Codegen 350XA supply

[milesy]

Re: Codegen 350XA supply

Maybe you need to adjust the probe compensation. Your scope should have a square wave output for that purpose - and if you're lucky it might even work. Otherwise, use a signal generator for this purpose. A soundcard capable of 96 or 192kHz set to generate a 1kHz square wave is fine for this.

Adjust the compensation trimmer on your CRO probe until the top and bottom of the square wave are parallel with each other and show no overshoot (peaks) or undershoot (rounding of edges).

yes the test generator works......never thought of adjusting the cro prope but makes sense........ill have a look tomorrow

[Th3_uN1Qu3]

Re: Codegen 350XA supply

as for the shape of the wave well i have to agree with you - it doesnt show on the rectified trace so the strange shape must be being caused by the CRO - but i have no idea why

Maybe you need to adjust the probe compensation. Your scope should have a square wave output for that purpose - and if you're lucky it might even work. Otherwise, use a signal generator for this purpose. A soundcard capable of 96 or 192kHz set to generate a 1kHz square wave is fine for this.

Adjust the compensation trimmer on your CRO probe until the top and bottom of the square wave are parallel with each other and show no overshoot (peaks) or undershoot (rounding of edges).

[milesy]

Re: Codegen 350XA supply

ok rievax your sort of right.......what i found was that there is a button on the cro to zoom in on a section of the wave - you can see it in the above picture marked TRIG VIEW PUSH

for whatever reason this button was activated even though it wasnt in its active state - so where i said i had it on 2v/div above what i was getting was actually peaking at around 30v - so it seems that my wave is correct .......... poor thing is getting a bit old now and is showing a few funny faults.

so if i said i was confused before - now i really am because i havent actually replaced anything and yet the supply seems to be operating correctly?????

as for the shape of the wave well i have to agree with you - it doesnt show on the rectified trace so the strange shape must be being caused by the CRO - but i have no idea why

[rievax_60]

Re: Codegen 350XA supply

I think your CRO might be causing the anode waveform to look asymmetrical, because the asymmetry isn't being reflected in the cathode waveform.
Does it change shape when you change the vertical position. Or try the other channel.
If the CRO was really set to set on 2v/div, the 12v output would be less than 2v. What was the 12v output measuring?
Better check the CRO calibration and that the inner vertical gain knob is set to is CAL position if it has that.

[milesy]

Re: Codegen 350XA supply

here you go rievax.......here is both anode and cathode waveform with no changes to CRO settings just moved the probe......set on 2v/div with a x1 probe


IMGP9597.jpg by Adrian Miles Photography, on Flickr


IMGP9596.jpg by Adrian Miles Photography, on Flickr

[milesy]

Re: Codegen 350XA supply

sure rievax ill post it up later on tonight....trouble is im not getting anywhere near the peak of 28v

[rievax_60]

Re: Codegen 350XA supply

I have just tried loading the 12v on my 350XA to 6A. Both waveforms change very little. The duty cycle increases by a few percent and the peak drops to 28v.
milesy, I am curious to see your 12v anode waveform with the 2Ω load.

[rievax_60]

Re: Codegen 350XA supply

My 350XA is a bit different. It has a F16C20 for the 12v rectifier. Its a TO-220 case 16A total dual fast silicon diode.
Silicon diodes take a small amount of time to turn on and off but this doesn't fully explain the large size of the spike.
It might be useful to see anode and cathode waveforms loaded and unloaded to see how much the amplitude changes. But its seems that the 12v rail is just being overloaded.

[b700029]

Re: Codegen 350XA supply

but when i plug in the 4 pin connector the MB voltage drops to 10v

Has this supply ever powered the mobo successfully? The 4-pin connector is +12V for the CPU buck converter and draws a lot of power.

[milesy]

Re: Codegen 350XA supply

ok looks like this one has me beat......ill move on to the next one - hope you guys can lend me some advice with this one

[c_hegge]

Re: Codegen 350XA supply

Ah, I didn't realise you had 5A diodes. they are usually only 3A. But still, 10A is insufficient.

Anyhow, if you're sure the diodes are OK, then you could have a bad winding somewhere on the transformer.

[milesy]

Re: Codegen 350XA supply

so the diodes are FR502 which according to the data sheet have a current capability of 5 Amps each - so realisticly i should be able to drag 10 A out of this supply

having a closer look at the wave form above and the picture shown above is with the CRO set at 2v/div and DC coupled.......doesnt that trace look a little small from what it should be if you refer back to rievax post on page 2 the trace he put up is with his CRO set on 20v/div........

if we assume the diodes are ok (as they test ok and seem to pass the voltage ok) then can i presume that the signal coming from the actuall transformer is too low???

[milesy]

Re: Codegen 350XA supply

the specs on the PSU say 12v 15 A....trying to find the data sheet for the diode to check the current capability

[c_hegge]

Re: Codegen 350XA supply

I'm still thinking try another rectifier if you haven't already done so. those two diodes on the bracket can only deliver about 6A combined. Any PC with a 4 pin CPU connector should have at least double that available, or more of it's a power hungry CPU.

[milesy]

Re: Codegen 350XA supply

back to more important issues

what are your thoughts on my supply hegge

[c_hegge]

Re: Codegen 350XA supply

Hi, c_hegge how are you ? I know you . We talked many times . May be not this place, May be somewhere .

Huh? Are you sure? The only members I have anything to do with outside here are Ratdude747 and Shovenose.

[milesy]

Re: Codegen 350XA supply

ok a strange state of play ..........lol (cant beleive this)

it seems that perhaps afterall there may be something wrong with the motherboard i was using for testing - to explain.

As of now......i plug the 20 pin ATX connector in to the MB and turn the PS on by shorting the green wire to ground. So at this point the PS turns on and the 12V is reading 11.75v - so great you say all is good.....but when i plug in the 4 pin connector the MB voltage drops to 10v and eventually the wire i have soldered to the regulator diodes melts of........woops so presumably there is a short on the MB.

so i took a 50W 2 ohm resistor and connect it to the 12v line of the PS (leaving the 4 pin connector off the MB) - again turn on the supply and this time the voltage drops to about 11V.....connect the CRO (which now also seems to be playing up a bit GRR) and looking at the trace without the resistor connected its all over the place.....connect the resistor and i get a nice clean rectified waveform expect for a large spike on the leading edge of the trace (see below) - oh and i should add this is without any addition caps connected on the front end - add the caps on the primrary and there is little if any change...


so im back to square one here and completely confused.



IMGP9588.jpg by Adrian Miles Photography, on Flickr

[Escort Eagle]

Re: Codegen 350XA supply

Hi, c_hegge how are you ? I know you . We talked many times . May be not this place, May be somewhere .

And milesy What is the result dude ? We are expecting for your result .

Thanks

[c_hegge]

Re: Codegen 350XA supply

I have not used 450V capacitors ratings before . I have always used 200V ratings . So, I am not sure Double Capacitors of 450V ratings are safe here or not !! Do it at your own risk !!

hi escort - why would you think that using 450v caps would be an issue?? i cant see a problem using a higher rated cap?

It won't be a problem, other than that 450V rated caps may not fit, as they will be physically larger.

[milesy]

Re: Codegen 350XA supply

If you touch the isolated scope you will get a live-earth short which will hurt a lot.

ahhhhhhhhhhhhhhhhhhhh