Solved: Yamaha AVR RX-V667 turns off with PS1 PRT 3L - TOP254PN IC gets very hot

[cprogrammer]

I guess no. Standby current spec is 0.2W (with no things attached) and 270mA means 1.5W not considering PSU losses. 270mA should not cause SMPS overheating and shutting down though.

Thank you so much for being patient and explaining things. 1.5W makes sense because I have HDMI control on in my settings. As per user manual, power consumption in standby state with HDMI control on and without any HDMI signal is 2.7W

I feel overjoyed.

Is the heat because I'm using 12v AC as input to LM317. My LM317 is one of those cheap Chinese counterfeits.

[wraper]

Different linear regulators will dissipate the same amount of heat given that voltage drop on them and current is the same. 12-5.5=6.5V  6.5V*0.27A=1.76W to dissipate. 6.5V*0.86A=5.6W which is a significant amount of heat.

[cprogrammer]

Now that I have an adjustable PS hooked up in place of the original 5.6 volts I can carry out an experiment to simulate PS1 PRT 3L as an academic exercise. I will keep on reducing the voltage and see what happens. I guess I should get PS1 PRT 3L.

The original problem was most probably TOP254PN shutting down when it got too hot. My replacement IC will take one more day to arrive as per the latest tracking information from the courier agency. So I can utilize tomorrow to carry out this test.

[cprogrammer]

Now that I have an adjustable PS hooked up in place of the original 5.6 volts I can carry out an experiment to simulate PS1 PRT 3L as an academic exercise. I will keep on reducing the voltage and see what happens. I guess I should get PS1 PRT 3L.

I indeed get PS1 PRT 3L when I reduce the output on the 5.6v rails.

[cprogrammer]

Yes, TOP258PN is a higher power version of the TOP254PN. It has lower Rds (1.7ohm vs 5.4ohm) and a higher current limit (1.65A vs 1A). It don't see why it could not be compatible.
I think it's worth trying to replace it. Be sure to test it first with the load disconnected in case it blows up
If everything fails, I bet substituting this part of the circuit with a decent 5V PSU should work fine.

My IC hasn't yet arrived so I was doing some reading on the TOP25x data sheet. This is what I came across. Should I be worried because of the lower Rds of 1.7 ohm vs 5.4 ohm. Should I put a 4 ohms resistance between pin3 of the IC and the transformer?

Thermal check – At maximum output power, both minimum and maximum voltage and ambient temperature; verify that temperature specifications are not exceeded for TOPSwitch-HX, transformer, output diodes and output capacitors. Enough thermal margin should be allowed for the part-to-part variation of the RDS(ON) of TOPSwitch-HX, as specified in the data sheet. The margin required can either be calculated from the values in the parameter table or it can be accounted for by connecting an external resistance in series with the DRAIN pin and attached to the same heat sink, having a resistance value that is equal to the difference between the measured RDS(ON) of the device under test and the worst case maximum specification.

Another thing I noticed. The Rds in off state (resistance between D and S pins) of TOP254PN is around 800 ohms. This almost proves that the IC TOP254PN has gone bad. Am I right in assuming that?

[wraper]

Should I put a 4 ohms resistance between pin3 of the IC and the transformer?

No

Another thing I noticed. The Rds in off state (resistance between D and S pins) of TOP254PN is around 800 ohms. This almost proves that the IC TOP254PN has gone bad. Am I right in assuming that?

Do you mean out of circuit? Did you measure it in diode mode? In what direction? MOSFETs have a body diode in opposite direction to normal "polarity" of MOSFET.

[cprogrammer]

Do you mean out of circuit? Did you measure it in diode mode? In what direction? MOSFETs have a body diode in opposite direction to normal "polarity" of MOSFET.

Yes, out of the circuit as I have already removed the IC in preparation for the new IC that will most probably come tomorrow. I measured in both directions. With the red pin of multimeter on D and black on S, it gives me around 800 ohms. If I reverse it, it gives me around 700 ohms. No, I didn't measure it in diode mode.

[wraper]

Do you mean out of circuit? Did you measure it in diode mode? In what direction? MOSFETs have a body diode in opposite direction to normal "polarity" of MOSFET.

Yes, out of the circuit as I have already removed the IC in preparation for the new IC that will most probably come tomorrow. I measured in both directions. With the red pin of multimeter on D and black on S, it gives me around 800 ohms. If I reverse it, it gives me around 700 ohms. No, I didn't measure it in diode mode.

If it's really in ohms and not kiloohms, it's not right.

[cprogrammer]

Do you mean out of circuit? Did you measure it in diode mode? In what direction? MOSFETs have a body diode in opposite direction to normal "polarity" of MOSFET.

Yes, out of the circuit as I have already removed the IC in preparation for the new IC that will most probably come tomorrow. I measured in both directions. With the red pin of multimeter on D and black on S, it gives me around 800 ohms. If I reverse it, it gives me around 700 ohms. No, I didn't measure it in diode mode.

If it's really in ohms and not kiloohms, it's not right.

It is in ohms and not kilo ohms. Exactly what I thought that it is not right and I think this is the reason why this IC gets super hot and shuts off after a min or so. The surprising thing is that the board is still giving out 5.64 volts output.

[cprogrammer]

Since I'm replacing TOP254PN with TOP258PN, I need some help. The specs of these two IC are almost the same, except higher wattage for TOP258PN vs TOP254PN and Rds on of 1.7 ohm for TOP258PN vs 5.4 ohm for TOP254PN.

Now on page 10 of the datasheet https://www.mouser.in/datasheet/2/328/tophx_family_datasheet-1511305.pdf for these IC, it says

On-Chip Current Limit with External Programmability
The cycle-by-cycle peak drain current limit circuit uses the output MOSFET ON-resistance as a sense resistor. A current limit comparator compares the output MOSFET on-state drain to source voltage VDS(ON) with a threshold voltage. High drain current causes VDS(ON) to exceed the threshold voltage and turns the output MOSFET off until the start of the next clock cycle. The current limit comparator threshold voltage is temperature compensated to minimize the variation of the current limit due to temperature related changes in RDS(ON) of the output MOSFET. The default current limit of TOPSwitch-HX is preset internally. However, with a resistor connected between EXTERNAL CURRENT LIMIT (X) pin (Y, E/L and M packages) or MULTI- FUNCTION (M) pin (P and G package) and SOURCE pin (for TOP259-261 Y, the X pin is connected to the SIGNAL GROUND (G) pin), current limit can be programmed externally to a lower level between 30% and 100% of the default current limit. By setting current limit low, a larger TOPSwitch-HX than necessary for the power required can be used to take advantage of the lower RDS(ON) for higher efficiency/smaller heat sinking requirements.

Now on page 19 of the datasheet, it illustrates how to use a resistor between the pin M and pin S to limit the current (i have attached the illustration as attachment 1). Now to limit the drain current of TOP258 to be the same as TOP254, I have to limit that to 33%. To do that I have to refer to Figure 55b on page which roughly puts 21K resistor between pin M and S. I have attached the figure as attachment 2.

On page 31 of the datasheet for Design checklist, it always says one can put a resistor between the drain and the source pin to limit the resistance. This is what is exactly quoted

Thermal check – At maximum output power, both minimum and maximum voltage and ambient temperature; verify that temperature specifications are not exceeded for TOPSwitch-HX, transformer, output diodes and output capacitors. Enough thermal margin should be allowed for the part-to-part variation of the RDS(ON) of TOPSwitch-HX, as specified in the data sheet. The margin required can either be calculated from the values in the parameter table or it can be accounted for by connecting an external resistance in series with the DRAIN pin and attached to the same heat sink, having a resistance value that is equal to the difference between the measured RDS(ON) of the device under test and the worst case maximum specification.

So as per my understanding, there are two ways to replace TOP258PN with TOP254PN.

1) Put a resistance between Pin D and Pin S to account for the difference between the Rds on of the two ICs. The difference is 5.4 - 1.7 = 3.7 ohms. My reasoning is that this will require me to put high wattage resistors.

2) Put a resistor between Pin M and Pin S and that will automatically limit the drain current to 33% of the typical drain current of TOP258. This allows one to use a higher capacity IC and lowers the heat sink requirement. I find this method better than 1) above as I can use ordinary resistor around 21K to limit the current to 33%. I already have a 22K resistor.

Little knowledge can be dangerous. I'm suspecting that could be the thing that will vapourise the yamaha power board when I replace TOP254PN with TOP258PN. Unfortunately TOP254PN isn't available here in India / lead time to order it is too long. Let me know if I'm on the wrong path.

[floobydust]

OP, ensure electrolytic capacitors C3707 47uF 16V, C3708 10uF 50V are perfect. I outright replace these because they are highly stressed, on 24/7 and are essential for the TOP switch IC to work properly. It is common on many SMPS for these get old and dry out, go low value and then the IC blows.

[cprogrammer]

OP, ensure electrolytic capacitors C3707 47uF 16V, C3708 10uF 50V are perfect. I outright replace these because they are highly stressed, on 24/7 and are essential for the TOP switch IC to work properly. It is common on many SMPS for these get old and dry out, go low value and then the IC blows.

Thank you so much. It makes sense. Few days back when this problem occurred and I started checking the values of all capacitors and resistors (I removed them one by one and tested), the only two components that did not match were C3705 (0.1 μF) which measured 0.006 μF) and C3708 (10 μF) which measured 7 μF).

I replaced C3705 which was a SMD cap with a ceramic cap (representative image attached below). 10 μF electrolytic capacitor was replaced with a 10 μF 63v.

C3707 is still measuring 47 μF. But I don't have a meter which tells me the ESR. So to be on the safe side I will replace C3707 too. Thank you for enlightening me of these two caps. Ever since this problem occurred, I have gained some knowledge in electronics. I think I have gained some knowledge of how a SMPS based power supply works. Few days back I didn't know what a MOSFET was. Even if I fail to repair the PSU, I have managed to make a another (transformer based)PSU using LM317, which I tested to work as a substitute for the SMPS based 5.6v supply. The only issue is that with this substitute, I have to plugin in two power cords to switch on the AVR.

[wraper]

There is no X terminal in in ICs with PN suffix, so you cannot attach a resistor to that. M pin allows to make current the limit dependent on input voltage, so it's not just adjusting the current limit. Current limit is a protection feature and should not matter during normal operation. So my suggestion is to not bother with that.

[wraper]

Also their current limits are not so wildly different to bother, 254PN has 1A limit 258PN has 1.65A limit.

[wraper]

1) Put a resistance between Pin D and Pin S to account for the difference between the Rds on of the two ICs. The difference is 5.4 - 1.7 = 3.7 ohms. My reasoning is that this will require me to put high wattage resistors.

If you place a resistor in series of either of them, it will not change the current limit as current is sensed completely internally within the IC. That quote from the datasheet is about something else. It's a way to measure a heat dissipation which would be with ICs with higher RDS(on)

[samnmax]

The lower Rds of the TOP258PN is actually a good thing, as it will generate less heat.

[cprogrammer]

Thanks to everyone in this forum. A big thanks and lots of gratitude to all of you @wraper, @samnmax, @floobydust. You folks really rock and were a great help to an ignoramus that i'm.

I had ordered 2 nos of IC TOP258PN. It came today. But before I put the IC, I replaced one more cap suggested by @floobydust. With bated breath I connected the power supply to this board without connecting it to the AMP. It was a total terror switching on the mains. I connected a voltmeter also to the 5.6v out. Switched on and absolutely no voltage. Feeling sad, I connected back the old malfunctioning IC and the I could see 5.7 volts. Connected the board and the amp switched on. But after a minute or so, the AVR turned off. As usual the diagnostic showed PS1 PRT 3L.

With no hope left, I removed the TOP254PN (the malfunctioning one) and connected the the other TOP258PN IC. Connected and switched on the mains and lo and behold, the AVR turns on and stays on. No more PS1 PRT 3L. The new IC doesn't at all get hot. So it looks like I got one bad IC and one good IC out of the two purchased. I will order two more so that this AVR lasts as long as it does. Had this AVR since 2011 and it has been my main system. Amazon firestick, Chromecast, PS4, my HTPC and mac mini are all connected to this AVR. Just didn't want to go through the trouble of buying a new AVR. This one is good enough for my needs.

The only difference is that I get 5.46 volts in standby state and 5.42 volts in powered on state. Earlier this used to be 5.64 volts in standby state. I don't remember the voltage in powered on state with the earlier IC.

But it looks like, the trouble is behind. Wonder what will happen with slightly below spec of the voltage. To be sure, I turned on the diagnositics by pressing INFO + Tone control + power on button.

DC is 50 (within range as per service manual)
PS1 / PS2 is 89/86 (withing range as per service manual).

[samnmax]

Glad we could help! Hope it lasts at least 10 years more