Hacking RMS into cheap multimeters? (feat. Cen-Tech P98764 / Mastech MS8229)

[slinkygn]

I was curious if anyone had tried "retrofitting" a non-RMS multimeter with an RMS IC, for example where the MM's chipset gives info for doing so in its datasheet.  I can't be the first to think of this, I'm sure, so I'm looking for any advice or warnings or such from those who have tried to navigate that.

In my specific case, I have a Cen-Tech P98764 (same as the Mastech MS8229), which is juuust capable enough to be kinda interesting and cheap enough to not fear tinkering with it, and while reading up on the RS232 hack for the device (https://www.eevblog.com/forum/projects/mastech-ms8229-rs232-hack/) I noticed that there was a link to the chipset's datasheet.  Light evening entertainment reading, right?  Figured I'd leaf through.  Couldn't help but notice that they included diagrams not just for average and peak AC rectification, but RMS as well (Step one: drop in an AD737 and a voltage regulator; step three: profit!)

Looks like AD737s are going for 10-12 bucks, so I figured it might be a fun if perhaps fruitless some-future-weekend project to try and see if it can be done, but honestly I have about a billion of those already waiting in line so it might be a while... figured in the meanwhile I'd see if anyone else had tried it, and if it was a viable thing or I'm kidding myself and it's a fool's errand.

[Kleinstein]

In principle one can do that extra added RMS chip, at least with many meters. Some more modern really cheap ones may not allow it and have the rectifier part inside the chip with no way to change it.
Economically it does not make much sense, as there are DMM with RMS in the $20 range.
If one is lucky, one could just replace the old rectifier part that with older chips sets can be separate from the ADC.
With more modern chips it depends on the circuit if it is possible to upgrade and if a positive or negative ouput is wanted.
One usually would have to adjust a gain factor, as the simple rectifier need a correction factor to read right with a sine wave.

[Terry Bites]

The LTC1968 is better and cheaper but not as cheap as crap meter with alleged RMS capabilities.

[strawberry]

RMS or true RMS

[Kleinstein]

The LTC1968 is better and cheaper but not as cheap as crap meter with alleged RMS capabilities.

There AD737 may not be a good solution for a handheld meter. It is made to also work with a rather high supply and this way is more something for a mains powered bench DMM, not so much for a handhelp with a 200 mV signal range.
Another weak point of the AD737 is that it's settling speed depends on the amplitude and with low amplitude it gets very sluggish.

For a battery powered meter I would prefer the AD8436. This one at least removes the dependence of the response time on the signal amplitude, but still suffers from a BW that depends on the amplitude. So one has to chose between a limited BW or nonlinear effects from variable BW.
The LTC1968/1967 don't have the amplitude dependent BW, and as it looks like less temperature effect, but the linearity at small values seems to be a bit worse. So I won't call it a clear winner on either side.

Some of the cheap meters use digital RMS as a 3rd method to do the RMS conversion. This should give good temperature stability, good linearity and a faster (FIR vs IIR type) response than the AD8436 or LTC196x (the AD737 is even slower). The down-side is often a limited BW (e.g. 1-10 kHz rang), but at least amplitude independent. The actual implementation may still vary, but much is part of the chip sets and thus no so much to go very wrong.

[slinkygn]

The LTC1968 is better and cheaper but not as cheap as crap meter with alleged RMS capabilities.

There AD737 may not be a good solution for a handheld meter. It is made to also work with a rather high supply and this way is more something for a mains powered bench DMM, not so much for a handhelp with a 200 mV signal range.

There are two reasons I lean pretty heavily toward the AD737:

1) The Fluke 177 and 179 use the AD737 for their true-RMS, which makes me feel comfortable about its applicability to task.
2) The datasheet for the MS8229's DMM IC (the Fortune Semi FS9721) uses the AD737 in the reference schematic for its true-RMS setup, simplifying the task of hacking one into this meter tremendously.  Even given that this is basically experimental and any results I get would not be something I'd trust to the degree that I'd trust a "real" true-RMS meter, using the reference schematic will remove a *lot* of variables and give the project a much higher chance of some success, I feel.

[Kleinstein]

For a bodge the AD737 has the defininately easier to solder case options.