Adventures adjusting the voltage of a notebook PSU

[robbak]

I had a failed notebook PSU that I had replaced (spliced the cable onto a good spare), so I broke it open to see what the problem was. Turned out it was simply broken solder joints on the output of the transformer, and re-soldering them got it working. Well, I had no use for another notebook PSU without a plug on it, but what I did have use for was a 12V supply to run LED strip-lights.

So, the output stage is based around a DAS001 chip by ST, (http://www.seekic.com/uploadfile/ic-mfg/20121129203352340.jpg) a dual op-amp with a voltage ref tied to the first op-amp's non-inverting input. The circuit follows from that- a voltage divider between the positive and ground driving the inverting input. The second amp is also connected to the vref, and is used for short circuit protection. The outputs of both op-amps go through resistors and (I assume) a twin-diode package to drive the opto-isolator.

So, I adjusted the voltage divider to make 12V (as it happens, the bottom half of the divider was made up of two resistors in parallel, and removing one was pretty close to what was needed. When powered up, the voltage was right. But, and this is doubtless no surprise to most of you, once connected to any load, the supply oscillated at about 1-2 Hz.

So, things I tried - removing one resistor made the current passing through the divider very small, so I put the resistor back and put a resistor in parallel with the top one instead. I increased the storage capacitance on the output, for no change. I thought it may be the short-circuit protection, so disabled that (no change). The last thing I tried was to see if I could remove some primary windings from the transformer, which made this a destructive teardown, as the glue in the transformer was considerably stronger than the ferrite. At least now I know how one of them is constructed!

Now, that is not the only notebook PSU I have languishing in my junk box, and I'd like to give this another go. I think the problem is that the PSU is ramping the voltage up too fast (it is aiming for 19V, not 12v, after all), overshooting, so the secondary circuit shuts it down too hard. If that is the case, what should I look to do to slow it down? What else could be causing the oscillation, and how would those be fixed?

[German_EE]

I suspect that some other part of the circuit (designed for a supply at 19V) is having problems at 12V. This is something I experienced a few months ago when, like you, I modified a supply for lighting operation. The output of the SMPS had the voltage divider but also just after the rectifier the circuitry around the opto-coupler was fed with the raw voltage.

[robbak]

I suspect that some other part of the circuit (designed for a supply at 19V) is having problems at 12V. This is something I experienced a few months ago when, like you, I modified a supply for lighting operation. The output of the SMPS had the voltage divider but also just after the rectifier the circuitry around the opto-coupler was fed with the raw voltage.

Thanks - that is very possible. I was investigating the chip that controls the primary side this afternoon, and if the optocoupler isn't working, short-circuit protection there would cut in, which does something that sounds very much like what I was seeing. So adjusting the dropper resistors to the opto is something else to try!

I've now got datasheets for chips that are at least very similar to the ones in this device, the NCP1200 on the primary and the TSM103W on the secondary.

[robbak]

So I had a go with a second one. First tip is these designers defy logic and physics when routing these boards! Both of the resistors in the voltage divider were illogically far away from the chip they were connected to. Once I had found that I had to find a resistor to put parallel with the top resistor, and this one was more tolerant than the first and required no other modifications.

I did see another place where dropper resistors could be adjusted. This design had a resistor network feeding pin 3, which is both the non-inverting input and the top of the voltage reference. Other devices may need these resistors adjusted.

For now, I have ordered myself a grab-bag of SMD resistors so I can make the adjustment permanent and fit it all back into it's box.