PWM for Power Control + Hello!

[C3H8O2]

Hi all. I'd like to first thank all of you for this forum which has provided some very good reading over the past couple of weeks.

About a year ago I quit smoking cold turkey after having a first rib resectomy because of Paget-Schroetter disease. Fast forward about 7 months and I was back to smoking again when I was working a few months overnight. If I had done my research on this forum beforehand I probably would have avoided buying a ridiculously overpriced/underfeatured starter kit (I'll let you guess which one. ). Been vaping for a couple of weeks now.

So, anywho... on to my current question.

Most mods and vv batteries I've seen use linear voltage regulation. Nothing wrong with that, as so many probably have proved. I like to tinker with circuits though, and was looking at using pulse-width modulation to control power delivered to the atmoizer ("wattage"... am I the only one who doesn't like that term, lol). Basically, switching full voltage to the atmomizer quickly (68 kHz is easily done with a 555 multivibrator), at a duty cycle from 100% down to however low still manages to vaporize the e liquid. At 100% duty cycle (basically equivalent to a switch between the battery and atomizer with no voltage regulation), using a 5v source, a 3 ohm atomizer would dissipate about 8.3 watts of power. So, would using a 75% duty cycle dissipate 8.3*0.75=6.2 watts, 60% dissipate 8.3*0.6=5 watts, etc., etc.?

I'm probably a very long way away from a homebuilt variable power mod, but in the meantime I can always play around with a 556, some power transistors, and nichrome wire on my bench. Thanks for any input you may have, and thanks again for this extremely informative forum.

 

[Bubba]

Most mods and vv batteries I've seen use linear voltage regulation. Nothing wrong with that, as so many probably have proved. I like to tinker with circuits though, and was looking at using pulse-width modulation to control power delivered to the atmoizer ("wattage"... am I the only one who doesn't like that term, lol). Basically, switching full voltage to the atmomizer quickly (68 kHz is easily done with a 555 multivibrator), at a duty cycle from 100% down to however low still manages to vaporize the e liquid. At 100% duty cycle (basically equivalent to a switch between the battery and atomizer with no voltage regulation), using a 5v source, a 3 ohm atomizer would dissipate about 8.3 watts of power. So, would using a 75% duty cycle dissipate 8.3*0.75=6.2 watts, 60% dissipate 8.3*0.6=5 watts, etc., etc.?

I'm probably a very long way away from a homebuilt variable power mod, but in the meantime I can always play around with a 556, some power transistors, and nichrome wire on my bench. Thanks for any input you may have, and thanks again for this extremely informative forum.

Soo....welcome to ECF

Also, I have no idea what any of this means, but it sounds impressive!

*Goes back to hitting rocks together to make fire

 

[Rodderik]

i'm not good enough to comment on the math but just fyi, most PWM mods on the market use the ~33.3hz but of course that doesn't make it right! but just thought it was worth a mention and may have some merit as to why so many use that frequency.

::shrugs::

 

[tnt56]

There was a link about a Provari VS Vamo on scope. The vamo reads out at 37htz and the provari read out at 800htz. Not sure if that's the answer you want but it's in the provari link. Sorry for spelling. Heres the link
http://www.e-cigarette-forum.com/forum/provari/381232-provari-oscope-vs-vamo.html

 

[C3H8O2]

Cool to hear there are commercial products out there using this already. Even 800 Hz is pretty low compared to most PWM schemes, but for heating up a coil of wire switching frequency really doesn't need to be that high I guess. Thanks for the feedback.

 

[UncleChuck]

Welcome to ECF!

As someone else mentioned, quite a few variable wattage devices are using a 33hz (roughly) frequency. Some people, me included, can detect a difference in the quality of the vape. To me personally, low frequency PWM (Like the 33hz chip) give a harsher hit, and some heads I have to keep at extremely low voltage to avoid burning. Some clearos I used to have worked OK on my Bolt, which is a straight battery voltage device.

If I used them anywhere over 3.2v on my 33hz chipped devices they would taste burnt. This wasn't a case of simply overpowering the clearo, as for one thing they were 2.1-2.3ohm heads in them, so running them around 4 volts should be fine. And because running them on the Bolt , which was putting out around 3.6-3.7v under load, did not result in a burnt taste, yet less power with the 33hz chip caused burnt flavor.

You can also hear the individual pulses when using some heads, which people describe as a rattlesnake sound. It appears that the majority of people cannot tell any difference in vape quality however, but if you ARE one of those people who can, I'd suggest using a much higher frequency than you planned.

I remember back in my R/C car days the speed control units, which are basically the same thing as the VV/VW circuits in PVs, would have frequencies of 2,000 3,000 and sometimes higher. These were small devices, not much larger than the board that's in PVs. They also had far higher power handing capability. 20amps, 30amps, etc which is huge power for vaping, as a 5amp limit on a PV would be considered very high power. And they weren't terribly expensive, you could get a good quality one for 60-100 bucks back in the day, I'm sure they are cheaper now.

I don't get why ecig makers can't release high frequency PWM devices (well, obviously provape can) seeing as it's possible to make a PWM device (ESC) that handles far more power, has other features than just power control, is about the same size, and similar cost. Must be laziness.

Anyway, once again welcome to ECF!

 

[C3H8O2]

Something to definitely consider, although an ESC control would be a lot more expensive that doing it myself.

Here's a circuit diagram I threw together for using a single 555 for PWM. Frequency is anywhere from 6.5 kHz to 1.2 kHz with a duty cycle range of about 56% to 99.9%. R1=470 ohms, R2 is a 1kohm pot, and C1=0.47uF. Doesn't provide the fixed frequency of the 556 circuit I was planning on but the range should be high enough.

Haven't decided on a transistor to use yet...

 

[Rader2146]

Most all VV/VW mods these days use a PWM controller. The linear regulators have become relics, although they can still be found in cheap BoxMods.

If you're the kind of person to "pull your own teeth" then a 555 can certainly make a fun project. Popular DIY regulators include the OKR-T/6 (~$9 @ Digikey or Mouser), or the LM2596 module (<$2 on eBay).

Frequency has no effect where power output is concerned. 10Hz or 10GHz doesn't matter. But I do think that there is a bit of validity to the argument that low freqs can affect taste. I certainly don't think that one can distinguish between the on/off cycles @ 33Hz as some have claimed, but overall taste may be a bit harsher due to the longer on time (more heat transferred per cycle).

 

[tj99959]

Guess I don't need to answer this Q

Just understand that PWM is a rather generic term when lacking defined parameters.

 

[C3H8O2]

Sounds cool. I do like to "pull my own teeth" though...

All I need is a potentiometer for this circuit... have everything else lying around... the lower frequencies and 5v supply should allow me to analyze it with a sound card.

 

[UncleChuck]

Frequency has no effect where power output is concerned. 10Hz or 10GHz doesn't matter. But I do think that there is a bit of validity to the argument that low freqs can affect taste. I certainly don't think that one can distinguish between the on/off cycles @ 33Hz as some have claimed, but overall taste may be a bit harsher due to the longer on time (more heat transferred per cycle).

You can easily hear the pulsing of the 33hz chip, it doesn't do it will all atty/cartos of all resistances, just some. It's what makes that rattlesnake sound everyone talks about, there are examples on YouTube if you have not heard the sound personally.

My theory on why it affects the vape is because you are getting high voltage spikes to the coil. Voltage spikes that are higher than the coil would normally handle. Theoretically, when there is no power being supplied between pulses, and the coil cools down, it should all average out to being the same temperature.

But I know for a fact it doesn't work that way. If I use a straight battery voltage device, a Bolt for instance, it's putting out around 4 volts under load with a fresh battery. I can put the exact same atty/carto on my Bolt, take a few vapes, then put it on one of my snake-chipped devices and it doesn't matter what voltage setting I put it to, 3.0, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 3.9, 40.0 etc. and it never matches the same performance as the Bolt, or Ego, or Provari (don't own one but have used one, and it doesn't suffer this issue at all)

The vape is harsher and vapor production is slightly lower/thinner for a given voltage. By that I mean, a straight battery voltage device running at 4 volts puts out more vapor than my snake chipped devices do at the same, or even slightly higher voltage. And the flavor is less clean, and the vape is harsher overall.

I mean, if you can actually hear the pulses, that shows you it's having a pretty big effect on the coil. Extremely high temperature, mixed in with extremely low temperature between pulses. This shows, for a fact, that the coil is NOT averaging out the heat, because if it was there would be a steady carto hiss, not a rattlesnake sound. The sound happens because juice is being vaporized very quickly, then it stops, then it starts again.

If the coil just soaked up the voltage spikes, and averaged the heat out so it was a steady temperature, like with a straight voltage device or Provari, then you wouldn't hear the rattlesnake sound.

 

[Rader2146]

We're on the same page, UncleChuck. I agree. What I was saying is that where watts are concerned, there is no difference in the power being delivered to the coil. A watt is a watt no matter how you slice it up. The difference is in the amount of heat that is transferred to the juice. Our coils are horribly inefficient as a fluid heater/vaporizer. Some heat is transferred to the air, some transferred to the wick, some to the juice. When you hit the juice with a [relatively] long blast of high power, there is more time for the juice to get overheated (more specifically, the wicking action is not fast enough to sufficiently cool the coil through enthalpy of vaporization). Thus you get the harsh taste from overpowered juice. Burnt juice, even if for only a portion of the time, is still burnt.

Oddly enough, it seems things have come full circle, but for different reasons than I originally thought. A couple months back before China started calibrating the new devices for V_RMS, I was swimming against the tide by saying that they should incorporate an output filter (like what the ProVari uses) instead of just changing the calibration. But then again, they could have also stayed with the higher freqs and no one would be any the wiser.

 

[UncleChuck]

We're on the same page, UncleChuck. I agree. What I was saying is that where watts are concerned, there is no difference in the power being delivered to the coil. A watt is a watt no matter how you slice it up. The difference is in the amount of heat that is transferred to the juice. Our coils are horribly inefficient as a fluid heater/vaporizer. Some heat is transferred to the air, some transferred to the wick, some to the juice. When you hit the juice with a [relatively] long blast of high power, there is more time for the juice to get overheated (more specifically, the wicking action is not fast enough to sufficiently cool the coil through enthalpy of vaporization). Thus you get the harsh taste from overpowered juice. Burnt juice, even if for only a portion of the time, is still burnt.

Oddly enough, it seems things have come full circle, but for different reasons than I originally thought. A couple months back before China started calibrating the new devices for V_RMS, I was swimming against the tide by saying that they should incorporate an output filter (like what the ProVari uses) instead of just changing the calibration. But then again, they could have also stayed with the higher freqs and no one would be any the wiser.

Ah ok sorry, guess I didn't need to explain it then

So the Provari uses a filter instead of simply a higher switching frequency, or do you think both? I've seen the output signal on the Provari and it's flat, so does that mean there is actually zero change or does it simply mean it's switching too fast to be picked up as a "peak and valley" signal?

In theory a much higher frequency should fix the rattlesnake issue, without the need for a filter, correct? I'm familiar with the basic concept of PWM, but not the specifics on how the boards on these different units are actually designed. I'm trying to figure out if filtering a lower frequency chip to flatten the signal out would be any different/better/cheaper/easier than just using a higher frequency chip in the first place? Any ideas?

Thanks!

 

[Rader2146]

Provari uses both; output filter and a higher freq. the filter is responsible for smoothing out the signal and making it appear to be a flat line, and for all practical e-cig purposes it is a flat line. You have to zoom WAY into be able to see the ripple (usually measured in mV).

Higher freqs would stabilize the temperature as the power would cycle faster than the coil could heat and cool. Basically the temperature delta between peak and valley would be much less as the freq gets higher.

The problem with filtering a low freq is that it usually requires larger components (inductor and capacitor). The higher the freq, the smaller components needed for the filter. At 33Hz the components wouldn't fit inside the tube without adding considerable length.

 

[UncleChuck]

Excellent explanation, thank you!

I figured it was probably the case that the peaks and valleys were small enough as to be rather insignificant for vaping with the Provari. I was under the impression that the frequency is dictated by the components on the board, yes? Just switching one out to a different value would change the frequency?

I'm trying to figure out why these cheap devices use 33hz chips instead of higher frequencies. Is it much more expensive to use components to achieve a higher frequency? I can understand if this is a result of cost cutting measures, but if they were just not putting any effort into designing the board I'd be a bit more upset about it, if you know what I mean. Although if there really isn't any cost difference between setting it up to run at a difference frequency, maybe we can hope these manufacturers will eventually released an updated device that runs higher.

Or even some sort of add-on filter device, something which screws on the 510 connection between the PV and the atty, to smooth the signal out. Does this sound like something that would be somewhat cheap to make? ($30 or so?) or would the size constraints you mentioned earlier make this a ridiculous large part?

Thanks again Rader, hope you don't mind me picking your brain

 

[Rader2146]

This is mostly guesswork, as I haven't actually played with any of the 33hz boards, but I have played with my 100hz VV Gripper board and there are only so many ways to get the desired result. The boards are basically 2 voltage converters paired together into one. A boost converter to raise the battery voltage up to a fixed 6v output, followed by a buck converter to chop up the 6v into the desired output to the coil. The boost converter does provide a flat, relatively clean, voltage. To do this is must use a much higher frequeny in order to keep the components to a reasonable size. Then comes the buck converter that operates at a much slower frequency to chop up the 6v. Reasons for the slow frequency would be slow sampling rate of the comparator, bandwidth of the processor (responsible for doing the math needed for VW), and cost savings.

I'd have to run a simulation to find the values for a filter that would work reasonably well with such a low freq, but I'd assume you would need about 2" additional length to fit the components. In contrast, at 150kHz the components will easily fit on the face of a dime. The components are pretty cheap, the cost would come to making something to house them.

 

[Dougiestyle]

DNA 12 & DNA 20 @ evolvapor.com

 

[C3H8O2]

Well, I threw together this board which dims an LED using a 555. All that's really needed to adapt it to PV's is use an NPN power transistor to switch current to the atomizer.

However, there's still no feedback. As the battery voltage drops so will the RMS voltage of the output.



Low - Medium - High

 

[C3H8O2]

Hey, could this be my first mod?

At a 1.4 kHz switching frequency, it gives a nice steady hit. I need to fine tune... a lot, lol. This was also the first cartomizer I fried (somewhere around 7 VRMS) and it made me want to puke.