Thought I was going insane - cheap DMM fail!

[paulca]

I was about to experiment with an op amp for the first time starting with comparator mode.

So I powered the chip and created a voltage divider for the reference with 2 1M resistors.  I was using my Extech to measure current to the breadboard so I pulled a cheapo off the shelf and checked the voltage divider.

According to it, 1.63V + 1.63V = 4.97V

Now I'm green enough to question myself at this point.  Did reality shift or did I get confused about kirchhoff voltage law.  Do I have another invisible 1M resistor in series? But I measured the voltage across both resistors in series it read 4.97V.  When I read either individually it read 1.63V.

So I checked with my Extech.  Pheww! Reality restored.  2.38V + 2.38V = 4.97V (still some error but at least it's close enough to be sane!).

Anyway I think this error factor tells me that 1M resistors might be a little too high for a voltage divider and for the meters to read correctly.  Maybe 10K or 4k7 would be better.

Schneider Electrics DMM IMT23002 = Fail

EDIT: Does this come down to the cheap meter having a 1M input impedence, so in fact I was testing 3 1M resistors in the divider the 3rd in the meter.  The smaller error from the Extech... maybe it has 4M input impedence?

[kalel]

I am not very knowledgeable on the subject at all, but I would assume that the multimeters input impedance has something to do with it.

You could always try the meter with a lower resistance divider, and see if it works fine.

[paulca]

I expect it's DMM impedance.   

The LM741 works as a comparator fine, but the next problem to solve this evening is why I show 1.23V in the output when the output should be LOW (which is 0V).  (on the good meter) and only 4V on HIGH.  (supply is 5V)

[woody]

Learning the art of Electronics, page 50 

[paulca]

A quick google suggests the LM741 will not go rail to rail.  It will go close to rail to rail, but ~1V off high or low seems normal.  S

Still it will work for my purposes.

[IanB]

Schneider Electric DMM IMT23002 = Fail

I have searched in vain for the specifications of this meter, but Google turns up nothing. That for me is a fail. Being unable to find a datasheet or specification sheet before purchase is just hopeless.

On the other hand, the measuring issue might not be a meter fail. If the meter is intended for electrical use it may have a Low-Z input to guard against ghost voltages. That would be appropriate for measuring power circuits but inappropriate for electronics. You have to choose the right meter for the application.

[wraper]

Cheapest multimeters usually have 1M input resistance, Those which are a bit more expensive have 10M input resistance in vast majority of cases. High input resistance mode (usually  >10 GOhm) is common only on expensive bench multimeters.

[wraper]

Schneider Electric DMM IMT23002 = Fail

I have searched in vain for the specifications of this meter, but Google turns up nothing. That for me is a fail. Being unable to find a datasheet or specification sheet before purchase is just hopeless.

https://media1.bauhaus.se/media/pdf/2148888B.pdf

This meter uses two fuses:
• Fuse for INPUT Inputs Protection: 250 mA / 600 V FAST fuse, Ø6.35 x 32 mm
• Fuse for “ 10 A ” Inputs Protection: 10 A / 600 V FAST fuse, Ø6.35 x 32 mm

At least it should be safe

[IanMacdonald]

Can also be if the opamp is oscillating. Any kind of pulses or mixed DC/AC are what separates the decent meters from the junk.

I suspect a lot of those 'over unity magnet motor' claims on YT arise through using cheap meters to measure pulsating currents. The experimenter is probably not actually trying to scam anyone, just doesn't understand the limitations of the test gear. When it seems like the output power is more than the input, they think they've found the holy grail. 

[wraper]

Can also be if the opamp is oscillating.

No, it 1M input resistance (in multimeter spec), therefore he gets 3:1 voltage divider once multimeter is connected.  And calculation confirms it perfectly.

[paulca]

Well the evening's tinkering was a success.  I replaced two complex transistor comparators with op amps and the used-to-be transistor/resistor only 555 circuit still works. Now minus 14 transistors.

Also taught me how dodgy breadboards can be.  It "works"...  if I poke things until the LED flashes properly.  Think I have lots of flakey connections around the place.

Still it's learning.  Fulfils it's purpose.

EDIT:  I'm going to give the breadboard(s) a bit of grace here.  When I tested the comparators in isolations they were always spot on.  The latch was a little flaky.  I built it originally using this guide:
http://www.instructables.com/id/Build-Your-Own-555-Timer/

I note he takes a short cut on the latch and I confess I didn't look into to it much.  He under cut the 555 design by quite a few components.  Maybe it's that which causes the circuit to be a bit flaky.

(also this evening I was playing theremin with an LED and an op amp balancing at threshold, all I needed to do was wave my hand within 2 inches of the breadboard and the LED lit, which continues to raise the ponder-ment about building a simple theremin.

[Brumby]

Can also be if the opamp is oscillating.

No, it 1M input resistance (in multimeter spec), therefore he gets 3:1 voltage divider once multimeter is connected.  And calculation confirms it perfectly.

 

I would call it 2:1 - but that's nomenclature.  The maths tells the story.

[Brumby]

Does this come down to the cheap meter having a 1M input impedence, so in fact I was testing 3 1M resistors in the divider the 3rd in the meter.  The smaller error from the Extech... maybe it has 4M input impedence?

The Extech will likely have 10M input impedance - but you are right.

If your cheap meter does have a 1M input impedance, then the math works out.  Just put 1M into all the resistors here and work it out.  R1 and R2 are your divider; RL is your meter....



Kirchhoff is safe.