I wanted to find out whether the meter can be used with lithium batteries instead of standard cells or not.
Problem with lithium cells ist the higher voltage. 4 lithium cells have up to 7.7V when they come out of the package. After a short time the voltage goes back to approximate 6.6V and stays a long time at this value.
Problem or not?
I could not find any hint to this question, so I started to explore the schematics.
First it looked good, all internal voltages came from low drop regulators.
+4V, VDD(3.3V) analog supplies and VDD_P (3.3V) for the 15V booster and the CPU.
So, no problem at all, none of the voltages can go to high.
But wait, what is that?
The voltage reference, ZD1, a ADR3412 reference regulator is connected directly to B+ via R94, a 0 ohm resistor. B+ is the battery voltage. And this is a problem in my opinion. According to the datasheet the input voltage range of the ADR3412 is 2.3V..5.5V, the absolute maximum is 6V.
That means even with alkaline batteries the voltage is above the input range and with new alkaline batteries also above the absolute maximum. With lithium cells it is way off from the allowed input voltage.
Populate R12 instead of R94 will supply the reference with +4V, so no problems in this configuration. But R94 is populated, not only in the schematics. I have checked this in my meter and actually R94 is populated. I will remove R94 and place it in the R12 position.I cannot understand why the reference is connected to B+, that makes no sense for me. Is this really a design flaw or did I overlook something?
I’ve been purging my lab of alkaline batteries, which keep leaking, and replacing them with Energizer lithium cells. The only thing I have that can’t handle lithiums is my 121GW, where its ADR3412 will be fed too high a voltage from fresh lithiums. I figured I might as well move the 0-ohm jumper from R94 to R12 to make the meter lithium-safe, as long as that’s not going to cause any other problems.
Before making the mod, I wanted to check on the forum. I’ve been pondering why the jumper was on R94 and not R12 in the first place. The potential reasons I’ve considered are:
- Perhaps AGND goes too high in some mode such that +4V minus AGND becomes less than 2.3V—the min Vin for the ADR3412 (I can’t quite follow the diagram for T1 to see if this is the case)
- Perhaps the +4V output from the NJU7741F-04 (U13) is too noisy, which might cause the 1.2V reference voltage to fluctuate
There were two other potential solutions I saw, instead of moving that jumper:
- Add a simple resistive voltage divider to feed perhaps 2/3 the battery voltage into the ADR3412
- Add a bodge wire to feed the 3.6V VDD output of the NJM2870F-36 (U1) into the ADR3412
A voltage divider would need to be chosen carefully such as to not impede the operation of the meter when using alkaline or NiMH cells. I don’t necessarily see any advantage to using VDD over +4V, although that regulator has an added “noise bypass” capacitor that the +4V regulator lacks, which
may reduce noise? Could also replace the +4V regulator with a “better” one if noise is a problem, but that’s perhaps getting excessive.
So I am thinking moving the jumper may be the best choice.
Thoughts? Has anyone else done this mod?
Reference links: