Testing WOPLS "automated" by Jer...

laatsch55

Administrator,
Staff member
Joined
Jan 14, 2011
Messages
74,126
Location
Gillette, Wyo.
Tagline
Halfbiass...Electron Herder and Backass Woof
#21
I see a difference when feeding the output to the input of the AP too Joe...cables and connectors ARE important...
 

oldphaser

Chief Journeyman
Joined
Sep 8, 2012
Messages
845
#22
Yeah Ed, was wondering.. metals do make a huge diff in some instances LOL.

Now a question.. I have two low pass filters that I can switch in for the reading. The below metrics are with no filtering, but yeah if I do put the 30Khz low pass filter, then at higher freqs the readings are much lower. So does that mean that the analyzer is seeing harmonics at higher freqs or is it common to have a low pass filter "in" when measuring?
Several years ago, I watched an interview with Justice Scalia and Ginsberg. In this interview Scalia talked about "originalism". "Originalism is a concept regarding the interpretation of the Constitution that asserts that all statements in the constitution must be interpreted based on the original understanding of the authors or the people at the time it was ratified."

In order for us all to compare Phase Linear amplifiers in a apples-to-apples manner, we need to use common terms, definitions and measurement techniques. That is why I have been on a quest to obtain as much reference material as I can. We also need to be aware that the terms, definitions and measurement techniques have also changed over the years. The industry standards during the era are very helpful in understanding these terms, definitions and measurement techniques.

In the Phase Linear 400 and 700/700B series 1 service manual it states the following:

step V-5 or 5-5:
Distortion: Monitor the left channel output with a THD analyzer. With an 8 ohm load connected to the left channel output, drive the left channel to (40V RMS for a 400) or (53VRMS for a 700) and verify a THD (later redefined as THD+N) of less than 0.25%. Repeat this procedure with the right channel. The same procedure may also be used to test for IM distortion of 0.25% or less

"NOTE: It may be necessary to utilize the high filter in the distortion analyzer if it is so equipped."


section VI TROUBLESHOOTING AND ALIGNMENT

THD analyzer adjustments.
1. Activate the 80kHz filter on the analyzer if it is so equipped and repeat THD measurement.
2. Examine test bench grounding system for possible ground loops. Ground the analyzer directly to the 400 or 700/700B output ground and not the load, scope or meter grounds. Use 16 gauge wire for the ground system.
3. Reverse the line cord polarity by removing the line plug from the AC outlet, reversing it and plugging it in again.


Therefore, all my measurements for THD (later redefined as THD+N) on Phase Linear amplifiers is currently with a 80kHz low pass filter.

NOTE: The industry standard IHF-A-201 (1966) Connection of Line Cord. "The line cord shall be connected for minimum hum on the highest-gain input and shall not be changed for any other test. One side of the power cord shall be grounded." The IHF-A-202 (1978) revision is the same with the exception that it says "When possible, the line cord shall...." This comment was probably added since we now have many polarized line-cords. If my understanding is correct, "HOT" is required to go through the power switch on those amplifiers so equipped.



I am also most particularly interested in 20kHz measurements at rated power and not so much 1kHz measurements as 20kHz is where many amplifiers have the most difficulty in meeting the manufacturer's specifications.


Signal-to-noise ratio measurements previous to the introduction of the series 2 amplifiers were "unweighted". That is to say flat within the bandwidth of the piece of test equipment you were using. Typically it was using a AC millivoltmeter and in that era it was approximately 5Hz to 500kHz. After 1978 most manufacturer's went to the "A" weighting measurement which will improve the signal-to-noise ratio measurement by 10dB or more.

NOTE: I have already previously discussed the different types of meters; averaging, peak and QMS peak. The IHF standard states: "the voltmeter shall have average rectifying characteristics and be calibrated to read rms voltage....."

For signal to noise ratio measurements: "the input is not left open, or unterminated." In the IHF-A-201(1966), the input terminals were shorted. Then in 1978 (IHF-A-202) a 1k ohm resistor was placed across the line inputs and this still exists in CTA-490-A R2008. For some other folks; "The usual practice is to leave the unit connected to the signal generator (with its low output impedance) set for zero volts."

"Digital multimeters often have adequate resolution for measuring small ac voltages, but will not necessarily have good accuracy at potentials as low as a few millivolts. Also, many digital multimeters have very restricted bandwidths. There are some exceptions, but most only offer good accuracy over a frequency range of only about 50Hz to 500Hz, which is obviously far less than the full 20Hz to 20kHz audio range. It is totally inadequate for most audio frequency testing."

Ed
 
Last edited:

oldphaser

Chief Journeyman
Joined
Sep 8, 2012
Messages
845
#24
I haven't been able to find a complete 8903A manual yet. I am guessing that it might be similar to the 8903B. If my memory serves me correct, the 8903B was also the foundation for the Audio Precision test systems early on.

Here are a few sentences taken throughout the 8903B manual:

Signal-to-Noise Ratio

"Measurement of the signal-to-noise ratio requires the use of the Audio Analyzer’s internal source. The Audio Analyzer simply turns the source (set to a specified level) on and off and measures the ac level for both conditions."

"Signal-to-noise ratio measurements are also filtered for improved repeatability and speed (1 reading/second typical), and automatic display rounding is provided. For accurate noise measurements, the Audio Analyzer uses true rms detection for both SINAD and signal-to-noise measurements. Most older instruments employ average detection which reads low for noise. The discrepancy can be 1.5 dB or greater and varies with the ratio being measured. For correlating results with past test data, the Audio Analyzer’s detector can be switched via special functions to an average responding configuration."

"When measuring complex waveforms or noise, a true rms detector will provide a more accurate measurement result than an average-responding detector that has been calibrated to indicate the rms value. For a sine wave, both the true rms and the average-responding detectors give correct rms readings. However, when the signal is a complex waveform, or when significant noise is present, the averageresponding detector reading can be in error. The amount of error depends upon the particular signal being measured. For noise, an average-responding detector reads low."

"The Audio Analyzer measures the true rms level on all ac measurements. True rms measurements assure greater accuracy when measuring complex waveforms and noise. For those applications where average detection is required, the analyzer can be switched to average-responding (rms calibrated) detection through special functions."



"For those applications requiring the use of an average-responding detector, select either 5.2 SPCL (for fast average) or 5.3 SPCL (for slow average) special functions."



I am not sure what kinds of numbers you will get from a 8903B when performing a unweighted measurement.
It appears for signal to noise ratio that has a range of 50Hz to 100kHz and an accuracy of +/-1dB.

An “A” Weighting Filter is Option 015, 055 for the HP8903B.

Ed
 
Top