Setanta
Member For 4 Years
Member For 3 Years
Member For 2 Years
Member For 1 Year
Member For 5 Years
Why?
The answer is very simple - digital multimeters (DMMs) are not designed to accurately measure low resistances, and at the levels used in coils for vaping, the readings you get from the resistance scale on a multimeter are grossly inaccurate to the extent they can be dangerous.
How Inaccurate?
NOTE: These figures assume the meter is working perfectly and being used exactly as specified by the manufacturer - the ‘best case’ scenario.
The minimum claimed error of some DMMs are as follows:
Cheap Generic DMM - ±0.55Ω
‘Budget’ priced DMM - ±0.25Ω
Industry Standard DMM - ±0.15Ω
Real World Consequences
Say you are a sub-ohm vaper and you build a coil which measures 0.3 ohms on your budget priced DMM. It’s low but nothing particularly radical. Let’s assume your setup gives a maximum of 4.0 volts across the coil after all losses are taken into consideration. Using Ohm’s law (V=IR) you figure out that this coil should draw around 13.3 amps from your battery. You’re using a 20 amp rated battery from a good manufacturer so everything is hunky-dory.
However, even a good quality DMM under perfect conditions can be out by ±0.25Ω, so your resistance could be anywhere between 0.55Ω and 0.05Ω. On a mechanical mod this leads to a possible current draw anywhere between 7.3A and 80A! Also note that lithium-Ion battery manufacturers consider 0.05Ω to be a dead short!
— If you want to see more information about multimeters read on, otherwise vape on —
On DMMs
The resistance function of a DMM is designed for use in situations where (typically) you are not measuring resistances much below 100Ω and don’t care too much about accuracy (which is fine for most uses in electrical engineering and electronics). The method it uses (known as the two-wire method) is inherently susceptible to all sorts of errors and is simply not suitable for measuring low resistances like those used in vaping coils. This is not meant as a a criticism of DMMs, just of their use in the wrong situation.
The Examples Above
I took the figures from the manufacturer’s data-sheets for three different meters and interpreted them as they would apply when used to measure coil resistance. The meters in question were:
ALLSUN DT830 Series - Available for <$10 - A generic DMM marketed under various names.
AMPROBE AM-240 - Available for around $50 - A decent quality budget priced DMM.
FLUKE 179 - Available for around $250 - The industry standard DMM from a top manufacturer.
Claimed Accuracy
First note that the claimed accuracy assumes that the meter is accurately calibrated and used properly. Also note that it does not include the many external sources of error such as lead resistance, contact resistance etc.
Secondly, in most cases, name brand DMMs normally operate well within their accuracy specifications - generic DMMs rather less so. When I checked out the DT-830 the specifications given by the manufacturer seemed reasonable for such a cheap device. However, many resellers claimed accuracies which were actually better than the $250 FLUKE!
How to read the specifications
There are two figures you need to take into consideration when calculating the accuracy of a DMM at a specific level, the resolution and the accuracy.
The resolution is the smallest amount the DMM is capable of displaying in a particular range - for the lowest resistance scale on a DMM this is almost invariably 0.1Ω. This immediately introduces an error of 0.05Ω into the equation due to rounding, a value which could prove dangerous for super-sub-ohm vapers in itself.
The accuracy is given in a rather cryptic fashion and the format varies somewhat depending on the source but it should always contain a percentage and a number. The AM-240 documentation lists the accuracy for the 400Ω range as ±(1.0%+2). What this means is it is accurate to within 1.0% of the reading shown ±2 least significant digits. The least significant digit is equivalent to the resolution so, in this case, it is 0.2Ω. If you add the 0.05Ω from above this gives you the minimum possible error of 0.25Ω. At sub-ohm levels, which is where the real danger lies, the percentage error is small enough that it can be ignored.
If you own a DMM then this information was probably provided with the device, If not, you can probably find it on the internet if you wish to check the claimed accuracy of your DMM.
Exceptions
Some, though very few, DMMs have a resolution of 0.01 ohms for low resistances and these are more accurate than the DMMs discussed above. The only one I could find a data-sheet for turned out to have a claimed accuracy around 0.08Ω. They still use the ‘two-wire’ method to measure resistance and hence I still wouldn’t recommend them for measuring small resistances.
Then there are LCR meters. These meters are designed to measure the inductance (L), the capacitance (C) and, critically for our purposes, the resistance (R) of a component. They are relatively specialised devices and hence rather more expensive. They often have a resolution of 0.001 ohms and an accuracy in the region of ±0.005 ohms when properly calibrated and used. However, these can be rather complicated to use properly and I would not recommend buying one for measuring coil resistance.
Other Options
If you’re using a mechanical mod, then the only other option for measuring coil resistance is one of the many purpose built coil resistance testers on the market. I have no first hand experience of these, nor have I been able to find a proper data-sheet or schematic for one. Without more information it is impossible to tell how accurate these devices are or even claim to be. What follows are pure speculation on my part. If anyone has better information on these devices I’d love to hear it.
Several resellers claim accuracy of between 0.2% and 0.4%. Given that professional level devices costing several hundreds of dollars cannot match this accuracy I find this claim to be pretty dubious.
It is likely that these devices use the four-wire (a.k.a. Kelvin) method to measure resistance. This method can get very accurate results but it depends on accurate calibration and stability. Without getting into the workings of the method, it uses a current source and a voltmeter to measure resistance.
If we just consider the percentage error alone, something like the AM-240 can measure voltage with ±0.5% error. This is already higher than what some resellers claim. Then we have to consider the calibration and stability of the current source. If we assume that something like the LT3092 is used for this purpose then the stability introduces another error of ±1%. If we assume the device actually receives calibration we can assume an error there in the region of ±1%. This gives a total of ±2.5%.
The resolution of these devices tends to be 0.01Ω. If we assume that a good quality display is used then there is also likely to be an error of 2 Digits, so the estimated accuracy turns out to be ±(2.5%+2) . This is an order of magnitude higher than the value claimed above.
That said, for a coil resistance of 0.3Ω, the error would work out to be 0.025Ω. which is an order of magnitude better than a DMM!. This assumes that the manufacturer actually cares about making a reasonably accurate device rather than making the cheapest one possible.
Even if we assume no calibration (±10%), poor stability(±5%) and a crappy display (±(2.5%+5)), the error for a 0.3Ω coil would be 0.1Ω - and that is still better than even the Fluke 179!
That said, without proper data on these devices I would still strongly recommend keeping your builds above 0.2Ω to avoid problems.
Setanta
The answer is very simple - digital multimeters (DMMs) are not designed to accurately measure low resistances, and at the levels used in coils for vaping, the readings you get from the resistance scale on a multimeter are grossly inaccurate to the extent they can be dangerous.
How Inaccurate?
NOTE: These figures assume the meter is working perfectly and being used exactly as specified by the manufacturer - the ‘best case’ scenario.
The minimum claimed error of some DMMs are as follows:
Cheap Generic DMM - ±0.55Ω
‘Budget’ priced DMM - ±0.25Ω
Industry Standard DMM - ±0.15Ω
Real World Consequences
Say you are a sub-ohm vaper and you build a coil which measures 0.3 ohms on your budget priced DMM. It’s low but nothing particularly radical. Let’s assume your setup gives a maximum of 4.0 volts across the coil after all losses are taken into consideration. Using Ohm’s law (V=IR) you figure out that this coil should draw around 13.3 amps from your battery. You’re using a 20 amp rated battery from a good manufacturer so everything is hunky-dory.
However, even a good quality DMM under perfect conditions can be out by ±0.25Ω, so your resistance could be anywhere between 0.55Ω and 0.05Ω. On a mechanical mod this leads to a possible current draw anywhere between 7.3A and 80A! Also note that lithium-Ion battery manufacturers consider 0.05Ω to be a dead short!
— If you want to see more information about multimeters read on, otherwise vape on —
On DMMs
The resistance function of a DMM is designed for use in situations where (typically) you are not measuring resistances much below 100Ω and don’t care too much about accuracy (which is fine for most uses in electrical engineering and electronics). The method it uses (known as the two-wire method) is inherently susceptible to all sorts of errors and is simply not suitable for measuring low resistances like those used in vaping coils. This is not meant as a a criticism of DMMs, just of their use in the wrong situation.
The Examples Above
I took the figures from the manufacturer’s data-sheets for three different meters and interpreted them as they would apply when used to measure coil resistance. The meters in question were:
ALLSUN DT830 Series - Available for <$10 - A generic DMM marketed under various names.
AMPROBE AM-240 - Available for around $50 - A decent quality budget priced DMM.
FLUKE 179 - Available for around $250 - The industry standard DMM from a top manufacturer.
Claimed Accuracy
First note that the claimed accuracy assumes that the meter is accurately calibrated and used properly. Also note that it does not include the many external sources of error such as lead resistance, contact resistance etc.
Secondly, in most cases, name brand DMMs normally operate well within their accuracy specifications - generic DMMs rather less so. When I checked out the DT-830 the specifications given by the manufacturer seemed reasonable for such a cheap device. However, many resellers claimed accuracies which were actually better than the $250 FLUKE!
How to read the specifications
There are two figures you need to take into consideration when calculating the accuracy of a DMM at a specific level, the resolution and the accuracy.
The resolution is the smallest amount the DMM is capable of displaying in a particular range - for the lowest resistance scale on a DMM this is almost invariably 0.1Ω. This immediately introduces an error of 0.05Ω into the equation due to rounding, a value which could prove dangerous for super-sub-ohm vapers in itself.
The accuracy is given in a rather cryptic fashion and the format varies somewhat depending on the source but it should always contain a percentage and a number. The AM-240 documentation lists the accuracy for the 400Ω range as ±(1.0%+2). What this means is it is accurate to within 1.0% of the reading shown ±2 least significant digits. The least significant digit is equivalent to the resolution so, in this case, it is 0.2Ω. If you add the 0.05Ω from above this gives you the minimum possible error of 0.25Ω. At sub-ohm levels, which is where the real danger lies, the percentage error is small enough that it can be ignored.
If you own a DMM then this information was probably provided with the device, If not, you can probably find it on the internet if you wish to check the claimed accuracy of your DMM.
Exceptions
Some, though very few, DMMs have a resolution of 0.01 ohms for low resistances and these are more accurate than the DMMs discussed above. The only one I could find a data-sheet for turned out to have a claimed accuracy around 0.08Ω. They still use the ‘two-wire’ method to measure resistance and hence I still wouldn’t recommend them for measuring small resistances.
Then there are LCR meters. These meters are designed to measure the inductance (L), the capacitance (C) and, critically for our purposes, the resistance (R) of a component. They are relatively specialised devices and hence rather more expensive. They often have a resolution of 0.001 ohms and an accuracy in the region of ±0.005 ohms when properly calibrated and used. However, these can be rather complicated to use properly and I would not recommend buying one for measuring coil resistance.
Other Options
If you’re using a mechanical mod, then the only other option for measuring coil resistance is one of the many purpose built coil resistance testers on the market. I have no first hand experience of these, nor have I been able to find a proper data-sheet or schematic for one. Without more information it is impossible to tell how accurate these devices are or even claim to be. What follows are pure speculation on my part. If anyone has better information on these devices I’d love to hear it.
Several resellers claim accuracy of between 0.2% and 0.4%. Given that professional level devices costing several hundreds of dollars cannot match this accuracy I find this claim to be pretty dubious.
It is likely that these devices use the four-wire (a.k.a. Kelvin) method to measure resistance. This method can get very accurate results but it depends on accurate calibration and stability. Without getting into the workings of the method, it uses a current source and a voltmeter to measure resistance.
If we just consider the percentage error alone, something like the AM-240 can measure voltage with ±0.5% error. This is already higher than what some resellers claim. Then we have to consider the calibration and stability of the current source. If we assume that something like the LT3092 is used for this purpose then the stability introduces another error of ±1%. If we assume the device actually receives calibration we can assume an error there in the region of ±1%. This gives a total of ±2.5%.
The resolution of these devices tends to be 0.01Ω. If we assume that a good quality display is used then there is also likely to be an error of 2 Digits, so the estimated accuracy turns out to be ±(2.5%+2) . This is an order of magnitude higher than the value claimed above.
That said, for a coil resistance of 0.3Ω, the error would work out to be 0.025Ω. which is an order of magnitude better than a DMM!. This assumes that the manufacturer actually cares about making a reasonably accurate device rather than making the cheapest one possible.
Even if we assume no calibration (±10%), poor stability(±5%) and a crappy display (±(2.5%+5)), the error for a 0.3Ω coil would be 0.1Ω - and that is still better than even the Fluke 179!
That said, without proper data on these devices I would still strongly recommend keeping your builds above 0.2Ω to avoid problems.
Setanta