Battery Voltages -- Surprise!

[Rocketman]

Thanks for the heads up. Thought I had found a $4 RMS meter

So , the eGo circuit drops about 100mv (resistive losses in the mosfet) and probably adjusts the duty cycle for charge and load.

So under load the approximately 11% error would be somewhat less just using the cheapo meter to measure RMS voltage on the DCV ranges?

Considering the range of +/- 0.2 ohms for most cartos and the accuracy of a cheapo meter on the ohms range, or even the voltage range, we are probably talking less than a 10% error in average watts, as measured with a cheapo DCV meter.
Maybe even better as the load goes up (like with a 2 ohm carto) and the duty cycle of the eGo increases.

Do you think it might even approach just a few percent error?
(and in a predictable error direction?)

 

[JW50]

4.4 watts on an eGo, and 4.56 watts on an 18650, both with a 3 ohm carto.
Sure seems to be a big difference in vapor for such a little difference in watts.

My number of 4.5 to 4.6 was based on an estimate of 3.7 volts, loaded, for the 18650. Perhaps an actual measurement on your 18650 would be higher. But even under most optimistic assumption, not likely to be higher than 5 watts. But, in any case, not sure vapor production has a direct proportionality relationship to watts.

 

[JW50]

Thanks for the heads up. Thought I had found a $4 RMS meter

So , the eGo circuit drops about 100mv (resistive losses in the mosfet) and probably adjusts the duty cycle for charge and load.

So under load the approximately 11% error would be somewhat less just using the cheapo meter to measure RMS voltage on the DCV ranges?

Considering the range of +/- 0.2 ohms for most cartos and the accuracy of a cheapo meter on the ohms range, or even the voltage range, we are probably talking less than a 10% error in average watts, as measured with a cheapo DCV meter.
Maybe even better as the load goes up (like with a 2 ohm carto) and the duty cycle of the eGo increases.

Do you think it might even approach just a few percent error?
(and in a predictable error direction?)

Just thinking a bit, max variation of DC volts from RMS volts would occur with minimum duty rate. Min duty rate is just off charger. In approximation, duty rate was 0.815. That translated to 1.108 multiplying factor to get to RMS. Watts are squared so factor 1.227 to watts. DC read would never be greater than RMS. So I think max variance is in the 10% range on volts and 20% or so on watts. Variance becomes less as one advances in the discharge cycle.

 

[Rocketman]

Just thinking a bit, max variation of DC volts from RMS volts would occur with minimum duty rate. Min duty rate is just off charger.

But doesn't it also adjust the duty cycle for load? No load would probably be the lowest duty cycle.
Heavy load would increase the duty cycle, maybe even close to 100% to try and keep the output regulated.

With a 1.6 ohm carto, what do you think the error (cheapo DCV indication) would be?
What if the duty cycle was only 96%? under load?

 

[Rocketman]

This duty cycle stuff is getting confusing

If the duty cycle goes up with load, then maybe the cheapo meter isn't that bad for making loaded voltage measurements of the PWM eGo after all.

Here's what I get with a 1.5 to 1.6 ohm mega dual coil carto, and a shot with a home made 50ma flashlight attachment.

 

[JW50]

But doesn't it also adjust the duty cycle for load? No load would probably be the lowest duty cycle.
Heavy load would increase the duty cycle, maybe even close to 100% to try and keep the output regulated.

With a 1.6 ohm carto, what do you think the error (cheapo DCV indication) would be?
What if the duty cycle was only 96%? under load?

In personal testing of an eGo variable voltage I did not find much change of duty cycle with changing load. What seems to me to be the case is that the electronics seem to adjust so that voltage output is constant throughout the discharge cycle and at cut off voltage the duty cycle is one. And since the relationship of RMS voltage to DC voltage is DC voltage divided by square root of duty cycle, RMS voltage gets closer to DC voltage as one gets closer to cut off. That is, just off charger basic voltage of batt is ~4.2 but duty cycling reduces that to ~3.4. (0.81 duty cycle). But as batt discharges, basic batt voltage say going to 3.7 volts, duty cycled voltage is still in 3.4 area (~0.92 duty cycle). Then further into discharge, basic batt voltage at ~3.5 and duty cycled voltage about 3.4 as well (0.97 duty cycle). And cut off voltage seems to be in that neighborhood of 3.4 volts at duty cycle of 1.0. That is, I don't think load has much, if anything, to do with the duty cycle "picked" by the electronics.

 

[JW50]

This duty cycle stuff is getting confusing

If the duty cycle goes up with load, then maybe the cheapo meter isn't that bad for making loaded voltage measurements of the PWM eGo after all.

Here's what I get with a 1.5 to 1.6 ohm mega dual coil carto, and a shot with a home made 50ma flashlight attachment.

Not sure I understand the two shots. Is one a loaded voltage (loaded with 1.5 to 1.6 ohm dual) and the other an unloaded shot? Or is the "other" loaded in some way other than the 1.5 to 1.6 dual?

 

[Rocketman]

The cheap meter results for the 3 ohm (25C resistance of 3.088 ohms) and the 1.5 ohm (25C resistance of 1.556 ohms) seem to indicate some sort of power control (on an almost full charge). The difference in 3.29 and 3.25 indicated on the cheapo meter says somethin is going on?
The I squared R result for those two cartos is mucho different

 

[Rocketman]

Not sure I understand the two shots. Is one a loaded voltage (loaded with 1.5 to 1.6 ohm dual) and the other an unloaded shot? Or is the "other" loaded in some way other than the 1.5 to 1.6 dual?

Sorry.

First one is LOADED with a mega dual coil. Second one is lightly loaded with a 50ma LED which is below the PWM control lower current limit.

 

[JW50]

Sorry.

First one is LOADED with a mega dual coil. Second one is lightly loaded with a 50ma LED which is below the PWM control lower current limit.

Not sure I'm on the mark here - but - in thinking situation through on the differing voltages observed at differing loads a bit more this occurs to me. Suppose the internal batt resistance is 0.187 ohms. If a 3 ohm atty is loaded onto batt that had unloaded voltage of 3.4, then loaded voltage across atty would be 3.2 volts and internal resistance drop would be 0.2 volts. Now say I load with 1.5 ohm dual. Drop across atty would be 3.02 and internal resistance drop would be 0.38. So when I look at meter reading voltage of 3.2 volt drop across atty of 3.0 ohms, it relates to output of 3.4 from batt. When I see 3.02 volts across a 1.5 ohm atty, that also relates to 3.4 from batt - and the differing drops across the attys had nothing to do with duty cycle.

 

[Rocketman]

If, per chance, the eGo has some sort of regulation built in, maybe just for the sake of argument, a PWM control circuit, to keep the output somewhat constant as the cell voltage drops from the full charge value to the 'just before cut off value, what parameter would this circuit monitor to perform this control?

The eGo does something, right? Of the various parameters a circuit could monitor such as power (aka Darwin), current, cell voltage, cell resistance, atty/carto resistance, output voltage (RMS ?), which do you think would be the easiest to implement?

If the eGo is trying to keep the output regulated by varying duty cycle, how would changing carto resistance effect this control?

If there was a pre-programmed voltage (simple?) and the control circuit varied duty cycle, would it also be affected by load resistance?

Or would it pick a different voltage to track and vary duty cycle in a different manner?

Just trying to figure out how this thing knows which carto I put on it


I got 3.29 with the 3 ohm carto, 3.25 with the 1.55 ohm carto and 3.42 with the LED.

 

[JW50]

The cheap meter results for the 3 ohm (25C resistance of 3.088 ohms) and the 1.5 ohm (25C resistance of 1.556 ohms) seem to indicate some sort of power control (on an almost full charge). The difference in 3.29 and 3.25 indicated on the cheapo meter says somethin is going on?
The I squared R result for those two cartos is mucho different

Continuing with thought above, suppose the PWM battery output unloaded is 3.5 and there is 0.115 ohms of internal resistance. If a 1.5 ohm load is placed on battery then voltage drop across the 1.5 ohm resistance is 3.25 volts (and 0.25 ohms across IR). Then if 3 ohm load is placed on battery (with same 3.5 PWM unloaded potential), drop across 3 ohm load will be 3.37 volts (and IR drop will be 0.13). So meter will read 3.25 volts across 1.5 ohm load and 3.37 across 3 ohm load - both arising from 3.5 volt, PWM, unloaded potential. In your actual case you say 3.25 across ~1.5 ohms and 3.29 across ~3 ohms. Considering a cheapo meter to start with - not really far from 3.25 and 3.37 volts.

 

[Rocketman]

Must be it.

Maybe I need a better meter

 

[JW50]

If, per chance, the eGo has some sort of regulation built in, maybe just for the sake of argument, a PWM control circuit, to keep the output somewhat constant as the cell voltage drops from the full charge value to the 'just before cut off value, what parameter would this circuit monitor to perform this control?

...

Heading way out of my field with that question. But first thought would be, no parameter, just a little guy of the board pressing a button when voltage goes too high. Just joking of course. Will give more serious thought to question shortly.

 

[WillyB]

My number of 4.5 to 4.6 was based on an estimate of 3.7 volts, loaded, for the 18650. Perhaps an actual measurement on your 18650 would be higher. But even under most optimistic assumption, not likely to be higher than 5 watts. But, in any case, not sure vapor production has a direct proportionality relationship to watts.

I can't exactly follow what you are saying, but here's an AW2600 18650 firing a Joye 510 that came in at 2.1Ω.

 

[JW50]

If, per chance, the eGo has some sort of regulation built in, maybe just for the sake of argument, a PWM control circuit, to keep the output somewhat constant as the cell voltage drops from the full charge value to the 'just before cut off value, what parameter would this circuit monitor to perform this control? ...

Actually this question goes to the details of circuit design, an area I have only a smidget of of knowledge. But, since your cheapo meter can measure average voltage it seems possible that a cheapo circuit is possible that uses average voltage as its parameter. Then some reference voltage is placed in circuit so that if the average voltage exceeds the reference voltage duty cycle decreased - if average voltage less than reference - duty cycle increased.

The eGo does something, right? Of the various parameters a circuit could monitor such as power (aka Darwin), current, cell voltage, cell resistance, atty/carto resistance, output voltage (RMS ?), which do you think would be the easiest to implement?

Again, no expert here. But I do think power regulation much more complex than voltage regulation or current regulation. If power regulation was cheap and easy I think there would be cheapo RMS meters everywhere. If one reads the theory behind how Fluke develops RMS readings - does not seem to me to be that easy. RMS is an abstract. It is not a fundamental something in a circuit like amps or volts or ohms - but a calculated thing.

If the eGo is trying to keep the output regulated by varying duty cycle, how would changing carto resistance effect this control?

Again, not expert. But, don't see why changing carto resistance has to change anything as far as regulation circuit is concerned. My suspicion is that there might be some very minor impact - but minor. If design is to keep batt output consistent with some reference - what may be downstream (the atty/carto) may be of no consequence to the circuit.

Or would it pick a different voltage to track and vary duty cycle in a different manner?

I suspect that the circuit board of the Joye eGo could have been different. But the choice made as the most marketable and "best" combination of things that users would use was the combination selected. Economics I suspect also had its place. If watts picked as the constant, cut-off would have to occur earlier if the watts commonly used by vapers outside of China (my perceptions here) were used. If cut off earlier, more complaints in that area. Lots of trade offs considered I suspect. Darwin picked watts but they did not do it on a commonly used and available battery.

 

[JW50]

I can't exactly follow what you are saying, but here's an AW2600 18650 firing a Joye 510 that came in at 2.1Ω.

OK. On this one it comes in at 7.06 watts. (3.85*3.85/2.1). My experiences and taste - pretty hot. But I read of some who claim to like 15 watts. To each is own. How was vapor production here?

PS - Comment on likely watt max range was related to a 3 ohm atty. Had your 3.85 measurement been on a 3 ohm atty - 4.94 watts.

 

[br5495]

When you hook a voltmeter, 4 dollar DMM, on the 20 volt DC range to an eGo driving an atty:
What does the meter read? I mean the display reading, not multiplied by duty cycle, not squared then square rooted ,
the actual meter reading?

Does it read 4.1 volts? 2 volts? 3.4 volts?
( I know what mine reads)
Just what math do you need to do to the cheap DMM reading to mathematically correct it?
Here's a shot of an ego, a carto, and super cheap DMM.
Could someone come up with an equation to mathematically correct this display reading (ignoring meter calibration inaccuracy) so I could calculate average power being delivered to the carto (3 ohm).
This meter obviously does not indicate RMS voltage. So it must be wrong. But, it is sort of square

BTW, what is the 'Crest Factor' of the eGo output into a resistive load?

My Fluke 73 (automatic) reads about 3.20 volts from my ego batteries with a 3.0Ω carto as a load. It also reads about the same voltage with a 3.2Ω load. But then, it reads about the same thing with a 2.5-2.8Ω load, indicating that the circuit is doing a fair job of regulating the voltage. This voltage will remain about the same until shortly before the battery shuts down for a recharge.

Since this meter is an automatic, there will always be very slight variations in the voltage each time it updates, so an exact reading is not possible. However, I believe an extended reading may be the equivalent of an average. Therefore, I have been using these voltages to calculate the average power using ohms law.

I keep a log of my daily juice consumption, so am familiar with the amount of power required to keep it within my set limits using other battery types. For all practical purposes, my daily juice consumption stays within these limits when the DC voltage from the eGo battery is used to calculate the power.

But then, I'm not all that much of a meter reader and some of my vaping habits may be the result of "seat of the pants" judgements. If so, I came about it honestly. I was a crop duster pilot for 45 years and didn't know any other way to do it. All I can say in defense of that statement is that I am still alive to talk about it, heheh.

 

[Rocketman]

br;
seat of the pants works better than most think

Getting a "feel" for how your e-cig works is probably the biggest shortcoming of those searching for that mysterious "sweet spot". I also get all I need from somewhat normal voltages with straight through 3.7 big battery stuff, but mostly because of charge life.

Bet you wish you could have vaped while flying, no worry about blowing the cherry oput


(the eGo ain't all that smart. It puts out about the same voltage no matter what you hook to it. Unless you go too low or too high in resistance)

 

[Stonemull]

the ego does PWM regardless of the load, any differences will be as pointed out above, due to the cell voltage dropping under load. This will affect the duty.
I have some CRO waveforms of ego somewhere, it starts off at about 85% duty cycle and drops within a 100mS or so to a duty based on the cell voltage under load. Frequency is around 85Hz from memory, this seems to have changed a bit in the later versions.
I don't think you can get a correct Vrms voltage reading with a cheap multimeter using ac V or dc V or any math tricks .. The method Rocketman indicated is a bit pointless .. square anything and take the square root and you end up with the same data you started with (unless it was negative, you will now have a positive result, unless you are using complex numbers in which case you end up with an i in the result).

Long story short .. you cannot accurately compare output power on mods that use PWM unless you know the actual cell voltage at the time of measuring the Vave, to find this you could make a little rectifier cct with a diode and a capacitor and measure the rectified peak voltage on the cap .. add about 0.5v for a cheap diode and you have Vp, measure the dc v before the diode and you have Vave, compare Vp to Vave and you should get a good value for the duty cycle. Using the dc and Vp you can calculate Vrms fairly well.