Author Topic: Multimeter Overclocking  (Read 1508 times)

0 Members and 1 Guest are viewing this topic.

Offline beenai2018Topic starter

  • Contributor
  • Posts: 38
  • Country: us
Multimeter Overclocking
« on: July 12, 2023, 05:04:20 am »
Hello



I have a "slow as paint drying, " extech ex430A especially on the low ohms range. It's terribly bad and rather unpleasant when a reading takes 6 seconds and the last digit still flickers.
The rest of the meter is not bad floor for my hobbyist purposes.
I would watch with envy as the Anengs would zip about and turn green with jealousy when flukes were fast and settled too in various reviews

I got a ex430A as a warranty replacement so decided to hack my old one. These meters have the 7721 chipset which are not great but they work. The Meters all have a 4 megahertz clock(maximum, apparently, lol) from the data sheet.

Since most multimeters are supposedly ratiometric , increasing the clock should just speed up the measurement(I thought). So went through my assorted junk box of crystals and I had 18, 16, 25, 3.57, 11.05 and 50 MHz.
 
My theory was any multiplier of four might be the best bet.Started out with the 16 and everything sped up. voltages, resistances were fine and got faster. Capacitances were multiplied by 4 and and currents were off on the amp range.

Since 16 was okay I tried the 25 and the meter still works on volts and resistances and it is zippity fast however currents have a weird issue where the value would lock on the correct value or  be incorrect at random.
Some crystals cause a multiplier on currents. Mostly looked like a y=Kx but some looked like a y=Kx+b.

Bottom line was the voltages and resistance is always worked fine however currents had an issue where it was a simple multiplier or at higher frequencies it was randomly off.
Just to give it a shot I tried the 11:05 Crystal and that worked perfectly. Voltages currents and resistances work and match with my other unmodified meter. I tried to test the lowest current ranges by supplying a few microamps and it was spot on. The capacitance has a approximately 3X multiplier. Temperatures measurements match. I'm not sure why an oddball Crystal that is not a multiplier of four works best. But it does and the meter is significantly faster.

 I will try a video when I get a chance, but yeah ,same feeling as pressing the turbo button ,lol.


Warning : Modifying testing equipment is dangerous. This can give false readings and a false sense of security. This can harm , maim or kill you.

« Last Edit: July 12, 2023, 05:56:50 pm by beenai2018 »
 
The following users thanked this post: thm_w

Offline LazyJack

  • Frequent Contributor
  • **
  • Posts: 260
  • Country: hu
  • Yeah, cool.
Re: Multimeter Overclocking
« Reply #1 on: July 12, 2023, 06:25:58 am »
How about the power consumption of the meter? I assume that could increase significantly with the clock speed increase.
 

Offline beenai2018Topic starter

  • Contributor
  • Posts: 38
  • Country: us
Re: Multimeter Overclocking
« Reply #2 on: July 12, 2023, 07:01:32 am »
Didn't measure the power consumption.Im running off rechargeable 9V which last for awhile.
Given that ,I would take the faster meter any day at the expense of power . Much better than waiting for the value to settle.
 
The following users thanked this post: wacd

Offline coromonadalix

  • Super Contributor
  • ***
  • Posts: 7001
  • Country: ca
Re: Multimeter Overclocking
« Reply #3 on: July 12, 2023, 10:48:02 am »
toss the mete away, i tried on some uni-t models,  yes youll get inconsistent measurements on some ranges ... tried many xtal values ...   even recalibrate them is no good
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Multimeter Overclocking
« Reply #4 on: July 12, 2023, 12:55:00 pm »
It will probably go horribly wrong if you just increase the clock speed. ADCs are designed around capacitances, etc.

Just get a faster meter and be done with it.
 

Offline GigaJoe

  • Frequent Contributor
  • **
  • Posts: 524
  • Country: ca
Re: Multimeter Overclocking
« Reply #5 on: July 12, 2023, 03:03:00 pm »
had play a lot with an oscillator change, some DMM behave when battery voltage drop.
most negative  - caps measurement need divide.   and beeper in ultrasound range.
my daily overclocked owon bt41 - speed , especially resistance range change,  very very comfortable ...
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 15151
  • Country: de
Re: Multimeter Overclocking
« Reply #6 on: July 12, 2023, 05:21:22 pm »
The resistance ranges are often slow due to autoranging. It takes some steps to get from the highest range to the low resistance range.

The ADC itself may still work kind of. Most of the modern DMM chips sets use some kind of sigma delta ADC and may also work with a somewhat high clock, though with some change in the performance (noise, INL, gain). One problem is that depending on the clock the 50/60 Hz suppression may no longer work and this would often cause quite some extra "noise".  Common integration times are something like 100 ms, 200 ms , 500 ms and one thus has a chance to find a higher suitable clock.
With a change in input bias and possible gain the readings may be something like 10.1-1% off, which may not be obvious, but could still be an issue.

Another point is that the supply current of the ADC would also go up, maybe proportional to the clock. So the battery may not last as long.
 

Offline beenai2018Topic starter

  • Contributor
  • Posts: 38
  • Country: us
Re: Multimeter Overclocking
« Reply #7 on: July 12, 2023, 05:39:59 pm »
It will probably go horribly wrong if you just increase the clock speed. ADCs are designed around capacitances, etc.

Just get a faster meter and be done with it.

This was the  premise of the experiment. "It will go horribly wrong and the meter (and the meter alone) will pop" .Lol

But to a large extent ,it actually works pretty well .
 

Offline beenai2018Topic starter

  • Contributor
  • Posts: 38
  • Country: us
Re: Multimeter Overclocking
« Reply #8 on: July 12, 2023, 05:45:15 pm »
toss the mete away, i tried on some uni-t models,  yes youll get inconsistent measurements on some ranges ... tried many xtal values ...   even recalibrate them is no good

My meter has a 4Mhz standard clock.
I tried a handful of crystals and settled on a 11.05 Mhz which is accurate on V,A,R,Temp . Cap and Frequency which are time based measurements   do need a  factor but its a linear one.

The speed at 25Mhz was really good but current measurements are off . Almost a y=  kx+b . Some crystals like 16Mhz cause a  k=yx relationship

 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 17518
  • Country: 00
Re: Multimeter Overclocking
« Reply #9 on: July 12, 2023, 05:46:28 pm »
Frequency

Oh, yeah... "frequency". Definitely not going to work afterwards.  :-DD

 

Offline beenai2018Topic starter

  • Contributor
  • Posts: 38
  • Country: us
Re: Multimeter Overclocking
« Reply #10 on: July 12, 2023, 05:50:22 pm »
had play a lot with an oscillator change, some DMM behave when battery voltage drop.
most negative  - caps measurement need divide.   and beeper in ultrasound range.
my daily overc
locked owon bt41 - speed , especially resistance range change,  very very comfortable ...
oh no . Forgot to mention this . The beeper is the best part .

The original ex430 has a very annoying beeper  |O (effective). And they doubled it up (2 beepers) for jack detection. Why not just use a diode OR?. Lol

Overclocked(and gets better at higher overclocks) , the beeper sounds like a soft "schnick" . Sounds very nice.

Helps loads when testing continuity.
 

Offline beenai2018Topic starter

  • Contributor
  • Posts: 38
  • Country: us
Re: Multimeter Overclocking
« Reply #11 on: July 12, 2023, 05:51:17 pm »
Frequency

Oh, yeah... "frequency". Definitely not going to work afterwards.  :-DD

Ha ha . Touche
 

Offline beenai2018Topic starter

  • Contributor
  • Posts: 38
  • Country: us
Re: Multimeter Overclocking
« Reply #12 on: July 12, 2023, 06:13:41 pm »
The resistance ranges are often slow due to autoranging. It takes some steps to get from the highest range to the low resistance range.

The ADC itself may still work kind of. Most of the modern DMM chips sets use some kind of sigma delta ADC and may also work with a somewhat high clock, though with some change in the performance (noise, INL, gain). One problem is that depending on the clock the 50/60 Hz suppression may no longer work and this would often cause quite some extra "noise".  Common integration times are something like 100 ms, 200 ms , 500 ms and one thus has a chance to find a higher suitable clock.
With a change in input bias and possible gain the readings may be something like 10.1-1% off, which may not be obvious, but could still be an issue.

Another point is that the supply current of the ADC would also go up, maybe proportional to the clock. So the battery may not last as long.

PLC is a thing I missed (didnt consider it )  when testing . However , both modified and  unmodified units compared well on the the most sensitive ranges and ,(my most) scary range of AC 120 .

Tests passed (My pass criterion is to be within 5-10% of each other for modified and unmodified meters)

DC
3 uA-4000uA

5-300mA

1A-3A (meter can do 10)

Resistances

1 ohm - 1 MOhm

Volts

10mV - 30V

AC

120V

Temperature ,Diodes Mode, Continuity Mode - Pass



Tests Failed

Hz(y=Kx) , Cap (y= Kx)


There are  definitely more tests (corner cases , ruggedness) etc  possible , but these basics are a good starter.






 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf