-resistor circuit in series with the gate.

The cap/res circuit will have a charge time which makes the gate of the MOSFET see a ramping voltage that starts low and goes up to 12v, where it will be saturated. Just make sure your charge time is slow enough, and you may need a resistor between the MOSFET and the ramping circuit (to make the path to the gate slightly higher impedance).

I personally would place another larger valued resistor (at least 10k, probably 20k or so) between the gate and ground. This provides a discharge path for the cap when the power supply is switched off.

>>163440
Simple solution OP,

http://www.youtube.com/watch?v=eHK83VsoYr0

>>163584
May I trouble you to draw a simple schematic, good sir?

http://www.youtube.com/watch?v=erZLnTGsJbc&feature=related

Solution, finally.

>>163519

So much FAIL.

>>163589
MFW

>>163584

This RIGHT HERE.

I've been trying to model this for a few minutes to show OP, but kept coming up with some fail even thought I know it'll work.

>>163587
Give me a few. I hope you don't mind hand drawn. I'm also going to breadboard it, but instead of a motor it'll be a multimeter.

>>163599
Looking forward to it.

11/10

Worst thread ever

It's like watching a committee of 5 year olds trying to do electrical engineering.

all the fags saying ''use a capacitor'' are mindlessly repeating advice they got from somewhere else, but it makes no fucking sense: the capacitor will only increase the startup surge.

what OP needs is a large inductor in series -- inductors resist a change in current, so the surge will be suppressed.

>>163606
Please explain further? Not an EE-fag, but I sure would like to learn.

> Please explain further? Not an EE-fag, but I sure would like to learn.

like I have the patience to start writing an explanation.... that's what wikipedia is for. the essential point is this: capacitors resist voltage changes and minimize them, inductors resist current changes and minimize them

>>163614
So we have something like this?

> So we have something like this?

exactly, tho your coil should show dotted lines to show it's iron core, or ferrite core, coz you need a pretty high inductance.

this page explains: http://www.electronics-tutorials.ws/inductor/inductor.html

>>163606
>' are mindlessly repeating advice they got from somewhere else,

Not many can be as self taught as you are.

>>163631
Ah, interesting read. Thank you!

OP you have to wire a 10Ohm 10W resistor to the 5V line on the PSU.

NTC.

That's all.

>>163637
wat

>>163639
I don't know the theoretical background but it works.
It needs to have a load on the 5V to properly drive the 12V line too.
Search around on google for PSU mods for explanations if you want.

I LOVE YOU GUYS!
I LOVE /DIY/!

>>163605
It's magnificent isn't it?

>>163606
It fucking worked!
Awesome, thanks!

I'm going to wind my own coil now, to see how small a coil I can get away with:)

Cheers!

>>163643
here is something:
http://web2.murraystate.edu/andy.batts/ps/powersupply.htm

Power supplies in today's computers are known as SWITCHMODE or Switching Mode power supplies and require a load to continue to operate after being switched on (the term switching mode actually applies to the technique of A/C to D/C conversion and not to the power up action). This load is provided by a 10 watt, 10 ohm wire wound load resistor (sandbar - about $0.80 at Radio Shack) across the +5 volt supply. While many of the newer power supplies will Latch_On without a preload, you will find that adding the resistor will (1) increase the measured voltage on the 12 volt rail slightly and (2) help stabilize the voltage level in this rail by minimizing voltage drop when the powersupply is loaded with a charger. Some inexpensive power supplies may fail if forced on without a load although the Design Guide states that the supplies should not be damaged if run without a sufficient load. The sandbar resistor has been zip tied to the case with a small amount of heat sink compound applied to the flattest side of the resistor . I will also take a file and remove any stamping flash that may remain around the ventilation slots. Without cooling, the resistor will get very hot and may fail prematurely; with this arrangement, the resistor will remain barely warm to the touch.

> I don't know the theoretical background but it works.

the background is this: in a PC supply, the only voltage that's actually regulated is the +5V, the other voltages simply follow along because of the correct turns ratio on the transformer. so, without a load on the 5V, regulation cannot occur and the power supply wont work.

however, this is old news. many newer supplies dont need that resistor.

>>163647
Wait, but my PSU handles the load just fine after it gets past the startup current, does your comment still apply?

running a dc motor on straight dc is very inefficient...PWM dat bitch

>>163649
>>163647
>>163645

I'm learning all sorts of stuff today. /diy/ truly is the best board.

>>163652
why.jpg
I don't even want to regulate its speed...

Anyway, should I use an air core for my coil, or not?
'cause after the motor has turned on, I obviously want as little losses from the coil as I can get, and those are just the loses from the resistance of the wire the coil is made up of; this means I want a relatively short length of thick wire, with as many windings as I can squeeze out, right?

>>163661
from >>163631
>exactly, tho your coil should show dotted lines to show it's iron core, or ferrite core, coz you need a pretty high inductance.

But whatever works, I guess..? Not him, btw.

>>163605
The learning process is sometimes a frustrating thing to watch for some people. For others it is like witnessing the beauty of birth.

>>163661
it makes the motor run constantly, which means you lose power to heat. Honestly, at such low voltage and current levels it's nothing to worry about. And you aren't using it in a situation where power conservation is an issue.

>>163661
So, this was an inrush problem and it only needed a ballast? Perhaps a switch or relay to bypass it once the motor is on and running?

>>163599 here
Sorry that took so long, but I had to do something else. There is a lot of *interesting* advice in this thread. Anyway, this is using the charge/discharge of a capacitor to drive the gate of a MOSFET. The power resistor and 12v LED are a dummy load (less than what the OP will be using, so the MOSFET needs a heatsink).

Sorry about the shaky hand, but this is actually a series of stills. The ramp speed is actually slightly slower, taking just about 5.5 seconds to reach 8.9

The voltage only goes up to 8.88 in this because it is coming from a 9v wall wart. The charge time is based on the capacitor value and the resistor.

>>163697
Wow! That looks awesome:)
Got a basic schematic for that circuit?

Thanks for taking the time to make that circuit and .gif; cheers!

>>163697
I'll draw the schematic in a sec.

>>163713
I think that's right. Test it first.

>>163723
A voltage divider to turn the mosfet on? I was thinking about using a zener diode

>>163725
Actually, there is a mistake in that - gimme a second.

>>163726
I think you should put a diode in parallel with the motor as well, from - to +

>>163725
>>163728
Corrected and improved! Anyway, this isn't The Way, but a way.

Also, except for the resistors, it is built from the parts I scavenged from a PC power supply.

>>163734
Much better. Just one thing, in your voltage divider, I think your math might be off. I don't know how many volts you can stick into the gate of the mosfet, but if I'm not mistaking it should be around 5?

Here's some voltage divider formulas. Oh, and if you need it to be 2 kΩ still, it's the ratio between R1 and R2 that counts.

>>163734
Alright!
Definitively saved:)
I finished winding my coil, and it works as i need it to, my motor starts without problems, but I might switch to your solution anyway, because it is so much more elegant:)
One question though, how much of the original 12V will actually get through to the motor?
With the circuit you built you lose about 0.1V, that's dependent on the resistances and capacitances used, right?

Thanks again:)

>>163740
>I don't know how many volts you can stick into the gate of the mosfet, but if I'm not mistaking it should be around 5?
Depends on the MOSFET - they generally switch on with as little as 1.5v (though not to saturation). You do raise a good point, though - a higher voltage will lower the on state resistance. The resistor between the gate and ground is to make the gate tend towards low. The value can be higher to aim for a higher gate voltage - 2k or even 5k might be a better choice.

I aimed for 3v because this one will pass up to 15 amps of current with a drain source voltage as low as 1v.

Most of the power MOSFETs I've come across tend not to reach saturation until around 10v, and will tolerate anywhere from -20v to 20v (with respect to the voltage applied to the source pin).

I didn't worry about the behavior of the voltage divider under load because MOSFETs are voltage operated, and the current through the gate is low enough to be ignored (for the PHP36N03LT the gate leakage current is 0.1 micro-amps).

>>163768
>how much of the original 12V will actually get through to the motor?
>that's dependent on the resistances and capacitances used, right?
Depends on the MOSFET, but the on state resistance is usually on order of mili-Ohms. IIRC they don't have the voltage drop that NPN BJT's do.

Grab the data sheet of any power MOSFET you were looking at, it will tell you.

>>163440
There is a difference between the steady-state current your motor draws, and what is called "start-up current". At start-up, especially with a mechanical load on the motor, the motor looks more or less like a dead short across it's leads. You need to determine what the start-up current requirements are, and use a supply that can handle it. There's nothing wrong with your power supply, it's doing what it's supposed to do: shut itself down when it detects a short across it's output(s), which, as I said, is what your motor looks like to it.

>>163734
>>163734
That looks like a soft-start circuit, which would be contraindicated for a motor because if it's got any sort of load on it, it might not have enough torque to start running.

>>163768
>because it is so much more elegant
Personally, I'd go with KISS instead. An inductor is the simplest possible solution, and it works well.
Always strive to eliminate as many parts as you can; it keeps costs down to a minimum, and there are fewer parts that can break and brick the entire thing.

Been looking at this thread, and now it is annoying me so much that i hand to answer.

you can handle this problem in lots of ways depending on how you want to use the motor and how you want to start it up.

Here is a few ideas..