My First Pulse Width Modulator Mod

[bstedh]

Well the SGV3 co-op is now over and done with. Man does that thing kick but. This PWM circuit really brings Dan's creation to life.

My noego PWM died on me. I will be taking it apart this weekend to try and figure out what happened before putting another together but I hope to have two or 3 of the switch bodies built and working this weekend. The dang thing just quit firing. The LED still lights but that's it.

 

[asnider123]

OK guys, I now have a good reason for resurrecting my PWM project. I built a triple 18650 mod and need a way to control the voltage without too much heat. Here it is:
http://www.e-cigarette-forum.com/forum/battery-mods/226319-wow-energizer-bunny-variable-voltage-mod.html

If the PWM works, it should provide the voltage control I need with much improved battery life. Now, I need to re-read all the posts in this thread to get caught up, you guys have been doing a lot of work since the original post.

Thanks for keeping it going!! I hope you have found the solution.

Alan

 

[bstedh]

Clutch's circuit has been working great. I have the two box mods that I have been running for a couple of weeks flawlessly. My NoEgo that I put one of the boards in is dead and I put too much epoxy in the thing to get it apart to trouble shoot =[. Didn't get around to trying another one just yet. But I will get it done. My future plan is to run 3 batt in series and regulate them to 7.75V and then PWM that down to about 3.5. It seems to me there is a huge difference in the vape when you pwm 8V compared to pwm 6V. It just seems to work better even though the Vrms is the same at the end.

 

[asnider123]

OK cool!! I am going cross-eyed trying to find the circuit diagram that has been successful, someone please point me at the post #

I waqnta power this mod with 3-18650's, so about 11-12 volts. I have on hand IRL3103 N-channel MOSFETS, 555's, an assortment of resistors, trimmer pots, and a few caps and diodes. Do I have what I need for this project? I need some counseling LOL

Alan

 

[asnider123]

Here is the circuit I am currently using, sorry it took me so long but real life kept getting in my way, lol. I laid it out like this to try to keep it simple for the non-EE types, it is pretty much how you would put it together on a pre-drilled board. I am also lazy and this was quick.

C1 is a .01uf capacitor (it has the number 103 on it)
R1 is 220 ohm 1/8 watt
R2 is 150 ohm 1/8 watt

I might to play around with the potentiometer a bit and go up to 300 ohm or maybe 500 ohm for more adjustability, but it is not necessary. This circuit works really well for dual coils and LR atties and cartos, I have been using it for close to a month now.

Hi Clutch .. will this run on 12v in my new 3-battery mod?

 

[asnider123]

Checked specs on the 555 and it has 18V absolute max supply V, so it should work fine. Ditto with IRL3103, 18v max. So, this circuit should work with triple 18650s .. I will get working on this next time I have a chance. It will probably be happier with a resistor at the 555 output - mosfet junction just to keep the gate voltage down, but I guess it makes little difference ??? Besides, blow a chip, replace it Got plenty!

Just need to know what you would suggest for resistor values to get the duty cycle down to the 30-50% range.

Thanks much for your help

Alan

 

[bstedh]

Checked specs on the 555 and it has 18V absolute max supply V, so it should work fine. Ditto with IRL3103, 18v max. So, this circuit should work with triple 18650s .. I will get working on this next time I have a chance. It will probably be happier with a resistor at the 555 output - mosfet junction just to keep the gate voltage down, but I guess it makes little difference ??? Besides, blow a chip, replace it Got plenty!

Just need to know what you would suggest for resistor values to get the duty cycle down to the 30-50% range.

Thanks much for your help

Alan

You will probably want to use a 500 ohm or 1K ohm POT to be able to get the duty cycle down low enough with a 12V supply.

 

[asnider123]

You will probably want to use a 500 ohm or 1K ohm POT to be able to get the duty cycle down low enough with a 12V supply.

Cool!!! I can do that! Next question: I have some space available for the 555 and FET, but it's about 4 inches away from the pot position, that won't effect the circuit, will it?

 

[bstedh]

Cool!!! I can do that! Next question: I have some space available for the 555 and FET, but it's about 4 inches away from the pot position, that won't effect the circuit, will it?

It shouldn't unless you are using really thin wire. Then the resistance of the wire may have some effect but I would think it would be un-noticeable.

 

[CraigHB]

When you get into the higher currents, wire resistance (and trace resistance) can become a factor. Looking at this chart, you can see that 4 inches of 24 ga wire has about 8mΩ of resistance. With 4 inches one way and 4 inches the other way that's a total of 16mΩ. With two amps, your're looking at almost 70mW of power lost in the wiring. Assuming 5V, that's almost a 1% loss. Not an unreasonable amount to give up, 1% isn't much, but losses increase with the square of current. At 3 Amps, losses are twice what they are at 2 Amps. In any case, I'd say you could reasonably give up the power even if you wanted to hook up the MOSFET 4 inches from the atomizer with 24 ga wire.

The other issue you may run into with long runs is the inductance increase and a small increase in capacitance. It won't effect the part of the circuit driving the atomizer, but for the gate input on the MOSFET, it could slow down response if switching freqeuncy is high enough. For example, if you wanted to place the MOSFET 4 inches from the 555 driver. Though, if you are running PWM frequencies under 20kHz, it shouldn't matter.

 

[asnider123]

When you get into the higher currents, wire resistance (and trace resistance) can become a factor. Looking at this chart, you can see that 4 inches of 24 ga wire has about 8mΩ of resistance. With 4 inches one way and 4 inches the other way that's a total of 16mΩ. With two amps, your're looking at almost 70mW of power lost in the wiring. Assuming 5V, that's almost a 1% loss. Not an unreasonable amount to give up, 1% isn't much, but losses increase with the square of current. Losses gets bigger fast as current goes up. At 3 Amps, losses are twice what they are at 2 Amps. In any case, I'd say you could reasonably give up the power even if you wanted to use 24 ga wire.

According to that chart, 1 ft of 24ga only has a total resistance of 0.02567 ohms (negligible loss), unless I read it wrong ??? Maybe I am reading it wrong, but I cannot see where you got your numbers???? Sorry. Alan

 

[CapeCAD]

According to that chart, 1 ft of 24ga only has a total resistance of 0.02567 ohms (negligible loss), unless I read it wrong ??? Maybe I am reading it wrong, but I cannot see where you got your numbers???? Sorry. Alan

He said 8mΩ for 4", that matches your 25.67mΩ for 1' (About a third)

 

[CraigHB]

Sorry if I didn't make that clear. 1 foot of 24ga wire has .02567 ohms or 25.67 mΩ, the "m" stands for a thousandth of an ohm or milliohm. People would think that ohms in the thousanths have little impact on circuits, but often, resistances that small can be quite significant, especially in high frequency or high current applications. Even with the PCB's I make for my mods, I routinely do things to save a few milliohms here and there.

One foot of 24ga wire carrying one Amp loses roughly 25mW or 25 thousandths of a Watt. As I said, that goes up exponentially with current. 3 Amps over that same wire dissipates almost ten times as much power. You are right, those amounts of power are insignificant, but you can start to notice losses once they hit double digit numbers and those 1% losses here and there do add up.

 

[asnider123]

He said 8mΩ for 4", that matches your 25.67mΩ for 1' (About a third)

I see .. you mean milliohms, boy am I dumb DUH

So, if I were to separate the pot from the rest of the circuit, do you mean I would have a big loss? This is not the power line, just part of the oscillation circuit. How about if I put the pot and 555 together and have just the FET remotely mounted? I think there would be room for the 555 where the 350T currently lives

 

[CapeCAD]

I see .. you mean milliohms, boy am I dumb DUH

So, if I were to separate the pot from the rest of the circuit, do you mean I would have a big loss? This is not the power line, just part of the oscillation circuit. How about if I put the pot and 555 together and have just the FET remotely mounted? I think there would be room for the 555 where the 350T currently lives

You're good, keep the power path short (fet/atty/battery) and the rest will be fine

 

[CraigHB]

If you can't keep the power lines short, I would say to keep your signal wires short as possible. Signal lines typically run to high impedance loads so they are more susceptible to parasitic effets (effects due to wire runs and circuit layout). So if you no option, you could put the pot and 555 circuit close together. You can probably get away with making the MOSFET remote, but there is that one signal line that goes to it. As I implied before, it should be okay as long as you are not running a really high PWM frequency. The lower the PWM frequency the better in that regard. If possible, physcially separate the power feeds from the signal line to limit any cross inductance.

In any case, there's probably no need to over-think it. It's easy to get into mental gymnastics in these discussions. Everything will probably work fine regardless of how you lay it out.

 

[asnider123]

You will probably want to use a 500 ohm or 1K ohm POT to be able to get the duty cycle down low enough with a 12V supply.

OK, I have the circuit almost breadboarded, should I still use the same R1 and R2 as in the diagram? I will be trying a 1k pot.

Also, I read that I need a pulldown resistor between source and gate, 47K I believe. If I use a resistor between 555 and the gate, will it keep control voltage low enough so as not to overdrive it?

 

[asnider123]

Aw heck .. this thing's working so wonderfully, I'm gonna put off the PWM until I start having probs with the linear reg

By asnider123 at 2011-09-25

 

[iuseaspork]

Can anyone tell me the frequency this is running at? I'm playing with cap values, but I just can't figure out what the freq is.

Also tried eliminating R1 and R2... Is this OK for almost full control over the duty cycle? Are there other things that they are doing in the circuit? I want almost full control to test both very high and ultra low resistance atties on this with high C rate lipos.


Circuit Simulator Applet

Without both R1 and R2:

http://www.falstad.com/circuit/#%24+1+1.0000000000000001E-7+24.46919322642204+43+9.92+43%0A165+256+192+368+192+2+0.0%0AO+384+256+400+256+0%0Aw+384+224+384+176+0%0Aw+256+224+256+112+0%0Aw+256+224+240+224+0%0Aw+256+320+240+320+0%0Ag+496+336+496+352+0%0Aw+240+320+176+320+0%0Aw+176+256+176+64+0%0Aw+176+64+208+64+0%0Aw+384+144+320+144+0%0Aw+320+160+320+144+0%0Aw+384+144+384+176+0%0Aw+256+288+240+288+0%0Ac+240+320+240+416+0+1.0E-7+4.946485595159474%0Aw+240+416+448+416+0%0Aw+448+416+448+336+0%0Aw+448+336+496+336+0%0AR+320+144+304+144+0+0+40.0+8.0+0.0+0.0+0.5%0Ad+240+224+240+288+1+0.805904783%0Aw+240+288+240+320+0%0Ar+496+304+416+304+0+47000.0%0Aw+416+304+416+256+0%0Af+416+256+496+256+0+1.51%0Aw+384+256+416+256+0%0Aw+496+272+496+304+0%0Aw+496+304+496+336+0%0Ar+496+176+496+240+0+3.0%0Aw+496+144+496+176+0%0AR+496+144+496+128+0+0+40.0+8.0+0.0+0.0+0.5%0Aw+256+112+320+112+0%0Aw+208+64+272+64+0%0A174+272+64+368+64+0+5000.0+0.9950000000000001+Resistance%0Aw+320+112+320+48+0%0Aw+368+64+384+64+0%0Aw+384+64+384+144+0%0Aw+176+320+176+256+0%0Ao+23+64+0+39+10.0+0.8+0+-1%0A

If the one without the extra resistors is ok... here is my parts list:



1 1 IRLR6225TRPBFCT-ND MOSFET N-CH 20V 100A DPAK 0 1.47000 $1.47
2 1 SBH-331AS-ND HOLDER BATT W/COVR 3AA ON/OFF SW 0 1.94000 $1.94
3 2 497-1963-5-ND IC OSC MONO TIMING 500KHZ 8DIP 0 0.42000 $0.84
4 10 CF18JT47K0CT-ND RES 47K OHM 1/8W 5% CF AXIAL 0 0.06500 $0.65
5 1 987-1067-ND TRIMMER 5K OHM 0.25W TH 0 2.02000 $2.02
6 5 490-5401-ND CAP CER 0.1UF 50V RADIAL 0 0.14000 $0.70
7 2 1N5819-TPCT-ND DIODE SCHOTTKY 1A 40V DO41 0 0.42000 $0.84
8 1 EG4722-ND SWITCH PUSH SPST-NO 2A 48V 0 4.59000 $4.59
Subtotal $13.05

 

[iuseaspork]

Stripboard design for whenever I get around to building it. Out of money... Ugh.

Anyone care to look it over for mistakes assuming you can read it. The colors are jumper wires.