Hi everyone,
What is the best/proper method for measuring amp output power?
And, once this is determined, can you really expect to see the db levels from published speaker/driver specs? That is, if the spec says 95 db/1w/1m, is this actual? Can you really rely on a sould level meter reading from say a hand held Audio Dynamics meter?
Thanks,
Rick
What is the best/proper method for measuring amp output power?
And, once this is determined, can you really expect to see the db levels from published speaker/driver specs? That is, if the spec says 95 db/1w/1m, is this actual? Can you really rely on a sould level meter reading from say a hand held Audio Dynamics meter?
Thanks,
Rick
Power measurement
Hi,
Inject 1KHz sine-wave.
Increase the input level until just before clipping starts to occur. The distortion is likely to be 2 to 5% at this point. Confirm it if you can.
Distortion onset rate depends on topology and design.
Use a dummy load, and measure the voltage across it.
P= V*V/R where V is the RMS value.
Be aware that DVM's will read a little low at 1KHz.
Cheers,
Hi,
Inject 1KHz sine-wave.
Increase the input level until just before clipping starts to occur. The distortion is likely to be 2 to 5% at this point. Confirm it if you can.
Distortion onset rate depends on topology and design.
Use a dummy load, and measure the voltage across it.
P= V*V/R where V is the RMS value.
Be aware that DVM's will read a little low at 1KHz.
Cheers,
Re: Power measurement
Ditto, except I use a THD meter and spec performance at 1% and 5%, and will sometimes do a number of frequencies to check performance at either end of the spectrum. It's hard to pick 1% THD on a CRO.
dhaen said:
Ditto, except I use a THD meter and spec performance at 1% and 5%, and will sometimes do a number of frequencies to check performance at either end of the spectrum. It's hard to pick 1% THD on a CRO.
Thanks guys.
So I use a signal generator input and an o-scope at input and across a 8 ohm resistor (dummy load) after the OPT? Measure V there (across resistor) and use the power equation, right? Then, that is true power to speaks?
Or, are you saying forget the OPT and measure V across dummy load (resistor) used in lieu of OPT? Then, convert to Pout considering trans ratio, etc?
Brett, I don't have a THD instrument--just have DVM, O-scope, signal gen-but thanks anyway. What's a CRO?
Assuming I get Pout, can I then expect to hear 95 db at 1 m at 1 watt using a speak of 95 db/m/w? That's pretty loud isn't it?
Thanks,
Rick
So I use a signal generator input and an o-scope at input and across a 8 ohm resistor (dummy load) after the OPT? Measure V there (across resistor) and use the power equation, right? Then, that is true power to speaks?
Or, are you saying forget the OPT and measure V across dummy load (resistor) used in lieu of OPT? Then, convert to Pout considering trans ratio, etc?
Brett, I don't have a THD instrument--just have DVM, O-scope, signal gen-but thanks anyway. What's a CRO?
Assuming I get Pout, can I then expect to hear 95 db at 1 m at 1 watt using a speak of 95 db/m/w? That's pretty loud isn't it?
Thanks,
Rick
Rick,
Yes and no 😉
Your speakers won't be the exact impedance quoted, so there'll be a small error.
CRO = cathode ray oscilloscope. - He must be old
I thought you might not have a distortion meter. You should buy or build one if you're serious.
I'll leave the loudness measurements for someone more expert than me to answer.
Cheers,
Yes and no 😉
Your speakers won't be the exact impedance quoted, so there'll be a small error.
CRO = cathode ray oscilloscope. - He must be old
I thought you might not have a distortion meter. You should buy or build one if you're serious.
I'll leave the loudness measurements for someone more expert than me to answer.
Cheers,
I use my distortion analyzer and generally check for power output and distortion at 100hz, 1K, and 20K.
Same method as others... scope and RMS meter.
Speaker output....
Is generally measured in a room set up for testing. Conditions can affect the output. Don't be surprised if the efficiency isn't quite the same as specified as you won't be able to duplicate the testing conditions.
The meter itself will have a bearing on the testing. Is it a type 1, or type 2 meter? Is the mic calibrated? Has the meter been certified this year?
I sincerely doubt that the meter in question is capable of this precise measurement. There are too many variables involved because you probably don't know what frequency/s was used. Without knowing how the testing was setup you will have a slim chance of being able to duplicate the test.
Be happy and enjoy your speakers.... I'm sure they are every bit as good as their specifications.
Joe
Same method as others... scope and RMS meter.
Speaker output....
Is generally measured in a room set up for testing. Conditions can affect the output. Don't be surprised if the efficiency isn't quite the same as specified as you won't be able to duplicate the testing conditions.
The meter itself will have a bearing on the testing. Is it a type 1, or type 2 meter? Is the mic calibrated? Has the meter been certified this year?
I sincerely doubt that the meter in question is capable of this precise measurement. There are too many variables involved because you probably don't know what frequency/s was used. Without knowing how the testing was setup you will have a slim chance of being able to duplicate the test.
Be happy and enjoy your speakers.... I'm sure they are every bit as good as their specifications.
Joe
Thanks all,
Here's where I am: I have "built" a few tube amps (SET and SEP) in the 1.4 to 4.5 watt range (12eh5 to 6bq5 to 6aq5). I have plans for more such as 6bm8, 6L6, 807, and 300b. Right now I really like the sound of the single tube amp (that is, the 12eh5 is biased at -3v and puts out reportedly 1.4 watts). Oh, almost forgot, I have a 45 SET, too.
I run a cdp straight into the grid of the 12eh5 (pentode) and output plate in series with the OPTs. (BTW, I have tried the 6cl6 but did not like it at all).
I run either some Altec 605a's closed boxes (103 db) or JBL Control 28 (92 dbl) with these amps. Not really satisfied with either. I want to confirm that I am getting all the power of the tube before buying more speakers (maybe pro speakers, looking at more horns). I really like the sound and imaging of the se 12eh5 with the JBL's but just not enough SPL.
Rick
Here's where I am: I have "built" a few tube amps (SET and SEP) in the 1.4 to 4.5 watt range (12eh5 to 6bq5 to 6aq5). I have plans for more such as 6bm8, 6L6, 807, and 300b. Right now I really like the sound of the single tube amp (that is, the 12eh5 is biased at -3v and puts out reportedly 1.4 watts). Oh, almost forgot, I have a 45 SET, too.
I run a cdp straight into the grid of the 12eh5 (pentode) and output plate in series with the OPTs. (BTW, I have tried the 6cl6 but did not like it at all).
I run either some Altec 605a's closed boxes (103 db) or JBL Control 28 (92 dbl) with these amps. Not really satisfied with either. I want to confirm that I am getting all the power of the tube before buying more speakers (maybe pro speakers, looking at more horns). I really like the sound and imaging of the se 12eh5 with the JBL's but just not enough SPL.
Rick
Rick, just in case you have a need to check your speakers rating you can do the pout calculation backwards to derive output voltage for 1 watt.🙂
Wattage
If your scope is calibrated and checked for accuracy, you can measure the p-p voltage and then divide by 2.82 to obtain the RMS voltage.
This voltage will also include the harmonic voltages. But it should be fairly close to accurate.
Then take the RMS voltage and divide by the load resistance to obtain the RMS current. Multiply the RMS voltage by the RMS current.
This should give you a rough estimate of the power output.
Hope this helps.
If your scope is calibrated and checked for accuracy, you can measure the p-p voltage and then divide by 2.82 to obtain the RMS voltage.
This voltage will also include the harmonic voltages. But it should be fairly close to accurate.
Then take the RMS voltage and divide by the load resistance to obtain the RMS current. Multiply the RMS voltage by the RMS current.
This should give you a rough estimate of the power output.
Hope this helps.
My 
If memory serves, the speaker efficiency with regards power transfer is a major factor. If you put out, or measure, 1 watt with a watt meter, or the non-inductive load, it may not be the same at the speaker.
I mean, you can have a speaker that is designed to handle 10 watts (1 inch voice coil, 5 ounce magnet) be 90dB/watt, and a speaker designed for 500 watts (2.5 inch coil, 10 pound magnet) be 90dB per watt. Will the 1 watt amp drive them to the same actual dB? Nope. That is because there is power loss in the bigger power speaker.
So, the rating really means that if you have enough power that the speaker processes 1 watt, you will get 90 dB. The 500 watt speaker may need 50-100 watts of amplifier power to give you close to 90dB.
In my opinion, if memory serves. Haven't read about this in 15 years. Could be starting another war! 😉
Gabe
If memory serves, the speaker efficiency with regards power transfer is a major factor. If you put out, or measure, 1 watt with a watt meter, or the non-inductive load, it may not be the same at the speaker.
I mean, you can have a speaker that is designed to handle 10 watts (1 inch voice coil, 5 ounce magnet) be 90dB/watt, and a speaker designed for 500 watts (2.5 inch coil, 10 pound magnet) be 90dB per watt. Will the 1 watt amp drive them to the same actual dB? Nope. That is because there is power loss in the bigger power speaker.
So, the rating really means that if you have enough power that the speaker processes 1 watt, you will get 90 dB. The 500 watt speaker may need 50-100 watts of amplifier power to give you close to 90dB.
In my opinion, if memory serves. Haven't read about this in 15 years. Could be starting another war! 😉
Gabe
RE:Could be starting another war!
Hi,
Gabe, I don't think that's correct.
Speaker effeciency as stated by the manufacturer is, or shall we say should, be the same regardless of its size or powerhandling.
Actually the poweroutput capabilty of an amp should be calculated by using a resistance of say 8,4 or 16 Ohm to get the rating.
You can measure it with a speaker connected but that will only tell you how much the amp is delivering with that particular load.
I don't know the formula by heart but I'm sure it's derived from Ohm's law.
Cheers,😉
Hi,
Gabe, I don't think that's correct.
Speaker effeciency as stated by the manufacturer is, or shall we say should, be the same regardless of its size or powerhandling.
Actually the poweroutput capabilty of an amp should be calculated by using a resistance of say 8,4 or 16 Ohm to get the rating.
You can measure it with a speaker connected but that will only tell you how much the amp is delivering with that particular load.
I don't know the formula by heart but I'm sure it's derived from Ohm's law.
Cheers,😉
Gabevee said:My
The 500 watt speaker may need 50-100 watts of amplifier power to give you close to 90dB.
I don't think anyone will be lining up to buy that driver. Don't know if this applies or not but the formula for wattage to db is: W=logx10. This is added to your 1 watt spl rating for db output at any given wattage. The formula for power is: power=logx20.🙂
OK, OK already, I did say if memory serves, doggoneit!
I also did a few searches on the web, and one place did say something similar to what I said. However, what the person said was that one has to look at the overall efficiency (I forgot the exact term). Remember that generally the speaker is tested for that SPL at 1kHz. It may not be 90 dB at other frequenies.
Now, I am sure we all know this.
Thanks for the correction. As I said, I read about this stuff 15 years or so ago, when I built my first pair of speakers,which BTW I still listen to. I haven't thought much about speakers since, except for the crossover networks, which I readily build and put into speakers I buy, improving their quality.
So... ignore what I said! 😉
OTOH, one must not think that either one watt is enough or that 2 watts will give one 180 dB. Doubling the power only gives a 3dB boost, hardly loud enough to detect. Perceptible doubling required 10dB, if memory serves (Uh, oh, I am in trouble again! 😉 ).
Later!
Gabe
I also did a few searches on the web, and one place did say something similar to what I said. However, what the person said was that one has to look at the overall efficiency (I forgot the exact term). Remember that generally the speaker is tested for that SPL at 1kHz. It may not be 90 dB at other frequenies.
Now, I am sure we all know this.
Thanks for the correction. As I said, I read about this stuff 15 years or so ago, when I built my first pair of speakers,which BTW I still listen to. I haven't thought much about speakers since, except for the crossover networks, which I readily build and put into speakers I buy, improving their quality.
So... ignore what I said! 😉
OTOH, one must not think that either one watt is enough or that 2 watts will give one 180 dB. Doubling the power only gives a 3dB boost, hardly loud enough to detect. Perceptible doubling required 10dB, if memory serves (Uh, oh, I am in trouble again! 😉 ).
Later!
Gabe
This is a rather old post so I'm hoping to revive it to get an answer to my recent experience and question.
So ...I have a question about this procedure.
I was using this method just yesterday to measure power on my cathode-biased 6V6 PP amp. (class AB1)
I injected 1khz and cranked it to just the edge of distortion/clipping point on my scope. Then I measured the RMS voltage across my 4 ohm dummy speaker-load.
However, my friend advised me that I cannot accurately measure power that way because the by-pass caps will keep charging up and increase the bias voltage reducing the effective output power.
And sure enough, when I put my voltmeter on the cathode voltage and I saw the bias voltage creep up over 2 volts!
So.. what is the proper way to dynamically measure power?
Last edited:
This is a big reason to go with fixed bias!
You can temporarily solder in some zener diodes to clamp the cathode voltage.
You can also put a normally closed momentary switch across your signal source so that you can blip the input and do a peak hold on the output.
You can temporarily solder in some zener diodes to clamp the cathode voltage.
You can also put a normally closed momentary switch across your signal source so that you can blip the input and do a peak hold on the output.
Just like you did.
The fact that cathode voltages rises is just a function of - cathode bias. And
this is the amp you have, this is the power you will get.
Building an amp with fixed bias will get more power, since the voltage drop
across the cathode resistor is gone. It's not all benefits, fixed bias also
introduces the risk of "redplating" as the feedback mechanism is gone.
In cathode based amps, if the current through the tube increases, the bias will
increase negating the current increase. When using fixed bias no such
mechanism exists.
- Status
- Not open for further replies.