K&S power supply problem

what is toroidal inductor? filter inductor ? if so, just replace the capacitor about 1000uf or more large ones into a good one. it sounds most likely this capactor was broken out.

if you need more help, you shold post the circuit here, so that people can analysis and give out a suggestion.

It's the image of high current toroidal inductor, manufactured by Pulse, part number : PE-51508

There's 2 capacitor that connected with this inductor, 8200uf/ 100v, both of capacitor was tested at our workshop and good condition.

I haven't circuit diagram, it's hard to find, there's 5 PWM board inside the power supply, 4 board regulator that 70vdc, +48vdc, -48vdc, +36vdc, +12vdc, -12vdc, +5vdc, come out from these board, and 1 board at the back of chase to connect other 4 board, and there's a step down transformer inside, it's about 36kg heavy for this power supply.

Cnpower, thanks for your input, i need more.

What meaning is " toroidal "?

From pic, it seems not to be filter industor. what size is it?

try to measure its output to see if there is any short circuit? and is the 36v normal?

the load would be about 3 ohms. and what input voltage of the industor? if only this industor hot, that means large current occur. check output if it has a good touching.

there is no more compunents at the capacitors , usually.

if no diagram, try take more photo. dont show us code, no use, because most of us dont know this machine.

Actually i don't know the meaning of toroidal, i read it from data sheet.

It's 58,42mm x 28,19mm x 58,42mm size.

We have checked the input and output, it's normal, 36vdc/12.8a ( input and output) the output is used for shaft motor (on the machine) and no problem with the motor because if we change with another power supply, the machine work properly.

I try to take more photo and come back with more photo.

I confused too, i almost give up with this power supply, as the measurement I'm sure it must be work, but it's different when we install it to machine, the inductor has burned twice.

Try replacing the capacitor. it's probably bad despite your testing otherwise. Make sure it is an exact replacement, a generic cap won't work.

If your toroid (donut), as in picture, is still burning up after replacing the cap then one of the PWM (pulse width modulator) boards is bad, and not properly regulating the pulse width. This seems unlikely as you are getting the proper output.

check wave as shown with an oscillascope. or replace your inductor new one., how do you check your capacitors? use multimeter resistance step?x100 or x1k

the output of boaed is usually has these components.

I too have working experience on K&S. I think while repairing power supply due to some component replacement the SMPS frequency went up. If you can tell what components you replaced I will be able to help you. What is the SMPS IC in the power supply?

It just implies that circular rings (usually Ferrite I believe) are used for inductors and transformers, no more, no less.

transformers of this type radiate far less electronic smog than conventional types, usually run quieter and are slightly more efficient.

I resent you calling the pure, pedigreed, with beautiful round shapes electromagnetic field that barely is felt from a toroid inductor in its switching operation, electronic smog. Smog implies some vulgar smoke particles attached to the microscopic water particles in fog. What a sacrilege....:-)

Whatever.......

But think of all the non toroidal transformers still made and in use and what they are emmitting!!

Still, the same beautiful round shapes electromagnetic field. May be the leakage is more than in a toroidal one but well contained and intensified in the iron or ferrite of the xfmr. I suspect that you had some bad experiences with EMI and blame, now, on all honest xfmrs....

I have built a few high grade pre amps and amps that needed a very steady supply to achieve the best response and a toroidal transformer was a basic requirement to assist in getting that high quality and correct reproduction of the music signal that I (and several others) were after....with as little electrical noise as possible.....

Careful design of the pcb layout was also needed to achieve that.

Its many years ago now but the design (mostly Lindsey-Hood if I remember correctly, with some other changes) was quite remarkable in its time. It was published in an electronics magazine that I used to take in the UK.

wow, what high level preamplifier did you build? not simple, how lower the noise? how wide frequency band? how much is its distortion? and voltage output and input?

I know german use switch power in audio equipment very earlier, but do you know switch transformer is very different from inductor? easpically toroiday type core as output filter inductor?

Cnpower, you right about the toroidal inductor, it must be replace with a new one.

I just made a experiment, i shorted some of winding of the inductor, until the voltage at 2 pins just 0,2vdc and pulse 48 vp-p, with 40 ohm load, the inductor became stable normally warm until almost 1 hour, but there's something else that made this power supply run abnormal, something that i almost missed, i check all of mosfet IRFP450 that been changed by my friend, he was changed with the lower quality from original ones, so the current that come out from 36vdc is less then the normal value.

The first time i test it with 100 ohm load, the power supply always disable, to make it enable it must more than 125 ohm load, after i replace with the same quality with original mosfet, it can enable until 40 ohm load.

Cnpower, I'm very thankful with your information and support, I'll be back with new report after i get a new toroidal, because it's hard enough to find.

This sounds like a circumstance that everywhere is fulfilled with acid, alkali and metal particles and humidity in the air. Andy , are you working in a chemical factory, or somewhere like this circumstances? when the leak magnetic field emitted, the thick or thin fog will be around you? thats very worst condition. But I was told that everywhere is clean in your country. whats matter? was this very impressive to you?

May be it is a switching power supply and so the inductors may be for that purpose.toroidal form is the tube form of a neumatic tire.-

Are you sure these capacitors output filter cap not rectifier input cap? where is the transformer? does it hot when the inductor was burned? has the inductor 2pins or 4pins? to be sure its not input common filter inductor. what do you mean "when i shorten the length of wire it is drop the out put voltage ( 36 vdc )"? is the wire that winding on the inductor? how about 18v power output? try to take off them. then to see if the indicator is burned? only been in saturation was the inductor burned. and through very large current. from pic, it seems core get into black. this core of the inductor is eventually not made of ferrite, its most likely Kool mu or mpp material from its color. check it, BTW, to see if there is something wrong with it. can you inspect the chips frequency? 20 or 30kc? how do you test voltage before the inductor?

I'm sure it's output cap, from switching transformer it used 2 parallel diac for +36v, then directly to inductor, this inductor only 2 pins, like normally dc filter, "when i shorten the length of wire it is drop the out put voltage ( 36 vdc )" i mean i cut the length of the wire that have 2 mm diameter about 15 cm and yes, it's the wire that winding on the inductor.

It's 48v, if i take them off, the power supply is going to st-by mode, and disable indicator lamp turn on and there only +5v, +12v and -12v come out from power supply, +36v and others are shutdown.

I'll check the freq that come out from the chips, and voltage before inductor is +36vdc but the voltage before switching transformer is 250vdc and switched by 2 mosfet IRFP450.

Thanks for your info and support, I'll come back with the result of frequency.

this sounds that your circuit is no problem but your core is failed after hot.

if itis ever over hot, no matter how you re-winding wire, its no use again.

if you would use ferrite material, do remember cut a gap on it. or it will easy saturation and cannt act an inductor.

try to find another mpp material core for this inductor, same size, same turns.

[i act its as 18 fom pic]

this problem is rare ocurring in cirbuit.

oh, I forget this,

before replace the core, try to add more turns on it, say, sqrt(2) times, so that you can increase L double. dont reduce turns.

if yet not work, then replace one new.

how about output rectifier diodes?

is it one or two?

I have found a schematic from unitrode, it's same with my power supply but D1 and D2 is DIAC, it's full bridge SMPS (switching mode power supply)

I continue about the frequency, it's 90,1 KHz from SMPS chips, same with 5v, 12v, 36v, 48v and 70v, all of them have same frequency 90,1 KHz.

I also measure the voltage between 2 pin of inductor (L at the schematic) with oscilloscope and resistive load about 100 ohm at 36v output, there's 4,6vdc and pulse 58vp-p 90,1KHz between 2 pin of inductor, is it right? or there's something wrong?

Can you explain to me why? as long as i know a inductor with 2mm diameter and about 50cm length it should be 0,... volt.

I remember 3823 has only one output. how to form full bridge switcher? besides, you have only two 450 mostransistor, where are the rest two?

if your two rectifiers is diac, how would you get dc at output? this is my first hearing.

100khz, good.

96 vpp is a bit higher, I think, but can be accepted. besides, the 4.6v votage drop is rather bigger, unnormal at all. thsi error might be caused by meter when you measure at dc steps.

check your second output. or use large power resistor of about 5ohm to get more true value..

check your diode again. are they broken through?

what mean is

Can you explain to me why? as long as i know a inductor with 2mm diameter and about 50cm length it should be 0,... volt.

try to post the waveform here.

and try more thick, say 2.6mm diameter wire on the core. to stand more current.

SMPS IC : UC3823N , PWM IC : UC3823DW

Replaced component : IRFP450 , toroidal inductor

Picture 1 is front view , picture 2 is top view , picture 3 is toroidal inductor that always burnt.

Problem description : tested off line OK , tested on the machine ( w/ loading ) in 15 minutes toroidal inductor get burnt.

Testing performed on the good condition machine.

As far i know when toroidal inductor in that position is acting as a noise filter but when i shorten the length of wire it is drop the out put voltage ( 36 vdc ). Is there any body can also help me to find the explanation ?

If you have changed UC3823 then there is definite chance that the switching frequency went up. My previous company was making TL3823...42; there is lot difference like: frequency, startup component values, etc. By changing R & C to reduce switching frequency by 10% you can solve this problem (pin 5 & 6)

Kvsubramanyam, i have not changed UC3823 yet, but I'll try to check the freq, hope you're right, thanks for your information and i hope you will share with me another possibility.

He he he . .....strange, lah. i guess i feeling more confuse. . Mr. Subra perhaps.

Dear Friend,

Verify your following comments:

1. K&S working with other power supply of other machine. Hence loads are OK
2. The problem is Toroid inductor (it is an Output filter) heating up if connected to machine. This toroid in some designs help transfer of energy to output.
3. Stand alone you are getting all voltages after repairing this power supply

Inductor heating up; that means power is dissipated in the inductor. More voltage drop across the inductor. Pure DC will not have any effect on any inductor. Hence AC in the form of ripple is heating up the inductor. Inductor "Impedance" depends on the frequency. Since nothing else changed, hence I came to the conclusion that due to some component replacement the "designed" switching frequency should have went up!

More knowledge can be had from: www.smps.us/

Hope your confusion is cleared.