Measurement Techniques

Measurement Techniques 

I guess this topic of measurement techniques and consideration would not be much interest to audiophiles but it is good for DIYers or electronics & electrical engineers like me that would want to understand more on measurement techniques versus psycho acoustics. 

My usual behavior is to measure the components before I use them on my DIY equipment. At least I can do some matching and understanding of the characteristics, and make sure it is working properly before wasting my time to troubleshoot it later when things did not go the way I want. This will ensure the equipment I build carry a certain basic quality with it. I’ve built a 300B amplifier that is matched between 2 channels and the values are very, very close to the designed operating points. 

There are some steps before those measurements are made. Like some will bake or run-in their resistors, capacitors and inductors before sending them for measurements as we all know, after run-in, the characteristics will stabilize and the value will be different compared to freshly built ones. Even our equipment like oscilloscopes will need warm-up to achieve stability and stated accuracy, why not the other components? 

Also, the built equipment itself needs some run-in, warm-up and maturity period before proper measurements can be made. We don’t listen to tube amplifiers when it is just switched on, or just built, or new tubes are installed, right? Therefore, measurements should also be done after that.  

Measurement techniques for transformer would be to load it at the intended operating conditions for a period of time, say an hour, before taking the measurements, as temperature will greatly affect the end results. Therefore, those measurements taken will have credibility or accuracy. Or else, it will just be rough estimates of the actual specifications. 

For output transformer, it is even more important to measure it at the actual operating conditions, where some output transformer manufacturers will state in their datasheet on how it is measured to perform to the stated specifications. It is true to some point that some will measure it under ideal condition where the driving source is almost perfect (measurement equipment usually behave that way so that real accuracy can be achieved) compared to the real usage scenario. Equipment used for reference or measurements will usually have very wide bandwidth, much wider than the device under test, and also have high input impedance, low output impedance, low output capacitance and etc to ensure the accuracy of measurement. 

The output waveform driven by signal generator with very low output impedance and low distortion will provide much better results compared to driving it with vacuum tubes or transistors intended for production/use. Even when driven by different vacuum tubes, the results will be different due to the electrical characteristics of different tubes will be different. Even same vacuum tubes but different manufacturers or different tubes from the same batch / manufacturer will behave differently. 

So, it’s good to have a reference point to compare your data with. Therefore, those measurements claimed by so-and-so could be biased. We need to take that with a pinch of salt. It could be used to their advantage. It is critical to understand it and take all those factors into consideration. If a standard reference point can be enforced, then the results will have more credibility, but it is not absolute though.

On output transformer measurement, the 1^st thing we would want to look at is the impedance versus frequency plot. As we all know, the output power and distortion are related to the load of the vacuum tube, imposed by the output transformer. We would want the output transformer impedance to be as flat as possible across audio frequency spectrum, so that the output power is consistent, and the distortion is also consistent. Imagine what happens if the impedance at various frequencies are vastly different. That would cause the output frequency response or power to be delivered to the speakers to be uneven. So, what the end user would hear is unbalanced sound => bad sound. 

Of course, such unevenness could be made to user’s advantage where one likes to tune the sound or match it to the speakers to get the ideal user preference. Or, it is made to match the vacuum tubes to be used with that output transformer to have flat frequency response. Therefore, matching is always very important in audio world. That’s why there are people that like tube amplifiers. 

That being said, still, the data measured in ideal environment will be very useful, as that shows how capable the components under test are, and how much improvement room you still have to make full use out of it. It’s better than nothing! Then, you won’t know whether it is your design or the component that is limiting the performance. Worse! 

We’ve a user that uses our precision series output transformer and here’s what he got when he measures the system output frequency response, in which we feel that the frequency response on the low side could be better but his driving stage is not that optimal due to compromise made to reduce the system size, anyhow, just for your reference.

• 8Hz -2.9dB
• 20Hz -1.2dB
• 50Hz -0.5dB
• 100Hz -0.25dB
• 500Hz -0.02dB
• 1kHz -0dB (reference)
• 5kHz -0.01dB
• 10kHz -0.01dB
• 15kHz -0.03dB
• 20kHz -0.04dB
• 30kHz -0.07dB
• 50kHz -0.1dB

      J&K Audio Design
    19/12/2013

Click here to find out what output transformers we have.
Click here to look at other products we offer and ordering details. 
Click here to find out what other interesting services we offer.
Click here to check out our price list.
Click here to return to latest publication.