I have a question I need to ask the amplifier gurus. It's been bugging me for quite some time now. I have researched it and still can't find an answer !
Q) Let's say we have a speaker system, to where a portion of the impedance
load goes to a very high value (say 75 ohms for discussion sake). It's an amplifier rated at 100 watts into an 8 ohm load. I know from ohm's law that the power output should be severely limited. Does the amplifier clip when asked to deliver more power than is available, at that frequency spectrum ?
Q) Let's say we have a speaker system, to where a portion of the impedance
load goes to a very high value (say 75 ohms for discussion sake). It's an amplifier rated at 100 watts into an 8 ohm load. I know from ohm's law that the power output should be severely limited. Does the amplifier clip when asked to deliver more power than is available, at that frequency spectrum ?
Most SS power amplifiers amplify voltage, and the resulting output current depends on
the load impedance, within the amplifier's capability. The voltage output remains
the same into a high impedance peak, but less current flows for signals at that frequency,
and so less power is delivered. Most speakers are intended to be driven by a voltage source.
Last edited:
This effect is very common in the bass region (say sub 100Hz) of a speaker. Fortunately the efficiency of the speaker at that frequency (expressed in terms of dB per Watt input) also increases in proportion.
> Does the amplifier clip when asked to deliver more power than is available, at that frequency spectrum ?
(Most) Amplifiers clip according to the INput VOLTAGE signal, not really by the load (if not too extreme) or Power.
> speaker system, to where a portion of the impedance load goes to a very high value
That's the Speaker Designer's problem. Not the amplifier designer, not the user.
As said, such impedance rises usually reflect driver efficiency areas higher than the average over the total bandwidth of the speaker. I had some E-V speakers with 2% woofers and tweeters that were 20% at 1KHz (but 1% at 10KHz). The crossover forced a huge impedance rise in the low range of the tweeter, so it just did not take much power, and the tweeter effective efficiency was ~~1% 1KHz-10KHz, a good match to the woofer (being more omni, it needed more output to sound flat on-axis).
(Most) Amplifiers clip according to the INput VOLTAGE signal, not really by the load (if not too extreme) or Power.
> speaker system, to where a portion of the impedance load goes to a very high value
That's the Speaker Designer's problem. Not the amplifier designer, not the user.
As said, such impedance rises usually reflect driver efficiency areas higher than the average over the total bandwidth of the speaker. I had some E-V speakers with 2% woofers and tweeters that were 20% at 1KHz (but 1% at 10KHz). The crossover forced a huge impedance rise in the low range of the tweeter, so it just did not take much power, and the tweeter effective efficiency was ~~1% 1KHz-10KHz, a good match to the woofer (being more omni, it needed more output to sound flat on-axis).
You have it back to front. If the load has unusually high impedance then the amplifier is being asked to deliver much less power than usual, not more. It is when the load has a low impedance point that the amplifier might clip if it has limited current capability, which could be due to inadequate design or deliberate current limiting as a protection feature.Scott L said:
Note that an amplifier is asked by the input signal to deliver a voltage to the speaker; it is then asked by the speaker to deliver a current.
Okay, let's consider a driver that is essentially in free air. The impedance peak is 75 ohms at 38 Hz (for discussion sake). We need all the power we can get, but for the sake of discussion, the amplifier is 100 watts at 8 ohms. What's the maximum power available at 38Hz ?
A voltage amplifier clips at a certain level of voltage.
Since power = VxV/R, then Vclip = SQRT(R x power) = 28.3Vrms.
Then power(38Hz) = VclipxVclip/75R = 10.7W, assuming the impedance at 75R is purely resistive.
THANK YOU !!
Gawd, it's like pulling teeth ! Less power is available to do the task. Now just imagine if something like 12 db equalization is also needed to make the design "work". Obviously, it can not.
Gawd, it's like pulling teeth ! Less power is available to do the task. Now just imagine if something like 12 db equalization is also needed to make the design "work". Obviously, it can not.
Yes, I do believe you are on to something here. A driver in free air doesn't have much of a "load" now does it ?
Commercial speakers are generally designed to be used with a voltage amplifier.
For an alternative view, see Current-Drive - The Natural Way of Loudspeaker Operation
Any impedance between a "perfect" voltage amplifier and the speaker terminals
will alter the frequency response at those terminals due to the voltage divider action
with the speaker system impedance curve. Tube amplifiers often show this effect clearly
due to their output impedance.
Last edited:
> Sorry but you have it upside down.
> Amplifiers clip when OUTPUT
Amplifier gain is essentially constant.
Amplifier supply voltage is nearly constant (except in severe use).
It will clip when he puts too much in the INput. Not really by what he does on the output. Yes, the actual pinch point is output clipping, but this is predicable from input voltage, not from output loading.
Yes, on single-tone he may clip on 1V input at 250Hz where the load is 7.8 Ohms, and he may be able to put in 1.3V at 55Hz where the speaker is 50+ Ohms. The input voltage for clipping changes much less than the output loading.
Upside down is me in my cellar prying beams off the ceiling onto my head. I may be a bit dazed but I know which way is up.
> Amplifiers clip when OUTPUT
Amplifier gain is essentially constant.
Amplifier supply voltage is nearly constant (except in severe use).
It will clip when he puts too much in the INput. Not really by what he does on the output. Yes, the actual pinch point is output clipping, but this is predicable from input voltage, not from output loading.
Yes, on single-tone he may clip on 1V input at 250Hz where the load is 7.8 Ohms, and he may be able to put in 1.3V at 55Hz where the speaker is 50+ Ohms. The input voltage for clipping changes much less than the output loading.
Upside down is me in my cellar prying beams off the ceiling onto my head. I may be a bit dazed but I know which way is up.
To decide whether a design is viable should be based on SPL output, not power delivered to the load.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.