Linkworld(?) LPJ19-25E

[stj]

Re: Linkworld(?) LPJ19-25E

dont forget the capacitors on the output will explode at 16v

how many amps at 24v do you need?

[bauto601]

Re: Linkworld(?) LPJ19-25E

dont forget the capacitors on the output will explode at 16v

how many amps at 24v do you need?

I'll just need about ~15V, that 24V was just the maximum voltage of the controller. If i run my heatbed at 24V (1ohm) it would draw 576W, and probably explode afterward together with the primary phase of the PSU

So i just need around 15V, i think that even 14V would be enough. I just like it to have some headroom.

Here they are also modding a psu with the WT7520 chip:
http://forums.reprap.org/read.php?340,317793,318935

The datasheet of it:


Can i archieve the incrase in voltage by adding a variable resistor to pin 16?

[momaka]

Re: Linkworld(?) LPJ19-25E

dont forget the capacitors on the output will explode at 16v

Not necessarily.

Electrolytic capacitors can usually take a few volts more than what they are rated for (called surge voltage, or something similar in datasheets - have a look). Chemicon KZE, for example, is rated for 20V surge for their 16V caps.

That said, you can still make the caps in the PSU pop even if you up the voltage by just a volt or two. The reason for that is because electrolytic caps "reform" to the voltage they have been used most often with - i.e. a 10V cap sitting on the 5V rail of a PSU will be reformed to take 5V more or less. If you suddenly put 9V to it, the cap may pop. So what you will have to do in such cases is slowly bring up the voltage on the capacitor. You can do it in 0.5V increments and let it sit for about 30-60 minutes with that voltage. Alternatively, you can take the cap out of the PSU and connect it in series with a 10-100 KOhm resistor and connect that to a 10V source. The series resistor will limit the current so that the cap won't overheat or short internally while it still hasn't been reformed yet. If you leave the cap like that for about 1-2 hours, it will be reformed into a 10V cap again. Then you can put it back in the PSU.

Can i archieve the incrase in voltage by adding a variable resistor to pin 16?

Possibly.
But PSU may become unstable or whine/whiste - if it does, that is a bad sign.

Usually all of the rails will be connected with resistors to a single feedback pin on the PWM IC. That pin will also be connected to ground with another resistor. You can raise or lower the resistance of that resistor to decrease or increase the PSU output voltage respectively.

If you decide to tweak the PSU, my suggestion would be to apply the series incandescent light bulb trick to avoid the PSU from blowing up if something goes wrong. See this for more info:
https://www.badcaps.net/forum/showpo...4&postcount=70

[bauto601]

Re: Linkworld(?) LPJ19-25E

I've made the following changes to the PWM chip:

- 50Kohm variable resistor from pin 16 (feedback) to ground to change the voltage
- Connected pin 4 (PT) to the ground to disable the UVP protection
- Cut pin 1, 2 and 3 from the chip. (they monitored the 3.3, 5 and 12V rail)

The psu is now working just as good as before but now i can change the voltage. I haven't increased the voltage to 15+ volts but it seems like the psu doesn't whine at any voltage below 15V.

I may going to cut the legs of the 3.3 and 5v rectifiers because they aren't necessary anymore.

[stj]

Re: Linkworld(?) LPJ19-25E

desolder them instead - then you have spares instead of wasting them.

[momaka]

Re: Linkworld(?) LPJ19-25E

If I remember correctly, usually this mod (variable resistor between ground and feedback pin) is okay for making the PSU voltage higher, but not lower. Lowering it too much with this mod may make it while. But since you are not doing that, I think it will be okay. I guess when you put it under heavy load, you will know for sure .

desolder them instead - then you have spares instead of wasting them.

+1

You can also remove the 3.3V choke/coil and other 3.3V circuitry to prevent the mag-amp circuit from messing something up.

[Th3_uN1Qu3]

Re: Linkworld(?) LPJ19-25E

Increasing the voltage is OK as long as you have adequately rated output capacitors, it will not affect stability. Technical background: the gain of the feedback loop increases, which actually makes it more stable than it was before.

However, there is one caveat: It will be more sensitive to line voltage changes as there will be less voltage headroom. If your load is fairly constant and there aren't many brownouts in your area it will be fine. With a nominal 12v rail you can go up to 17-18v before running out of voltage from the transformer. 14-15v is the max you can safely get while still maintaining regulation.

The most popular ATX PSU mod is converting it to 13.8v for running ham gear or car audio amps, a lot of people have done it and it works fine for everyone, so you have little to worry about.

[bauto601]

Re: Linkworld(?) LPJ19-25E

Increasing the voltage is OK as long as you have adequately rated output capacitors, it will not affect stability. Technical background: the gain of the feedback loop increases, which actually makes it more stable than it was before.

However, there is one caveat: It will be more sensitive to line voltage changes as there will be less voltage headroom. If your load is fairly constant and there aren't many brownouts in your area it will be fine. With a nominal 12v rail you can go up to 17-18v before running out of voltage from the transformer. 14-15v is the max you can safely get while still maintaining regulation.

The most popular ATX PSU mod is converting it to 13.8v for running ham gear or car audio amps, a lot of people have done it and it works fine for everyone, so you have little to worry about.

I've upgraded the capacitors on the 12v rail to a mix of capacitors that i had laying around:
1x 1000uf 36V
1x 560uf 25v
1x 470uf 36v

All the other capacitors are 10v ones so they can hanle the increased voltage just fine.

I'm running the power supply at ~18V now (~200W load) and it's nice and stable. The fans (which are on the fancontroller of this unit) are running a lot louder now, so i'm going to buy a resistor for them tomorrow so the unit will be more silent and my fans are going to live longer than a week

[Majestyk]

Re: Linkworld(?) LPJ19-25E

A passive PFC made of true iron!

The Chinese use to sometimes "replace" iron and copper by cheaper materials, as you can see here:

http://www.overclock.net/t/1246964/h...made-of-cement

and even worse here

http://www.pcpop.com/doc/0/294/294565_all.shtml

It´s unfu#*ing belivable...

[bauto601]

Re: Linkworld(?) LPJ19-25E

I know it's an old thread but i just wanted to give a status update after two and a half years of operation.

The power supply is still working perfectly fine. It's running at about 17.6 volt with a load. It's just perfect in combination with the heated bed of the 3D printer. It only needs about 2 minutes of warming up time to reach 90 degrees celcius.

The fan controller is still working, i didn't expect that with 3 fans on it with an input voltage of 17.6 volt. I guess Linkworld made it quite decent for their standards.

And thank you guys for sharing all this knowledge.

[Th3_uN1Qu3]

Re: Linkworld(?) LPJ19-25E

The "fan controller" is just an 1 Amp PNP transistor, a NTC thermistor, a diode and a resistor. Not much to fail there, especially if the transistor is housed in the larger TO126 case.

Nice to see that it is still working. Thanks for the update.