Repair ZP120

Fantastic, thanks Tim(?). 
not often a cheap repair with Sonos!

i’ll go and find a bit of wire now!

Kindest regards,

Andy

I’m a bit stumped on a ZP120 and am looking for help.  The zp120 had a bad varistor, between VPRI and VMID), with lots of scorched bits when it caught fire.  It took out the PWM chip, the fuse and a board trace.  I’ve replaced those but still can’t get it working properly.

The attached image is from the underside of the zp120 board, below the large capacitors.

I have circled two small circuits in green.  The one on the left looks like a bleeder resistor that is intended to discharge the large 1000uf capacitor that sits between VPRI and VMID (about 150V).  The 68k and 240k resistors provide 308k of resistance across the capacitor.  I thought the small “p1.c” diode is some kind of snubber diode, but I’m not so sure any more, based on what’s going on in the other circuit...

The similar circuit one on the right seems to bleed the 150V between VMID and PGND.  That circuit also has the addition of D16035, which sits between the bottom legs of R16082 and D16041.  This additional chip is a 19V zener diode in a SOT-23 package.  (MMBZ5249B).  Pin 3 of this zener diode (which is presumably supposed to be at 19v when the circuit is working property) is the voltage that drives the PWM chip.

My board reads 0V at that junction.  I have replaced the zener with no luck.  When I provide 16V from an external source, the PWM operates as it’s supposed to.  The PWM chip needs 15V to get started, but operates ok with as little as 9V… if I remove my external 16V, the circuit continues to operate (and somehow the circuit now internally generates about 12.5V at the same junction of R16082 and D16041).

I can’t figure out how this circuit works.  Specifically, the junction between R16079 amd R16082 reads about 110V when powered up (this seems right given the resistor divider created by R16082 and R16079).  I understand that the zener is supposed to clamp that to 19V, but what is the P1C diode (D16041) doing?  If it is a regular diode then it wouldn’t pass any voltage to the zener.  If it isn’t a vanilla diode, then what is it?  I’m guessing D16041 isn’t working but I don’t know what to replace it with.  It does not read a voltage drop in either direction when tested (the other P1C reads 0.68V).  I can’t find a reference to that marking anywhere.

Hi, can you sketch the circuit around these diodes out for me?

I would imagine if the other one gives .68v, then it is probably a general purpose silicone diode. 

As I say, sketch it out and I'll have a think.

Kind Regards,

Andy

Sure.  The attached is how I think the circuit is organized.  I don’t quite ‘get’ how the 19V zener is providing 19V in this, but working devices do in fact read 15-18V when running.

D16040 reads with 0.68 forward voltage drop.  I wonder if there is a reverse voltage avalanche breakdown property that is allowing some current to pass to the cathode of D16041.

I had a typo:  I mislabelled the zener.  It is D16035 on the board (not D16041).

So I gave up on trying to solve the problem with the existing circuit.  I ended up replacing the 240k resistor at R16082 with a resistor divider network.  I used 15k+4.7k+100k, with one leg of the 100k tied to the 59k resistor, and one leg of the 15k tied to GND.  The junction between the 4.7k and 100k resistor should then end at about 18V (ie approx (15+4.7)/(15+4.7+100+59)*sqrt(2)*115).  I tied that to the VCC input of the PWM.  I used ¼ watt resistors for all three of the resistors.  The lower resistance at R16082 (~120k instead of 240k) drops the voltage at the junction between R16079 and R16082 to 109V.  The current flowing through the resistor divider is 0.9 milli-amp when the PWM draws no power, and <1.1mA when it acts as a short circuit.  As a result, power dissipation for the largest resistor (100k) in the resistor network is under 1/10 watt (making the ¼ watt resistors overkill)

The circuit works fine now.  I didn’t time it, but it may take a couple of seconds longer to boot up.  There is a 100uF capacitor on the PWM VCC, and that needs to charge up before the PWM VCC line exceeds the 16V it needs to start oscillating.  Otherwise, there is no issues with the circuit.

Maybe this only works so well since the PWM chip is tolerant of voltage fluctuations on its input (it needs to start up with >16V, but then will operate with any voltage from 8-24V).  The original circuit may have provided a more stable VCC, but it’s just not needed given there is only one chip drawing less than a milliamp.

Hi, sorry for not coming back sooner.

looks like the 4 resistors are acting as a voltage divider off the 320v rail. The 19v zener is then just working in the normal way. If any of the resistors in the chain is out you wont get 19v. I thin d16041 is there to stop feedback once the pwm chip is providing its own supply.

if you got it working ok as you described, then i guess thats fine. If not, then check the resistors.

Kind regards,

Andy

Hello, im from Germany and need your Help.

These two capacitors are defective. do you know the data for these?

TH16001 and TH16002.

Thank You

Max

Hi Max.  Those are both NTC thermistors (not capacitors).  They can be replaced with part # SCK-054.

They act as 5 ohm resistors at room temperature.  As they warm up, their resistance drops.  This prevents a large amount of current inrush at power up.

In a pinch, you could try replacing them with jumpers, but I would only do that if you are powering the board with a dim-bulb (as described earlier in this message thread).  In that case, the incandescent light bulb will limit inrush current.  

Hi Tim, thank you. I have now ordered two of them. Can I destroy them by changing the AC voltage from 230V to 115V on the 230V grid?