Frequently there are speakers mentioned that are labeled as 'amp killers'. I own a couple pairs of these speakers, although certainly not the worst offenders.
My question is - what kills the amplifier, is it prolonged high volume music at low impedance (seems likely as heat would build quickly). What about transients - like a drum solo at loud volume that the protection never engages? How to prevent this besides the obvious replacement of speakers.
I understand that current multiplies as resistance drops, my question can not be answered simply by 'current kills amps' - if that's the answer, what's the mechanism.
Before I knew any better I used to run a pair of Infinity Quantum 5's (watkins 12" woofer) and a pair of RSIIIa's together at extremely loud volumes (live levels) during drum solos - both of these speakers operate between 2 and 4 ohms through much of their range. I did this using both a pioneer SX-1250 and an SX-1050. I have never (yet) had an amplifier failure - although I regularly push the equipment to its limits. My guideline used to be to turn the volume down a hair when the protection circuitry engaged. I have recently added an Adcom 555II (which I did kill with the same speakers, then repaired) into the mix to take the load off but I am curious which is more damaging to the amp - long term high power output which the protection circuitry will cut short, or instantaneous high current which the protect seems to ignore (is this by design?).
I know the devices derate with heat - does that imply that they can drive much higher currents for short periods of time as long as temperature is controlled (more heatsink area, better interface?)
Is it safe to assume that the output transistors are the weakest link in the chain excluding the power supply? More so with 30 year old transistors?
It seems that the common failure mode is the output transistors fry and take the drivers with them, correct? If the output transistors are replaced with a modern higher current device (MJ21193/94) that doesn't cause instability, will that add safe operating area to the entire amp, possibly protecting the drivers as well? or do you then end up overstressing the driver transistors?
My untrained brain thinks that by substituting a higher power transistor that is known to work well in this application the primary failure point should be eliminated and my fears shall be lessened. It also seems that the factory protection would protect while the outputs still had lots of headroom left adding even more protection.
I don't have any plans to ever go back to powering a wall of speakers with a reciever but I would like to be able to drive it to its limits with one set of speakers without fear of damage to the amp. And those limits would be with lower impedance speakers than the amp was perhaps designed for.
Both of the receivers mentioned above are still in daily use and both have been fully refurbed within the last few years with all new caps, new power supply transistors and zeners, new protection relays, transistors and diodes. In the case of the 1050 it was mostly preventative after a protect relay failure and the relay driver transistor failure, the 1250 lost the 24 volt power supply and as such I fixed that problem then rebuilt the rest of the board with modern components.
I prefer using the internal amplifiers 90% of the time, due at least partially to the speaker switches and protection circuitry to keep the speakers from melting in case of an amp failure.
My question is - what kills the amplifier, is it prolonged high volume music at low impedance (seems likely as heat would build quickly). What about transients - like a drum solo at loud volume that the protection never engages? How to prevent this besides the obvious replacement of speakers.
I understand that current multiplies as resistance drops, my question can not be answered simply by 'current kills amps' - if that's the answer, what's the mechanism.
Before I knew any better I used to run a pair of Infinity Quantum 5's (watkins 12" woofer) and a pair of RSIIIa's together at extremely loud volumes (live levels) during drum solos - both of these speakers operate between 2 and 4 ohms through much of their range. I did this using both a pioneer SX-1250 and an SX-1050. I have never (yet) had an amplifier failure - although I regularly push the equipment to its limits. My guideline used to be to turn the volume down a hair when the protection circuitry engaged. I have recently added an Adcom 555II (which I did kill with the same speakers, then repaired) into the mix to take the load off but I am curious which is more damaging to the amp - long term high power output which the protection circuitry will cut short, or instantaneous high current which the protect seems to ignore (is this by design?).
I know the devices derate with heat - does that imply that they can drive much higher currents for short periods of time as long as temperature is controlled (more heatsink area, better interface?)
Is it safe to assume that the output transistors are the weakest link in the chain excluding the power supply? More so with 30 year old transistors?
It seems that the common failure mode is the output transistors fry and take the drivers with them, correct? If the output transistors are replaced with a modern higher current device (MJ21193/94) that doesn't cause instability, will that add safe operating area to the entire amp, possibly protecting the drivers as well? or do you then end up overstressing the driver transistors?
My untrained brain thinks that by substituting a higher power transistor that is known to work well in this application the primary failure point should be eliminated and my fears shall be lessened. It also seems that the factory protection would protect while the outputs still had lots of headroom left adding even more protection.
I don't have any plans to ever go back to powering a wall of speakers with a reciever but I would like to be able to drive it to its limits with one set of speakers without fear of damage to the amp. And those limits would be with lower impedance speakers than the amp was perhaps designed for.
Both of the receivers mentioned above are still in daily use and both have been fully refurbed within the last few years with all new caps, new power supply transistors and zeners, new protection relays, transistors and diodes. In the case of the 1050 it was mostly preventative after a protect relay failure and the relay driver transistor failure, the 1250 lost the 24 volt power supply and as such I fixed that problem then rebuilt the rest of the board with modern components.
I prefer using the internal amplifiers 90% of the time, due at least partially to the speaker switches and protection circuitry to keep the speakers from melting in case of an amp failure.
You are right on most of your assumptions. The term you are looking for is SOA (safe operating area) of the output devices.
You mentioned the famous mj21193/4 , they have among
the highest SOA of all the outputs.
But , that might also throw the current limiting circuitry
out of spec. What I am saying is that device substitution
is a little more involved than swapping out to a larger
device.
The 3 mechanisms are current gain , SOA , and the voltage rating
of the device.
You are right that SOA is dependant on temperature (A bigger
HS )
For example , if you change a device with a current gain
of 25 for one with 50+ , your bias will increase making for
a hot amp.
It is not like replacing a light bulb.
OS
You mentioned the famous mj21193/4 , they have among
the highest SOA of all the outputs.
But , that might also throw the current limiting circuitry
out of spec. What I am saying is that device substitution
is a little more involved than swapping out to a larger
device.
The 3 mechanisms are current gain , SOA , and the voltage rating
of the device.
You are right that SOA is dependant on temperature (A bigger
HS )
For example , if you change a device with a current gain
of 25 for one with 50+ , your bias will increase making for
a hot amp.
It is not like replacing a light bulb.
OS
In this case the power supply seems to be able to source a good amount of current. It is a large torroidial transformer w/ 44,000uf per channel capacitance (2x(2x22,000)) and seperate rectifiers for each channel.
This receiver is (was) rated as 165wpc @ 8 ohms, 200 @ 4 ohms <.1% THD. It weighs 65 pounds. I am not looking to increase power output, just make it bulletproof. Currently I measure ~400 watt peaks with music (drums) into a 4 ohm load on both channels before clipping is visible on the scope (114v pk-pk).
It seems that it will put out well over its rated power almost indefinitely at 4 ohms - I assume that eventually the heatsinks will not be able to deal with the heat, or maybe eventually it will short circuit and die.
This receiver is (was) rated as 165wpc @ 8 ohms, 200 @ 4 ohms <.1% THD. It weighs 65 pounds. I am not looking to increase power output, just make it bulletproof. Currently I measure ~400 watt peaks with music (drums) into a 4 ohm load on both channels before clipping is visible on the scope (114v pk-pk).
It seems that it will put out well over its rated power almost indefinitely at 4 ohms - I assume that eventually the heatsinks will not be able to deal with the heat, or maybe eventually it will short circuit and die.
No, this is the old Pioneer receiver. They measure pretty similarly at 8 ohms, the adcom has an advantage maintaining output voltage at 4 ohms - below that the Pioneer checks out (as it is designed to). I suspect the adcom would drive a dead short for a little while before failure while the pioneer is probably teetering on the brink of destruction at 3.9 ohms.
The adcom works perfectly but has no protection. 90% of the time I don't need the headroom of the adcom and would prefer to have the speakers protected against amp failure, and not have the heat of the adcom.
Although in reality the speaker pieces are still available on e-bay, but I have gone this far without damaging them . . .
The adcom works perfectly but has no protection. 90% of the time I don't need the headroom of the adcom and would prefer to have the speakers protected against amp failure, and not have the heat of the adcom.
Although in reality the speaker pieces are still available on e-bay, but I have gone this far without damaging them . . .
CBRworm said:
Hi
The power supply (Transformer) is the weakest link. You only gain 0.8 dB going from 8 ohms (165 watts) to 4 ohms (200 watts) per channel.
If the power supply were robust, it would deliver 165 watts in an 8-ohm load and, 330 watts per channel @ 4 ohms. That would equate to a 3 dB gain.
Cheers!
Hi
One thing that can kill an amp is a very reactive speaker. At some frequencies the impedance can become very low and/or exhibit a phase change. This change in phase of current wrt voltage can be very hard on outputs, BJT's especially. On a resistive load, the maximum current flows when the voltage is at maximum. Thus there is minimum voltage left across the output transistor for that period of time. But if the phase of the speaker impedance is 45 degrees (could even be greater than that) then the maximum current flows when the outputs have half (or more) of the rail voltage across it for that period of time. Since P=V*I, the instantaneous power in heat dissipated by the die of the transistor is very high and can exceed the de-rated SOA........then POOF the transistor will melt.
One thing that can kill an amp is a very reactive speaker. At some frequencies the impedance can become very low and/or exhibit a phase change. This change in phase of current wrt voltage can be very hard on outputs, BJT's especially. On a resistive load, the maximum current flows when the voltage is at maximum. Thus there is minimum voltage left across the output transistor for that period of time. But if the phase of the speaker impedance is 45 degrees (could even be greater than that) then the maximum current flows when the outputs have half (or more) of the rail voltage across it for that period of time. Since P=V*I, the instantaneous power in heat dissipated by the die of the transistor is very high and can exceed the de-rated SOA........then POOF
the transistor will melt.
I guess it's between not many and not any.OMNIFEX said:
You will get many amps that will gain >=1.5dB as test resistance halves.
You will find quite a few that will gain >=2dB as test resistance halves.
You will find very few that will gain >=2.5dB as test resistance halves.
Even power amplifiers that have a regulated supply cannot manage +3dB as the test resistance is reduced to half the specified minimum speaker resistance (reactive load is much worse than a resistive load).
AndrewT said:
What about Mark Levinson and, Krell?
This amplifier offers less than 1 dB increase from 8 ohm to 4 ohm.
Not taking the program content, the impedance dips of the frequency based on the cabinet, in addition to how long and, loud is needed, into account, having a 1 dB gain is not worth the torture this receiver will face when driving 4 ohm nominal loads.
It is safer to say this is an 8 ohm amplifier which can handle impedance dips of 4 ohms. Old Receivers like these just burn themselves up since they offer no thermal protection.
Cheers!
Ok, that makes sense. The speakers I have dip down to 2 ohms at some points and are definitely hard to drive. This may be due to phase change, etc. Each speaker has 2 10's with a 1,200 uf cap in series with the input as well as a couple of inductors across the output to the woofers.
So that goes back to my original question, if I were to put in more modern output transistors like MJ21193/94 would they be more tolerant of a highly reactive load and less likely to 'POOF' or will they still die the same death, or will the load be put on the driver transistors.
The outputs in the amp from the factory are 4 pairs of NEC 2SC600 and 2SD555 (or maybe I have that backwards).
I am not concerned with the power supply or its lack of ability. My concern is to keep the amplifier components from letting the smoke out. The power output level is fine.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.