Magnavox DRV power supply SMPS

[momaka]

Re: Magnavox DRV power supply SMPS

Are you talking about diode "Dclamp"? If yes, its anode is connected to pin 5 of the transformer and its cathode (stripe) is connected to parallel resistor-capacitor combo, "Clamp" and "Rclamp".

At least I don't see which other diode you're talking about. But please specify if it's something else. It's very important that you get the clamping network wired correctly in order to prevent blowing your ncp1200 chip.

Also, try opening the picture with Windows Picture Viewer and zoom in. The junctions should be a lot more clear that way. Anywhere where you see three or more wires coming together with a DOT over them, that's a junction. If wires cross each other with *no* dot, then there's no junction.

[johnfin]

Re: Magnavox DRV power supply SMPS

I meant resistor 2106 on the way up to the clamping section

[momaka]

Re: Magnavox DRV power supply SMPS

One side of resistor R2106 connected to pin 3 of the transformer and clamping circuit. The other side of resistor R2106 goes to pin 8 on the NCP1200 chip. That's it.

Make sure to change the value of resistor R2106 according to the diagram too. I don't know why I didn't erase the designator, but I should have to prevent confusion. In any case, it's a 22 KOhm resistor (the original schematic has 390 KOhm for R2106). So just beware of that.

[johnfin]

Re: Magnavox DRV power supply SMPS

...."One side of resistor R2106 connected to pin 3 of the transformer and clamping circuit. The other side of resistor R2106 goes to pin 8 on the NCP1200 chip. That's it.
also goes to A/C input, right

[momaka]

Re: Magnavox DRV power supply SMPS

Yes, correct.

[johnfin]

Re: Magnavox DRV power supply SMPS

frankenboard

[momaka]

Re: Magnavox DRV power supply SMPS

frankenboard

Not bad, though.
Just try to keep most wires as short as possible. Also, maybe make the grounding wire a bit thicker. Other than that, looks ready for a test, I suppose.

[johnfin]

Re: Magnavox DRV power supply SMPS

So for our first test am I leaving Q2101 and R2234 as original?

[momaka]

Re: Magnavox DRV power supply SMPS

Yes, let's try everything original and see what happens.

Don't forget the series incandescent bulb trick for current limiting! See if you have something in the range of 100-300 Watts.

[johnfin]

Re: Magnavox DRV power supply SMPS

Any special output measuring points. Maybe somewhere on the high or low side after the transformer. I wouldn't want our modification results to be thrown off by something on the low side

[momaka]

Re: Magnavox DRV power supply SMPS

^ Just the voltage on the 5V rail output (i.e. the EV+5V), as that is the main output rail and also what's used for feedback. Maybe the 12V rail (EV+12V) too, if you have two multimeters.

If we get nothing on the output, then we might need to take more measurements to try and figure out why the mod is not working. Again, this mod is still in "beta" version, so it not working correctly the first time (or at all) in not unlikely. Probably will need some revisions, but we have to test it as is to see if that's the case. Maybe we get lucky we designed everything correctly (ha! that rarely happens in the EE world of circuits) and it all works. Either way, at least nothing should blow up or burn up though, as newer PWM ICs like this can be pretty good at protecting things.

[johnfin]

Re: Magnavox DRV power supply SMPS

Well it went up in smoke, or at least the ncp1200. Turned it on, light came on full brightness, then a pop and smoke. Chip was hot so I assume its bad. Any way to check it? I will triple check the wiring again.

[momaka]

Re: Magnavox DRV power supply SMPS

Well it went up in smoke, or at least the ncp1200. Turned it on, light came on full brightness, then a pop and smoke. Chip was hot so I assume its bad. Any way to check it? I will triple check the wiring again.

Time for an autopsy and circuit adjustments then.

The only way the NCP1200 chip can go bad is if the main switching transistor Q2101 or its source resistor R2108 are bad (Q2101 with short-circuit between Gate to Drain/Source or R2108 open-circuited). So remove Q2101 and R2108 and check them.

That said, if the NCP1200 looks overheated even a slight bit, it's better to replace it. As I suggested earlier in the thread, when building these kinds of mods, it's a good idea to get *at least* two of these controller and any other specialty parts. That way, if one blows up, you won't have to wait again (and pay for shipping again) for the part to arrive.

Also, I now see there are a few possible improvements we can make to the circuit - particularly in protecting the NCP1200 chip. Since its Vcc is rated for up to 16V max, limiting the voltage on a few of the signal pins to 15-16V should work to protect the IC - that would be the CS and DRV pins. In addition to that, maybe replace R2108 with a 1 to 1.5 Ohm, 2W resistor (flame-proof metal film or fusible type of resistor only). This will set the over-current protection to a lower level, so that the NCP1200 chip will self-limit sooner. And lastly, in order to further protect the NCP1200 from the main transistor, Q2102, you could build the driver circuit show in Figure 26 ("Improving Both Turn-On and Turn-Off Times") on page 13 of the NCP1200 datasheet.

For starters, though, again make sure that Q2102 and R2108 are 100% good. Better yet, replace Q2102 with the new MOSFET you bought and also change R2108 with a 1 to 1.5 Ohm, 2W flame-proof resistor. Then check all your wiring again and give this another try. In particular, check the Snubber/Clamping network you built and verify that the diode and resistor in it are good (take out of circuit to verify). Detailed pictures of your complete project, before you power it on, will be extremely helpful too in order to see if there are any mistakes. And of course, do have a look at the modified circuit I made again, and make sure that your add-on board is not only wired the way it should be, but also that any other parts or paths that need to be removed from the original circuit already are. So please compare the original and modified circuits very carefully.

[johnfin]

Re: Magnavox DRV power supply SMPS

Ok, before we dig in, I replaced the ncp1200 and replaced the mosfet with the one you mentioned before. Plugged it in and the light did not come on and no voltage at output. 120vac at input. So lets get from there and start measuring. Let me know where.

[momaka]

Re: Magnavox DRV power supply SMPS

So was the original MOSFET bad as I expected? I'd like to know, as that can tell more why the circuit/mod failed. Simply replacing a part with a new one without telling us if the old one was bad doesn't always tell the full picture.

Also, did you check R2108 to make sure it's not open-circuit?
While at it, check R2101 and R2106 for open-circuit as well.

You should get a quick flash (even if very faint) from the bulb due to the input capacitor charging up, regardless if the PSU has any output or not.

[johnfin]

Re: Magnavox DRV power supply SMPS

I have replaced that mosfet before I believe. I appears to be bad
2108 is open (hard to believe), need to check with analog meter
2101 ceramic whopper 9.9m
2106 smc .38m

so no light at all, tried in pitch black.

2108 is .68 ohms right? I need to buy one or find a 2 ohm and 1 ohm and put them in parallel

[momaka]

Re: Magnavox DRV power supply SMPS

I have replaced that mosfet before I believe. I appears to be bad

Appears?? Should be easy to tell with a regular digital multimeter on diode test.

2108 is open (hard to believe), need to check with analog meter

Given that there was a light bulb in series with the power supply... yes, unlikely for that resistor to have become open-circuited with this circuit mod. But it could have become open-circuited before you made the mod circuit - which would explain why the NCP1200 failed, as the shorted MOSFET probably backfed high voltage into the CS and/or DRV pins of the NCP1200.

So yes, double-check that resistor. Though I don't see why you would need to do it with an analog meter. A cheapo digital DMM showing 1-2 Ohms across the resistor (as most cheap DMMs usually can't go down to less than 1 Ohm resistance) means it's probably good.

2101 ceramic whopper 9.9m

9.9m? is that 9.9 milliOhms (mOhms) or 9.9 MegaOhms (MOhms)

2106 smc .38m

Same goes for this one as the above one.

2108 is .68 ohms right? I need to buy one or find a 2 ohm and 1 ohm and put them in parallel

Yup.
Both options will do. Though I suggest getting all three: that is, a 2-Ohm, a 1-Ohm, and a 0.68-Ohm. Reason being is on the next run, you could try the 2-Ohm resistor first. This will raise the sense voltage at the CS pin on the NCP1200, so the NCP1200 will self-limit sooner, in case there is something wrong with the circuit. This may result in low voltage output on the PSU outputs. But for testing the mod circuit, that's fine, as this should protect the NCP1200 a little better. Then if that doesn't work, you can use the 1-Ohm resistor, and if not, then finally the 0.68 Ohm.

Again, just make sure these resistors are flame-proof.

[johnfin]

Re: Magnavox DRV power supply SMPS

which type are flame proof and what wattage for these,?

[budm]

Re: Magnavox DRV power supply SMPS

MOX: https://www.mouser.com/Xicon/Passive...h8z0?P=1z0zls8
1W or 2W for testing.
https://cdn.badcaps-static.com/pdfs/...24bbb17a37.pdf

[johnfin]

Re: Magnavox DRV power supply SMPS

So what could have taken down the BIG resistor? Would be good to know so it doesn't happen again. Bought parts, waiting....more down time.