DC on both meters on lowest range, 20 V.

AHA, the keystone part of the puzzle. This blows Phil's statement out the water which is rare, there is no TRMS in DC.

Now you are down to the meters being plain old wrong, after all one just be, right ?

First of all realize that 100% LED compatible dimmers, which could be used on multiple lights must be the raw PWM. ON OFF ON OFF. If you smooth it off the a DC voltage, LEDS in parallel might fire at slightly different voltages and screw ya all up. So PWM is fine with LEDs as long as you never exceed the Imax (current rating) As such your meters on DC are actually measuring AC in the wrong mode.

It will help to know how the digital voltmeter works. Now when a voltage is applied to a DVM it is of course divided by the range control and whatever.. After that it goes to a comparator. The other input to the comparator is a sawtooth wave, generated quite accurately. When your input voltage and the voltage from the sawtooth cross, the comparator changes states and puts out an output. The exact time that happens is used to stop or read a counter to know just how much the internal voltage has risen.

It is easy to see that each DVM can have its own timing, or speed. higher speeds read faster and they are more expensive if they are accurate.

Now if you want a DC measurement, which could be useful, then you must feed the meter DC, not AC. As you can see frigs it all up. So, an extremely simple way would be to get like a 5K resistor and like a maybe what guys, 68uF ? Take the resistor from the output, connect it and the probe there, and the cap and the other end of the cap to ground. yyou just made an R/C filter.. This should make the meters read the same.

That is because they are measuring DC on the DC range.

I can't remember when the last time it was I had to make an R/C filter like that but I know I did a couple of times.