the 10µF was the highest value I had laying around for now. If necessary, I can order a bigger cap later.
I was referring to a missing screenshot, which you’ve since provided.
Since there are now two variations in circulation—either a “larger” capacitor directly at the buffer’s output or smaller capacitors per channel acting as buffers for the EXOR gate—I will test and document both options using FFT measurements.
For simplicity's sake, a large capacitor would be preferable, but the version with locally mounted capacitors also has its advantages.
I'll be able to test the latter this weekend, since I don't have a “large” single capacitor here right now—except for a standard leaded electrolytic capacitor.
There are two ways to detect the presence of the 7.63 kHz interference: either using FFT, in which case you will definitely be able to detect it.
Or, as shown, a simple and practical method is to reduce the vertical gain and observe the trace displayed on the screen.
You can't expect a flat line even with maximum suppression of the 7.63 kHz; the sensitivity at 1 mV/div is simply too high for that (500 µV/div is no longer a reliable indicator in this case).
But then there's “natural noise.”
No matter which version is ultimately chosen, I hope it's clear that we shouldn't remove the buffer circuit, as Owon did.
That was (supposedly) helpful in avoiding the very obvious 39 kHz.
But it’s counterproductive if you also want to suppress the 7.63 kHz by increasing the buffer capacity of the 2.5 V rail.
This is due to the specified maximum load capacitance of 10 µF for the reference chip.
I've now also taken a look inside the second Scope—which I bought from Voelkner this time—and, as expected, it's in the same condition as most of the others here.
It also has the old firmware installed—a good opportunity to document the current state before updating to the latest version, so that after the update, I can see what has changed.
Home
Help
Search
Login
Register
