More I just found:
VR5 on the KAS is actually 5k, not 2k.
This one is troubling since it’s a time-constant: all the 0.01?F caps that feed the 3.3MΩ resistors into the output buffer S/H are actually 0.033?F - the same as the sample caps. I’m not 100% sure exactly when these paths are for. They basically a low-pass filter (Tc = 0.01e-6 * 3.3e6 = 33m sec for values in docs and 109mS for the actual value on the board) across the sample caps. I’d guess this is to remove any glitching when the analog switches are switching. However, I don’t know how bad the glitching can be and if the time-constant was purposely increased to fix an problem or if it was just simpler for the assembly-line person to just make the whole row of caps the same 0.033?F parts from the bin!
Ideas? Comments? What’s on your KAS board (a rejected Capitol One catch-phrase…)?
David
