Author Topic: DualParaMos-555PWM v1.0 pcb (Read 80391 times)

david4500 · « **on:** March 08, 2015, 06:08:24 PM »

https://oshpark.com/projects/k3tN98E9

Parts: https://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=bc14af5dff

In a 1590G:

Maybe room for a tact fire switch in a 1550P:

Much thanks for help the guys from http://www.reddit.com/r/BoxModders/ https://www.facebook.com/groups/boxmodders/. The board is based on their 555 pwm mosfet with p-fet schematic & diagram.

Breaktru · « **Reply #1 on:** March 08, 2015, 06:17:11 PM »

Great job David. Thanks for sharing with us

Zanderist · « **Reply #2 on:** March 12, 2015, 08:41:12 PM »

What software are you using for the visuals and schematic?

I see an issue of not knowing what the duty cycle is. Maybe an LED that dims as according to the duty cycle.

I also think for some reason that having it outputting from pin 7 won't work, I say this from experience with what I've built. I ended up using pin 3 to get the duty cycle.

I'm thinking about switching the diodes out for schottky diodes, I have them already but the pain about them is they are 2 times the size of a normal diode. My fault for not paying attention when ordering.

david4500 · « **Reply #3 on:** March 12, 2015, 10:49:48 PM »

Diptrace for the pcb design, schematic and 3d preview.

SolidWorks for the render of the enclosure and pcb.

I followed a diagram which was stated to work. The guys on BoxModders had two different diagrams/schematics. I followed the one that used a pfet. That may be why the circuit is connected differently to pin 7 instead of pin 3.

TheModfatherInc · « **Reply #4 on:** March 19, 2015, 08:48:52 AM »

David! of course youre here LOL

david4500 · « **Reply #5 on:** March 19, 2015, 07:48:57 PM »

Quote from: TheModfatherInc on March 19, 2015, 08:48:52 AM

David! of course youre here LOL

I certainly am. Nice to see your work being posted here. All of these pcbs really seem to be taking off. I'm anxiously awaiting this pcb from oshpark... should have it by the end of the week.

david4500 · « **Reply #6 on:** March 24, 2015, 12:53:18 AM »

Got these boards today

Discovered an error on the pcb. The pattern for the p-channel mosfet, the gate and source are backwards from the actual component. Oops. I was able to get the board to function by flipping the pfet belly side up and bridging solder from the legs to the pads. The oshpark project has been deleted until I make a correction.

Fix with upside down pfet here:

Test fire:

https://www.youtube.com/watch?v=85d_IIGTmCk

On the low end of the 10K pot, the whine was audible and gave me a headache. I have some 100K and 50K pots on order to lower the frequency.

Emex · « **Reply #7 on:** March 26, 2015, 06:43:16 PM »

I just want to understand this project correctly....

from what I read/understand, being a mosfet based circuit, its going to deliver the full voltage from the batteries/power source but it uses pulse with modulation to change the duty cycle?

So at 50% power, and a 3.7v source (under load), the 510 connection would be seeing the full 3.7v but with an output that looks like this on an oscilloscope? _|-|_|-|_|-|_|-|_|-|_|-|_|-|- (forgive the crappy looking attempt at a square wave)

I understand the PWM is generating a duty cycle wave, but what about the frequency of that wave?

I always think of my first VV/VW mod which was the Vamo, and I believe was also a PWM based device, but the frequency was pretty slow, and you could tell/hear the duty cycle and it was kind of a funny vape at times. Im guessing a high frequency PWM would "smooth that out"?

Thanks

Very interesting stuff. David4500, I am following all your work, very impressive.

drmarble · « **Reply #8 on:** March 26, 2015, 08:51:25 PM »

Actually, the batteries are in series so it 'sees' 7.4 volts which is modulated with the 555 timer to give a lower average voltage. If you modulated a 3.7 volt input you would get a pretty anemic vape. This is a nice simple solution for an efficient buck converter.

Emex · « **Reply #9 on:** March 26, 2015, 10:02:45 PM »

Quote from: drmarble on March 26, 2015, 08:51:25 PM

Actually, the batteries are in series so it 'sees' 7.4 volts which is modulated with the 555 timer to give a lower average voltage. If you modulated a 3.7 volt input you would get a pretty anemic vape. This is a nice simple solution for an efficient buck converter.

Thanks for the correction, but it was only an example to help ask the question.... Regardless of what the battery voltage is, its always going to deliver that same source voltage to the coil regardless correct?

So a 50% PWM duty cycle is still firing at 7.4v in its on state, then 0 volts in its off state... essentially regulating the power via PWM though duty cycle (lower average voltage as you said), not via directly though supplying a steady bucked voltage to the coil right?

So that leads me to my next question... quality of vape. What type of frequency of PWM are we talking about, is it still "smooth" like the OKR or Mech, or can you hear/feel the PWM like a vamo?

I assume a higher frequency would provide a better result, but david4500 post says he feels it would be of benefit to lower the frequency via changing the pot value...

david4500 · « **Reply #10 on:** March 27, 2015, 06:07:58 AM »

Frequency is 1.44khz (1440hz) set by the 10K pot and 0.1uf capacitor

http://www.wolframalpha.com/input/?i=1.44%2F%2810*0.1%29%3DFrequency+in+khz

OKR is 600Khz. A Vamo is around 33hz.

Emex · « **Reply #11 on:** March 27, 2015, 06:23:54 AM »

O wow, that's much higher than I thought it was going to be.

Thanks for also referencing the OKR and Vamo... really helps put things into perspective.

Out of curiosity, what has been you favorite board so far between all the designs you have tried?

CraigHB · « **Reply #12 on:** March 27, 2015, 05:18:31 PM »

There's quite a big difference between a DC-DC converter like the OKR and a PWM regulator. Converters use PWM but the output is purely DC. PWM regulators output a pulse that varies in length. The reason the frequency is high for a converter is because the output is filtered to DC and it takes much less capacitance and inductance to filter a higher frequency.

For PWM regulation the frequency is not particularly critical. When using PWM to power an atomizer it's low because it keeps any noise out of the audible range. PWM regulators can squeal when frequencies are higher. Also, it's better for efficiency when the frequency is lower. MOSFETs are much less efficient at higher frequencies.

PWM regulators do have some big advantages in that they are simple and highly efficient. However they are limited in maximum output by peak input voltage and of course they are not DC so there's limitations in what they can power, motors and heating elements are perfectly suited for that type of regulation.

Emex · « **Reply #13 on:** April 03, 2015, 04:51:59 AM »

I got an idea for a pass-through style box using this board... I thought it would be a fun way to use PWM, but my supply would be 12 volts.

What in this circuit needs to change? Is it just a matter of modifying some of the cap and resistor values?

Thanks for the help!

TheModfatherInc · « **Reply #14 on:** April 03, 2015, 09:17:00 AM »

You will have to confim with David but I am fairly certain the only changes needed will be the ones he's making to the design to correct his mistakes. The pcb of course needs to be populated with components that are rated for the job. For example. Don't use 16v caps with 4s input, etc

TheModfatherInc · « **Reply #15 on:** April 03, 2015, 09:17:58 AM »

I run my pwm up to 15v input without trouble

david4500 · « **Reply #16 on:** April 03, 2015, 01:21:36 PM »

Newest version. Issues fixed.

https://oshpark.com/shared_projects/4JRetcg2

You can use up to 4 cells in series (16.8 volts). The timer is the lowest rated components at 18 volts. Everything else is rated for 20v or higher.

TheModfatherInc · « **Reply #17 on:** April 03, 2015, 01:27:40 PM »

I populate all mine with 50v but o wanted to let you confirm any limitations of your own design

Visus · « **Reply #18 on:** April 04, 2015, 05:24:24 PM »

Very nice.

Unbelievable with right battery and fet it can be ran upwards to 900watts+ lol holy shiz

David I love the sound lol, its like the spooling of an electric rc car esc..

Yeah it would get annoying after a while.. maybe

I know its switchable but are the bats wired in parallel or series for this board and can insert a charger when parallel?

TheModfatherInc · « **Reply #19 on:** April 04, 2015, 08:38:51 PM »

The sound never gets old all I vape is pwm
To it

Zanderist · « **Reply #20 on:** April 15, 2015, 02:18:00 PM »

I'm gonna post this question in this thread,

What is the interest in using 8.4 volts? I feel things get a little scary beyond 4.2 volts.

I going to put this idea out there I think you could use a 556 to add a 10 ( or whatever you like) second cut with the additional 555 time circuitry. I only bring this up because if it's running at 8.4 volts how long does it need to be on to produce satisfying vapor?

I think for your next trace maybe make a tap point to solder in a voltmeter to know the status of the batteries.

CraigHB · « **Reply #21 on:** April 17, 2015, 11:11:59 AM »

In terms of PWM, peak voltage determines maximum output voltage. With the series cells typically used, maximum output voltage would be minimum battery voltage with 100% duty cycle. Series cells can be run down to 6V to make use of all the available battery capacity. 6V is a reasonable voltage maximum for a regulated mod.

For an unregulated mod it's possible to use series cells to obtain higher power with no detriment. If you write it out in terms of wattage, it's obvious doubling the voltage and doubling the resistance provides the same power compared to a given voltage and resistance. It's always better to use higher voltage to obtain higher power since currents are reduced resulting in less power loss due to resistance in the current path. The problem is that most atomizers are not designed to accommodate a coil twice as long or coils in series. In order for things to work in a similar manner, the coil has to be twice as long or in series instead of parallel for dual coils.

Ian444 · « **Reply #22 on:** April 17, 2015, 07:10:58 PM »

Quote from: CraigHB on April 17, 2015, 11:11:59 AM

For an unregulated mod it's possible to use series cells to obtain higher power with no detriment. If you write it out in terms of wattage, it's obvious doubling the voltage and doubling the resistance provides double the power compared to a given voltage and resistance.

Fixed it for you

and thank you for your never-ending help and advice on this forum, you are a great asset here. I have to read your posts several times as there is so much in-depth info in them.

Emex · « **Reply #23 on:** April 17, 2015, 07:56:45 PM »

Quote from: Ian444 on April 17, 2015, 07:10:58 PM

Fixed it for you and thank you for your never-ending help and advice on this forum, you are a great asset here. I have to read your posts several times as there is so much in-depth info in them.

You are both right...

1 ohm + 3.9v = 3.9A / 15.21 Watts
2 ohm + 7.8v = 3.9A / 30.42 Watts

I think the point he was trying to make was that higher voltages you see more efficiency/ less loss as you have less loss via resistances, and you can actually downsize the wiring/gauge requirements when you start using higher voltages.

CraigHB · « **Reply #24 on:** April 18, 2015, 03:12:20 PM »

I do make mistakes sometimes, thanks for correcting that.

Take two coils in series each one Ohm, series resistance is 2 Ohms, 7.4V with 2 Ohms is 27W, current flow is 3.7A

Take the same two coils in parallel, resistance is 1/2 Ohm, 3.7V with 1/2 Ohm is 27W, current flow is 7.4A

That's the point I was trying to make, same power with half the current. As power loss increases with the square of current it makes a big difference. With series coils there's much less power loss between the battery and load. There's normally not a lot of resistance there but when you get into higher amperages, small resistances can become significant. For example a resistance of a tenth Ohm in the current path between battery and coils would result in a power loss of 7.4A squared times 100 mOhms for 5.5W with the parallel coils. With the series coils it's 3.7A squared times 100 mOhms for 1.4W.

Another advantage with removable cells in terms of series versus parallel is you don't have to be concerned with the potential for meltdown if one of the cells is installed backwards or if there's a big charge mismatch. Parallel cells have to be protected from that occurrence. It's not an issue with series cells.

david4500 · « **Reply #25 on:** April 18, 2015, 06:03:42 PM »

Newest version. Now slightly smaller.

https://oshpark.com/shared_projects/KVKnPnxr

Zanderist · « **Reply #26 on:** April 18, 2015, 07:02:37 PM »

What could you recommend on soldering SMD components?

I had a nightmare with them when I didn't pay attention to the size I brought (capacitors the size of pinheads) during my first attempt at my take on the 555 timer. I end up using through hole but defeated my intention of space saving.

My problem with SMD is I don't think I have what I need to pull it off

Emex · « **Reply #27 on:** April 18, 2015, 08:42:16 PM »

I just use a decent temp controlled soldering iron and do it by hand... I do have a hot air station but I haven't actually used it yet for soldering, just de-soldering components and heatshrinking.

The right tips do help... just watch some more vids on youtube on the process.

david4500 · « **Reply #28 on:** April 19, 2015, 05:40:13 AM »

Flux, plenty of flux. I just use a hakko 888d station set to 660F and use the stock chisel tip. Watched a few youtube videos on "hand soldering surface mount".

If you want some decent sized smd caps or resistors for your own board, try 1206 or 1210 case. They aren't super small. The smallest I've used is 0805 and I found them alright to work with.

http://www.resistorguide.com/resistor-sizes-and-packages/#SMD_resistor_size

CraigHB · « **Reply #29 on:** April 19, 2015, 02:49:28 PM »

I've done 0403 (1005 metric), those are difficult. 0603 (1608 metric) is what I use most of the time. I solder everything by hand with a fine tip. For the Hakko station I find the T18-BR02 tip works the best with very small parts. It's mainly about the tools, a fine tip, fine wire solder, a pair of sharp tip tweezers, good magnification, and liberal application of flux. A narrow chisel tip should work well down to 0603. The shorter the tip the better.

dc99 · « **Reply #30 on:** April 19, 2015, 05:28:29 PM »

Craig, have you ever used hot air? I have an old weller and one of the el'cheapo's with hot air. Ive never tried.

Ian444 · « **Reply #31 on:** April 19, 2015, 05:31:53 PM »

I was also wondering if solder paste is needed to get solder under the big drain pad on the back of the mosfets? Or does the wire solder (with plenty of flux) wick in there easily?

david4500 · « **Reply #32 on:** April 19, 2015, 06:31:51 PM »

Craig, do you have any PCB fabricators to recommend aside from Oshpark?

Ian, I apply some flux to drain pad, lightly tin, apply more flux, place the mosfet, dab flux on the drain tab, tin the tip the iron (chisel tip), apply to the tab and pad and dwell for a few seconds while soldering. You can see the solder melt along the sides and hopefully underneath the drain as well. Although paste would probably be best... I just don't have any.

CraigHB · « **Reply #33 on:** April 20, 2015, 12:06:07 PM »

Quote from: dc99 on April 19, 2015, 05:28:29 PM

Craig, have you ever used hot air? I have an old weller and one of the el'cheapo's with hot air. Ive never tried.

Well, the "proper" way to do SMD is with reflow, either hot air or some kind of oven. Thing is I've been doing electronics projects since the time of discrete transistors and through-hole components and never really graduated from hand soldering. So no, I don't typically use reflow soldering at all. I have done it a few times in a frying pan to remove components, but not for assembly. I don't have a hot air station, but it's on the list of tools to buy.

Quote from: david4500 on April 19, 2015, 06:31:51 PM

Craig, do you have any PCB fabricators to recommend aside from Oshpark?

I've been using Osh Park for years now. I started using them when it was just a guy taking co-op orders from DorkBot PDX. I've tried a few others before, but the prices with Osh Park are the best I've found. I'm not loyal to them in any way so if I find someone cheaper, I'd use them. Can't imagine anyone can beat those prices though. You wouldn't believe what I've paid to have PCBs made in small quantity before.

Hades3123 · « **Reply #34 on:** April 20, 2015, 04:09:41 PM »

Quote from: david4500 on April 18, 2015, 06:03:42 PM

Newest version. Now slightly smaller.

https://oshpark.com/shared_projects/KVKnPnxr

Would it be possible to pay someone to do this? I can solder, I just don't know how the hell to solder for SMT/SMD components. I would gladly pay someone for the work, if they could make one or two of these for me. How much would you charge, if it were possible to do this. I've seen the digikey parts list post. It's about $8 not including S&H for the components for this.

Parts List, (SMD/SMT)
http://www.digikey.com/classic/Ordering/AddPart.aspx?WT.z_cid=Shared_Cart
https://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=bc14af5dff

OSH Park - DualParaMos-555PWM v1.4
https://oshpark.com/shared_projects/KVKnPnxr

Awesome design btw David. Very nice.

As far as the soldering SMD/SMT components, I believe it looks possible, but... I could probably do it, but I'll just pay someone that has more experience then myself, to do it this time.

https://www.youtube.com/watch?v=Jpj3tilIaik

dc99 · « **Reply #35 on:** April 20, 2015, 04:51:16 PM »

Quote from: CraigHB on April 20, 2015, 12:06:07 PM

Well, the "proper" way to do SMD is with reflow, either hot air or some kind of oven. Thing is I've been doing electronics projects since the time of discrete transistors and through-hole components and never really graduated from hand soldering. So no, I don't typically use reflow soldering at all. I have done it a few times in a frying pan to remove components, but not for assembly. I don't have a hot air station, but it's on the list of tools to buy.
Thanks Craig. I was just curious. Since I have a crap load of Davids boards here I might give it a shot.
I've been using Osh Park for years now. I started using them when it was just a guy taking co-op orders from DorkBot PDX. I've tried a few others before, but the prices with Osh Park are the best I've found. I'm not loyal to them in any way so if I find someone cheaper, I'd use them. Can't imagine anyone can beat those prices though. You wouldn't believe what I've paid to have PCBs made in small quantity before.

Hades3123 · « **Reply #36 on:** April 20, 2015, 06:36:51 PM »

Zanderist · « **Reply #37 on:** April 21, 2015, 10:38:38 PM »

While I am will to attempt it, I am not willing to take the risk of money wasting the parts in a soldering massacre.

Although I am thinking what I could do is mount one side of the board with the parts and tact them in with solder paste and then place into a conventional oven.

Maybe release a through hole version for the noobs.
Also maybe through in an "On' indicator LED.

Hades3123 · « **Reply #38 on:** April 21, 2015, 10:52:39 PM »

Quote from: Zanderist on April 21, 2015, 10:38:38 PM

While I am will to attempt it, I am not willing to take the risk of money wasting the parts in a soldering massacre.

Although I am thinking what I could do is mount one side of the board with the parts and tact them in with solder paste and then place into a conventional oven.

Maybe release a through hole version for the noobs.
Also maybe through in an "On' indicator LED.

I'd like to see that.

Hades3123 · « **Reply #39 on:** April 21, 2015, 11:02:21 PM »

. . .

shandy27 · « **Reply #40 on:** April 25, 2015, 05:03:50 AM »

would it be possible to add some sort of feedback to adjust the duty cycle as the battery voltage goes down. i'll be watching this with interest it looks a really cool project.

CraigHB · « **Reply #41 on:** April 25, 2015, 01:11:38 PM »

It's kind of tricky to do that with a 555 timer since that's an analog circuit. You'd have to set up a network of components to adjust the charge and discharge rates of the capacitor based on battery voltage. I wouldn't know where to start with something like that. When things start getting out of hand with analog circuits it's usually easiest to simplify by going digital. Then it's only a matter of coding a task rather than ending up with an unruly network of analog components. That's pretty much the reason micro-controllers were invented in the first place.

shandy27 · « **Reply #42 on:** April 25, 2015, 02:07:13 PM »

thanks, looks like i'll have to bite the bullet and start learning to program the arduino.

CraigHB · « **Reply #43 on:** April 25, 2015, 02:25:41 PM »

It's actually pretty easy to do a PWM driven mod with Arduino or any other MCU. Most have a peripheral PWM module that only takes a few lines of code to control.

shandy27 · « **Reply #44 on:** April 25, 2015, 03:35:35 PM »

the less code for me to do the better lol. sorry for the thread hijack but do you have any tips for learning programming micro controllers.

TheModfatherInc · « **Reply #45 on:** April 25, 2015, 04:23:23 PM »

I bought my last run through seeedstudio.com and have been pleased with the boards

Zanderist · « **Reply #46 on:** April 25, 2015, 10:59:12 PM »

Quote from: shandy27 on April 25, 2015, 03:35:35 PM

the less code for me to do the better lol. sorry for the thread hijack but do you have any tips for learning programming micro controllers.

You could get some practice by looking into C programming.
Use the book C for dummies.

For Arduino I started with the book
Beginning Arduino by Micheal McRoberts

Start a thread about it though.

CraigHB · « **Reply #47 on:** April 26, 2015, 04:44:21 PM »

Yes Arduino is pretty much the defacto starting point these days. Though you probably would not want to limit yourself to that platform. A lot of people do start on Arduino and stick with it for the duration. You can do that if you want, but you don't have to.

doobedoobedo · « **Reply #48 on:** April 26, 2015, 07:00:39 PM »

I'll second arduino. It's easy so long as you have some basic C. Understanding programming helps a bit too

.

Ian444 · « **Reply #49 on:** April 27, 2015, 02:25:46 AM »

I started with Picaxe and then went to Atmel ATtiny micro's using assembly language programming, as its very similar to Picaxe Basic. I could never wrap my head around C, I did try... Picaxe is a gentle introduction to micro's. They will do pwm and sample voltages, but they have speed limitations, which may or may not be an issue. I found Picaxe very easy to get going with despite zero previous experience.