Battery Voltages -- Surprise!

[Rocketman]

I'll try and convince you
The cell is not magic (only the controller chip).

Think about ANY Li-ion e-cig. As you vape the cell voltage will fall. Refer to the load tests performed by the flashwarts on one of those 'other' forums . The loaded cell voltage falls and at some point we need to have a way of shutting it off to keep from overdischarging it. The falling cell voltage is what gives most unregulated e-cigs their characteristic "SAG".

Even the cell inside the eGo will sag and as it gets past 90% or so of charge capacity it drops like a rock. The Li-ion cell is pretty flat, but still sags, and the discharge curve will have a 'knee' towards the end.
Even the eGo magic chip can't keep that from happening.

IF, the eGo increases Duty Cycle as the cell falls off in an attempt to keep the vape somewhat consistent the Duty Cycle has to be increased faster than the cell falls off or the vape would fall off. Eventually it does anyway.

Try this with loaded voltage and duty cycle and see what a correction like this does:
4 volts peak, 85% duty cycle,
3.9 volts peak, 88%,
3.8 volts peak, 90%,
3.7 volts peak, 93%,
3.6 volt peak, 96%,
3.5 volts peak, 99%.

That would be flatter than an unregulated e-cig, I think.

 

[jimho]

No insults intended on my part. Looking for what is accurate. 3.6 or 3.7 - not a major deal. Most of volts read or derived headed toward "power". R is usually in that "power" and R will vary easily 10%. One can only hope the errors of R cancel out the errors in volts.

Why do you need to know the output with any accuracy? (straight question)
Correct- R will vary by 10% as you heat it up.... nobody ever discusses thermodynamics of an atty-or the efficiency of heat transfer from the wire/wick to the juice- they come into play too.

If you want accurate, get an RMS meter- nobody is saying that it provides any value other than academic...The main point (and we all agree) is that a cheapo DMM doesn't report RMS voltage properly and can't easily be made to do so- so becareful when COMPARING YOUR OUTPUT TO OTHER DEVICES.

I'm not insulted, I just don't understand what the struggle is about. Why does anyone care what the RMS voltage is other than to accurately calculate the power and/or make some comparison between different manufacturer's items - and in that case- I'd say that accuracy does indeed matter. Also, your calculation makes asumptions about the duty cycle which require either a trace or an RMS meter - unless someone told you what the duty cycle was, you couldn't make those assumptions...

So what is it you are after here and why? What's the point of the discussion?

Once you decide on your RMS approximation, you're going to crank it through to power- and then what are you going to do with that?

 

[Rocketman]

Not that this is what the eGo is using as a correction curve, but notice how 0.5 volts of cell sag is reduced to 0.21 volts of RMS sag, while the average voltage goes up 0.07 volts with this example.

volts/duty cycle/RMS/Avg

4.00 85 3.69 3.40
3.90 88 3.66 3.43
3.80 90 3.60 3.42
3.70 93 3.57 3.44
3.60 96 3.53 3.46
3.50 99 3.48 3.47

 

[Rocketman]

Once you decide on your RMS approximation, you're going to crank it through to power- and then what are you going to do with that?

Pick the worst of the wife's egos to tear apart


Oh, and try to get to the magic 1000 posts in this thread


and jim, straight answer:
how any threads and posts in just this forum have been made about eGo, clones, Rivas, 3,4 volt regulated, 3.7 volt unregulated, this is better than that, etc.

How many of those posts have NO purpose in life?
The discussions continue though.

Next you will probably see youtube videos about this is better than that

 

[jimho]

I'll try and convince you
The cell is not magic (only the controller chip).

Think about ANY Li-ion e-cig. As you vape the cell voltage will fall. Refer to the load tests performed by the flashwarts on one of those 'other' forums . The loaded cell voltage falls and at some point we need to have a way of shutting it off to keep from overdischarging it. The falling cell voltage is what gives most unregulated e-cigs their characteristic "SAG".

Even the cell inside the eGo will sag and as it gets past 90% or so of charge capacity it drops like a rock. The Li-ion cell is pretty flat, but still sags, and the discharge curve will have a 'knee' towards the end.
Even the eGo magic chip can't keep that from happening.

IF, the eGo increases Duty Cycle as the cell falls off in an attempt to keep the vape somewhat consistent the Duty Cycle has to be increased faster than the cell falls off or the vape would fall off. Eventually it does anyway.

Try this with loaded voltage and duty cycle and see what a correction like this does:
4 volts peak, 85% duty cycle,
3.9 volts peak, 88%,
3.8 volts peak, 90%,
3.7 volts peak, 93%,
3.6 volt peak, 96%,
3.5 volts peak, 99%.

That would be flatter than an unregulated e-cig, I think.

I agree... who are you trying to convince?

 

[Rocketman]

All the poor saps that don't have a meter because we tell him he needs an expensive one.
and those that keep making expert commentary about stuff that, well I'll stop there


and ah heck, I learn stuff here also.

 

[jimho]

Pick the worst of the wife's egos to tear apart

Well- who cares- just ask her which one she wants to sacrifice- she'll know - the one she doesn't like is the right one...



Oh, and try to get to the magic 1000 posts in this thread

and jim, straight answer:
how any threads and posts in just this forum have been made about eGo, clones, Rivas, 3,4 volt regulated, 3.7 volt unregulated, this is better than that, etc.

I don't know- don't read them, don't care - but if that's what you want to compare, you should do it with the right tools- not approximations- otherwise, you're adding to the noise because your data will be suspect.

How many of those posts have NO purpose in life?
The discussions continue though.

Do you think you'll end them? Not a chance...

Next you will probably see youtube videos about this is better than that

There already are-

Have a good night gents...

 

[JW50]

Not that this is what the eGo is using as a correction curve, but notice how 0.5 volts of cell sag is reduced to 0.21 volts of RMS sag, while the average voltage goes up 0.07 volts with this example.

volts/duty cycle/RMS/Avg

4.00 85 3.69 3.40
3.90 88 3.66 3.43
3.80 90 3.60 3.42
3.70 93 3.57 3.44
3.60 96 3.53 3.46
3.50 99 3.48 3.47

I end up a bit different when I use your cell volt/duty assumptions. I get:

Cell Duty RMS Avg
4.0 0.85 4.34 3.40
3.9 0.88 4.16 3.43
3.8 0.90 4.01 3.42
3.7 0.93 3.84 3.44
3.6 0.96 3.67 3.46
3.5 0.99 3.52 3.47
I think the average should be flat at about 3.23 to 3.25 (a loaded voltage that I have typically seen on a cheapo volt meter). So, using the cell voltage and average as starting assumptions I get this:

Cell Duty RMS Avg
4.0 0.81 4.44 3.24
3.9 0.83 4.28 3.24
3.8 0.85 4.12 3.24
3.7 0.88 3.95 3.24
3.6 0.90 3.79 3.24
3.5 0.93 3.64 3.24
Do have a meter (relatively cheapo) that will measure duty rate. But not much confidence in read out. Duty rate varies too widely on nearly instantaneous basis. But, did get a read of 0.816 duty rate on an eGo just off charger, loaded at 3 ohms. Average volts measured 3.25 (and AC volts measured 1.28).

 

[Rocketman]

RMS greater than cell voltage?
That's almost like the GPS saying "turn left, here"

Think of the 50/50 DC example:
4 volt peak X .707
and
4 V X .5

Your 81% is with a 3 ohm, maybe 85% with a 1.6 ohm?

mostly flat average or mostly flat RMS, which one will give mostly flat heat?
(the marketing hype selling point of the eGo)

 

[JW50]

RMS greater than cell voltage?
That's almost like the GPS saying "turn left, here"

Agree - makes no sense. Let me try again.
Rocketman says:

Not that this is what the eGo is using as a correction curve, but notice how 0.5 volts of cell sag is reduced to 0.21 volts of RMS sag, while the average voltage goes up 0.07 volts with this example.

volts/duty cycle/RMS/Avg

4.00 85 3.69 3.40
3.90 88 3.66 3.43
3.80 90 3.60 3.42
3.70 93 3.57 3.44
3.60 96 3.53 3.46
3.50 99 3.48 3.47

I don't end up a bit different when I use your cell volt/duty assumptions. I get:
Cell Duty RMS Avg
4.0 0.85 3.69 3.40
3.9 0.88 3.66 3.43
3.8 0.90 3.60 3.42
3.7 0.93 3.57 3.44
3.6 0.96 3.53 3.46
3.5 0.99 3.48 3.47
I think the average should be flat at about 3.23 to 3.25 (a loaded voltage that I have typically seen on a cheapo volt meter). So, using the cell voltage and average as starting assumptions I get this:
Cell Duty RMS Avg
4.0 0.81 3.60 3.24
3.9 0.83 3.55 3.24
3.8 0.85 3.51 3.24
3.7 0.88 3.46 3.24
3.6 0.90 3.42 3.24
3.5 0.93 3.37 3.24
Do have a meter (relatively cheapo) that will measure duty rate. But not much confidence in read out. Duty rate varies too widely on nearly instantaneous basis for confidence. But, did get a read of 0.816 duty rate on an eGo just off charger, loaded at 3 ohms. Average volts measured 3.25 (and AC volts measured 1.28).

Flatest heat would key to RMS. However, that is not consistent with my observations of eGo at just off charger, mid-way through discharge cycle and just prior to cut off. My observation was constant cheapo voltage (average voltage). Keying to average voltage makes the heat more consistent on eGo compared to non-PWM batts. But still, some fade of heat as one advances into discharge cycle. For certain fade on the non-PWM batts into cycle.

 

[JW50]

...
Your 81% is with a 3 ohm, maybe 85% with a 1.6 ohm?
...

Will try different ohm loads on my untrustworthy duty cycle meter. I'm thinking I'll get similar duty rates for both. But - we'll see.

 

[Rocketman]

The 85% I was referring to was the much earlier post where a very good measurement was made showing 3.98 peak volts from a 900mah eGo loaded with 1.6 ohms I believe, and that was 'loaded' including the effects of cell resistance, wiring resistance and the output mosfet on resistance. 3.98 volts peak, to the atty.
The duty cycle was measured at 84.7%, if I remember correctly.

That's just about 8.4 watts from an eGo.

Not my measurements.

 

[Stonemull]

hi peeps, had a good weekend ? summers on the way here

Had a play with a spreadsheet to prove it to myself 100% ..
The error between Vave and Vrms depends purely on duty cycle with these PWM waveforms.

Vrms = Vave * √(100/dc)

where dc is a percentage ..
so 0 error at 100% up to less than 1/3 the correct reading at a 10% duty (√(100/10) = 3.16)
rarely going to see much of a duty under say 70% so error in voltage will be under 20%.

wattage is based on the square though so errors really accumulate quickly. for eg.. a 70% duty cycle results in 42% more power in true (rms) wattage into an atty than what calculations with a average DC value would give.
But .. you lose the squares a square roots ..
do the calculations using the Vave for power . then multiply by 100/dc to get the real power.

eg .. if you had a reading of 3v and 70% duty into a 2Ω atty.
P = (v^2/r) * 100/dc
= 9/2 * 100/70
= 4.5 * 1.42
= 6.43W

 

[Rocketman]

Yea, rub it in about "summer almost here"

(we gonna make it to 1000 posts after all)

Stone,
as the Duty cycle heads upwards of 85% as the cell discharges, and BEFORE the cell voltage starts falling like a rock at the end of the charge, do your numbers show the dif between Vavg and Vrms getting less?

I agree any error in Voltage measurement is squared when you try and calculate Power from an inaccurate voltage.
What do you get for a 'fresh off the charger' eGo for power? and is the Power the same, as delivered to a 3 ohm and a 1.6 ohm load?

Rocky


I have upgraded to the new e-cig RMS Power Calculator

 

[JW50]

Will try different ohm loads on my untrustworthy duty cycle meter. I'm thinking I'll get similar duty rates for both. But - we'll see.

Am having difficulties making readings mentioned. Device for testing atty loaded batts is apparently shorting out. No apparent short when not on batt but when on batt irregular results. Any suggestions on best way to make device for testing with atty load?

Did test unloaded eGo just off charger with my "unreliable" relative cheapo meter. I got 3.528 volts on DC scale, 0.208 volts on AC scale, 28.7 hz and duty rates of 89.1%, 85%, 84.2%, 23%, 21% and 86.6%. Duty rate just unreliable in my opinion.

Then at this link: Artisan's Workshop - Ego Booster.m4v - YouTube I copies an oscilloscope screen of the testing of a loaded ego. That screen is attached below:

[/URL] Uploaded with ImageShack.us[/IMG]

 

[JW50]

...

Vrms = Vave * √(100/dc)

where dc is a percentage ..
so 0 error at 100% up to less than 1/3 the correct reading at a 10% duty (√(100/10) = 3.16)
rarely going to see much of a duty under say 70% so error in voltage will be under 20%.

...

I think Vrms = Vave * √(100/dc) is same as Vrms = Vave / √(dc) where dc in latter equation is duty expressed as faction. (e.g. 81% = 0.81 as faction)

 

[jimho]

you guys know you want it!!!

Specs:
DSO-2150 USB_Hatek Electronic co.,Ltd._http://www.hantek.com.cn

Get it cheap on ebay:
DSO2150 150MS/s PCBase USB Digital Storage Oscilloscope | eBay

 

[Rocketman]

you guys know you want it!!!

How well does that work? For a couple hundred, that would be nice.
Would take less Kitchen counter space than mine.

 

[Stonemull]

Not sure I agree here. Rectify a positive or zero wave form and you get the same positive or zero wave form back. Just like squaring a square root. Also disagree that one can not get a correct Vrms using a cheap multimeter using ac V/dc V. But both numbers needed and knowledge that the ac V is determined from a full cycle and variance from the non-rectified average. Calculation is not particularly straight forward - but calculable nonetheless. The use of the two reads not much different than using other means in determining Vp and using it in connection with dc.

You didn't read my post thoroughly, I mention adding a capacitor and a diode, the cap will charge to the peak voltage and retain it while the ecig is off, something like a 1uF would do the job easily if the only load is a multimeter.
So this rig measures peak voltage and another reading or second meter gives average voltage.

maths not that tricky, working a real world example might be .. Vave = 3V, Vpeak = 3.1V (as measured above)
assuming a 0.6v drop for the diode, will be pretty close. then the real Vpeak = 3.7v
duty = 3.0/3.7 = 0.81

rms correction = 1.11 = 1 / √0.81
RMS volts = 3 * 1.11 = 3.3V

 

[JW50]

you guys know you want it!!!

Specs:
DSO-2150 USB_Hatek Electronic co.,Ltd._http://www.hantek.com.cn

Get it cheap on ebay:
DSO2150 150MS/s PCBase USB Digital Storage Oscilloscope | eBay

I think you're right. But still have to figure out a reliable device for capture of input from loaded eGo (or any e-cig batt). Any suggestions or advice? Would love to see wave forms from an eGo with differing loads. Wondering if duty changes with different loads. Suspect peak voltage changes at different loads but would love to see it on a scope. Could a DSO2150 display a power wave form of voltage times current?