Viability check, request for comments

I'm planning the following:
A system consisting of the following parts

  • ATtiny 1634 cpu( 12MHz, 15 i/o in this design, 12 analog, 2UASART, spi etc)
  • LDO regulator 500mA 0,4V dropout
  • USB programmer
  • 30 column breadboard
  • Soldering area similar to breadboard with additional ground planes for improved performance and realiability

These parts will be organized into 2 modules:

  • A Lab module with CPU, LDO, Programmer and breadboard. est pricing about 35 EUR
  • A application module with CPU, LDO and solder area, est prining about 15EUR

The Idea behind this is that development will be done on the lab module. This will minimize the connections between different board on the workbench since the CPU, LDO, programmer and breadboard are integrated. Also there are programs like Fritzing which enable you to go from a schematic to a breadboard design and ensure consistency.

Once development is finished the components and jumpers can easily be moved to the solder area of the application module. The CPU can then be programmed with the programmer from the lab module. This would give you a working and tested design in a very short time.

Hi, that chip is inferior to a mega328 in terms of flash, ram etc. But it does have 12 ADC inputs. Is that why you chose it?

Why not simplify your design by making a module similar to a pro mini? Then standard usb to ttl serial adaptors could be used for programming. Also the module could be plugged into a breadboard for prototyping, then the same module can be soldered directly onto stripboard or off-the-shelf breadboard style pcbs.


The choice of CPU depended on a number of factors. Primarily price and usability. The flash and RAM is quite adequate, at least for my needs. And 12 analog definitely played a role in the selection process.

Maybe some clarification is at hand: The Application module is just one single pcb containing both the CPU, LDO and solder area. The lab module has the cpu, ldo, programmer and breadboard on the the same base plate (self contained) and female headers for the i/o connections

Well, doubt the price difference between the mega328 and tiny1634 would be significant if you plan to have even a small batch manufactured, but I can imagine the extra adcs would be important in a lab environment. (Not sure how a tiny is more "useable" than mega....)

I think I understand your description of the two modules, but my suggestion of a pro-mini style design would need only one module in place of the two you propose and use more off the shelf components like usb serial adaptors and breadboard style pcbs.

Maybe you would say the more self contained version would be more robust in a lab environment...


I often use more that 1 USART and 8 analog input Hence usability. The drawback i see in my design is the 12MHz clock, a bit on the slow side.
And yes, one of the ideas was to make the system robust Both the lab module and the application modules are self contained. I have been doing a number of designs with the Nano/Micro/Mini style arduinos on breadboard style pcb’s. It’s a quick and convenient way to work. Lately i have been using Tiny 84 in dip packages directly on those boards.
So the design is more a development of that (with the addition of a power supply) than trying to replace a mega 328.

The idea was once the design is finalized and the components are soldered onto the application module (only one pcb with CPU, LDO and application specific components) the lab module is free for the next project.

What im thinking of is adding some SOIC prototype area so you can use these directly since DIP’s are getting more and more scarce and i want to get away from modules and breakout boards since they increase cost and decrease robustness.

Im attaching a image of the application module PCB