From playing with the programmable Asgard driver and using each 7135 at a time (total of 8, so 12.5% steps), I don't think it would be particularly useful like that. You'll need at least one and maybe two low modes that are less than a single full-on 7135, and then something like 1-2-4-8-chip modes. The jump between 75%/6 chips and 100% is barely noticeable.
What MCU are you guys planning to use? I see a 20pin SSOP mentioned but then the diagram shows an 8-pin device. The text looks like Z8F04 to me...if it is I should point out that the max input voltage is only 3.6V.
Actually MCU Pin 1 on post #7 is a PWM. So you can set the lowest mode from 80uA to 350mA or in steps of 1/4196 (using 16 bit PWM). You can set 1 or two modes in less than a single full-on.
Example modes:
Moonlight = 1mA >
Low = 175mA >
Low Med = 1 AMC7135 (full ON) >
Med = 2 AMC7135 >
Med HI = 4 AMC7135 >
High = 8 AMC7135 >
You can program any mode from 80uA to #AMC7135*350mA Ex. 8*350 = 2800mA).
or if stacked to 11, 11*350mA = 3850mA. I think I like this setup versus one PWM controlling all 8 AMC7135.
The original request was to individually turn on each 7135 from BHSimon. Initially I thought about using a 20pin SSOP but it looks like it’s too big for the board
PPtk’s input is to use binary addressing of 7135s which is really a great idea! We can use then an 8-pin MCU.
The diode drops the Vcc to the MCU to 4.2 - 0.7V which is around 3.5V. Testing at MFG usually test them at higher level than 3.6V so it’s OK. I have actually tested it at VBat=4.5V using a bench PS. See my DIY project here.
Learn to use AVR chips… an 8K Tiny85 is like $.65 in onesies… also a LOT more people are equipped to use them. Sure, they’re not quaint and steam driven like the Zilog chips, but they are rather nice.
I was thinking the same thing. I started doing a PCB layout last night. I used an Attiny25 SOIC SSU as a placeholder until I figured out what chip tivo532 was actually using. Give me an hour when I get home (in 5hours ) and I will have it finished.
Of course if someone beats me to it then let me know so I don't waste my time!
- Matt
EDIT: Oh, and the other advantage to the AVR is that so many people around here program them using the SOIC clips that on board break out points won't be needed which will save a tonne of space on the PCB.
It was tricky enough fitting everything on a 17mm PCB (with a 1mm thick GND ring so only 15mm of usable space either side) especially when you take into account that there are 4 groups of individual signal lines to connect back to the MCU (normally only 1).
Removing the need for programming points makes it MUCH easier.
Good points/suggestions. It just that I’m familiar with the Zilog chip. With the problems I’m seeing members modifying and programming code, I find it much easier on Zilog tools. It’s a complete free windows IDE with a full blown C-compiler/debugger:
1. Edit
2. Rebuild All
3. Download Code
4. Stop and disconnect
I think for me the price is not that much compare to the other parts of the flashlight. And for the advantages to me (maybe not for the others here).
As for the test clip, I think it can be done. Except I opted not to open the driver by having the programming pins accessible on the spring side.
AVR Studio is the free development environment for all the AVR chips. It is VERY comprehensive and people swear by it. I HATE IDE’s, though. I do all my work outside of AVR Studio except for running the STK500 programmer that I use.
I use the WINAVR compiler. It is a full blown C compiler based upon GCC. WINAVR is no longer being developed and has been superseded by another package (I forget the name).
You like WINAVR, pyro? I found the generated code to be significantly larger and slower than that generated by AVRStudio itself, and a LOT larger and slower than that generated by IAR Workstation. I swear by IAR Workstation for both AVR and STM8..