Yamaha Stagepas 600i - Rchan keeps blowing

[Journey11]

Hi All,

I have this stagepas 600i that has stumped me. Should have been a simple fix but ... here's the story.
Apparently the right channel blew when my client decided to use a set of home entertainment speakers with this unit in place of the speaks that come with it. On the right channel he needed to extend the speaker wire so he cut the connector off and spliced the extra length of wire and apparently used a different gauge of wire. I'm gonna guess it was probably bigger than the original cable since the original cables are thin pieces of junk. The extra length was not much so cable resistance shouldn't have been an issue. In the manual from Yamaha, they warn against using any other speaks cables other than what is supplied. (thin junky crap). There is/was a poster in one forum that stated that these amps are very sensitive to changes in speaker wire gauge ... I call BS.
Refer to diagrams included for the following descriptions ...

So I replaced the blown components ( both right chan power trannies Q445 & Q446 and their respective current limiting resistors R472 & R473 Marked in red on pics). This was round 1. I did a cursory check of other supporting components to look for more blown items ... found nothing. Returned the main brd to him (1st go round I didn't have the whole unit to reassemble and test myself ) He reassembled powered up and in about 5 seconds the channel blew again.

He then brought the whole unit back to me for round 2 ... ding ding ding ! After inspecting the brd again I found 1 more component that blew (R477 Marked in red on pics) Note here ... not sure if I missed that in round 1 so couldn't determine if that caused the 1st repair job to fail.
So replaced all blown components a 2nd time. I then checked again for more defective parts but this time really looked it over very very well ... again found nothing. Reassembled and powered up and again in 5 secs POOF ! ... really pissed now ...

Round 3 ... ding ding ding !
Again replaced all blown items but this time around a different resistor than R477 went ... this time it was R476 (Marked in red on pics) on the opposite side of the power channel. So this is where I'm at now.
So before reassembling and test ... I really really really went to town looking for other blown parts.
In the amp schematic components marked in light purple were all triple checked and measurements compared to the Left working channel.
Inserting diagrams here and below pics I have my questions and suspicions on causes for failure.


After speaking with the owner about how it possibly blew ... he mentioned that the splice may not have been done safely and the lines shorted and it blew. Ok so that makes sense.
After 1st repair ... I'm assuming he used the original speaks and wires to test.
When I powered up after 2nd repair ... no speaks connected ... blew again.

After close inspection of the 0.022 resistors on the power lane ... I noticed the spacing between the solder pads was very thin. This lead me to a speculation ... If either one of these resistors was misaligned enough to basically cause a short across their landing pads ... this would cause a failure which would lead to a blow out again ...
Would I be correct in this assumption ??? ME thinks so. It is POSSIBLE I did not align one or both of these resistors and this was what caused my repair to fail.
Another question ... lets say those two resistors were properly installed ( no short across the landing pads ) ... what else in this amp config would be causing the power transistors to turn on full blast and burn out the power lane ?.
When I found R476 and then R477 had blown I wondered how and why these were going as well.
After round 1 when I found R477 blown open, I suspected I missed it and that's what caused it to fail the 1st time. But after 2nd replacement I checked everything thoroughly and it blew again. So I think I can safely assume that an open R477 was not the cause of the 1st repair failure.
This my last set of components and do not want to have to order parts again ... hence this post for some advice.

I know this is a long post but wanted to be as clear as possible here. Any help here will be much appreciated as I really want to get this thing fixed and it really shouldn't have been such a cranky SOB. If I haven't been clear on something here please do ask ...

 

[PETERDECO]

It sounds like you covered the bases quite well. With these direct coupled circuits, anything can be the culprit. Intermittent semiconductor or capacitor shorting and causing full conduction of the outputs, etc. Most amps I have found have the proper biasing to the outputs (about .55V) even with the outputs removed. Maybe try monitoring the output transistor bias across emitter and base (with output removed) to see if there is an intermittent surge. If so, work your way toward the input stage by stage until the surge disappears.

 

[Journey11]

It sounds like you covered the bases quite well. With these direct coupled circuits, anything can be the culprit. Intermittent semiconductor or capacitor shorting and causing full conduction of the outputs, etc. Most amps I have found have the proper biasing to the outputs (about .55V) even with the outputs removed. Maybe try monitoring the output transistor bias across emitter and base (with output removed) to see if there is an intermittent surge. If so, work your way toward the input stage by stage until the surge disappears.

Hi Peter,
Thanks for that quick response.
When you say " with output removed " to measure transistor biasing ... I'm not sure how a person should/would do that.
The only way I can think of what you're suggesting ( so as not to have any current running through the transistors ) would be to remove the 0.022 resistors that connect to the high voltage supply ?
And because this thing is so modular ... might be a challenge to hook it all up outside the box so I can actually turn it on.

 

[Journey11]

It sounds like you covered the bases quite well. With these direct coupled circuits, anything can be the culprit. Intermittent semiconductor or capacitor shorting and causing full conduction of the outputs, etc. Most amps I have found have the proper biasing to the outputs (about .55V) even with the outputs removed. Maybe try monitoring the output transistor bias across emitter and base (with output removed) to see if there is an intermittent surge. If so, work your way toward the input stage by stage until the surge disappears.

Hi Peter,
I was talking to a friend of mine and was reading your reply and telling him about it and something dawned upon me about an intermittent as you mentioned. It can't be because 2 times in a row it blew immediately within 5 seconds each time. Whereas an intermittent component failure would have had two differing time spans before it would blow. wouldn't you think. 5 seconds of full over current or over voltage would be about right to pop those components.

 

[PETERDECO]

Yes it sounds like a short. I forgot you don't have the complete unit and only the board. I would use an ohmmeter and check the semiconductors, particularly those KTA transistors from emitter to collector, for shorts and compare the resistance readings with the good channel.

 

[Journey11]

Hi Peter,
I do now have the whole unit ... since the 1st repair failed he brought the whole system.
After replacing all bad components for a 3rd time ... I have been checking and rechecking everything I can think of and of course comparing each measurement with the counterpart on the working channel.
Both transistors ( the new ones ) measure more or less the same as the good chan.
But just to refresh my knowledge base about these types of designs ... those very low resistance high wattage resistors (0.022 ohm)
are for current limiting ? Or current sensing for feedback ? What function ?
And also I haven't come across a design like this that uses such a very low value. Most are closer to 1 ohm.
For example a Crown amp I'm working on now uses 0.4 ohm.

 

[Martaine2005]

For such a low resistance, a guess would be current sense.

Martin

 

[PETERDECO]

Yes they look like part of a current sensing scheme. As the drain on Q445 conducts heavily, it looks like it will turn on Q443 which appears to turn on Q444 also. The output of Q444 is labeled OCP which, I'm guessing might mean Over Current Protection.

 

[Journey11]

Ah yes good observation ! I didn't see that label til you pointed it out.
So I'm sitting here wondering whether I should slap it all back together and see if if pops those components again.
If it does ... I'm gonna put it all in a wood chipper !

 

[PETERDECO]

Years ago I was given a very expensive power amp which had a blown channel and had gone through several repair shops without success. All of the semiconductors tested OK, yet every time you turned it on, the outputs fried. The fix? I replaced every single transistor, bias diode and coupling cap in the bad channel. Worked like a charm. Judging by your schematic though, that would be a huge undertaking.

 

[Journey11]

HI Peter,
Whoa WHoa WHOA ! ... don't such scary things ! It would be a long drawn out nightmare ...at minimum.
Not sure what else a guy can do to protect those components while actually powering up and take maesurements ?
and I'd have to charge the poor guy who owns it a brazillion pesos ...

 

[PETERDECO]

You can wire an incandescent light bulb in series with the power cord. If it glows brightly when turning on the amp, the amp is drawing a lot of power and still needs troubleshooting. I would use a 100W bulb which would limit current to about 1A (assuming you have 110V). 1A shouldn't blow up the outputs even if they are in full conduction from a shorted component.

 

[Journey11]

HI Peter,
After looking at it again ... It wouldn't be too too bad. I put a shopping cart together for most of the semi's and some caps, the parts cost would be less than $40. So I may just go ahead and do that.

 

[Journey11]

Hi Peter,
I have another brain twister for ya ... it's Behringer BX4410 Ultrabass bass amp. I'll start another thread for it so keep an eye for it.

 

[Journey11]

Hi Peter,
I had an idea to power this unit up without risking blowing my output again ... let me know what you think.
On each output transistor at the gate, there is a 4.7 ohm resistor where the signal comes in. (R470 and R471) ...
What if I were to remove them so as to (kinda isolate ?) keep these transistor from turning on at all.
Then I can look at what is happening from that point backwards to see if and or what is causing them to switch on fully and over current.
Another question in my mind was about the lightbulb trick ... If I were to do this, to wire up the bulb I'm assuming you wire the bulb in series on one of the AC input rails ? I also have a variac here I can use as well ... would this be another option to limit risk ?

 

[Journey11]

Another idea ? ... remove the 0.022 ohm current sensing resistors (R472 & R473) so no current flows through the channel .... then start taking measurements to see what's what ... ? Say starting at the mains Gate pins and go backwards from there ...
Since the unit ran even after the right side blew ... I think it will do the same after removing the resistors ... then various measurements can be done ... also wondering about cold vs hot resistance for the bulbs. Out of curiosity, I measure cold resistance for 60 watt and 100 watt bulb ... 60 was around 32 ohm and to my surprise 100 was less than half that at 13.2 ohm. So with these low values cold ... this may not be effective at current limiting at initial startup ??? Remember, it took only about 5 seconds or less for the transistors to blow ... thoughts ???

 

[Journey11]

I took another look at those Curr limit resistors again. Attached some pics. Now that they are out of circuit ... still like to hear yer thots on powering up without them installed for measurements at the gates of the mains. Would like to see what voltage if any is there.
As for the pics ... I think one can see if these are NOT centered and aligned perfectly ... it would be so easy to get a short or maybe an HV arc if they were misaligned just enough ... but had a brilliant idea ... gonna solder them in UPSIDE down so absolutely NO chance of shorting or arcing. Look at how wide the spacing is on the top vs the bottom.
Definite problem solved not knowing if they are shorting across the landings.
I also almost forgot to replace R476 ... only noticed that when I decided to remove R472 and 473.
Anyway, like I was saying in previous posts here above ... I think I can safely power up with these 2 removed and get an idea what's going on at the gates of the mains. However taking them out will disable the OCP signal but that should be fine since no current can pass. Unless the whole damned thing won't function properly with ALL feedback signals available and bla bla. And also realized another kinda major problem ... when the unit is assembled ... some of my critical test points are under the pcb ... DOH !
Anyway, I'll be watchin for your reply here ...

 

[PETERDECO]

Do not remove R470 and R471 and expect the outputs to draw no current. Those transistors are FET's and can stay in full conduction without voltage on the gates to turn them off. You can try the variac to limit the voltage and compare readings with the good side, unless like you say it might need full power to turn the supply on. I would remove the output transistors (to keep them from blowing) and see what voltages are on them, particularly from source to gate and compare with the good side.

 

[Journey11]

Ya I kinda figured those were the poorest choice ... what about the 0.022 curr limiters ? they are OUT now. Removing the trannies again might just damage the pads and via's ... might be ok to do the 0.022's ? Those removed will shut down the OCP circuit which might screw up the main CPU but all I want to see is the voltages across the gate source like you said.

 

[PETERDECO]

It looks like the .022's will cut the power to the whole circuit. I don't think you will get any readings on the gates.