Failing JBL PRX635's

[Gypsy Kiss]

Originally Posted by ChiroVette

lol You do realize that what the tech support guy told you was EXACTLY what agedhorse has been saying to you about this issue in the thread, right?

Seriously, the JBL guy himself basically told you that there is absolutely no need for a fan and that its ONLY function would be to make you "sleep better," thus acting like some electronic placebo. Again, that is precisely what aged has been telling you.

Really?, I mean really?? Come on man, the amps run too hot when just sitting there with no signal running through them???? that just dos'nt make sense??. With the fans running the amps get luke warm ( How I would suspect they are suppose to run)...I don't really care if the pope himself tells me that its normal, I'm not taking the chance. Never in my 25 years of active gigging and running my own PA have I ever come across an amp module that runs as hot as this 635, NEVER, ,,,Sounds awesome but runs too hot. My fans are a simple solution to any overheating problems, and when I can book the boxes in with my local tech I will get them to permanently mount a fan somewhere on the amp . I don't see how wanting your gear to run cooler is a bad or wrong thing???.

[Al Poulin]

I just noticed some warnings at the end of the 635 user manual :

 

 

-The PRX635 should not be played at high volumes for any period of time OR left idling, since the combination of lack of adequate cooling and poor heat dissipation may cause thermal shutdown and/or possibly burn the hands of curious children or monkeys.

 

- When playing any kind of music that is performed with any type of instrument and by any type of person AND as a precautionary measure, the 635 should be turned off by the user after 20 minutes of use to avoid automatic shut down.

 

- For most reliable operation, the 635 should always be used outdoors, in shaded areas, preferably in winter time AND in a cold country such as Canada.

 

- DJs should not use the 635 AT ALL since they seem to be of the opinion that a professional active speaker should be designed to be able to cope with operating conditions that are ''not ideal'', such as playing music at higher volumes that a home system.

 

- The back of the 635 should not be used as a George Foreman grill unless you have purchased the specially designed conversion kit. Item mumber "BBQ635".

 

- JBL assumes no responsibility for loss of enjoyment of your guests or audience because of the failure of your 635 and the frequent ''cool down'' periods that may be required for regular operation. You should be happy that you have been able to purchase and use a JBL speaker. That is really all that is important. Please do not post your negative reviews of thermal failure reports on the web as they will automatically be dismissed, mostly because your are probably a DJ if your 635 failed...

 

Al

[agedhorse]

Originally Posted by Al Poulin

I just noticed some warnings at the end of the 635 user manual :


-The PRX635 should not be played at high volumes for any period of time OR left idling as lack of adequate cooling or heat dissipation may cause thermal shutdown.

- When playing any kind of modern music or rock, or country, or raggae etc... and as a precaution, the 635 should be turned off by the user after 20 minutes of use to avoid automatic shut down.

- For most reliable operation, the 635 should be used outdoors, in shaded areas, preferably in winter time AND in a cold country such as Canada.

- DJs should not use the 635 AT ALL since they seem to be of the opinion that a professional active speaker should be designed to be able to cope with operating conditions that are ''not ideal'', such as playing music at higher volumes that a home system.

- The back of the 635 should not be used as a George Foreman grill unless you have purchased the specially designed conversion kit. Item mumber "BBQ635".

Al

Nice Al.

[Gypsy Kiss]

Originally Posted by Al Poulin

I just noticed some warnings at the end of the 635 user manual :


-The PRX635 should not be played at high volumes for any period of time OR left idling, as lack of adequate cooling and poor heat dissipation may cause thermal shutdown and/or possibly burn the hands of curious children or monkeys.

- When playing any kind of music that is performed with any type of instrument and by any type of person AND as a precautionary measure, the 635 should be turned off by the user after 20 minutes of use to avoid automatic shut down.

- For most reliable operation, the 635 should always be used outdoors, in shaded areas, preferably in winter time AND in a cold country such as Canada.

- DJs should not use the 635 AT ALL since they seem to be of the opinion that a professional active speaker should be designed to be able to cope with operating conditions that are ''not ideal'', such as playing music at higher volumes that a home system.

- The back of the 635 should not be used as a George Foreman grill unless you have purchased the specially designed conversion kit. Item mumber "BBQ635".

- JBL assumes no responsibility for loss of enjoyment of your guests or audience because of the failure of your 635 and the frequent ''cool down'' periods that may be required for regular operation. You should be happy that you have been able to purchase and use a JBL speaker. That is really all that is important. Please do not post your negative reviews of thermal failure reports on the web as they will be dismissed, mostly because your are probably a DJ if your 635 failed...

Al

LOL!!!

[Al Poulin]

Originally Posted by agedhorse

Nice Al.

I knew you'd like it Andy. I modified a little since you read it though. :-)

[Gypsy Kiss]

It sounds to me like JBL have used the same amp design in the 635 as they do in all the PRX stuff which is fine but being a 3 way box maybe it has a little more heat generated than the other models and this is the tipping point, it's the no buffer zone. I think JBL really needed to add extra fins to the 635 or dare I say it a $5.00 fan

[agedhorse]

Yeah, really.

There are class D topologies that run hotter when idling and do not increase in temperature (or even some parts run cooler) under load. You are comparing old school, old technology to completely different operating methods. There is absolutely nothing at all in common between class D and conventional linear amps, including the way they operate and their thermal dissipation curves.

I worked on a design a few years ago, troubleshooting why fan cooling part of the class D amp caused massive failures. Turns out that a thermal stabalization feedback component was being cooled too much and at a gradient unrelated to the operation of the amp itself. This caused the amp to increase the switching speed and run away failing in the process. I don't think this will happen with the PRX, in part because of the averaging effect of the heatsink, but this example shows that sometimes a result that does not seem to make sense actually makes all kinds of sense once you understand the technology and the particular design approach. In fact, this reminds me of a problem with an old school linear class AB amp that also had thermal runaway problems, again a sensor coupled to the main heatsink on a CFP output stage in a fan cooled amp would sometimes blow up when the amp became hot. It wasn't quite that simple because increasing the cooling speed of the fan made the problem worse in some cases. The problem is that the thermal feedback was tracking the wrong parameter and as long as the fan cooling was very slow or the amp was allowed to equalize, or the output stage was allowed to run hot there was no problem but as soon as it was cooled down it caused the driver that was the real cause of the problem to run away with uncontrollable drift. As long as the output stage was hotter than the driver the sircuit was perfectly stable.

Another example? Linear or lateral MOSFETs used in a classic linear output stage must be biased much higher than any bipolar transistor, resulting in the FET output stage running much warmer at idle. This is normal and necessary because of the transition from positive to negative temp.co. versus bias of the MOSFET. This also is not a defect or a bad design, but entirely necessary and causes no harm (there are hundreds of thousands or even millions of these output stages in successful operation globally)

25 years of active gigging doesn't give you the perspective of 35 years of designing amplifiers does.

Originally Posted by Gypsy Kiss

Really?, I mean really?? Come on man, the amps run too hot when just sitting there with no signal running through them???? that just dos'nt make sense??. With the fans running the amps get luke warm ( How I would suspect they are suppose to run)...I don't really care if the pope himself tells me that its normal, I'm not taking the chance. Never in my 25 years of active gigging and running my own PA have I ever come across an amp module that runs as hot as this 635, NEVER, ,,,Sounds awesome but runs too hot. My fans are a simple solution to any overheating problems, and when I can book the boxes in with my local tech I will get them to permanently mount a fan somewhere on the amp . I don't see how wanting your gear to run cooler is a bad or wrong thing???.

[agedhorse]

 

Originally Posted by Gypsy Kiss

It sounds to me like JBL have used the same amp design in the 635 as they do in all the PRX stuff which is fine but being a 3 way box maybe it has a little more heat generated than the other models and this is the tipping point, it's the no buffer zone. I think JBL really needed to add extra fins to the 635 or dare I say it a $5.00 fan

 

Assuming (and it's a big assumption that even I wouldn't think of making, knowing the class D architecures that are possible and the multitude of potential failure mechanisms that exist) that it is a thermal problem. It could be something entirely unrelated (and likely is) because JBL/Crown is very thorough in their design and proof testing. Hell, we don't even know what percentage of products fail?

[agedhorse]

Originally Posted by Al Poulin

I knew you'd like it Andy. I modified a little since you read it though. :-)

I don't know Al, you guys up there seem to think salt and pepper are exotic spices

(my wife's family certainly thinks so... at least you don't eat fermented fish)

[Al Poulin]

Andy - I admit, that just like MOSFETs, I can sometimes be a little biased. But at least I'm not bipolar like those transistors.

[Gypsy Kiss]

Originally Posted by agedhorse

Yeah, really.

There are class D topologies that run hotter when idling and do not increase in temperature (or even some parts run cooler) under load. You are comparing old school, old technology to completely different operating methods. There is absolutely nothing at all in common between class D and conventional linear amps, including the way they operate and their thermal dissipation curves.

I worked on a design a few years ago, troubleshooting why fan cooling part of the class D amp caused massive failures. Turns out that a thermal stabalization feedback component was being cooled too much and at a gradient unrelated to the operation of the amp itself. This caused the amp to increase the switching speed and run away failing in the process. I don't think this will happen with the PRX, in part because of the averaging effect of the heatsink, but this example shows that sometimes a result that does not seem to make sense actually makes all kinds of sense once you understand the technology and the particular design approach. In fact, this reminds me of a problem with an old school linear class AB amp that also had thermal runaway problems, again a sensor coupled to the main heatsink on a CFP output stage in a fan cooled amp would sometimes blow up when the amp became hot. It wasn't quite that simple because increasing the cooling speed of the fan made the problem worse in some cases. The problem is that the thermal feedback was tracking the wrong parameter and as long as the fan cooling was very slow or the amp was allowed to equalize, or the output stage was allowed to run hot there was no problem but as soon as it was cooled down it caused the driver that was the real cause of the problem to run away with uncontrollable drift. As long as the output stage was hotter than the driver the sircuit was perfectly stable.

Another example? Linear or lateral MOSFETs used in a classic linear output stage must be biased much higher than any bipolar transistor, resulting in the FET output stage running much warmer at idle. This is normal and necessary because of the transition from positive to negative temp.co. versus bias of the MOSFET. This also is not a defect or a bad design, but entirely necessary and causes no harm (there are hundreds of thousands or even millions of these output stages in successful operation globally)

25 years of active gigging doesn't give you the perspective of 35 years of designing amplifiers does.

This kinda sounds just like what Marshall said to there customers who bought the DSL amps when they came out in 2000. These amps had an inherent runaway bias problem due to bad mother boards not being insulated correctly ( look it up, plenty of stuff out there)..It wasn't until 2008 that Marshal admitted there was a problem , I see the same thing here with JBL. I betcha 10 - 1 the next line of JBL PRX 700?, 800 ? Whatever,if they don't change the amp design come with fans.. Betcha..

The amp runs too hot "For me" end of story, been using them 4 nights a week for a month now and no problems at all, they sound fantastic and they stay cool...

[Gypsy Kiss]

 

Originally Posted by agedhorse

Assuming (and it's a big assumption that even I wouldn't think of making, knowing the class D architecures that are possible and the multitude of potential failure mechanisms that exist) that it is a thermal problem. It could be something entirely unrelated (and likely is) because JBL/Crown is very thorough in their design and proof testing. Hell, we don't even know what percentage of products fail?

 

Agreed, look man, I love JBL gear ( well most of it, the JRX stuff not so much love) but when an amp is too hot to touch for more than a few seconds without burning your hand ,something is wrong . Your totally right mate, I have,nt the faintest idea how digital amps are designed, but 60 degree + operating temps with no form of added cooling is just asking for trouble in my book,??? Dunno??? but I will continue to run fans on my boxes for that piece of mind..

[Dookietwo]

Only guessing and I don't know of course but I wonder if putting a fan on the heat sinks is stressing components more than they should be? Running the amp very hard and thinking everything is ok when the "Inside" of the amps components are running way too hot with no way to monitor this doesn't sound like a good idea.

As an example.
If Device X is designed to run at 60 volts at 60 degrees and the heat sinks will just take care of that everything is fine.
If you put a fan on the heat sink Device X is now run at 80 volts. The heat sink fins with the fan on them wick the heat and everything seem to be ok on the "outside" but Device X is still running at too high of a voltage.
Sooner or later Device X will fail and the user will complain saying I never ran the amps even hot......
Driving a car too its max and putting a big fan on it so it stays cool doesn't mean there is no damage going on to the car on the inside.

Run the box within its means or get more rig for the gig if that isn't enough. There is more than the tempature of the heat sinks on the outside to consider. Its what the device inside sees which damages the electronics in some cases.

Dookietwo

[Gaffle]

Originally Posted by agedhorse

You are certainly entitled to your own opinion, even if it isn't accurate.

What's inaccurate about what I said? I'm not guessing here or going off of what I read on the web, I previously owned and experienced the excessive heat and shutdowns firsthand. I also had to open up one of the brand-new-out-of-the-box PRX635's and re-crimp the pair of unattached lead wires on the midrange. (great quality control JBL).

My DJ comment stands as accurate, by far the majority of whining has come from that segment of the market. In fact, for another manufacturer of power amps that I am affiliated with, the failure rate of amps sold into the DJ market is much, much higher that the identical amp sold into other markets. The only difference being the users.

I could just as easily say your opinion isn't accurate but you have work experience firsthand so you're entitled to it.

edit: oh, I see that you are a DJ... sorry if my comments offended.

Don't worry, I'm not easily offended. I know there are some nimrods out there who don't know jack about proper sound reinforcement and sound quality; I'm not one of them.

- The back of the 635 should not be used as a George Foreman grill unless you have purchased the specially designed conversion kit. Item mumber "BBQ635".

Good one.

BTW: I did recommend a pair of PRX635's to another DJ who was looking for a total sound solution w/o the need for subs. Her gigs ranged from small parties to weddings. She didn't have any shutdown issues within the first 6 months and I believe she's owned them for at least 1 1/2 years. I haven't heard her experience lately with them.

[OneEng]

Originally Posted by Al Poulin

I just noticed some warnings at the end of the 635 user manual :


-The PRX635 should not be played at high volumes for any period of time OR left idling, since the combination of lack of adequate cooling and poor heat dissipation may cause thermal shutdown and/or possibly burn the hands of curious children or monkeys.

- When playing any kind of music that is performed with any type of instrument and by any type of person AND as a precautionary measure, the 635 should be turned off by the user after 20 minutes of use to avoid automatic shut down.

- For most reliable operation, the 635 should always be used outdoors, in shaded areas, preferably in winter time AND in a cold country such as Canada.

- DJs should not use the 635 AT ALL since they seem to be of the opinion that a professional active speaker should be designed to be able to cope with operating conditions that are ''not ideal'', such as playing music at higher volumes that a home system.

- The back of the 635 should not be used as a George Foreman grill unless you have purchased the specially designed conversion kit. Item mumber "BBQ635".

- JBL assumes no responsibility for loss of enjoyment of your guests or audience because of the failure of your 635 and the frequent ''cool down'' periods that may be required for regular operation. You should be happy that you have been able to purchase and use a JBL speaker. That is really all that is important. Please do not post your negative reviews of thermal failure reports on the web as they will automatically be dismissed, mostly because your are probably a DJ if your 635 failed...

Al

LOL!

I particularly like the BBQ635 part

[Gypsy Kiss]

Originally Posted by Dookietwo

Only guessing and I don't know of course but I wonder if putting a fan on the heat sinks is stressing components more than they should be? Running the amp very hard and thinking everything is ok when the "Inside" of the amps components are running way too hot with no way to monitor this doesn't sound like a good idea.

As an example.
If Device X is designed to run at 60 volts at 60 degrees and the heat sinks will just take care of that everything is fine.
If you put a fan on the heat sink Device X is now run at 80 volts. The heat sink fins with the fan on them wick the heat and everything seem to be ok on the "outside" but Device X is still running at too high of a voltage.
Sooner or later Device X will fail and the user will complain saying I never ran the amps even hot......
Driving a car too its max and putting a big fan on it so it stays cool doesn't mean there is no damage going on to the car on the inside.

Run the box within its means or get more rig for the gig if that isn't enough. There is more than the tempature of the heat sinks on the outside to consider. Its what the device inside sees which damages the electronics in some cases.

Dookietwo

I have enough rig for the gig man . My gear does not work hard and I don't push it hard. These boxes are oven ready at idle man, like literally fry an egg and they don't get any cooler from that point. I have a tiny 3" desk fan blowing air over the back of the fins and this is enough to bring the heat down to what I would call an acceptable level. I would love to have a no bull chat with one of the Engineers that helped to develop this model and ask him/her why would you design a amp that is running at full throttle all the time????? It don't matter how well designed the amp is , if something gets too hot it will cease to work , weather it be a active speaker, a car, a light bulb... or a wet soggy moldy old cabbage, if it gets too hot it stops.

[Mutha Goose]

 

Originally Posted by agedhorse

There are class D topologies that run hotter when idling and do not increase in temperature (or even some parts run cooler) under load...

 

I have been designing Class D amplifiers for automotive applications for about 10 years now (read restricted air flow and excessive ambient temperature environments). I have litterly millions of my designs in the feild (you may even have one in your car/truck). I can COMPLETELY back Aged up on this. Effiency in the lowest 1/4 of load dissipation is horrible. Far worse than most other topologies. By the time you reach 50% of rated output, effecieny increases into the area of 80% of its maximum (variations per dseign of course) and you begin to see the benifits.

[agedhorse]

 

Originally Posted by Mutha Goose

I have been designing Class D amplifiers for automotive applications for about 10 years now (read restricted air flow and excessive ambient temperature environments). I have litterly millions of my designs in the feild (you may even have one in your car/truck). I can COMPLETELY back Aged up on this. Effiency in the lowest 1/4 of load dissipation is horrible. Far worse than most other topologies. By the time you reach 50% of rated output, effecieny increases into the area of 80% of its maximum (variations per dseign of course) and you begin to see the benifits.

 

Exactly. There are some topologies that address the efficiency at idle (specificly for EnergyStar and the EU equiv.) but these amps use all kinds of tricks to control switching losses at low signal levels, sometimes using adaptive risetime algorithems on the PWM signal to reduce losses but then RFI/EMI become an issue. With class D amps, there are no free lunches... in fact lunch is pretty expensive as that's the cost of admission to this topology.

[abzurd]

Originally Posted by OneEng

LOL!

I particularly like the BBQ635 part

I've had mixed results. When I first got mine it cooked everything so fast I was constantly burning things. I mounted some fans on it and usually have some ice handy in case I need to slow the cooking down even more.

The rotisserie attachment is pretty good though. We put a chicken on it right before we start playing and it's ready about the end of the first set. I'll brush on some BBQ sauce during a guitar solo in the last few songs. We really look forward to the break.

[Dookietwo]

Originally Posted by abzurd

I've had mixed results. When I first got mine it cooked everything so fast I was constantly burning things. I mounted some fans on it and usually have some ice handy in case I need to slow the cooking down even more.

The rotisserie attachment is pretty good though. We put a chicken on it right before we start playing and it's ready about the end of the first set. I'll brush on some BBQ sauce during a guitar solo in the last few songs. We really look forward to the break.

I think I was at one of your shows.

Was this you throwing the chickens in the back of the amps?

Or maybe it was this show?

Dookietwo

[Gypsy Kiss]

 

Originally Posted by agedhorse

Exactly. There are some topologies that address the efficiency at idle (specificly for EnergyStar and the EU equiv.) but these amps use all kinds of tricks to control switching losses at low signal levels, sometimes using adaptive risetime algorithems on the PWM signal to reduce losses but then RFI/EMI become an issue. With class D amps, there are no free lunches... in fact lunch is pretty expensive as that's the cost of admission to this topology.

 

But "why" would you want a digital amp to run right on the edge?, and why do no other active boxes ( at least all of the ones I have used/seen) get as hot?. My Ev's have no fans , actually no external heat fins nothing , but will run all night and only end up luke warm. I fully get that the numbers say "it works" but why hang by your toes on the edge of the cliff when obviously you don't really have to???..That's the part that gets me.

[Mutha Goose]

Earlier you stated that the amps should not run hotter when running idle. So we corrected that mis-statement... to that you reply with a different argument about a different point. Are you trying to learn more or argue?

I don't think ANYONE is saying that an amp should be run right at the edge of SOA. What he is saying is that unless you know the subtle design targets, details, and operating ranges, it may not be a good idea to actively alter its environment. I can also tell you that too hot to touch MAY not nessessarily be indicative of a problem. Many amps will burn you, period! But they are designed to be run that hot. Heat by itself is not a problem. Heat that exceeds the design/components is. But designs/components can be and are designed specifically for high temperature operation.

[abzurd]

Originally Posted by Dookietwo

I think I was at one of your shows.

Was this you throwing the chickens in the back of the amps?

Or maybe it was this show?

Dookietwo

LOL ... the JBL is a much more elegant solution.

The popcorn finale at the end of the night is a lot of fun too. With the fans flipped around it's like a confetti cannon.

[Gypsy Kiss]

Earlier you stated that the amps should not run hotter when running idle. So we corrected that mis-statement... to that you reply with a different argument about a different point. Are you trying to learn more or argue?

Always trying to learn more... So when do you reach the point of "too hot" for a digital amp? or even , "wow, that's getting pretty hot" maybe we should allow a little headroom?? ... Why did JBL design an amp that runs so hot you can fry an egg on the back???...My point is why would you want to run any amp that hot? even if the numbers say it works, show me another digital active speaker that runs that hot??? Just trying to get my head around the logic behind the design.

[lonotes]

Originally Posted by abzurd

LOL ... the JBL is a much more elegant solution.

The popcorn finale at the end of the night is a lot of fun too. With the fans flipped around it's like a confetti cannon.

Can you reverse the fans with a remote footswitch? Or do you have to physically move the fans? Is it just plain popcorn, or can you do caramel corn too? Is it just corn, or can you load it with rice and do that whole Quaker "Shot From Cannons" thing, and have breakfast?