Yamaha RX-V483 dead after thunderstorm

[mmalka]

Hello,
I am troubleshooting a Yamaha RX-V483 AV receiver that became completely dead following a thunderstorm. There is no standby operation.
I have the Yamaha RX-V483/HTR-4071 service manual and have been troubleshooting the OPERATION (2) AC-DC standby power supply rather than replacing parts at random.
The primary side is alive, but the standby SMPS is not producing a useful secondary output. The most unusual finding is that IC541 (STR3A453) VCC is approximately 34.5V, whereas the Yamaha service manual indicates approximately 16.58V for my U model.
Equipment: AstroAI DT132A True-RMS DMM. I do not currently have an oscilloscope or bench power supply.
Measurements/results so far:
AC input: ~121.7 VAC
Main fuse: good
C5419 primary bulk voltage: ~163.9 VDC
C5419: 215 µF (specified 220 µF)
R5404 startup resistor: 3.02 MΩ (specified 3 MΩ)
R5403: 1.1 Ω (specified 1 Ω)
C5420, IC541 VCC capacitor: 51.2 µF (specified 47 µF/50V)
Voltage directly across C5420: 34.5 VDC
IC541 pin 2 VCC → pin 3 GND: 34.67 VDC
IC541 pin 4 FB/OLP → pin 3: 0.264 V (manual shows ~2.928V)
IC541 drain-to-ground resistance: several MΩ, polarity dependent; not shorted
R5407: 45.5 Ω (specified 47 Ω)
D5403: 0.493V / OL
D5404: 0.606V / OL
D5406: 0.661V / OL
D5407: ~0.661V / OL
D5414: 0.494V / OL
D5415: 0.609V / OL
D5409 secondary dual diode: symmetric readings; no obvious short
Secondary voltage is extremely low; approximately 1.04 VAC across D5409 outer terminals
IC544 (TL431): REF→A 0V, K→A 2.3V
IC543 optocoupler: no obvious short
T5401 accessible windings have continuity and isolation; no open winding found.
D5401 is particularly puzzling.
The service manual identifies D5401 as UDZV15B (15V Zener). It measures 0.705V / OL in ordinary DMM diode mode, which of course only verifies its forward junction and does not test its Zener breakdown voltage.
I have positively identified D5401 from the PCB silkscreen. With the receiver unplugged and the primary capacitor discharged:
D5401-L → C5420− / primary GND: 0 Ω
D5401-R → C5420+: ~4.9 MΩ, rising
D5401-R → IC541 pin 2 VCC: ~4.9 MΩ, rising
D5401-R → IC541 pin 1 S/OCP: ~4.59 MΩ
D5401-R → C5420−: ~4.59 MΩ
This is confusing me when I compare the physical PCB connectivity with my interpretation of the service-manual schematic.
At this point the evidence seems to put the problem in or around IC541 and its VCC/startup/auxiliary/control circuitry. The primary DC supply is present, but the secondary is essentially dead and the feedback circuit consequently does not appear to be regulating.
My main questions are:
What would cause the STR3A453 VCC to reach ~34.5V instead of ~16.6V?
Does that measurement itself strongly suggest a failed STR3A453, or could another component in its VCC/auxiliary circuit cause it?
How should I interpret D5401's PCB connections relative to the schematic?
What would be the most diagnostic next measurements before replacing IC541?
Is there anything in the measurements above that points to a different failure mechanism?
I am comfortable taking resistance, diode, capacitance and carefully selected live-voltage measurements, but I only have a DMM, so I cannot currently examine the switching waveform.
I cannot attach the Yamaha service manual for being too large, but it can be found at   https://www.manualslib.com/manual/3194262/Yamaha-Rx-V483-Htr-4071.html

Any guidance on narrowing this down to the actual failed component would be greatly appreciated.

[GraemeG]

Here is the power supply circuit from the manual. This might save others time trying to track it down in the huge manual! I haven't had time to actually look at it yet.

[MathWizard]

So was it plugged in and the standby power was on, or was the main power also on during the storm?? Well I'm jumping to conclusions here for fun, I made it 2/3 the way through your list but never looked up any chips yet. But how often can an over-voltage of the mains, while driving the primary side switchers, result in an over-voltage high enough on the secondary side to break stuff ? With the low voltage on the IC541 pin 4 FB/OLP pin, I'd wonder what that the feedback network is doing and what opto-controller network is also probably tied too. I'll look at the datasheet latet, I'm playing F1 2018, driving the old Ferrari 2002 V10 cars

[Bud]

How come the secondary is dead if you measured voltage across TL431 ?

[GraemeG]

I think MathWizard might be on the right track, there is definitely something odd with the feedback.

I've attached the datasheet for the STR3A453 PWM chip. Looking in section 10.1 at the startup operation, it shows that there is an internal constant current source to initially charge it's Vcc capacitor C5420 and it should start switching when it reaches 15V (typical). Once switching has started then the auxiliary winding from the transformer keeps C5420 charged via the diodes D5403 and D5414. As you mention, the schematic shows the voltage on the cap should be about 16.58V. Given that it is around 34.5V on Vcc and the FB pin is much too low at 0.26V instead of the correct 2.9V, I can only assume the switching circuit is actually working but I'm guessing the feedback signal is missing so it is running as hard as it can to try to lift the feedback signal. The absolute maximum rating for Vcc of the chip is actually 32V so you are lucky it hasn't smoked!

D5401 on the feedback (FB) pin makes sense as the absolute maximum rating for that pin (from the datasheet) is 14V so they used a 15V zener to try to protect it. Yours is probably OK as it still meters like a normal diode. If it measures "open" from K to A then it isn't shorting out the feedback signal.

I have a couple of concerns. Measuring "1.04 VAC across D5409 outer terminals" is not valid. According to the circuit diagram, these 2 terminals are connected together (the 2 diodes are in parallel) so the voltage should be 0 although you may be reading some noise or ripple? The two secondary windings of the transformer are connected in parallel so you need to measure the AC from pins 8 & 9 to pins 10 & 11 of the transformer or from the negative of C5422/C5423 to one of the outer terminals of the diode. But a simpler option will be to measure the DC voltage on C5422/C5423 which should be about 5.5V according to the schematic. Keep in mind that the SMPS chip runs at about 65KHz and the waveform won't be sinusoidal so depending on your multimeter you may not get sensible AC readings.

Another problem is the 2.3V K-A across the TL431 reference. This should be around 4.4V (from the schematic 5.418-0.986V) and the reference pin should be 2.486V. This implies the opto-isolator LED is not lit and so there is no feedback. The reference pin should get it's voltage via R5414, R5415 and (0 ohm link) R5416 and then it is measured by the TL431 across R5431. R5431 could be shorted or the other resistors could be open, or even L5401 could be open, but in that case you will measure a high voltage, maybe 10V or more, on C5422/23.

Going back to the primary side, I just realised that the voltage on the FB pin is generated by the STR3A453 chip itself. A block diagram in the datasheet shows it is supplied internally from Vcc and normally the output transistor of the opto-isolator will pull this voltage down to reduce the output. With the voltage at just 0.26V the chip is either damaged internally around the FB pin or the opto-isolator is driven almost to saturation and pulling the voltage down or C5403/D5401 is shorting the signal. But in this case, if the chip was working correctly, the Vcc voltage wouldn't be so high.

That is all just a long way to say I have concerns about the chip being damaged internally around the FB input. But I am also confused that the secondary doesn't appear to have correct or excessive voltages given Vcc is so high. So maybe there is something wrong on the secondary side. And of course there could always be more than one problem!

Can you get a measurement of the secondary voltage on C5422/23? Also across the opto-isolator diode (shown as 0.986V on the schematic). And as a sanity check maybe try an AC voltage reading across pins 5 & 6 of the transformer. This should correspond to the high Vcc reading and, allowing for my previous comments on the high frequency and non-sine wave shape, 34.5V rectified should be something like 25VAC across the transformer winding. I'd like to be sure the high Vcc is coming from the transformer and not from the internal chip startup circuit. Maybe the chip is not actually running at all? But then why is there some voltage on the secondary?

A lot of this is just me "thinking out aloud". I might be totally off the mark!

[mmalka]

It was fully powered on, and being used, while watching TV. The TV survived. There was lightening nearby (could never find out where exactly), the lights flicked but never went down. Damages: Frontier DVR, Frontier ONT, 2 routers, 1 switch, irrigation controller, 1 printer, 2 LED mirror power supplies, and the Yamaha receiver. Not fun...
As no power is currently reaching the main board, I have been focusing on the standby board, but I am no expert.

Your point about the feedback network is where I’m looking now too. The primary bulk supply is present at about 164VDC, but IC541 has a very abnormal combination of VCC = 34.5V and FB/OLP = 0.264V.
I’m going to make some more direct measurements on the secondary feedback circuit, particularly the actual 5.5V rail and the optocoupler LED, and will post the results.

[mmalka]

Good point. Calling the secondary “dead” was too strong.
There is clearly some secondary-side voltage because I measured about 2.3V K-A on the TL431. What I really meant is that the secondary does not appear to be producing the expected regulated ~5.5V standby rail.
I haven’t yet measured the DC voltage directly across C5422/C5423, which is a much better test. I’ll do that next and post the result.

[mmalka]

Thank you — this is extremely helpful, and the datasheet explanation clears up several things.

One point I noticed after reading the STR3A453 datasheet more closely: I think the FB behavior may actually be the opposite of “running as hard as it can.” The datasheet says that as the optocoupler sinks more current, the FB/OLP voltage decreases and the peak drain current is reduced. For the STR3A453, oscillation starts entering standby/burst behavior around the 3V region, and the oscillation-stop threshold is only about 1.53V.

Since mine is sitting at only 0.264V, that would seem to indicate that either the optocoupler is pulling FB very hard toward ground, something external around FB is dragging it down, or the FB section of IC541 itself has been damaged.

The VCC reading is also difficult to explain. I measured 34.5V directly across C5420, not just at the IC pins. The datasheet gives 32V as the absolute maximum VCC and about 29.1V as the OVP threshold. Mine is the STR3A453, not the “D” version, so as I understand the datasheet it should latch off when OVP is reached. A steady 34.5V therefore seems particularly abnormal.

I agree completely about my D5409 measurement. Measuring AC between the outer terminals was not a useful way to determine the secondary output, especially with a ~65kHz non-sinusoidal waveform.

My next measurements will be:

DC directly across C5422/C5423 to establish the actual secondary rail voltage.
Voltage directly across the IC543 optocoupler LED, which Yamaha shows at about 0.986V.

After that I can check the T5401 auxiliary winding at pins 5–6 if necessary. I’m a little cautious about interpreting that AC measurement with my DMM because of the switching frequency and waveform shape, but it should at least give us another clue.

Also, your explanation of D5401 makes sense. It is indeed on the FB/OLP network, not the VCC supply. Mine measures 0.705V forward / OL reverse in diode mode, although of course that does not verify its 15V Zener breakdown characteristic.

I’ll post the C5422/23 and optocoupler measurements next. Thanks again — this has helped narrow the diagnostic path considerably.

[GraemeG]

The datasheet says that as the optocoupler sinks more current, the FB/OLP voltage decreases and the peak drain current is reduced. For the STR3A453, oscillation starts entering standby/burst behavior around the 3V region, and the oscillation-stop threshold is only about 1.53V.

Since mine is sitting at only 0.264V, that would seem to indicate that either the optocoupler is pulling FB very hard toward ground, something external around FB is dragging it down, or the FB section of IC541 itself has been damaged.

The VCC reading is also difficult to explain. I measured 34.5V directly across C5420, not just at the IC pins. The datasheet gives 32V as the absolute maximum VCC and about 29.1V as the OVP threshold. Mine is the STR3A453, not the “D” version, so as I understand the datasheet it should latch off when OVP is reached. A steady 34.5V therefore seems particularly abnormal.

D'oh! I didn't pay enough attention and got the feedback behaviour inverted. Yes a lower voltage should reduce the output not increase it! And I missed the OVP issue. Given the original cause (power surge) I think there are enough doubts about the way the chip is behaving that I'd be ordering a replacement 

Another random thought, if the optocoupler is pulling the FB voltage down there will be a voltage drop across R5405. If there is no voltage drop then the chip itself is not providing the voltage on the FB pin or C5403 is shorted (less likely).

[mmalka]

Update — two open fusible resistors found in the feedback/control area

I made a few more measurements based on the suggestions here.

R5405 and R5409 both measure open circuit (OL).

I initially questioned the readings because these SMD parts are very small, so I verified probe contact by touching both probes to the same solder termination on each resistor. That gives essentially 0 Ω, confirming good contact. Measuring from one end of the resistor to the other gives OL on both.

According to the Yamaha parts list:

R5405 = 100 Ω, 1/16 W fusible resistor
R5409 = 100 Ω, 1/16 W fusible resistor

Yamaha lists both as the same part, ZN108500.

I also located and tested C5403, which is the capacitor directly associated with the IC541 FB/OLP circuit.

Specified: 2200 pF / 100 V
Measured capacitance: 2.4 nF
Resistance across it: 4.64 MΩ

So C5403 appears normal and is certainly not shorted. The service manual identifies C5403 as 2200 pF / 100 V.

D5401 also previously measured about 0.705 V forward / OL reverse, so there is no obvious short there either.

The important question now is what the open R5405 means for the strange IC541 readings.

We previously measured:

IC541 VCC = ~34.5 V
IC541 FB/OLP = 0.264 V
Yamaha schematic expectation is approximately 16.58 V VCC and 2.928 V FB/OLP

R5405 appears to be in the optocoupler/FB path. If it is open, then as I understand it the optocoupler should no longer be able to pull the FB pin low through that path. The STR3A453 datasheet says that when secondary feedback current disappears, FB normally rises rather than collapsing toward zero.

This makes the 0.264 V FB reading seem even more suspicious of IC541 itself, especially since C5403 is not shorted.

On the other hand, because R5405 is definitely open, it could have caused loss of regulation and possibly contributed to the abnormal VCC condition before IC541 went into protection or was damaged.

My questions before replacing anything are:

1- Would you replace R5405 and R5409 first and then retest the supply before replacing IC541?
2- Could an open R5405 by itself plausibly explain the 34.5 V VCC, even though it does not seem to explain the 0.264 V FB?
3- Since these are fusible resistors and both are open after a surge event, should I assume they opened because of another fault and check anything specific before powering up with replacements?
4- Would you replace IC541 at the same time, given that its VCC has already exceeded its 32 V absolute maximum?

I would prefer to identify the original cause rather than simply replace parts and risk opening the new fusible resistors again.

[GraemeG]

Curious! I didn't consider the resistors being open. R5405 and R5409 are in series with the optocoupler IC543 so all of them have been hit by a substantial surge that can only have come via IC541 the SMPS chip.

Perhaps IC541 is in some sort of over voltage shutdown that is causing the FB voltage to drop but I'm very suspicious that it is damaged. If you are feeling adventurous you could temporarily replace the 2 resistors with ordinary 100 ohm devices and see if the unit powers up.

At this stage I would replace IC541, R5405, R5409 and IC543. Even if the optocoupler or SMPS chip are still working they have all been stressed by the power surge. It's much less likely that components on the secondary side have been damaged but you can verify that once the primary is working. The STR3A453 is not a common device but there seem to be a couple of sites online where you can order small numbers.

https://www.alibaba.com/product-detail/Original-New-in-Stock-PMIC-AC_1600713453957.html
https://www.axis-part.com/product/detail/store/8745514/STR3A453.html

The TLP785 optocoupler seems widely available. I haven't looked for the resistors but fusible resistors are also widely available so if you can't get the exact part you should be able to get an equivalent. I haven't checked shipping costs and there may be other vendors so do your own search. I would order at least 2 or 3 of all these components to allow for testing!

[aeg]

STR3A455 is a beefier version of STR3A453 and is in stock at Arrow. This might be less risky than buying from Alibaba. The parametrics are slightly different, though, and I'm not 100% sure at a quick glance that it's a drop-in replacement. Arrow also has STR3A453D which is like the STR3A453 but with hiccup type rather than latching shutdown.

[GraemeG]

Good call @aeg

The STA3A455 appears to be an exact drop in replacement and is probably a better device to use in place of the 453. It has a lower on resistance and higher power capability so should run a little cooler in circuit. Probably no significant difference but if the 455 is easier to get I would go with that for the replacement. The D version should work perfectly too but as you mention aeg it will hiccup in an overload situation. I would go for a non-D version if I could get one but wouldn't hesitate to use a D if that was easier to get.

Arrow seem to have them although you have to buy 4 minimum https://www.arrow.com/en/products/str3a455/sanken-electric.html
Also Veswin and maybe a few other sites plus the ever present but sometimes questionable Aliexpress.

Try to get all your components on the one order so you minimise the shipping costs. Shipping is likely to be more than the component costs.

[mmalka]

Thanks a lot everyone — this has been very helpful!

The finding that both R5405 and R5409 are open seems especially significant now that I understand they are both in the IC543 feedback path. Since IC543 and those two fusible resistors would all have been exposed to the same abnormal event, I agree that replacing only the resistors and trusting the other components is probably not the best approach after a lightning/power-surge failure.

I also checked C5403, which measures about 2.4 nF versus 2.2 nF nominal, and about 4.64 MΩ in resistance mode, so it does not appear to be shorting the FB node.

Considering the other measurements — particularly IC541 VCC at about 34.5 V, above its 32 V absolute maximum, and FB/OLP at only 0.264 V — I am now planning to replace:

IC541, IC543, R5405 and R5409 together.

I think I will skip the temporary ordinary-100Ω-resistor experiment. I understand why it could be useful diagnostically, but since I don't have a current-limited bench supply/variac setup and the existing parts are specifically fusible, I'd rather restore the proper protection before powering the board again.

The STR3A455 suggestion is particularly useful. After comparing it with the STR3A453 family information, the non-D STR3A455 appears to have the same basic control/protection characteristics and pinout, with the main difference being the more capable/lower-RDS(on) MOSFET. Since Arrow has genuine STR3A455 stock, that looks considerably safer to me than taking a chance on marketplace STR3A453 stock.

So at this point I am leaning toward using the non-D STR3A455 from Arrow for IC541 rather than the STR3A453D or an independently sourced STR3A453. Update: Just quoted with Arrow and the shipping cost from Japan kills it. The challenge is that I cannot find the 453 or the 455 in stock at any reputable distributor in the US.

Arrow also appears to carry the Toshiba TLP785, so I'm trying to source everything from one supplier to avoid paying several shipping charges.

The remaining issue is R5405/R5409. Yamaha identifies both as 100Ω, ±5%, 1/16 W fusible resistors, part ZN108500. From measurements against a ruler in my board photo, the component body appears to be approximately 2.0 × 1.1–1.2 mm, so it looks like an 0805 / 2012 metric package.

Before I order, does anyone see any reason not to use the non-D STR3A455 in this circuit, or any parameter difference that we may have overlooked?

Also, if anyone knows a suitable currently available 100Ω 0805 fusible resistor equivalent for ZN108500, particularly one stocked by Arrow, I'd appreciate the part number.

[GraemeG]

Your plans sound reasonable to me.

Just for my curiosity, how much is Arrow's shipping? In Australia I have to build up a AUD$60 order to avoid a AUD$24 shipping charge from Mouser and several of the other large suppliers. That's why I usually head to Aliexpress for the non-critical bits.

Fusible resistors are not so common in the size you need. I found https://www.digikey.com.au/en/products/detail/kamaya-inc/FRC20C2A101JTP/15181242 although it is a 1/10W version. I also see references to this product https://www.mouser.com/datasheet/3/1099/1/SEI_FRC.pdf but it is in a tiny 0402 package and doesn't seem to be stocked.

Hopefully someone else can help. I have only had a quick scan so there may be another supplier with the exact equivalent.

Failing an exact match I would happily use the 1/10W components.

[mmalka]

Thanks — that helps a lot.

Arrow's shipping was actually the deal-breaker for me. The STR3A455 parts themselves were only about US$11.32 for four, but Arrow quoted US$51.99 shipping from Japan, plus a small tariff and tax. That made the total over US$64, which doesn't make economic sense for this repair.

I had independently found the same Kamaya FRC20C2A101JTP at DigiKey, so your recommendation gives me more confidence in using it. DigiKey shipping to me is only about US$5, so I’m leaning toward ordering a few of those and first replacing only R5405 and R5409.

Since those two resistors are definitely open, this would be the cheapest way to see whether the original STR3A453 and TLP785 survived the surge before I spend more money on the harder-to-source parts.

If the supply voltages return to normal after replacing the resistors, great. If IC541 VCC is still around 34.5 V or FB remains around 0.264 V, then I’ll know I need to proceed with IC541/IC543 replacement.

I agree that the 1/10 W rating is not an exact match for Yamaha’s original 1/16 W part, but since it is still specifically a fusible resistor, I’m comfortable using it for this repair/diagnostic.

[Kean]

I don't know anything about them, but I saw via Octopart that a HK/Shenzhen based company called Axis Part list stock of both STR3A453 and STR3A455 (also mentioned by GraemeG in post #10 above).
They also seem to offer EMS shipping which might make buying some more viable compared to Arrow/Verical.

You could maybe also offer any extras for sale to people local to you via buy/sell section of this forum or on ebay.  It might take a little while to move, but you could help others and possibly recoup your costs.

Unfortunately their company name is not one of the easiest to search for and check reputation.  I'd guess they would be less risky than AliExpress, but for a part like this I wouldn't be too concerned about trying AliExpress as a source if it was reasonably cheap and there wasn't another viable option.

[aeg]

Arrow's shipping was actually the deal-breaker for me. The STR3A455 parts themselves were only about US$11.32 for four, but Arrow quoted US$51.99 shipping from Japan, plus a small tariff and tax. That made the total over US$64, which doesn't make economic sense for this repair.

Yikes, that's awful. They give free shipping with a $100 order, so maybe that's an option if you are one to occasionally order a lot of parts.

[mmalka]

I don't know anything about them, but I saw via Octopart that a HK/Shenzhen based company called Axis Part list stock of both STR3A453 and STR3A455 (also mentioned by GraemeG in post #10 above).
They also seem to offer EMS shipping which might make buying some more viable compared to Arrow/Verical.

You could maybe also offer any extras for sale to people local to you via buy/sell section of this forum or on ebay.  It might take a little while to move, but you could help others and possibly recoup your costs.

Unfortunately their company name is not one of the easiest to search for and check reputation.  I'd guess they would be less risky than AliExpress, but for a part like this I wouldn't be too concerned about trying AliExpress as a source if it was reasonably cheap and there wasn't another viable option.

Thank you but Axis is charging too much for the parts and shipping to the US is $62, what makes it not feasible.

[Kean]

Thank you but Axis is charging too much for the parts and shipping to the US is $62, what makes it not feasible.

Fair enough.  Often they wont honour the prices shown on their site, which are just there to draw you in.  But sometimes you'll get a sales rep in a good mood who will be amenable to help with cheaper shipping even if it is more work.  Probably less common these days with all the extra paperwork for tariffs and related charges for US & EU customers.

In that case, I'd probably take the risk on the cheapest eBay or AliExpress part that is compatible and doesn't look too dodgy.  It may not be an optimal replacement, but will at least help you move forward with the repair and see if there are more issues to solve.

[MathWizard]

Ok I'm still too lazy to read all the posts, but at least I'm not having a drink like the last time...Yeah so if there's no switching going on, and there's 34V on the Vcc of switcher IC, that would sound like it's cooked internally, and that voltage was coming through from the high voltage on pins 5-8. Otherwise if the transformer was ok, and not switching, there should be no voltage for the Vcc pin (from the outside anyways).

So that FB pin was actually an output pin, but it measured low, even though the fusible 100 ohm R5405 was damaged? So the IC543 wasn't pulling it low ? If so then yeah there's probably more internal damage. And I guess some high current out of the FB pin, with the IC543 turned on must have cooked the 100R resistor.

I'd guess the bit of voltage on the secondary side, was that coming from the separate low voltage rail with the IC523 7812 12V circuit ? Again I didn't read all of it here or look at all the schematics GraemeG posted, but it reminds me of similar circuits I've worked on.

But yeah when I first read your thread, I missed the 34V on that Vcc pin, that's not good at all, especially if no switching is going on.

[mmalka]

Thanks — I think your interpretation is increasingly consistent with what I’m seeing.
One detail I found in the Sanken datasheet is that the STR3A453 has an internal startup-current circuit connected to pins 5–8, so even without transformer switching the IC can charge the VCC capacitor from the high-voltage side. However, it should normally start around 15 V, and the OVP threshold is around 29.1 V. My measured ~34.5 V is above even the 32 V absolute maximum, so I agree that this looks very suspicious internally.
The other point that concerns me is FB. R5405 is definitely open, so IC543 should no longer be able to pull the FB node low through the normal feedback path. C5403 also tested good at about 2.4 nF and is not shorted. Yet FB remains only 0.264 V. According to the datasheet, with loss of optocoupler feedback the internal FB current source should tend to make FB rise, not sit near zero.
So the combination of 34.5 V VCC and 0.264 V FB, with R5405 open, seems to strengthen the case that IC541 has internal damage.
I’m not sure normal FB current could have opened the 100 Ω resistors, though. My understanding is that normal feedback current is very small, so perhaps an internal surge/failure caused abnormal current through that path.
Regarding the 2.3 V I measured around the TL431 secondary feedback circuit: I checked the Yamaha schematic and IC523 is the separate 7812 regulator fed from the main power transformer, rather than from T5401, so I don’t think it normally explains that standby-side voltage unless there is some backfeed path.
At this stage I’m still considering replacing the two confirmed-open fusible resistors first as a low-cost diagnostic step, but I agree IC541 is looking increasingly suspect.

[GraemeG]

And don't forget the optocoupler. If there was enough current to blow the resistors then the optocoupler has at least been stressed, if not destroyed. I would replace it even if it appears to be working.

[mmalka]

Thanks — good point. I have only ordered the two replacement fusible resistors so far, not the optocoupler.
I will try to go to a local hackerspace next week where they have proper soldering equipment and an oscilloscope. My plan is to have R5405 and R5409 replaced first, then test IC541, IC543 and IC544 before deciding what else to replace.
Given your comment, I’ll make sure IC543 gets checked carefully. If it shows any sign of damage or degraded operation, I’ll replace it rather than trust it just because it isn’t shorted.
I’m still trying to avoid replacing IC541 blindly because the STR3A453 is relatively difficult/expensive to source in small quantities, but the 34.5 V VCC and 0.264 V FB readings do make it look increasingly suspicious.
I’ll report back with the measurements after the resistor replacement and bench testing.