Dick, I was just looking at your picture, and I noticed that the indictation is 10 x better than the spec! That would be a hidden treasure!
Then I noticed you have the analyzer set to measure 2nd harmonic only.
Does that mean that the 2nd is (much) lower than the others? That could explain the difference between what you show and the spec.
Do you have a spectrum of this?
Jan
The 725D is in analysis mode. That means it looks at the harmonics and ignores the noise. These issues always make comparisons on measurements harder.
I think Richard did a few tweaks to his. I need to document and post the tweaks to my KH4400.
The KH44XX are quite good and some judicious opamp swapping and other tweaks can get as much as 10 dB better. Keep in mind that very secondhand examples still fetch $300+.
The KH4400 uses a JFET for the AGC, the KH4403 uses an analog multiplier. The KH4403 also has a balanced out.
I think Richard did a few tweaks to his. I need to document and post the tweaks to my KH4400.
The KH44XX are quite good and some judicious opamp swapping and other tweaks can get as much as 10 dB better. Keep in mind that very secondhand examples still fetch $300+.
The KH4400 uses a JFET for the AGC, the KH4403 uses an analog multiplier. The KH4403 also has a balanced out.
yes, of course. certain R replaced with trim pots and opamp changes are the primary changes. [Dont ask for the specifics/details as i am in Asia and the K-H is sitting on the shelf in California.]
Just basic mods are used. Also some extra C on the PS rails. The K-H allows selection of freq vs a single freq board.
The Panasonic analyzer has no filters in play.... the 725 does have. Both are showing FFT sum for thd value using the first 5 harmonics. Both are shown in THD measurement mode (not 2H or 2fo).
THx-RNMarsh
Last edited:
Now if you only had a distortion-less amp for the active setting. Idea, it might be trivial to program a miniDSP to fix the gain error of only the residual on the passive filter, it really is only an inconvenience the error vs harmonic is known almost exactly ahead of time.
The caps are C0G 1%, the resistors Mini-MELF 0.1%.
My mistake, the filter is not twin-T - it is the Hall network.
http://www.mouser.com/Semiconductor...ransistors/MOSFET/_/N-ax1sf?P=1z0y3dtZ1yiaumt
Google can make anything possible. I know absolutely nothing about these parts, however I intend to find out more. http://www.ti.com/lit/ds/symlink/csd25404q3.pdf
The package is human hostile.
Seems, that these are not depletion mosfets. Ids=0 if Vgs=0. Mouser's mistake.
Unfortunately...