...
This transistor clamping circuit on the additional PCB is obviously the circuit, which Dave had explained in his videos #1157 and 1158.
...
Another remark: the use of the 1N4007, as well as a TVS at that point both affect negatively the calibration of the 5M and 50M Ohm ranges by their leakage currents. Only without them, these ranges may work linearly.
Frank, 1158 shows the mezzanine boards that are used on the ones I just purchased. These are not the same parts that are used on the current revision, nor is it the same PCB.
I remember your posts about the leakage. I've never ran any experiments as I didn't think it would be helpful to run on the prototype with the other changes I had made but would not be at all surprised.
I can't comment on how any of these changes effect the electrical robustness of the meter. If I wanted to analyze the most recent revision, I would need a current schematic for starts. Still, like the UT181A, the has a decent front end design, the layout needs some help. That's why I run the tests.
PS Afaik Dave said "we are not going to spin a different pcb revision, we'll fix problems in software". Turned out either my memory is bad, or they changed their mind.
I just paid for two brand new meters and received the floor sweepings that are now apparently 2 PCB revisions old. I wasn't told anything about them being old stock when I made the purchase or I wouldn't have placed the order. I got fucked once. It won't happen again.
I seem to remember Dave posting something to that effect. Searching, here is one about the the switch. The brand new, out of the box meter certainly had some intermittent problem.
https://www.eevblog.com/forum/testgear/new-eevblog-branded-multimeter-coming/msg1405035/#msg1405035I wasn't able to find those posts about the revision and firmware.
As far as how to proceed. We have two old meters that do not represent the current production. Sure I can test them but like the prototype, if they do poorly, the response will be we addressed that in the new revision.
I can't see wasting too much time on the details. For example, Frank mentions the non-linearity. The new meter will behave differently.
If the new design addresses the shim, cycling the switch may not prove useful. Then again, I am up to seeing how it holds up without spit shinning the contacts. It could prove entertaining but keep in mind that the fall back is always going to be, we addressed that in the next revision.
I still plan to keep the one meter as a virgin, if anything just for a reminder of business practices. As for the second meter, feel free to comment on what you would like to see done with it. I'm not at all apposed to just saying fuck all and run the transient tests to failure and cycling the switch just for the entertainment.