ULTIBOX : Ecig Boost mod with lcd screen , 18650, Constant power. OPEN SOURCE
- Thread starter pailpoe
- Start date
- th_trl_thread_readers 0
- Status
Hello Scubabatdan,
I use for my Board : CADSOFT EAGLE CadSoft - Home of CadSoft EAGLE PCB Design Software- Best PCB Design Software
And for 3D render --> Eagle3D en:eagle3d:eagle3d [www.matwei.de]
This software (eagle) is not complicated to use. I recommend
.
I use for my Board : CADSOFT EAGLE CadSoft - Home of CadSoft EAGLE PCB Design Software- Best PCB Design Software
And for 3D render --> Eagle3D en:eagle3d:eagle3d [www.matwei.de]
This software (eagle) is not complicated to use. I recommend
.
...looking good! I may just hold my pending Digi-Key and madvapes orders!
...assembled?
For consideration as an enclosure...
Project Enclosure (4x2x1") - RadioShack.com
...assembled?
For consideration as an enclosure...
Project Enclosure (4x2x1") - RadioShack.com
Hello Scubabatdan,
I use for my Board : CADSOFT EAGLE CadSoft - Home of CadSoft EAGLE PCB Design Software- Best PCB Design Software
And for 3D render --> Eagle3D en:eagle3d:eagle3d [www.matwei.de]
This software (eagle) is not complicated to use. I recommend
Pailpoe, don't let Dan's modesty fool you. He is our resident Raidy of sorts. He has designed several attys with the genisis concept and is very proficient at circuitry and metalworking. The reason that I said you should meet is that you said you weren't a mechanic.
I would like to see the 2 of you get together and put something together with your batt/board design coupled with an encosure/atty of Dans design........Of course we are all here to help.
Dan you can thank me later for volunteering your time. LOL
Last edited:
Give me the enclosure design and I will make an enclosure out of aluminum or stainless steel for you.
Dan
Very interesting. Thanks so much for posting.
I'm working on a mod that's similar, but my approach is different in various ways. For example, I'm using voltage control with current sensing only incorporated for user feedback and overload detection. I'm using a LiPo instead of an IMR round cell. My LCD display is a lot different. Of course, overall my parts list quite a bit different, but I'm using the same digital pot, but 10kΩ instead of 50kΩ and I'm using a similar charger controller, but the 73833 instead of the 73832.
I like how you solved the issue of feedback compensation by using that particular PWM controller. Though, you give up the ability to use synchronous rectification which is a biggie for me.
The data sheet for the Microchip PWM controller states, "outputs up to 5W" yet you are claiming 20W. How do you get around that limitation?
You are stating 7-8A, but you don't state whether that is output or input. I assume that's input? 7-8A output would be quite lofty for that circuit. You are also stating 10V. 5A at 6V output is the best I can do myself. I hit my PWM controller's overload protection with voltage or current any higher.
Output capacitance is low, only 30uF. Isn't output ripple going to be a problem? I used 200uF on my last mod. I'm attempting to use 100uF on the latest one and I feel that's low.
All of the return current from the converter is flowing through the ground plane under the control and sensing circuits. I use an isolated ground plane to keep high currents away from those circuits. You don't have any issues with noise?
I don't see a voltage regulator anywhere on the schematic yet you show 3V Vdd on the pot and MCU. I don't know of a way to use straight battery voltage without thowing off the ADCs (as battery voltage falls off) other than using an external voltage reference. I don't see one of those either. You're also showing a 5V clamp on the charging input, but I don't see where that 5V source is coming from.
How are you getting the resolution you need out of the signals across the current sense resistor? Those signals are only in the tens of mV. I'm using an op-amp in differential mode to amplify the signal for the MCU's ADC.
I see your design as somewhat lossy. You've got a 20mΩ current sense resistor for the PWM controller and another 10mΩ resistor for current sensing on the atty low side. Then there's a diode rectifier. I use a 10mΩ current sense resistor on the low side and a MOSFET rectifier. My PWM controller is voltage mode so it doesn't require a current sense resistor. You're using a small 3uH inductor so I'd expect to see some notable loss there as well (DCR goes up quick with higher inductance). I use a 1uH low profile SMD part with very low DCR myself. It took me a long time to find one with the specs I wanted. It seems you're liberally sacrificing efficiency in your design. Have you tested it on the bench? What kind of efficiency are you getting?
How's the idle draw? For me, that was a big source of issue to get it under 10µA on my mods. Takes some smart component choices and meticulous coding. Always have to pay close attention to it.
I don't see any fuses. You can sense over-load conditions in code, but I wouldn't rely on that 100%. If sensing fails and there's a short, well that's a big problem.
The PIC18F MCUs are nice parts as far as 8 bit µcontrollers go. I'm using a PIC24F myself. It's also available in a SSOP 28 pin package. Cost is similar. It's much more powerful since it's 16 bit and has a full featured accumulator. I find it much easier to code.
In any case, that's quite a design to make publicly available. I would never do that with the one I'm doing. That's very gracious of you.
I'm working on a mod that's similar, but my approach is different in various ways. For example, I'm using voltage control with current sensing only incorporated for user feedback and overload detection. I'm using a LiPo instead of an IMR round cell. My LCD display is a lot different. Of course, overall my parts list quite a bit different, but I'm using the same digital pot, but 10kΩ instead of 50kΩ and I'm using a similar charger controller, but the 73833 instead of the 73832.
I like how you solved the issue of feedback compensation by using that particular PWM controller. Though, you give up the ability to use synchronous rectification which is a biggie for me.
The data sheet for the Microchip PWM controller states, "outputs up to 5W" yet you are claiming 20W. How do you get around that limitation?
You are stating 7-8A, but you don't state whether that is output or input. I assume that's input? 7-8A output would be quite lofty for that circuit. You are also stating 10V. 5A at 6V output is the best I can do myself. I hit my PWM controller's overload protection with voltage or current any higher.
Output capacitance is low, only 30uF. Isn't output ripple going to be a problem? I used 200uF on my last mod. I'm attempting to use 100uF on the latest one and I feel that's low.
All of the return current from the converter is flowing through the ground plane under the control and sensing circuits. I use an isolated ground plane to keep high currents away from those circuits. You don't have any issues with noise?
I don't see a voltage regulator anywhere on the schematic yet you show 3V Vdd on the pot and MCU. I don't know of a way to use straight battery voltage without thowing off the ADCs (as battery voltage falls off) other than using an external voltage reference. I don't see one of those either. You're also showing a 5V clamp on the charging input, but I don't see where that 5V source is coming from.
How are you getting the resolution you need out of the signals across the current sense resistor? Those signals are only in the tens of mV. I'm using an op-amp in differential mode to amplify the signal for the MCU's ADC.
I see your design as somewhat lossy. You've got a 20mΩ current sense resistor for the PWM controller and another 10mΩ resistor for current sensing on the atty low side. Then there's a diode rectifier. I use a 10mΩ current sense resistor on the low side and a MOSFET rectifier. My PWM controller is voltage mode so it doesn't require a current sense resistor. You're using a small 3uH inductor so I'd expect to see some notable loss there as well (DCR goes up quick with higher inductance). I use a 1uH low profile SMD part with very low DCR myself. It took me a long time to find one with the specs I wanted. It seems you're liberally sacrificing efficiency in your design. Have you tested it on the bench? What kind of efficiency are you getting?
How's the idle draw? For me, that was a big source of issue to get it under 10µA on my mods. Takes some smart component choices and meticulous coding. Always have to pay close attention to it.
I don't see any fuses. You can sense over-load conditions in code, but I wouldn't rely on that 100%. If sensing fails and there's a short, well that's a big problem.
The PIC18F MCUs are nice parts as far as 8 bit µcontrollers go. I'm using a PIC24F myself. It's also available in a SSOP 28 pin package. Cost is similar. It's much more powerful since it's 16 bit and has a full featured accumulator. I find it much easier to code.
In any case, that's quite a design to make publicly available. I would never do that with the one I'm doing. That's very gracious of you.
Looks like there's a mcp1700 3V regulator
This Pic does have a fixed internal voltage reference (1.024V, 2.048V, or 4.096V)
Looks like it's originated from the usb connector
I agree.
As this is noted as not final, I'm interested in watching the progress.
Last edited:
@scubabatdan
Thanks
A lot of question
, good
-->See the datasheet for more informations
For me 750Khz, --> reducing the size of capacitors.
--> 10mA
Thank you for your comment.
Thanks
A lot of question
, good--> 5K (502) for me, not 50K
This chip does not limit the power output, Input current limiting with R6 on my schematic . this resistance only serves to protect the transistor. R6=0.020 --> Imaxinput=7A
-->See the datasheet for more informations
Yes input, of course. 20-25w seems to be the max for me (Pspice simulation).
The most important, What's your frequency ?
For me 750Khz, --> reducing the size of capacitors.
No at this moment, Power transfer between L2 and C7,C8,C9 . to put them close to each other.
3v MCP1700 regulator on my board and 5v from USB connector. I use internal ADCreference = 1,024v.
For current sensing : Rshunt = 0,1 Ohm , Vref=1,024v, Conv = 1024
--> 10mA
DCR = 17mOhm for my inductor. too much for you ? It seems reasonable
Yes.
Why do you want 10uA ?,
The protection is the current shunt if the software CRASH.
Good idea. I'll have to look at this range.
Why not ?
Thank you for your comment.
I cant wait to see this thing in action! When do you think you will have the first one together?
What type of device is needed to interface the programming port? Where might they be found? Other than installing updates or programming by others, it is likely that non-programming end users (but perhaps tinkerers) will get involved with its use?
- Status
Similar threads
