DMM giving different readings between mv and ma

This isn't the first driver I've built, so I'm feeling kinda stupid that I have to ask this...

I'm building this driver to work on 1.6A to give an M140 a long life. The problem is thus:
When DMM is hooked up in series to measure current, it reads 1.43A.
When I measure volts across the 1Ohm/5Watt resistor on the dummy load I get 1.77V.
The measurement of the resistor in ohms is 2.1 with the probe in ohms at 1.1, so I think it's safe to say the resistor really is 1Ohm and measured current across the resistor is 1.77A.

What can cause this difference and which one is accurate?

I'd rather not go below my spec goal, but I would HATE to kill the diode.

Thanks for your help in advance!

As the shunt heats up, the resistance increases slightly. This makes the voltage drop on it go up. Your ammeter is much more accurate than your typical 1 ohm shunt.

Cyparagon said:

As the shunt heats up, the resistance increases slightly. This makes the voltage drop on it go up. Your ammeter is much more accurate than your typical 1 ohm shunt.

Click to expand...

I believe you are right. The difference is proportionally about the same when testing on higher and lower currents last night.

If the ammeter is less prone to error, why do we have so many dummy loads that use the "measure V across the resistor" to measure the current? rather than measure the actual amps in the circuit?

Thanks again Cyparagon

dethlore said:

why do we have so many dummy loads that use the "measure V across the resistor" to measure the current? rather than measure the actual amps in the circuit?

Click to expand...

Laziness and ignorance, I suppose. People don't know any better, and it's easier to poke a resistor than to put the ammeter in circuit.
Arguably, your ammeter is doing the exact same thing, but the shunt in a typical ammeter is not only more precise, but much less effected by temperature variations.
Most will tell you it's "good 'nuff," but your case suggests otherwise, doesn't it? :undecided:

Cyparagon said:

Laziness and ignorance, I suppose. People don't know any better, and it's easier to poke a resistor than to put the ammeter in circuit.
Arguably, your ammeter is doing the exact same thing, but the shunt in a typical ammeter is not only more precise, but much less effected by temperature variations.
Most will tell you it's "good 'nuff," but your case suggests otherwise, doesn't it? :undecided:

Click to expand...

This case even may not suffer too much; I've read and seen the graphs where M140s can be pressed fairly easily over 1.8A with proper cooling... but... when I build another one at 2.0A or when I build my first S06J, this will be very important!

I'm going to modify my dummy load to have hookups in-series for my ammeter.

Thanks for your help! I'd give +1 but it says I need to spread it around before I can give you more. haha

Sorry, one more question... regarding the dummy load for the M140.
I know from the many other threads that you test the 445nm diode with 4 diodes instead of 6 because the 1Ohm resistor drops the rest to the 4-4.5V.

I have a constant current/voltage source set at 4.7V and 1.57A. When I measure across the resistor on the dummy load set to 4 diodes (1N5401, 3A), I get 1.3 V. Measuring the current just before the load, I get 0.93A.
Now... if I bypass (remove) the 1ohm resistor, I get the correct current of 1.57A, but the voltage measured across the load is now only 3.5V. If I move the jumper up one, I get 4.5V, but the current now measures at 1.1A.

So ... if measuring with the ammeter and checking voltage as well, should I get rid of the 1 ohm resistor and set the load at 5x 1N5401 (1V drop at 3A), in which case I need to set the constant voltage higher to compensate?

EDIT: I just realized that the resistor being in circuit or not doesn't really matter as long as I'm measuring the voltage drop across all of it. The goal is to simulate the voltage drop of the LD. So I increased voltage and am now measuring 5.2V across the load and am getting 1.57A. I also measured volt drop across the 1 ohm and am getting 1.59V (converts to 1.59A).
Here's the interesting part. If I set the dummy at 3 diodes keeping the 1 ohm, I measure 4.5V across the load and still getting the 1.57A... and the same goes at 2 diodes and 1 diode for that matter. Current is constant and there is enough voltage supplied for the load to take what it needs. The rest (0.7V) I guess just gets burned off as heat, right?

I never bothered with using a test load like Cyparagon said, I just short the driver through the ammeter. If it's a true constant current regulator, it shouldn't hurt it.

Things said:

I never bothered with using a test load like Cyparagon said, I just short the driver through the ammeter. If it's a true constant current regulator, it shouldn't hurt it.

Click to expand...

Yep you're right. I just shorted it with the ammeter and it gave me exactly what I was looking for. I've always liked using a dummy load... mostly because I still feel like a beginner in building my own drivers... and partially I guess was because it gave me something else to build myself!

Well, I'm feeling really good right now. I have a stable driver tested from 1.6 to 3.5 Amps running on 2*18650. Few more tests to run for longevity... but... COOL!

This is really my first real venture in designing my own driver. I've always followed the linear driver instructions before... but wanted to design something on my own with a different chip.

I think some of you have forgotten some of the good reasons for using a dummy load and resistor:

1. A dummy load keeps the meter essentially outside of the circuit, and many are easily attached to a driver for measurement. With an ammeter, unless you're using clips or some fixed setup, you have a great potential to fry your circuit due to accidental disconnection. In the case of a buck/boost CC driver, a disconnect breaks the feedback mechanism in the circuit, causing an overload. In the case of live measurements of an operating laser diode, you risk frying the diode.
2. The voltage meter on many multimeters is more accurate than the ammeter, and has a greater range than the ammeter. This is especially true on cheap meters. The parts needed to make an accurate dummy load cost about a dollar, versus the cost of a decent meter.
3. Using an ammeter requires switching the leads to the current-measuring plugs, potentially leading to the user forgetting to switch them back when voltage is measured, causing the fuse to be blown due to a short. No such catastrophic failure occurs when in voltage measurement mode and connecting to any part of the circuit. The concept of switching leads for different modes was the most difficult lesson I had to drill into newbies learning how to measure voltages/currents in circuit; they can learn all that after they've built some laser circuits that work.
4. The voltmeter doesn't require switching between different plugs for different ranges. A current outside the range of the low current mode (~200-400mA) can lead to a fuse being blown; however, sticking with the high-current range (< 10-20A) requires a larger shunt resistance, potentially making it less ideal, and less accurate.
5. If you're building the chain of diodes to "mimic" the voltage of your load, it's not a major step to put a resistor in there as well. If resistor accuracy is severely affected by heat, buy a better resistor with better tolerances. They're not expensive or hard to obtain. Probably cost less than a blown out fuse.

Now if you're confident in your ammeter measurement skills, by all means go ahead and stick with that form of measurement. However, dummy loads are a very appropriate tool to suggest to most people who don't wish to deal with the problems associated with ammeters and their operation, or simply don't have the equipment necessary for a decent measurement. The only reason I bother with all the hookup required of connecting an ammeter to a driver, is if I want to see live current updates as I'm adjusting a pot.

Bionic-Badger said:

I think some of you have forgotten some of the good reasons for using a dummy load and resistor:

1. A dummy load keeps the meter essentially outside of the circuit, and many are easily attached to a driver for measurement. With an ammeter, unless you're using clips or some fixed setup, you have a great potential to fry your circuit due to accidental disconnection. In the case of a buck/boost CC driver, a disconnect breaks the feedback mechanism in the circuit, causing an overload. In the case of live measurements of an operating laser diode, you risk frying the diode.
2. The voltage meter on many multimeters is more accurate than the ammeter, and has a greater range than the ammeter. This is especially true on cheap meters. The parts needed to make an accurate dummy load cost about a dollar, versus the cost of a decent meter.
3. Using an ammeter requires switching the leads to the current-measuring plugs, potentially leading to the user forgetting to switch them back when voltage is measured, causing the fuse to be blown due to a short. No such catastrophic failure occurs when in voltage measurement mode and connecting to any part of the circuit. The concept of switching leads for different modes was the most difficult lesson I had to drill into newbies learning how to measure voltages/currents in circuit; they can learn all that after they've built some laser circuits that work.
4. The voltmeter doesn't require switching between different plugs for different ranges. A current outside the range of the low current mode (~200-400mA) can lead to a fuse being blown; however, sticking with the high-current range (< 10-20A) requires a larger shunt resistance, potentially making it less ideal, and less accurate.
5. If you're building the chain of diodes to "mimic" the voltage of your load, it's not a major step to put a resistor in there as well. If resistor accuracy is severely affected by heat, buy a better resistor with better tolerances. They're not expensive or hard to obtain. Probably cost less than a blown out fuse.

Now if you're confident in your ammeter measurement skills, by all means go ahead and stick with that form of measurement. However, dummy loads are a very appropriate tool to suggest to most people who don't wish to deal with the problems associated with ammeters and their operation, or simply don't have the equipment necessary for a decent measurement. The only reason I bother with all the hookup required of connecting an ammeter to a driver, is if I want to see live current updates as I'm adjusting a pot.

Click to expand...

BB, thanks for reminding me of the dummy's purpose. For safety sake and not wanting to kill any diodes, I will be using both methods to make sure I'm not missing something. You are right, if you are putting on all those diodes, adding that resistor is pretty easy, and it's hard (for me) to forget about proper usage of the ammeter when dealing with a $50+ diode. I always use alligator clips - part for keeping that good connection and part so I have free hands and can watch or quickly disconnect if I've done something stupid and the parts start glowing or exploding!

Ok guys, I hope you will bear with me here, while I try to get better educated.

I've got a pretty good RadioShack DMM (~$35) 42 range, auto-blah blah. It has the standard 400mA/10A plugs. So, it's not a HF $5 piece of crap.

Is it accurate enough to get an accurate reading of the current used? Do I not use my 10W/1Ohm resistor in the load when measuring current?

I've had eratic readings while measuring voltage across the resistor, and as the X-Boost heats up rapidly, the current reading drops pretty fast. Also, I never EVER see the 1.9A I am trying to pump out of this driver.

I appreciate your help

That meter is probably decent enough to get an accurate reading. Make sure you use that 10A plug. I also don't think your dummy load is to blame at this point.

Is your XBoost properly heatsinked? Also, are you providing power from some high quality cells that are able to consistently produce enough current?

Thanks B-B. It's a reasonably accurate meter, and I am reading voltage across the resistor, rather than current directly, and when I do measure current, it's always in the 10A fused port.

The X-boost is heatsinked, but admitedly, probably not enough at this current. I mean, it drops in less than a minute.

I'm using 3100mAh Panasonic batteries, so it shouldn't be them. I really need to buy a decent power supply.

Related to measuring across the dummy load:
Because the voltage drop of a resistor varies depending on how much current you’re driving, setting the load to 4* 1N5xxx diodes may not exactly be accurate if you are pushing say 3 Amps through the circuit… unless I’m understanding this wrong.

Example: driving 5.3Volts at 3Amps.
If I measure voltage across the load with 4* N54xx diodes set (drops 1V each at 3A), I am getting 5.3V. If I measure voltage across the resistor, I am getting 1.3V (1.3A of current), but if I set the load down to 3* diodes, I measure 5.3V across the load but am now getting 2.3V across the resistor. Setting the load to use 2* diodes gives me 5.3V across the load and 3V across the resistor (yes I realize this requires the resistor to be rated at least 18W). Measuring current with the ammeter gives me the same readings as Voltage across the 1 ohm…

So would setting the dummy load to 2* diodes instead of 4 make an accurate dummy load based on these readings?

^Looks like it. Maybe you should invest in a beefy rheostat and use that as your test load.

Bionic-Badger said:

The voltage meter on many multimeters is more accurate than the ammeter, and has a greater range than the ammeter.

Click to expand...

That does you absolutely no good when you use a 10% power resistor which drifts rather drastically with heat. 10% power resistor + 0.0000001% voltmeter will always be vastly inferior to a 1% ammeter. If you can get a <1% 0.1 ohm resistor for cheap (can you? Maybe, I dunno), my qualms pretty much go away. But like I said, that's what an ammeter is anyway. No one uses those, though. It's always a junk 1 ohm.

I'm honestly not sure how I feel about the other points. Part of me wants to say "if someone demonstrates carelessness with a multimeter or doesn't feel it necessary to learn how to use their equipment properly, why are they qualified to use a dangerous device such as a laser?" and part of me wants to appreciate protection against user-error. :yabbem:

Cyparagon said:

^Looks like it. Maybe you should invest in a beefy rheostat and use that as your test load.

Click to expand...

Consider it a Father's Day present from me to me!

Cyparagon said:

That does you absolutely no good when you use a 10% power resistor which drifts rather drastically with heat. 10% power resistor + 0.0000001% voltmeter will always be vastly inferior to a 1% ammeter. If you can get a <1% 0.1 ohm resistor for cheap (can you? Maybe, I dunno), my qualms pretty much go away. But like I said, that's what an ammeter is anyway. No one uses those, though. It's always a junk 1 ohm.

Click to expand...

Which is why I like to use all methods (ammeter in series, volts across load, volts across 1ohm). I like to be as thorough as I can - measure 3 times cut once - idea.

Cyparagon said:

I'm honestly not sure how I feel about the other points. Part of me wants to say "if someone demonstrates carelessness with a multimeter or doesn't feel it necessary to learn how to use their equipment properly, why are they qualified to use a dangerous device such as a laser?" and part of me wants to appreciate protection against user-error. :yabbem:

Click to expand...

And this is precisely why I ask questions when in doubt. I think I understand a concept, but then I experience something I've never seen before and I want to make sure I DO still understand right and be a better DIYer.

Took my quassimoto laser in to University to get power metered. My physics professor was really excited to play with this laser, but at the same time was VERY surprised that anyone can buy these as they can easily be used as a weapon.

So I'm very grateful for this forum and the LOADS and LOADS of support and information I get from it. When I started this hobby to build my first red laser, I knew next to nothing about lasers or electronic circuitry.