ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => SR Forum Archives => LAB: The Classic Live Audio Board FUD Forum Archive => Topic started by: Dave Barnett on October 11, 2005, 02:09:47 PM

Title: Amp failures?
Post by: Dave Barnett on October 11, 2005, 02:09:47 PM
Why is it that when I lose a channel of an amp, it's never during a show, but immediately upon system power-up in the afternoon?

I really hate that feeling of "uh, that was working fine last night".
Title: Re: Amp failures?
Post by: RicBlade on October 11, 2005, 02:15:52 PM
Define "lose a channel". Are you blowing a fuse? Popping a breaker? Or is some other component completely failing requiring repair? If so, what? That could be clue.

I would think powering up is probably the most stressfull moment in the life of an amplifier.

Your power supply may also be questionable.
Title: Re: Amp failures?
Post by: Riley Casey on October 11, 2005, 04:05:56 PM
Failure at turn on is most often due to " in rush current " when the sudden charging of the large power supply capacitors ( in traditional DC power supplies ) and the charging of the various other caps and circuit elements draws the greatest differential current.  The various semiconductors are also being subjected to a sudden application of their full supply voltages.  It's the roughest time in the life of of those poor little electronic components which is why there is a strong school of thought that leaves electronics on all the time in places like recording studios and computer labs.
Title: Re: Amp failures?
Post by: John Roberts {JR} on October 11, 2005, 06:58:25 PM
Dave Barnett wrote on Tue, 11 October 2005 13:09

Why is it that when I lose a channel of an amp, it's never during a show, but immediately upon system power-up in the afternoon?

I really hate that feeling of "uh, that was working fine last night".


If I could choose at first afternoon check is not the worst time to fail. While cold start inrush is perhaps a little stressful for the reservoir capacitors I don't think I would call that the worst stress the amplifier ever sees.

I guess it's logical that if your amps worked for the first few hours of a gig and nothing has changed they should keep on working. So the operative reason for failure is "something changed". Perhaps  in transit something that was almost touching got bumped and is now ready to release smoke, or perhaps the mains voltage which was loaded down at night is several volts hotter in late afternoon (often other way around if lots of commercial activity nearby). Another mechanism is thermal cycling. Every time a product heats up and cools down the metal guts expand and shrink again... sometimes things fatigue or micro-creep. At cold turn on will usually be the extreme case for shrinkage (trust me honey the water in the pool is cold).

Who knows maybe it did fail last night,  Shocked

Inrush current is understood and designed for. My home brew  living room amp has a truly old school power switch with a warm-up position formerly used to get the tube heaters going before you apply plate voltage. In my solid state rework I use the heater position to charge up the power supply capacitors through a current limiting resistor. In full on position the resistor is shorted out and the speakers finally connected (using relays). No turn-on/off pops and TV picture doesn't get small at turn on. Very Happy        

JR
Title: Re: Amp failures?
Post by: Dave Barnett on October 13, 2005, 03:22:59 AM
To elaborate --

(1) This is in a fixed installation, nothing is getting moved.

(2) The amplifiers in our rig are all Crown Macrotech 5002, 3600, and 2400.  The Macrotechs all have some sort of internal power up sequence, so it's not the same as an old amp where you're hitting the transformer, diode bridge, and filter caps as soon as you throw the switch.

(3) I agree that the afternoon is the most convenient time to deal with this problem, but it's still frustrating.

(4) The failures I experience are generally blown channels, i.e. failure of multiple output devices.  Tuesday's failure marks the fourth time this 5002 will have to be returned to Crown for service.

If the amplifiers were failing in service, at least then I could point to specific conditions and say "that's why it failed".  Like for example guest BE slamming the hell out of it.
Title: Re: Amp failures?
Post by: John Roberts {JR} on October 13, 2005, 09:39:09 AM
Dave Barnett wrote on Thu, 13 October 2005 02:22

To elaborate --

(1) This is in a fixed installation, nothing is getting moved.

(2) The amplifiers in our rig are all Crown Macrotech 5002, 3600, and 2400.  The Macrotechs all have some sort of internal power up sequence, so it's not the same as an old amp where you're hitting the transformer, diode bridge, and filter caps as soon as you throw the switch.

(3) I agree that the afternoon is the most convenient time to deal with this problem, but it's still frustrating.

(4) The failures I experience are generally blown channels, i.e. failure of multiple output devices.  Tuesday's failure marks the fourth time this 5002 will have to be returned to Crown for service.

If the amplifiers were failing in service, at least then I could point to specific conditions and say "that's why it failed".  Like for example guest BE slamming the hell out of it.



I have no logical explanation for output devices failing at cold turn-on.

Multiple device failures will be pretty common if any one fails it will take others with it since they typically fail as a dead short. The typical failure vector for bipolar power transistors (sorry I don't know what's inside those models) is either over-voltage or over temperature. Since it isn't likely to be too hot from a cold start, what about voltages.

The power supply voltage at cold start will be a little higher due to the power transformer being cooler. More commonly discussed is how PS rails sag after amp heats up but inverse of that is higher rails at cold start.

That said I could see that worst case for voltage stress would be to clip a cold amp with no speaker load. With speaker load connected, voltage across output devices will still be more than hot amp.

I repeat that this may be exacerbated or probably trumped by mains voltage swings which could be higher or lower (my guess) than at night, but every location is different and should be metered.

JR
Title: Re: Amp failures?
Post by: Dave Miller on October 13, 2005, 09:55:46 AM
This is not directly amp-failure-related, but I once had to sort out an install where the supply breaker occasionally tripped due to turn-on surges. Replacing it with one that had greater surge-handling capacity cured the problem, but I did wonder why it only happened infrequently.

Apart from the fact that the surge current was probably only bordering on the breaker trip threshold, I thought of two other explanations:

(a) ambient temperature - if everything is cold (eg no building heating) then the resistance of copper (including that in the amp's transformer as JR mentioned) is lower, so switch-on surges might be higher

(b) the exact point on the 50Hz (60Hz in the USA) mains cycle that equipment is switched on. If this is near the peak, then wouldn't the surge be greater.

I wonder if either of the above could contribute to occasional power-on failures. Or I am talking nonsense??

Dave
Title: Re: Amp failures?
Post by: John Roberts {JR} on October 13, 2005, 10:16:56 AM
Dave Miller wrote on Thu, 13 October 2005 08:55

This is not directly amp-failure-related, but I once had to sort out an install where the supply breaker occasionally tripped due to turn-on surges. Replacing it with one that had greater surge-handling capacity cured the problem, but I did wonder why it only happened infrequently.

Apart from the fact that the surge current was probably only bordering on the breaker trip threshold, I thought of two other explanations:

(a) ambient temperature - if everything is cold (eg no building heating) then the resistance of copper (including that in the amp's transformer as JR mentioned) is lower, so switch-on surges might be higher

(b) the exact point on the 50Hz (60Hz in the USA) mains cycle that equipment is switched on. If this is near the peak, then wouldn't the surge be greater.

I wonder if either of the above could contribute to occasional power-on failures. Or I am talking nonsense??

Dave



Power transformers go through a bit of a dance when they are first turned on since they are AC devices but depending upon when on the start up waveform they hit there will be a varying DC component making the transformer primary look more like a short than a winding for the first milliseconds.

This is an issue for big power designers wrt inrush, breakers, and power switches. AFAIK it does not impact amplifier output stage reliability, but inrush current can vary widely and randomly if power switching is not synchronized to input waveform.  

JR
Title: Re: Amp failures?
Post by: Andy Peters on October 19, 2005, 04:23:51 AM
Dave Barnett wrote on Thu, 13 October 2005 00:22


(1) This is in a fixed installation, nothing is getting moved.

(2) The amplifiers in our rig are all Crown Macrotech 5002, 3600, and 2400.  The Macrotechs all have some sort of internal power up sequence, so it's not the same as an old amp where you're hitting the transformer, diode bridge, and filter caps as soon as you throw the switch.

(3) I agree that the afternoon is the most convenient time to deal with this problem, but it's still frustrating.

(4) The failures I experience are generally blown channels, i.e. failure of multiple output devices.  Tuesday's failure marks the fourth time this 5002 will have to be returned to Crown for service.

If the amplifiers were failing in service, at least then I could point to specific conditions and say "that's why it failed".  Like for example guest BE slamming the hell out of it.


Meter the mains.  Perhaps with everything turned on at night during a gig, the draw is sufficient to cause a sag that goes away in the afternoon.  

Maybe the power company is doing odd load-management things; perhaps they expect higher usage during the day for most customers (air conditioning on in office buildings, whatnot) so they stiffin things up during the day and the mains rises too much.  Then at night, they crank it down a bit?

Maybe Gene Pink's theory about the powercos gradually raising the mains voltage in order to force you to buy more light bulbs has now been expanded to forcing you to fix power amps?

-a
Title: Re: Amp failures?
Post by: Iain_Macdonald on October 19, 2005, 04:35:50 PM
Dave,

I guess this is all rather academic, since you have probably sent it back for repair by now. I know we are two nations divided by a common language. But looking at #3 on your list it would seem to indicate that it is just one amp that is going wrong.

Questions:



If it is just this lone amp, there are many answers, driver circuit, power supply especially VZ switching, cracked earth/neutral print on a board etc etc. I think it would be very weird for a connected device(spkr), to take out the amp at switch on. Do make sure that when you send it back for service that you ask them to put it through some thermal cycles. Rather than just a continuous soak test. I have seen a weird problem of instability when someone used the jack socket with the XLR to link out to another chan/amp. I can't remember exactly which combination, and I think the amp was in bridge mode. It would be worthwhile looking at what happens to the rack power i/p when you switch on the rack/amps. I am making the assumption that the amps are not left on with the level control at max, and then the breaker switched on. No I didn't think so !!! Also look and see if someone tried to do something "clever" with the rack power for the 5002. I would personally think it strange that power problems blew the o/ps on these amps because they have so many protection circuits. Also if you had ultrasonic/infrasonic signals then you would notice unusual operation of the signal/ioc led, but never say never. Whilst it would be unusual for thermal problems to be the cause. I would check the temp of the rack near the 5002. Sometimes in  installation racks, depending on type and configuration, the bottom amp does not get enough airflow.

my 0.2 Euro

Iain



Title: Re: Amp failures?
Post by: Tim Duffin on October 19, 2005, 08:23:47 PM
The Crown MA-series does have a built in 7Hz filter.  So you can discount infrasonics from messing with the channel blowing.  I did however have the same thing happen to me with a macro-tech many times, it turned out that one of the comparators was not functioning correctly and did not allow the amp channel to operate.
The channel was not blown (burned junctions in the output devices) but rather that the channel never activated, leading me to think that it was blown.  
It was finally fixed by Crown and returned, but your problem might just be that the amp is "messed up" and cannot be repaired--I've known a few cars like that...hmmmm.  I would try another one in its slot and see if the problem repeated or if there were any strange side effects.

T