With tube amps, this does not matter that much. The main difference will be in the sound, the 4R tab will have more bass and less treble, the 8R tab more treble and less bass. This is off course avery broad generalisation. In reality you need to test both to find what fits you the best, and neighter will do any damage or change the sound very drasticly.
with solid state amps, this does matter a lot more as their is no output transfo, and the wattage change depending on the tab. The main issue may be that the psu can't handle it on high volume and the amp gets overcharged.
with solid state amps, this does matter a lot more as their is no output transfo, and the wattage change depending on the tab. The main issue may be that the psu can't handle it on high volume and the amp gets overcharged.
Depending on how your tube amp is constructed, you might actually get a little more output voltage across 8R speakers.
You will almost always get at least a little more voltage into 8 ohms. Less voltage drop across the tube, less voltage drop in the resistance of the transformer, and less droop in the power supply at full load. That means less than a 3 dB drop. Is it 1 dB or 2dB, you pretty much have to measure it or at least run through some math.
An issue often overlooked is that if plate current drops at full voltage swing due to higher impedance, the screen current goes up and you have to keep an eye on it. You can usually drop the screen voltage (which means less g2 power dissipation) without losing too much of the voltage swing you gained by raising impeadance. If it was running conservatively in the first place you might be ok. If you tended to melt the screens down and short them out it may actually get worse unless you drop Vg2.
An issue often overlooked is that if plate current drops at full voltage swing due to higher impedance, the screen current goes up and you have to keep an eye on it. You can usually drop the screen voltage (which means less g2 power dissipation) without losing too much of the voltage swing you gained by raising impeadance. If it was running conservatively in the first place you might be ok. If you tended to melt the screens down and short them out it may actually get worse unless you drop Vg2.
I highly doubt that---even in a solid state amp, it's usually less---for instance, a Peavey NV400 puts out a quoted 210 watts into 4Ω & 130 watts into 8Ω. That's ~ 2db. With a tube amp, it's gonna be less.
Doubt it is specifically designed for exactly 4 ohms. Tubes will have a desired impedance in which they will put out max watts at a specific plate voltage. You have to measure it.
I've built guitar amps that put out the most watts into about 6 ohms, 4 or 8 ohms was close.
If a mismatch is employed the primary impedance is affected with a valve amplifier.
Say for instance you have a push pull EL34 amplifier, the anode impedance that the valves are designed to work with is 6k. If you place a 4R load on the 8R tapping, the effect is to almost halve the load impedance because if your transformer has a 25 - 1 ration, the ratio does not change so the anode load will reduce. The opposite applies to an 8R load on the 4R tapping.
All that happens is the point at which distortion happens changes.
Say for instance you have a push pull EL34 amplifier, the anode impedance that the valves are designed to work with is 6k. If you place a 4R load on the 8R tapping, the effect is to almost halve the load impedance because if your transformer has a 25 - 1 ration, the ratio does not change so the anode load will reduce. The opposite applies to an 8R load on the 4R tapping.
All that happens is the point at which distortion happens changes.
True for a ideal voltage source.
The product of power times impedance is a constant (above '800')
Amplifers in real life are a tiny bit less then ideal yet close to perfection.
Depending on design and topology you get different values.
I follow Kay in his argument.
Solid state amps are close to perfection as far as nearly zero output impedance.
Valve amps are very far from it - as John DH says. Another diyAudio member, Printer2, has also measured output impedances up to 50 ohms from the 8-ohm tap of a valve guitar amplifier. 50 ohms is so big compared to an 8 ohm speaker impedance that this means valve amps are closer to acting as a current source than a voltage source.
DF96 was exactly right. For a perfect voltage source, output power doubles when you halve the load impedance. For a perfect current source, power halves when you halve the load impedance. For any actual power amp, you'll get something between these two extremes.
-Gnobuddy
The small-signal output impedance is NOT the large-signal (power output) impedance!!
Naked pentodes may be 50 Ohms, small-signal, on the speaker winding, but if appropriately designed the Most Power will happen very near 8 Ohms (or whatever). It will fall-off either side. Since tube curves are "soft" it won't follow any exact rule, like a solid-supply transistor amp making double power in half impedance. (And even this only goes so far, until PT sag spoils the trend, or protection cuts-out).
Naked pentodes may be 50 Ohms, small-signal, on the speaker winding, but if appropriately designed the Most Power will happen very near 8 Ohms (or whatever). It will fall-off either side. Since tube curves are "soft" it won't follow any exact rule, like a solid-supply transistor amp making double power in half impedance. (And even this only goes so far, until PT sag spoils the trend, or protection cuts-out).
I was thinking about this issue earlier, and I think it almost makes no sense to define an output impedance for a device that's being hustled from cut-off to saturation every half-cycle. There isn't "an" output impedance, there is an impedance that varies continuously during every audio cycle, and varies a lot.
There are geek jokes about "the science of variable constants", but if ever there was a case of "constants" that aren't constants at all, it's gotta be any of the major vacuum tube parameters of a guitar amp output valve that's being driven deep into overdrive!
It's a different story with small signals, when there is much less variation of a parameter during a cycle of audio signal, and so it's reasonable, and makes mathematical sense, to find an average value and put that in a data sheet.
-Gnobuddy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.