Here is link to Keysight 34401A schematics

And there's extensive analysis of various portions of the 34401A in AoE III, plus more coming out in the x-Chapters. These are part of our "Designs by the Masters" series, pp 342-347, 918-922 and 894-896. In x-Chapters, pp 353-355.

Nice, should be reading it more often. I have edition 1 and 3, but have not yet have had time to read edition 3

Cheers

Klaus

Aha, you can start by reading those pages, they're were fun to explore and write up, and should be fun to read. The HP Agilent engineers were really on a roll. Looking at specs, they've continued the process with their newer DVM versions. They maintain or improve the accuracy for 1 PLC measurements, while cutting the auto-ranging time to 5ms, and found ways to make 20us 4.5-digit measurements with the same circuitry (that's for the 34465A; I think it was 800us for the 34401A). But they no longer publish schematics, so we can't figure out how they did it. But of course, competitors who take the time to reverse engineer, may be able to figure it out.

My gripe about the 34401A is that the VF display kicks a huge amount of noise into the input terminals. They partly hid that in software by having AC measurements below some level jump to zero.

Yeah, if you publish your schematics in the service manual, the Chinese will clone your device in about 30 days, give or take a few hundred days depending on the economics of the product in question and whether or not you gave them a sneak peek before it went into production.

On the other hand, if you *don't* include schematics in the service manual, the Chinese will clone your product in about 30 days, give or take a few hundred days depending on the economics of the product in question and whether or not you gave them a sneak peek before it went into production.

Big two thumbs up for the "How they did it" sections in AoE3. It'd be good to see more of them in future editions or X-chapters.

-- john, KE5FX

We have a 34410A, a close relative of the 34401. I've never checked if it suffers from the VFD noise issue but one thing that caught me out is that the AC volts function has some sort of negative offset, which the software pegs at zero, until it has warmed up. This is what I got when I logged the ACV measurement, from a cold startup, with the input terminals shorted:

The first non-zero reading is after about 15 minutes. We have two of these DMMs, I should try the other one sometime.

I was testing a low-noise PSU and wrongly thought it was below the measurement limits of the instrument, until I did the same test later in the day and found several millivolts of noise.

The data sheet says the readings are only guaranteed to be within spec after a 90 minute warmup. Lesson learned: Always read the small print!