Why is XPROTECT is low?

[willnotwish]

I think that the XPROTECT signal is low on power up. This is why the blue light on the front panel on/off switch is flashing.

The question is: why?

[willnotwish]

My current understanding is that there are three reasons why XPROTECT is low:

1. VPRE+ and/or VPRE- are lower than expected (maybe below 5V when they should be much higher, e.g., 40V). Call this XP_SRC_A.
2. VD+ and/or VD- are lower than expected (not sure of approximate range but similar to 1.) Call this XP_SRC_B.
3. The "fan monitor" circuit triggers a fan fault. Call this XP_SRC_C.

These signals are OR-ed together via open-collector outputs. XPROTECT is pulled up to the 5V+SBU rail via a 10K resistor mounted on the POWER PACK assembly.

[willnotwish]

The main POWER PACK ASSY (actually the amplifier board E) is connected to a couple of auxiliary secondary windings on the main transformer in order to access signal sources 2 & 3.

The POWER PACK ASSY supplies 5V+SBU to the transformer auxiliary PCB. It also carries the XPROTECT signal, which travels via the POWER PACK ASSY, through the COMPONENT ASSY, and thence to the front panel where it is (presumably) read by a microcontroller which signals the standby power board to cut off the mains AC supply to the PSU transformer primary via an isolating relay.

The origin of the XPROTECT signal is the VD level detector output, to which is ORed the fan monitor signal. XPROTECT then travels to the POWER PACK, where the VPRE level detection signal is similarly ORed.

If any of these three signals are low, XPROTECT goes low, the protection circuit is activated causing the power light to flash and the unit to shut down.

[willnotwish]

Initially, the unit was found to start up briefly when the power button was pressed. The display lit up and the volume level was shown, but only for a few seconds (two, maybe) before the unit shut itself down again.

It was subsequently necessary to wait about 60s before trying again. Pressing the power button during this "cool-off period" had no effect.

[willnotwish]

Disconnecting the 6-way connector linking the POWER PACK and the power transformer auxiliary meant that the unit would start; XPROTECT remained high.

Given that XPROTECT is pulled up to the +5V standby rail on the POWER PACK assembly, it is reasonable to assume that XP_SRC_A is not the culprit.

[willnotwish]

However, with the 6-way connector still disconnected, an independent 5V (from a lab psu) supplying a "replacement" 5V+SBU, and XPROTECT pulled up by a temporary 10K resistor (simulating the POWER PACK pull up), XPROTECT remained high with the fan connected, regardless of whether the fan was driven or not.

Simulating a fan failure by disconnecting it physically caused XPROTECT to go low. Otherwise, it stayed high.

[willnotwish]

These two observations are at odds with each other.

More investigations are needed.

[willnotwish]

It would be preferable to isolate the three sources and reconnect the cable linking the transformer auxiliary and the POWER PACK. Each source could then be added one at a time to see which pulls down XPROTECT.

It would appear at this stage that XP_SRC_C (fan detect) is working as designed.

[willnotwish]

Having reassembled the whole unit, I now find that is refusing to start at all, with the relay on the standby SMPS board clicking repeatedly, and my dim lamp tester flashing brightly at regular intervals (every second or so). Previously, it was powering up for a few seconds before shutting down as a result of XPROTECT going low. Now it appears to be drawing too much current, thus forcing the SMPS to shut down.

I think it would be wise to power the main transformer directly from my isolated variable supply, in order to test whether there is a short on the POWER PACK assembly or not. It may be the case that the POWER PACK is fine, but that my dim bulb tester is causing the problem because the inrush current is too high, causing the input voltage to drop and the standby SMPS not to power up cleanly. I have seen this before when dealing with the Samsung TV's SMPS.

[willnotwish]

Turns out the startup issue wasn't an issue at all, once the dim-bulb tester warning lamp (DBTWL) was bypassed. Right now I am running the unit from the main IEC input at 240V with the DBTWL bypassed.

[willnotwish]

There is no issue with XPROTECT from the POWER PACK ASSY, as witnessed by https://github.com/willnotwish/pioneer-repair/issues/2#issuecomment-1193321587

I have broken out the connection between the POWER PACK ASSY and TRANS 3 ASSY W in order to investigate XP_SRC_A, XP_SRC_B and XP_SRC_C.

There appears to be nothing wrong with XP_SRC_A as the POWER PACK ASSY is functional. We know that if the XPROTECT signal is disconnected, it will be pulled high (to 5V) by the POWER PACK ASSY and the unit will start up.

[willnotwish]

The XPROTECT circuitry is shown here

[willnotwish]

XP_SRC_B (the REGB "power good" signal) needs checking. I believe (from previous checks) that XP_SRC_A is OK.

[willnotwish]

I believe I have confirmed that XPROTECT (as shown in the circuit diagram below) is the problem.

[willnotwish]

How does this circuit work? It looks a bit complicated.

XPROTECT is an open collector output from the bottom transistor of the Q4942 NPN pair.

[willnotwish]

RT3P22M equivalents: PUMB1, RN2903

Pinout

[willnotwish]

RT3N22M

[willnotwish]

Redrawing the circuit may make it easier to understand.

[willnotwish]

[willnotwish]

My suspicion is that Q4941 (the PNP pair) has a faulty transistor (pins 3, 4, 5).

I may be able to check by comparing it to the other transistor in that package (pins 1, 2, 6).

[willnotwish]

The check is difficult to do with the package in place. I think I'm going to have to remove it. I will attempt to duplicate its functionality with larger discrete components on a breadboard.

I think any general purpose PNP transistor with a gain of about 100, a max current of 100mA and a max voltage of 50V should do it. The bias resistors are nominally 22K.

The tricky bit is breaking out from the original PCB to the breadboard. Very fine wire is needed.

[willnotwish]

Note that both Q4941 and Q4942 packages are symmetrical: it doesn't matter which way round they go. This is convenient, but now I have removed Q4941 I don't know which of its two transistors is which. I guess I'm looking for any kind of difference between the two. It may be the case that both are bad.

I will need to solder Q4941 to a breakout PCB (if I have one small enough) to get access to the individual pins because the package is tiny.

[willnotwish]

I think Mouser has stock of Q4941. Farnell is out of stock at the moment.

This is incorrect. Farnell has stock.

[willnotwish]

I used a 0.65mm pitch breakout adapter. It was tricky though, because Q4941 was a bit too narrow.

[willnotwish]

Having broken out Q4941 to a breadboard, I find it works as expected. :1st_place_medal:

Specifically, 5V is present at the collector (pin 3) of the PNP transistor. This pushes up the input voltage at pin 2 to +5V_UST, thus turning the other PNP transistor off. XPROTECT is therefore HIGH.

I don't know why it didn't work before. Perhaps there was some sort of dry joint.

The test will come when I solder it back. That will have to wait for a bit because I am away until next weekend now.

[willnotwish]

I also need to test the unit as a whole to check it works OK. So far, all I have been able to do is power it up and check the display. I haven't actually used it as an amplifier!

[willnotwish]

See https://www.youtube.com/watch?v=KZG172eeeR0 for some amusement.

[willnotwish]

Amp works OK but there is a buzzing sound from the speakers. See https://github.com/willnotwish/pioneer-repair/issues/4