Product Roadmap

From Free Charge Controller wiki
Jump to: navigation, search

For lack of a better label, here is a proposed ‘product roadmap’ that makes sense to me:

  1. Complete 5-amp board
    1. Roll changes of current sense circuitry and minor repairs into a v3.0 board
      1. Include headers for proposed daughter board
  2. Develop daughter board for additional 5 amps each
  3. Develop single 20-amp board
  4. Develop 20-amp daughter board

Each stage of the development (1 through 4) builds upon the other and represents a major change in hardware. There will also be a significant software component in each stage. However, they do not all have to be developed serially. The more members of this group that step up to take on responsibility could begin to develop several of these stages in parallel. Let me break down each stage for you:

1. Complete 5-amp board

The v2.0 board was a huge success, in my opinion. I threw a lot of circuitry together and it all worked, except for the current sensor. There are only two serious layout issues that need fixing as well. However, there is just no way to get around using surface mount (SMT) parts. This board makes use of a lot of new technology whos industries have given up on thru-hole parts. The current sensor is a perfect example. I think the time has come to embrace the fact that if this project ever wants to have ambitions of being able to compete with off-the-shelf charge controllers, it needs to embrace surface mount technology.

I’m planning to roll the above changes into a new v3.0 board. Taking a nod from the community (Jonathan and Gavin in particular ;-) I’ve realized that I need to plan a path to higher current levels. This is the real driver of the proposed road map. I think the most intelligent way to do this without a complete redesign is to add ‘plugs’ or ports to the main board to allow daughter boards to be attached for additional current flow. The current microcontroller has enough I/O to control four buck converters (1 on board + 3 daughter cards) for a total of 20 amps.

I’m hoping that what I’ve learned from the v2 board will allow the v3 to be the last major hardware revision before we have a product that people will trust using. Trust is the key word in that last sentence. The board will need to be validated for reliability. Based on the fact that my day job is as a reliability engineer, this shouldn’t be a huge hurdle. However, it will take time. Before moving on to the other stages of the road map, I want to validate this design by examining it with a thermal camera, operating it at high and low temperatures, and other tests for reliability. I need to have a strong foundation before moving on the future stages of the product. Note however: this does not mean that aspiring members of this community couldn’t begin work on those future stages ahead of me.

Before personally moving on to the future stages, I may also want to make one finale major revision to the hardware. So far there hasn’t been many complains about the voltage limit of 40 volts. That limit is based on voltage limit of the MOSFET driver IC in the buck converter. If we want to handle more power, that means we should support more amperage and more voltage. I’ve come across a buck converter driver IC made by International Rectifier that can drive a buck converter at hundreds of volts. Once we get the current design validated, I may pursue this change before moving on to the other stages.

2. Develop 5 Amp Daughter Board

I’ve been reading a Microchip app-note on how to drive parallel buck converters in a synchronous method. Once the main board is working reliably, it shouldn’t be difficult to design a ‘daughter’ board that can attach to the main controller board to add another five amps. This would essentially require only the buck converter and sensor circuitry to be duplicated off-board. Only a PWM signal and sensor connections need to bridge the gap between boards. However, the high frequency of the PWM signal could present a problem. I think we need to work together as a group to find the best solution to create a reliable, cost effective method for creating this connection.

3. Develop single 20-amp board

At this stage, it is assumed that the reliability of the main board has been validated and we are very comfortable running 5 amps over a wide range of environmental conditions. Our buck converter should also have a higher voltage threshold at this point. We can now take what we’ve learned in accomplishing those goals to expand the main board to be able to handle 20 amps. Coil-craft, the company that makes our 5-amp coil for the buck converter on the v2 board, also makes a 20-amp coil. This will require driving the buck converter at around 100 Khz instead of 20 Khz. This part of the project will mostly focus on modifying the main board to handle the excess current, buck converter component value changes, repercussions of higher drive frequencies, and making it backward compatible.

4. Develop 20-amp Daughter Board

This stage will build upon what we’ve learned in stages 2 and 3. If we can accomplish those two goals, it shouldn’t be difficult to expand the daughter board to be able to handle 20 amps. This would mean we could control 80 amps with a controller and three daughter boards at potentially several 100 volts.

To conclude, I think this roadmap makes a lot of sense. It’s a route for a natural evolution to a highly customizable and reliable product. Ultimately, it will be up to everyone who takes action to determine the ultimate path of this project.

Taking action is the key. We have a lot of talent connected to this mailing list. If we can work together to make a reliable product that everyone has the right to use, sell, and modify we will have done a large part in helping the world to adopt a sustainable energy infrastructure. Those who take action and contribute to this product will have their names literally attached to this product (as per terms of the Attribution-ShareAlike license) as it evolves. That is a lot of street credit.