|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
73.225.74.165
In Reply to: RE: Tube matchin posted by neolith on November 21, 2015 at 06:55:42
Noise is always difficult to measure accurately, because there are so many possible sources which may confound the measurement.
First, the test rig must not contribute significantly if you want the tube noise. That means a quite power supply, and checking the noise from bias and plate load resistances to be sure they are all small relative to the expected noise.
Second, the signal level is small; you need a really quiet amplifier with a gain around 1000 (60dB). This raises the tube noise to a level that can be seen on a measuring device such as an oscilloscope.
Third, you really need a spectrum analyzer, since tube noise has a white noise component plus a "1/f noise" component. The spectrum analyzer is also good for detecting small levels of hum and other interferences, which would otherwise distort or dominate the measurement.
Here's an interesting thread on triode noises:
Follow Ups:
Paul, just to add to what you've said, the external 60 dB amplifier is really the key to this. It's also relatively simple to construct. If one selects wide bandwidth opamps with very low noise specifications (nV/hz), two can be connected in series, 30dB gain each, in order to produce a very high quality amplifier. In real world terms, it's possible to construct an amplifier like this with a simple regulated supply (or just a bipolar "cleanup" circuit) that will enable viewing a noise floor below 100 nanovolts. Input this to a PC soundcard with FFT (spectrum analysis) software, and your system will be orders of magnitude more sensitive than what's needed to analyze tubes.
--------------------------
Buy Chinese. Bury freedom.
Agreed. I would use batteries for the preamp, to ensure against power supply noises.
"I would use batteries"
Yes, especially if the preamp will only see occasional use. Also, if one desires to make measurements down to DC, a split voltage source is required. Two 9V batteries are much less expensive than two regulated supplies. BTW, if two opamps with +30dB gain each are chained, only the first needs to be low noise. The second (and whatever follows) will have virtually no impact on the noise floor measurement.
--------------------------
Buy Chinese. Bury freedom.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: