Repair ZP120


The amp has no power. Is there an internal fuse for this unit?

Has anyone taken one apart?

407 replies

Userlevel 1
Badge +3
Also check the resistance of the pads for the collector emitter on the PCB.
Userlevel 4
Badge +5
I opened up a non-functioning ZP120. The image shows what I found (it looks like a damaged capacitor to me). This is located just downstream of the rectifier. I looked at other images in this thread, and (while it's hard to tell), none of them seem to show a capacitor in this location -- I wonder if this was somehow dislodged from elsewhere and landed in the right place to blow the fuse?? In any event, as soon as I touched it, it dislodged from the circuit board.

In any event, I bridged the blown fuse with a 32W bulb, and the device booted up just fine (the bulb goes bright and then dims as the capacitors charge up). I get voltages on the board, although they differ slightly from the ones that gruv2ths documented.

As soon as I try to play anything I get the white-off-orange-off cycling lights. The sonos app shows the music is playing, but the volume gets reset to about 15% a few seconds after I increase it.

Coincidentally, I have another connect that has the exact same symptoms (I don't recall if it is a ZP120 or ZP100). The behaviour is the same, and I get a low level of volume on the RCA jacks from it. I have not attempted to open that one up yet.

I'm eager to get these working, and am fairly competent with electronics. I have meters and an oscilloscope and am looking for help to troubleshoot this, if anyone is able to offer help on what to look for.
Userlevel 1
Badge +3
timc995, see below, its a ceramic cap. No idea the PN or parameters. Looks like it is just before or in parallel with the common core inductor. Your primary side voltages may be a little different due to the bulb, it is making a voltage divider with the unit, the thing to watch is the secondary side DC voltages.
This goes for everybody troubleshooting their Sonos Amps, make sure your CPU_GND test point is shorted to the GND36 test point. In the attached picture, you will see a small yellow jumper on the bias converter transformer secondary. That jumper is makes sure the returns are connected together. They need to be.
If they are not, the 3.3V and the 8.5/15V outputs won't be grounded together which meas lots of confusion for the processor.

I think what we are seeing here with most or all of the no power fails on the Connect:AMP is the AMPs inability to handle power surges. There is no active circuit on the front end of the power supply to handle surges. The passive filter does the best it can, which sometimes is not good enough. And since every surge is different the failures are all different. I have not seen the cap you show in the picture blown before.
Userlevel 4
Badge +5
In looking at those two points on the circuit board (where the capacitor blew), I was measuring 278VDC. SO, I guessing that the capacitor was just a filter, I installed the largest capacitor I had that fit the space. I used two 250V 1000pF in series (so 500V 500pF equivalent). I also replaced the fuse with a 5A 250V one.

After this, I had the exact same problems when I booted it up with the 32W dim bulb. It would boot fine, but as soon as I played any music I got the white/orange flashing, and the volume would reset to about 15%.

I was able to see that the 36V test point would jump up to 36VDC then gradually drop to 15VDC over about 2-3 seconds. The fault would come on right as it dropped below about 20VDC. The 278VDC would also drop... to something well below 100VDC. I surmised that my 32W bulb just wasn't passing enough juice, so I removed it. Then the 278VDC rose to 305VDC, and all my problems disappeared. It's rock solid now, and I'm quite a happy camper.

I was even able to effect the entire repair without removing any of the circuit boards. 🙂 Thanks gruv2ths!

Now, on to the ZP100 that's been failing. That one has a bad channel. It either doesn't play from the left channel with no fault, or it throws a fault (white/orange flashing).
Badge
Hi gruv2ths,

I received the power board today. I added my logic board, misc. screws and cover, etc. The amp performed fine with one notable exception. There is noise on both channels. In addition to the general noise, there is a clicking at a speed of exactly two clicks per second. The noise exists whether the amp is muted or not. However, after removing a music signal the noise disappears after several minutes (I assume the amp turns itself off in this case because I don't even here the base amp noise). The noise is much more than the amplifier noise floor. The noise exists when the amp is the only product on my SonosNet. However, when I add my Play 5 the noise becomes noticeably louder. The noise also spikes a bit when I adjust the slider volume via my controller. Any idea what the issue could be?
Badge
To clarify my previous post. The presence of noise is more consistent than I thought. When either the amp is muted and/or the music is paused, the noise disappears after several minutes and the amp is absolutely silent. Otherwise the noise exists. I have the amp hard wired by Ethernet cable to my Bridge which in turn is hard wired to my router.
Userlevel 1
Badge +3
Hmmm.. This is a new one for me.
Does the AMP work at high volumes, even though the clicking exists?
Can you take a video with your phone and upload to YouTube so I can hear it?
So a couple of things:
Have you tried removing and re-seating the riser card?
Have you tried using your riser card as opposed to the one that i shipped with the power board?
Does your phone app indicate that there is a firmware update available for any of your system?
Badge
The noise sounds to be a constant level. it's a crackling sound. The volume of the amp works fine - low to high volume. At modest to high volumes the noise is masked of course. My system is update to date re s/w. I'll try reseating the riser card as well as using mine. I can also send you a video with a recording. However, I'm going out of town until after New Year's so I likely won't be able to post anything further until mid next week. Thanks for the suggestions. I'll pursue next week.
Userlevel 1
Badge +3
Also, are you using the aux/analog input or a streaming source?
If this is a analog source, it could be your source or the RCA cables. Likely grounding between the two.
When you push play from your controller, you are enabling the 36V supply which powers up the amplifier section.
That supply dies after a while once you push stop. Sounds like the noise is following the 36V supply on and off.

Let me know what you find once you're back. Happy New Year!
Userlevel 4
Badge +5
Re my ZP100...

The problem it had was that it was not getting any output on the left channel, and within the app it was detecting a fault. This fault caused it to reset the volume to low levels. This made it unusable as an AMP and also not particularly useful to use the RCA outputs. (I'm not sure -- maybe I could have set those to fixed output -- not sure).

In any event, I opened it up today and couldn't see anything wrong with it. However, during my poking around inside, I must have bridged something... I think It was from the AC to DC board to the chassis as I tried to measure the DC voltages there. There were fireworks and popping. I quickly unplugged it.

The interesting thing (to me) is that it is actually more usable now than before! I took the whole thing apart looking for blown capacitors or anything similar. The only thing I found was the both STA505 chips were quite obviously blown... they both literally had pins missing/blown off (see the image). I figured it was very dead and reassembled it to prove it. Lo and behold... It booted up the first time. Not only that, but it stopped throwing faults! There was no audio on either speaker (not surprising given the amp chips were dead). BUT... both RCA outputs were working. SO despite my boneheaded mistake, I now have a device that can at least be used as a sonos connect (no AMP).

I'll probably try to replace the AMP chips. I have two STA508s. These are the chips that are used in the ZP120, but they are pin for pin compatible with the STA505s that are used in the ZP100. I imagine that even if I am successful in replacing the AMP chips (no small task), there is a high probability that there are now other problems on that board, given the fireworks. On the other hand, the blown pins on the amps look to be only the positive and negative supply pins (4,5,6,7,12,13,14,15). If the board traces survived, maybe all the damage is isolated to the amp chips.
Userlevel 2
Badge +1
Hi folks, I've got a ZP120 with blown power supply too. When powered up I smell "semiconductors". I've been testing an found both Schottky diodes (D16002 and D16006 text: VC3 type: VS-30BQ015PbF) are broken (virtually shortened 0.06V anode-kathode on the diode tester and 0.46V in reverse direction). It seems these are between VPRI and the drain of the Mosfets on the back. Of course I can replace both but just wonder if this is a sign of the main problem. A bit suspect both have blown. The mosfets seem to be okay and also the other pair of diodes in the same circuit (U3J - MURS360T3G; ultrafast rectifiers). Any suggestions?
Userlevel 1
Badge +3
I don't have a unit open in front of me. Can you take a pic of your unit and circle the parts in question.
I know them better visually then by ref-des.

They sound like the 36V supply rectifiers. Could be them or you have a short on the 36V rail. I have seen both problems.
Remove the diodes and see if the parts are shorted.
Userlevel 2
Badge +1
Gruv2ths, thanks for the quick reply. See picture attached. I have lifted one side for both diodes and 36V rail has no shorts.
Userlevel 2
Badge +1
Checked a bit more....smell wasn't semiconductor but electrolyte. The 33uF 400V capacitor died (C16102). See photo....I think I will just replace all of them. Clearly some ageing of parts.
Userlevel 1
Badge +3
Hey man, no problem, I just happen to be at my computer. I think those are opposite the 36V supply FETs.
I think those are protection diodes for the FETs, but not too sure. I would think since they are primary side and shorted that you would be blowing fuses?
This is not so common in my experience that they are indicative of another problem. This is probably the first layer of your onion. LOL
Are you using a dim bulb in series with your 110Vac? You should be if you are not.
Remove the shorted parts and see happens. Do you have a short between the 36V test point and the RTN36 test point?
Userlevel 1
Badge +3
I think that is the vbias cap to the integrated pwm chip. Not common but happens.
2 different circuits from the diodes you showed earlier. Another layer of the onion..
Userlevel 2
Badge +1
First thing I checked was the fuse but that's still fine. Out here we have 230VAC 🆒
Need to check for a bulb, these things are getting hard to find with LED taking over and thr old bulbs no longer on sale.
So far I see no shorts. I just ordered all the parts, including 2x 1000uF. They seem a bit swollen too. I will report back once I have replaced the parts. That will be 2019 for sure 🙂
Userlevel 1
Badge +3
OK cool, the bulb is worth the work to get running, will be helpful for anything off line you test.
I have had a hell of a time getting those 1000uf caps in the right case sive. A little wider and they are too wide, same goes for the height.
Looking forward to hearing how the parts change works out.
If you don't mind me asking, where are you? I am in Texas.
Userlevel 2
Badge +1
Gruv2ths, I measured the 1000uF and have ordered SLPX102M200C4P3 CORNELL DUBILIER - diameter 25mm, height 45mm. Not really high quality but you are right, space is tight on the pcb.
By the way I'm in the city Leiden in the Netherlands. Also have 2x broken Play-5's lying here (1x PSU ticking and 1x amp stages blown up) . Ever done some work on those?
Userlevel 2
Badge +1
Gruv2ths, sorry one more question. Do you have the winding diagram for the transformer PQ3230 in the 36V section? I think mine is dead (would explain the broken Schottky diodes). On the primary side I only find one good winding whereas there are 3 pins with wires attached. On the secondary side no connectivity between the thick wires which run into the pcb. Will be tough to find a replacement...maybe need to take it apart and wind a new transformer myself. Thanks!
Hi Gruv2ths,

I tried both your riser card and mine with no change in the noise level issue. The noise exists whether the analog inputs are used or the source is digital streaming. I posted a video in YouTube. Hopefully you can access it at:

https://www.youtube.com/watch?v=l2rIuxriv_s

I'd like to get my hands on another logic board (i.e. the top board) to see if that would eliminate the noise.
Userlevel 2
Badge +1
Piperdog,

Have you ruled out a problem with the low pass filter (LPF) of the class D amplifier? The noise you hear may be "switching" noise leaking through the LPF...the electrolytic capacitors may need replacing.
Userlevel 1
Badge +3
Pietje Puk - Let me make some measurements on a 36V converter transformer that have here. The continuity should be the same on mine as yours, different turns ratio though I think? (given the lower mains voltage here). If it's the transformer, all is not lost. With the bias converter transformer, when they go open usually once removed you can see where the winding burnt off the termination pin. Fingers crossed that if your transformer has blown open that it can be repaired the same way.

Piperdog - Thanks for the youtube, that helps. Very weird! Wondering if you had a power surge, if maybe some of that over voltage made it through the bias power converter and up to your logic card. If you have a microscope, can you look for cracked ceramic capacitors? I suspect noise that should have been damped is getting in and causing that weirdness.
Pietje, he is using a known good power board so hopefully the Class D stuff is ok.
Userlevel 1
Badge +3
Pietje, here is what I measured.
Let me know if it’s confusing
Userlevel 2
Badge +1
Gruv2ths thanks. I get the same so transformer should be fine then. Need to peel back the onion a bit more. Parts have arrived so will start replacing the diodes and capacitors. Also managed to fix one of the Play5 (1st) gen which I bought recently. Just had to replacw the 10 Ohm resistor in the power supply on the primary side.

Reply