I've been studying and experimenting with the readout setup on a 7A26, in order to convert it into this current probe plug-in that I mentioned recently. I'm quite familiar with the 7K readout system, having a number of custom plug-in projects in the works, but I never really appreciated the versatility of the automatic scaling capability. I have the 7A26 set up now with temporary mods of resistors tacked onto the channel 1 RO board, and toggle switches to work the desired display ranges. One of my questions in the other post was about down-ranging instead of up, so that say, you can indicate all the steps at ten times higher sensitivity, as opposed to ten or a hundred times less for an attenuating voltage probe. After refreshing my brain on the readout stuff, I see that the normal readout can shift only up to 10X or 100X, and there is no 1000X per se, and no 0.1X either. But, I also found that the system is versatile enough to handle this, by simply shifting the prefixes as needed. It still figures out what to do, and is still limited to basically three decades of easy (as in using the probe code) ranging 1X, 10X, and 100X, relative to whatever the base numbers are. If you need to go beyond in either direction, you just change the base prefix. I think this is pretty slick.
For this current probe plug-in, the base readout range is 5A-5mA per division, numerically the same as the original 5V-5mV, modified to read "A" instead of "V," with a resistor changed or tacked on the board. For the big high current probe type, the X10 and X100 are already available by switching in the right code resistors - this will be internal, not via the probe code ring, which will be used instead to detect when a P6021 is connected. So, the following readout ranges are planned:
Big probe: 500A-500 mA/div, 50A-50mA/div, and optionally 5A/5mA/div - this lowest range won't work well near the lower settings, unless amplified, since the probe isn't sensitive enough..
P6021: 5A-5mA/div, and optionally 500mA-500uA/div - this is the "tricky" one, needing effectively divide by 10.
I thought at first it may be quite complicated, but it turned out that I just had to add a divide by 1000 mode with resistors and relays or other switch devices telling it to put up the "u" instead of "m" prefix, and forcing it to 100X, and voila, 100 x 1/1000 = 1/10, and the readout figures it all out. I could also get its two lower decades if needed, by using X10 or X1. Altogether, three or four resistors, and some switching, provides all four display ranges, and one resistor makes it say "A," and no surgery (run-cutting) or even part removals are needed - everything is an add-on (thanks to the need in this case, to increase code currents in certain modes, and not needing to reduce any of them).
That's for each channel, of course. Not bad, eh?