Magimix M 100 Coffee Machine PSU

All the regulation needs to be is zener diode and resistor, if that. The low voltage appears to be derived from a series resistor and the large flat blue capacitor. See what the circuitary is around the low volatge C near all the white wires?

Be careful when tesing with it live mains can bite and mains applied to the wrong place in the LV side (probe slip) will probably kill it completely.

Thanks Dave,

The white wires are for.. the two pin socketed pair on the right - temperature probe mounted on the boiler unit

so do you think the LV is coming from the big blue cap, and the small resistor between it? that one measures 0.455M while in circuit (off).

I don't have a capacitor test option on my meter :(

Not sure how to tell the value of that cap either :-\ - here is a picture, can you help!?

Ta!

Looks like 680nF at 275V to me.

TOJ.

it might be a discrete one. Most likely a shunt reg.

no, far too high

I'd follow the pcb to see where it goes after/before the C. You'll need to fault-find & replace something. Even if you fit a wallwart to provide 5v you still need to fix something.

NT

In effect yes, trace out the board as a circuit diagram and you'll find that you have Live > high wattage resistor > big blue C > rest of circuit. The small high value resistor is to bleed away residual charge on the capacitor. After the big blue C you will probably find some retification and smoothing (the LV C next to the white wires?).

Toby was thinking very hard :

There is a rather cheap and crafty way to provide low voltage power, using a capacitor to buck the AC voltage - and that is what I think I am seeing here. I think the large blue cap marked 275v is the cap which bucks most of the mains, hence its voltage rating, the two black diodes are the rectifiers and the two maroon diodes next to them might be zenor diodes to regulate the supply.

Of course it is just a guess and I might be very wrong. These things are very difficult ot work on, when they have mains supplies mixed in with the low DC voltage you are trying to test.

En el artículo , snipped-for-privacy@care2.com escribió:

It's a 470k resistor, so a reading of 0.455M in circuit looks ok to me.

En el artículo , Dave Liquorice escribió:

There's 4 diodes in the bottom left hand corner fed by the big blue C and the 470k resistor. Two of them seem to form a rectifier with the output fed to pin 14 of the PIC, which I assume is Vcc. The other two diodes might be 5V zeners.

I'd check the diodes first, then lift one leg of the big grey resistor marked 254 and the 470k resistor and check they have the correct values out of circuit.

I'd be wary of that heatsink on the triac being live with power applied.

I would have agreed with you 100% until 12 months ago when the humdity switch in the bathroom failed to turn the fan off on occasions. Eventually traced to low volts at the trip reference - and the cause was the large cap:-

But very high for a typical CR PSU. Xc= 1/ 2 pi f c = 0.0047 Mohm = 4k7 R= 470k So total impedance of the series pair is apx 470k. On 240v-5v that allows v=ir, i=v/r = 0.5mA.

Its not impossible, but its too low to power much, and the CR impedance rat io is a bit senseless.

NT

Yeah but it's across the C as a bleed resistor not in series. The big grey high wattage jobbie is in series...

Yes, so its that grey R thats significant in the psu

NT

Just make sure it's an X2 one.

Okay, I have finally lifted some components, and it looks like the capacitor is the culprit here!

I had another DMM with a capacitor setting, and (out of circuit) it is showing as just under 138nF.

I have had a look through my various boxes of bits, but don't have another one, so will order one and report back!

Using a plain 600v one would be ok too

NT

Just to report back, I have now changed the capacitor, and it seems to be working properly now :-)