I have just finished my second amp, a custom designed stereo ipod amp. The design was for 3.5 watts / channel and I'd like to see what I'm actually getting. I have a 1KHz sine wave mp3 file which gives me 0.71Vrms at the input. I'm thinking I can put a 0.1 ohm resistor in series with the speaker and measure the Vrms across the resistor and across the speaker. Doing the math to convert the voltage across the resistor to amps, is it simply V x A ?
While it is not a speaker, I use a resistive load for power measurements as speaker impedance can pretty much be considered off specified value at almost any frequency.
Measure the resistor with a precision ohm meter, or measure by feeding an accurately measured voltage to get your cal value.
From there, I adjust the output power while looking at the waveform on a scope and increase input till I get the onset of clipping. This lets me measure max continuous power.
I also like to look at distortion with a spectrum analyzer (Audio Tester) at 1W out.
Then I ignore my measurements and go listen to it.
Measure the resistor with a precision ohm meter, or measure by feeding an accurately measured voltage to get your cal value.
From there, I adjust the output power while looking at the waveform on a scope and increase input till I get the onset of clipping. This lets me measure max continuous power.
I also like to look at distortion with a spectrum analyzer (Audio Tester) at 1W out.
Then I ignore my measurements and go listen to it.
I rig the amp to a resistive load and crank up the input signal until I get 2~3 % THD on the output (measured with a distortion analyzer). Then I measure the output RMS voltage and use P = E^2/R to find the output power. I use the same method to measure output power at 1 % THD and measure the THD at 1 W out. I perform this measurement at mid band (1 kHz) with spot checks at lower and higher frequencies. I usually also run a distortion vs power vs frequency sweep.
Now, why 2~3 % THD? Well... In my 300B SET with a source follower grid drive on the 300B, I have lots and lots of grid current available. But the 300B isn't designed for A2 operation as far as I know. I have noticed that once significant grid current starts to flow, the THD basically falls off a cliff. That tends to happen around 2~3 % THD, so I stop my measurements there. At that point the output waveform is also rather ugly on the o'scope.
If you don't have a dedicated distortion analyzer, you can use a computer sound card and some software. TrueRTA comes to mind.
As far as calculating output power on paper, I'd just look it up in the datasheet for the output tube. Or perhaps dust off RDH4 and look in Chapter 13 (AFAIR).
~Tom
Now, why 2~3 % THD? Well... In my 300B SET with a source follower grid drive on the 300B, I have lots and lots of grid current available. But the 300B isn't designed for A2 operation as far as I know. I have noticed that once significant grid current starts to flow, the THD basically falls off a cliff. That tends to happen around 2~3 % THD, so I stop my measurements there. At that point the output waveform is also rather ugly on the o'scope.
If you don't have a dedicated distortion analyzer, you can use a computer sound card and some software. TrueRTA comes to mind.
As far as calculating output power on paper, I'd just look it up in the datasheet for the output tube. Or perhaps dust off RDH4 and look in Chapter 13 (AFAIR).
~Tom
Last edited:
@astouffer
Meter is a VC99, spec for AC volts frequency response is 40-400 Hz.
I'm assuming I can't use this multimeter to measure frequencies in the KHz ?
Let's forget about clipping and % distortion for a minute. I just want to know
how I can get a rough measure (+/- 10%) of the power being delivered to the speaker.
As TheGimp mentioned, I'll actually be using an 8 ohm 5 Watt resistor.
Meter is a VC99, spec for AC volts frequency response is 40-400 Hz.
I'm assuming I can't use this multimeter to measure frequencies in the KHz ?
Let's forget about clipping and % distortion for a minute. I just want to know
how I can get a rough measure (+/- 10%) of the power being delivered to the speaker.
As TheGimp mentioned, I'll actually be using an 8 ohm 5 Watt resistor.
Last edited:
Connect your voltmeter across the speaker. Play your test audio. Measure the RMS volts on the speaker. Compute the power using P = E^2/R where R is the speaker impedance.
A 400Hz tone would be much better to use. If you can't obtain one, maybe an audio editor like Goldwave could be used to "stretch" your 1kHz sinewave to 400Hz.
This isn't precision, but should give a useful rough measurement.
A 400Hz tone would be much better to use. If you can't obtain one, maybe an audio editor like Goldwave could be used to "stretch" your 1kHz sinewave to 400Hz.
This isn't precision, but should give a useful rough measurement.
The Bias & B+ are informed by the amp datasheet or service manual.
Example in a big Triode amp:
0.80mA x 900V x Transformer= output Watts
The transformer value is a Primary/Secondary ratio I unknow.
Some expert would help??
You obviously have a computer... That computer has a sound card. It may not be the best quality equipment, but it's generally good enough. Use that along with the free or demo spectrum analyzer and o'scope software around and you have a distortion analyzer and an oscilloscope.
Be careful with the sound card, though. They tend to have max input voltages on the order of 5 V, so you may have to build a 10:1 voltage divider so you don't fry the sound card when measuring the amp output.
~Tom
Don't use the speaker for the load as you will have no idea as to what impedance it is.
Here's the simple answer:-
Use an 8.2R, 5W resitor as the load. Get your speaker and add a 330R resitor in series with it.
Now put the speaker/330R resistor combination in parallel with the 8.2R resistor.
The amp will now see an 8R load.
You can then put your meter across the output the amp and wind up the sine wave until you can hear it starting to distort (it will be fairly quiet), back off the level so it no longer distorts and read the meter (should be RMS volts).
You now have the RMS Volts and the correct load impedance so you can work out the power.
Regards
Henry
Here's the simple answer:-
Use an 8.2R, 5W resitor as the load. Get your speaker and add a 330R resitor in series with it.
Now put the speaker/330R resistor combination in parallel with the 8.2R resistor.
The amp will now see an 8R load.
You can then put your meter across the output the amp and wind up the sine wave until you can hear it starting to distort (it will be fairly quiet), back off the level so it no longer distorts and read the meter (should be RMS volts).
You now have the RMS Volts and the correct load impedance so you can work out the power.
Regards
Henry
I believe the speaker impedance is 8 ohms. The question is, at what frequency? I'm also assuming 400Hz is the frequency used by the speaker manufacturer.
A suitable multimeter can give a "ballpark" distortion indication: Take a peak reading of the sine wave. Compare that with an RMS reading of the same sine wave. The latter should be 0.707 of the former. If it isn't, that's a good sign that the signal is clipping.
A suitable multimeter can give a "ballpark" distortion indication: Take a peak reading of the sine wave. Compare that with an RMS reading of the same sine wave. The latter should be 0.707 of the former. If it isn't, that's a good sign that the signal is clipping.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.