Dear all
I already have a working model for my robot but I would like someone with expertise to advise what improvements (i.e. component recommendations) can be made and help me to design a circuit board that can be printed.
Thanks in advance
What is the budget and time frame?
1 Like
That was my line ...
For reference on the task:
Since you already have a working robot, I assume you know what's connected to what. If you can take the trouble to learn, you might find you can quite easily turn out your own PCB. KiCad is an excellent PCB production software that has pretty comprehensive design aids built in. It's free, and has an excellent support forum.
I would rather hear what someone quotes instead of saying I have X budget and being charged up to the maximum budget, in terms of time frame, ideally I would like it to be done in a couple of weeks.
Thank you. I'm sure that I could learn the software, however I am mainly seeking guidance through expert advice, not just layout suggestions but recommendations on whether I should/could change components/modules. And if I considered to manufacturer a product, what would be the best materials/components to use in terms of cost, reliability, performance and security i.e. software security.
Quotation very much depends on your goal: hobby project and industrial/medical/life support projects don't require the same level of security.
The first observation I would make is this:
Did you notice that GPA7 & GPB7 on the MCP23017 should not be used as inputs despite what the configuration registers allow. As stated by Microchip, it can lead to SDA signal corruption or even malfunction in the host bus under some conditions.?
The SPI version of the chip (MCP23S17) doesn't seem to have the flaw, you can daisy-chain them or use their address or use CS directly or via a demuxer (74HC138/74HC154 or similar)...
It will depend on if you are dealing with an amateur/hobbyist or a pro. It is unlikely though you will hear from a pro as most require a minimum 13 week project and have to overcharge for anything less than a year. Since a top pro working for one of the big 3 would be earning around $1M annually you can see you are probably not going to get them for much less than $400k.
Good luck.
This has always been a hobby project, my comment regarding security was referring to using the ESP32 in the event I wanted to produce a product for resale i.e. would the ESP32 be suitable in terms of someone wanting to hack it to steal my code. I am aware that hackers are very savvy but part of the reason why I wanted to enlist an expert would be to provide advice on the components that I use.
Thank you for comments regarding the MCP23017, I was not aware of this.
Thanks for your comments, however I was not anticipating someone who commands a fee of anything near $400k - 1 million to be reading my post on this forum! 
I wouldn't have thought that it would cost much more than a couple hundred pounds to draw a printable circuit based on my schematics.
Thank you.
1 Like
The reason people ask about budget is generally not so they can charge you the maximum, but rather to gauge the scale of what's acceptable.
If you came to me and asked me to design a "very simple" PCB, but your budget was only $50, I'd politely decline immediately before even spending the time to look at your requirements. But if you had a budget of $5,000, I might say, "OK, I can do that for $450, and by the way, here is a list of suggested changes that could make it easier to manufacture in quantity, but those changes will raise the price by $350."
1 Like
I was simply trying to point out the HUGE difference in a Pro and an Amateur. I personally know guys at FB making over $750k and by chance ran across an article on Tech salaries a few months ago. I retired in 2000 and only earned about half that amount plus the commisions on my two employees.
You can get lots of free, near-expert advice on component/module selection here on the forums. Complete with counter-arguments and a level of justification that it's probably difficult to get out of an "expert" consultant (who is probably mostly going to recommend things that they are personally more familiar with.)
Designing a PCB will be more expensive, and highly dependent on the size and complexity of the overall project. Are you talking a 2-layer 100x100mm board, or a 10 layer 250x350mm board? Carefully designed to eliminate EMC and Noise issues, or mostly a wiring-harness replacement?
The whole "design for professional manufacturing", including things like FCC and UL compliance and approvals, custom injection molding for packaging, supply chain decisions and vetting, and etc, takes a whole different breed, and is likely to be very expensive.
The current state of your project: "it works, but maybe I want to change everything" is extremely dangerous, both from your point of view (costs can keep mounting up) and from the would-be consultants point of view ("I want the customer to be happy and the product to be a success, but they're burning up money and effort fiddling around...")
1 Like
Thank you for your feedback.
A 2-layer board to eliminate EMC and Noise issues and all of the components and modules take up a lot of space. Generally, it would be nice to clean things up.
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.