I’m paying way too much to get PCBs made.
How do I know this? I’ve done it before. I know how many boards I’m making (quite a few!). I know the costs. I know the tradeoffs. And not only that, I’ve gotten quotes from China. I’ve seen how little I can pay (15-20% of what I will end up paying per board).
But that’s not the whole story of course (one sentence statements of my idiocy are left for my Twitter account, this is much more long form). No, this is actually a combination of factors, mistakes and conscious decisions that brings me to this point.
So first, the reveal: I will be using OSHpark to manufacture boards for Contextual Electronics. Here’s why:
- I know James (Laen) and trust him — I have met and had beers with Laen. I’ve bought boards from OSHpark before and had (small) issues quickly resolved. I worked with him for the boards I sent to the Open Hardware Summit. I can gChat with him from time to time. These things go a long way in business, in my book.
- I believe in OSHpark — You probably have been seeing more and more purple boards pop up due to the awesome $5/sq in service for a 2 layer board with soldermask and silkscreen. I can’t imagine how that would have changed my learning process when I started making PCBs and I get to hear how positively this has impacted people in the process of learning. Added bonus: Made in the US of A, baby!
- I got a bulk rate — I was a bit fast and loose with the amount of material I designed into the BenchBudEE. I wanted to have a lot of large mechanical features, I wanted to teach using a 4 layer board and I wanted to be able to plug other boards into the Arduino headers (like a display on top) while still being able to access the auxiliary circuits. All of this added up to 15 sq in of PCB material. Using the $10/sq inch service from OSHpark, that would have been $150 for 3 boards. I am paying less due to the number of CE members getting boards. If you contact them, you might be able to negotiate a discount for bulk orders as well.
- It includes logistics — That is a shipped price. Now, before you tell me how easy it is to put PCBs into an envelope and slap a stamp on them, allow me to tell you: I don’t care. I’m not in the logistical business. Many many people are better at this than I am and I don’t want to get into it. In fact, I’m paying a premium in order to avoid it. Aside from getting set up and thinking about shipping and international issues (customs, transit times, etc), I’m also handing of some of the PCB manufacturing worries (working directly with the fab). As the title implies, I’m in the knowledge business now. I will ship and build hardware for myself, of course…but then I’ll teach others how to do the same for themselves. It’s also much more likely that CE members will be doing low quantity manufacturing and need to learn more about that at first than the high volume stuff that would end up consuming my time.
- It saves me head space — I’m a chronic worrier. I know myself and I know I will stress over perfecting the logistics. Instead, I should spend that time developing more class content for members, including the upcoming Session 1B of Contextual Electronics. Plus, working with people who are good at what they do (like Laen) helps me sleep better at night.
Will I definitely use an outside company to ship boards forever? No, it’s possible that I change my mind in the future and decide to pull this kind of service in house. I have seen my podcast co-host do the same with his shipping, as he works on fulfilling his Kickstarter campaign. This has been an interesting lesson and I’ve talked with Dave quite a bit about his struggles and how it’s paying off. He’s building a foundation for many future hardware campaigns (probably crowdfunded).
Note the differences though: Kickstarter campaigns ship everything at once, in batches (in a best-case scenario). I plan to continue developing and teaching new courses on a regular basis (in batches). However, if I want to allow people to sign up for past courses asynchronously (whenever they want), I may want to continue to offload shipping single units to individuals. Having a service like OSHpark or another fulfillment house do this on a per-person basis (even if at higher cost), could make lots of sense for me so I don’t have to keep inventory. Technically this could already happen as the OSHpark service would allow a student to order boards and all work we do for Contextual Electronics will be licensed as Open Source Hardware. However, negotiating longer term contracts and having distributors (or fulfillment houses) keep them on-hand will completely remove this from my list of tasks.
From a business perspective, I am still making a very healthy margin, as I am selling a “knowledge product” (yak). However, if there were price increases due to changes in business conditions, this could be passed along to members as a “cost of doing the course”. Obviously I will work to reduce that burden on them and the trend points to the cost of PCBs continuing to fall.
It kind of sounds weird, doesn’t it? A hardware guy complaining about making and shipping hardware? But in a world of tight timelines, constrained resources (see also: my time) and my desire to continue developing new stuff, this seems to make sense for me. I would love to hear your thoughts in the comments.
Well, it’s my first day out on my own and I have already had a rejection. Whee!
In reality, it was a contract job that I bid on without too much expectation of getting the work. I definitely was not counting on this work as part of my survival in my new jump to self employment. Mostly it was an interesting problem that I would have enjoyed working on and I would have been able to work with some good people; this is my main disappointment with not getting the work.
However, I did get feedback on my proposal, which is great. A lot of times, people will just blow you off and you later get the, “Oh, we went with someone else” a few months down the road. The feedback was positive, that the proposal was well enough done, which was reassuring because I haven’t submitted many in the past. The reality of the situation was that they decided to go with a completely different architecture for the project, which I had no chance of doing well. As such, their decision to not go with me was a good one.
I suppose the biggest disappointment after “not getting the work” was that I failed to convince them to see my solution as the best solution (really the point of any proposal). I knew this would be an uphill battle from the beginning as the project manager seemed to favor the other way of architecting the solution. But how do you change someone’s mind when they have been envisioning a particular type of solution from the beginning? This isn’t isolated to this situation, I have experienced a preconceived notion by managers in the past.
I could complain about a manager making a decision in a vacuum, but the truth is, I don’t know if this decision was made in a vacuum. In fact, it could be that the solution that was asked for is in fact the best way to go forward. If I really wanted (or needed) the work, I suppose I could tailor a solution to the manager’s preconceived notion, quickly iterate on it and show how it ultimately will fail as a final solution and then be ready to propose the alternate (better) solution. Is this realistic as a contractor? No. I think from the outset if you plan on failure, you’re going to hurt your reputation. Plus, as I mentioned above, I have no way of actually creating the (preconceived) solution, so this wouldn’t have even been an option for me. I think the only real way for me to win this (long term) would be to develop this solution on my own for fun/educational purposes and then be at the ready to offer it up later as a possible solution, assuming the preconceived method does ultimately fail. In reality, I have other things to get done, so I’ll just wait and see if they call and I’ll start on the work then.
So yeah, first day out of the gate and I’ve already had a rejection! It feels good to get it out of my system…I’m sure no more will happen in the future…right? RIGHT? Riiiiiiiiiiiight.
This topic was discussed in greater detail in Episode 184 of to The Amp Hour. This article is more about the personal feelings that have accompanied this (ongoing) transition.
Starting Friday, for the first time in my adult life I will be without a job; at least a job as I’ve always defined it. I have been employed as a full time electrical engineer (in various roles, the most relevant being analog ones) for 8 years straight. Prior to that I had been working as a co-op for 10-20 hours a week for another 2 years. All told, I would estimate I have worked roughly 17,000 hours up until this point (in an official capacity). While that would qualify me for the “expert” category if you listened to Malcolm Gladwell and the people he was quoting in that book, it feels like I’m just getting started. It feels like I have a TON to learn and the backlog of things I need to know in the future keeps growing. And you know what? That is one of the best parts of engineering! The real story here is that I have been working for the past 8 months on an electronics education program that is currently running. It’s called Contextual Electronics.
The program launched in January and as of this writing, we are in the middle of the 4th week out of 8 of Session 1A (where we are designing a board). Session 1B (where we build the board) will start in March, a couple weeks after the conclusion of Session 1A. It has been a spectacular group of people who decided to sign up for the program and my decision to leave my full time job is largely based around wanting to serve them better. The 6 months before the program started were dedicated (on nights and weekends) to developing the 8 week curriculum. I want to be able to be forward looking with future curriculum and to better serve them on a day to day basis. But that’s not all I’ll be doing! I’ll also be working on:
- A medical startup — This is a fledgling project with a friend, but one that I am very excited about. It’s in a new field for me as well, resetting the clock on my 10,000+ hours. It is unlikely my “nights and weekends” schedule will change any time soon and I’m ok with that, especially if I get to work on something that excites me like this project does.
- Consulting — I have had a consulting company for about 4 years now, called “Analog Life, LLC” (sounds familiar, eh?). That was normally part time, smaller jobs when I could find them. This will be a larger part of my life now as it will help to support my family more than being a source of income to build out my lab. If you have analog work you need done (or really any electronics projects you’d like to talk about hiring me for), please feel free to contact me.
- Giving talks — This is a new one for me, but I am currently scheduled to do two talks this year: One at the HKN Student Leadership Conference and one at SolidCon. Very different venues, excited about both (and always open to more!).
- Running meetups — I started a local meetup group in Cleveland alongside my friend Martin Lorton. It’s a great opportunity to network and an even better excuse to drink beer and talk about nerdy stuff. It’s called Charged Conversation (shamelessly stolen from a list of possible taglines for The Amp Hour)
- The Amp Hour — Nothing new or different, Dave and I will continue producing podcasts, ranting about what upsets us in the electronics industry and asking people much smarter than us to come talk to us. I’m excited about the potential a looser schedule has for The Amp Hour. Running a show with someone in a timezone 14 hours ahead of you can be restricting…and now that work schedules aren’t a problem (Dave is also self employed), we don’t have that restriction. Just families and the need to sleep at some point.
I also have a running list of projects and other things that have been put off as part of spending my free time on Contextual Electronics. While I don’t think I will have “free” time in any way, shape or form, I do hope some of the following will start to work their way back up the priority list (in no particular order):
- Machining — Remember how I used to write about my new toy, my Taig mill? I started with it about a year ago and then decided I wanted to work on a part time gig that could turn into a full time gig (which it has!). Of particular interest is using my mill to create circuit boards by routing copper clad FR4. This could have lots of benefits for my consulting.
- Run classes in person around the world — This fulfills two interests: Getting out into the world and teaching some of the Contextual Electronics content in person and also doing it in different areas (and seeing the world in the process). This one takes more planning and will be a longer term goal.
- Writing — Sure, everyone says they want to write more, but I do already…it’s just not very visible. I do daily updates for Contextual Electronics and also did so throughout the Beta Process. I would just like to move that same consistency onto this site or others. As much as I already say online, I have a lot more opinions to air!
- Different Podcasts — I stepped away from The Engineering Commons to work on Contextual Electronics (and even went back to talk to them about the program). I am regularly thinking up new podcast topics and would like to explore different formats.
- Developing more products — This is part of developing the curriculum of Contextual Electronics (each session produces an Open Source Hardware project), but I would love to do a start to finish project with the sole goal of commercializing it (without a large corporation backing me).
- Working in different industries — This would be part of consulting, but I would love to branch out a little bit. I have been focused mainly in analog for industrial applications in the past, but also enjoy FPGAs and occasionally have been known to program a microcontroller. I would also like to try different industries, maybe working on a large scale art installation or similar. This is a very nebulous goal, but the main point is I want to continue learning by doing.
- Sleeping/getting healthy — To say my health has declined would be a bit of an understatement. I have definitely gained weight and I can feel my mind getting cloudier from less rest. I think that was another thing that helped edge me towards this decision. I have been keeping the following schedule for the past 8 months:
- Up at 8:00 am
- Get to work at 9:30 am (flexible schedule)
- Short/no lunch
- Leave work at 6:30 pm (on a good day)
- Dinner with family 7:00 pm
- Start working on CE or recording The Amp Hour at 8:00 pm
- Collapse into bed at 2:00 am. Lather, rinse, repeat.
- Playing music — This one has fallen completely off the map and represents true leisure time for me. I used to play in bands and I really miss it. I doubt I’ll ever get back to where I was as a drummer, but I would love to be active musically again. If I can, it will be a sign that I’m moving towards a simpler, more balanced life.
I think one of the things that excites me, is that this inadvertently moves me closer to a post I wrote about 5 years ago called “Engineering The Perfect Day“. Some of my personal goals have definitely changed and music has definitely faded in my priorities, though I still like the feel of that post. It feels like what would be a good life. Only time will tell if it will actually be my future. More than anything, I made good on what I’ve often said about my future when people asked me, “What’s next?” I always respond that my next employer would be me. I don’t know if I’ll be my last employer, but given my increasing restlessness with corporate engineering culture, it seemed like an inevitability. That said, I am sad to leave my current job (which I’ll possibly talk about in the future, as the company is also a good one and I’d like to promote their end goals). I believe I’m leaving under good conditions, though I had asked to stay on part time and was denied for now (more about that in The Amp Hour episode linked at the top). But since I’m leaving and because I want to capture the emotions I’m feeling during this process, I thought I would do a quick list about the things I will and won’t miss.
Things I’ll miss
- Co-workers — Always first on my list, I wouldn’t be working there if I didn’t enjoy the people I work with. I’ve learned a lot from this group and learned a bunch from the ones at the job before. For the first time I won’t have any co-workers and coming to terms with that will be odd. I suppose Twitter folk will be my office mates (as well as my slightly less verbal dogs).
- Structure — I was warned about this from consultants in the field now and ones that have come back to the workplace: the lack of structure can be difficult at the beginning. I can see why. It feels like I have an extended vacation coming up to work on whatever I’d like. In reality, I need to purchase food and keep my house in the future.
- Large scale projects that require skills beyond my own — We talked about this a little bit on The Amp Hour a few weeks back. Large companies often work on larger projects and the accompanying budgets. This can present (though doesn’t guarantee) more interesting problems to be solved. It also allows me to lean on skilled co-workers to help me along on new tasks.
- Seeing my work out in the world — Similarly, when the big projects get out into the real world, they can have some impact. Engineers (myself included) take great pride in pointing at a piece of hardware and saying, “I designed/tested/built that”.
- Non-billable time — As I move into consulting, I recognize some of the finer points of how to bill clients and give them The Warm Fuzzy Feeling™ (one of my favorite articles on the internet). This also means I won’t have as much non-billable time that corporate jobs allow for training and researching new topics. Sure, I track my time at my job, but as a salaried person it is more based upon the result of my work than the number of hours on the project.
- Leverage over vendors — As a hardware person, this can be a big one. Having a large company name behind you (and the associated dollar amounts) can really get a vendor’s attention. Now that I’m a shop of one, that will be much more difficult when I need to talk to someone inside a chip company. Hopefully I might be able to use my previous industry connections and my role on The Amp Hour to garner a little extra attention. Every bit helps when you’re in a bind.
- Significant resources (travel, prototypes) — Basically having a budget from my employer and being able to use it to build the stuff necessary in any hardware business.
Things I won’t miss
- Corporate culture (from all sides) and hierarchy — Blah blah blah, Chris doesn’t like large corporations. But more than experiencing an Office Space environment, it’s more about the slowness that large companies move. If there is a new technology, it’s less often the case that a large company will be a first adopter. As a small consultant, I can take that risk (either on products I design for sale or with a client’s approval on their project).
- Time restrictions — While non-billable time is nice, it does also mean that I am sitting at the office when I could be out working on a podcast or developing more course material. One of the key parts of having a larger workforce is having flexibility in your workers so they can jump onto new projects as they pop up. This also can mean that workers get a little restless when underutilized.
- Commuting — Working in Cleveland in the winter, this comes into sharper relief. I bought a house near my first company for a short commute and promptly changed jobs (brilliant!). Now my 25 mile commute is a drain on the wallet, my car, my sanity…and most importantly my time. My new commute will be down the basement stairs with a much lower chance of slipping and sliding on the snowy roads.
- Profit restrictions — While a salary is a nice guarantee to a steady amount of money throughout the year, it is also by definition is a limited upside. You’ll never earn more than your salary (plus a possible bonus, also usually limited). As a consultant, and more so as someone starting a business, there is “unlimited” upside, meaning there is a potential for making more money. Sure, there is also risk of making less (or no) money. I think Paul Graham captures this idea well in his article about wealth.
- Job roles — Just like having available labor in their workforce, large companies like defined roles so that it is easier to identify who should take on work as it comes down the pipeline. They also like that there is efficiency in one person becoming a subject matter expert (the theory being they get stuff done quicker). Yes, I like analog electronics. You could even say I love the field. But as mentioned above I want to branch out and try new things. And these days, it feels like engineers need to branch out in order to stay flexible and employable (even as a consultant).
Noticeably absent from both lists
- Meetings — These will still happen in my new ventures and may even become more common as I need to get to know new clients and assure them I can do the work. The best case scenario is to minimize the stupidity of all meetings by keeping them short and to the point.
- Benefits — While healthcare was one of my largest concerns going into this decision, it’s really not as big of a deal. I pay either way. You do too! I now will be responsible for paying my corporate premiums as part of any healthcare plan I’m part of of; if I take COBRA, I’ll pay what the company paid for me while I worked there. I’ll still need to pay my personal premiums and put money into an HSA. Yes, it’s a significant cost, but the entire amount paid towards healthcare can be viewed as a lump sum, regardless of who pays it. If my company didn’t pay part of my health insurance now, I would expect them to pay me more per month to match my compensation package to the rest of the market. Now I expect my personal company to pay for my health insurance. The big difference between this year and last is I now have an option if COBRA is overpriced or (in the past) wasn’t guaranteed coverage: I can switch to Obamacare. I like Mr Money Moustache’s take on the new regulations.
- Having a dedicated place to do work – Yes, I also complain about loud co-workers near my cube at work and the fans that blare in the servers near my desk, but my home office will be no different; I’ll have distractions that range from puppies wanting to go outside and play to wives who ask me to do laundry during conference calls (and I’ll listen because she has been so supportive of me!). Unless I get an isolation tank and a waterproof laptop, I likely will have some kind of distraction to complain about.
Really this is all just the beginning of a journey on my own. I’m really appreciative to everyone who listened to my nervous whining about making this jump and still reassured me. I’m glad my wife puts up with my crazy electronics work and my crazier schedule. I’m very grateful to all the people I’ve worked with in Cleveland over the years and am optimistic that people in other parts of the country are as friendly and helpful. And I’m happy you read all the way to the end of this overly detailed brain dump. Hope to do more of this soon. Thanks to Theen Moy for the cover picture of the caps.
I’ve been sour on education lately. Given my entry into the field, I felt it prudent to explain myself.
In general, I’ve been well served by my own education. I went to a great public school system where I grew up and I went to Case Western Reserve University (CWRU) for my undergraduate degree in EE (I have neither a Master’s nor PhD). I had decent grades (solid B+ student throughout my tenure) and I had a good experience with my classmates. So what gives? Why do I complain about education?
One reason is that the real value that came out of CWRU was the co-op program. It was run well and it was my two jobs as a student (taking 7 months off from school at a time) that truly taught me about the nature of electronics. In fact, the hands on stuff is where it all came together (as it often does). And then, looking back towards the theoretical teaching that I learned I started scratching my head as to why it was taught in such a way.
Because it’s always been taught that way.
That’s about the best I can come up with. Sure the math matters (really, it does). And learning about the equations behind resistor dividers or op amp circuits does a lot for students’ eventual need to solve problems. But why were we learning it? What were these mystical devices? Why did I only see them drawn as symbols on a page and not hold one in my hand. Most importantly, why didn’t anyone seem to care about that fact as they showed me that V=IR?
The missing piece was that the University setting started ramping (for electronics) around the time that long-time hobbyists entered the school. Many students entering the electronics programs in the 50s and 60s were at home tinkering with radios and kits while they were younger, so the program teaching the math behind these components fit into the mindset of the incoming students. Some students still enter with this background, of course, but even in the event that a student has been working on hobby electronics for 15 years before entering University, the focus will not be on the practical. Why?
I’ve heard other reasons, that the practical stuff is too specific. You can’t go and teach someone how to use a specific piece of software or a specific bench tool, because the material would be quickly dated (obviously they had never glanced at the tools being used by academia, as those are often ancient in all but the highest research areas). Instead the focus once again moves to the math and the equations that govern ideal behavior. An important concept, but not nearly as useful when used in isolation.
My opinion is that the system feeds back upon itself; students are expected to become PhDs who are then expected to teach students. Coursework is constructed to fit that end goal. The ones who survive are the ones who did the best at mastering the theoretical and the ideal. No room is left for the practical and even if there were, the ones teaching are not well versed in it (with the possible exception of the ones that are hands on in research labs). However, in an engineering school, very few are going to become professors (tenure track positions continue to decline while the number of PhDs continues to grow) and many more are going to enter industry where the practical skills are expected. Schools assume on the job training will take care of the rest (or senior project type courses), while employers continue to cut training budgets and have ridiculous job requirements for entry level positions.
So I decided to build my own thing. I want to give people the context they need to build electronics.
My initial idea for Contextual Electronics was formed out of frustration. Why doesn’t anyone teach this stuff? Schools don’t. Workplaces don’t. Online, very few do (and they’re often scattered across the web, piecemeal). OK, so that looked like there was a need. But what should I teach?
- PCB Design — This was a skill I used often but rarely was taught in any manner. It was more of “figure it out”. I did, but I made some costly mistakes along the way.
- Analog — I had a lot of trouble with signal chain stuff when I left school, regardless of the fact that I took classes in it throughout school. This also was a topic not as widely covered with online resources. Seeing as this has been my main area of employment since then, I thought it worthwhile to share my struggles and how I got past them.
- Tool building — At the suggestion of a friend, I decided to work on something that would be a useful tool on my bench. This allowed me to teach something that has direct applicability, but also something that will be flexible enough to build upon later.
- Supply chain — One of the biggest shocks to the system when leaving school and thinking I knew how design worked (or even in my co-op days) was the “other stuff”. Turns out when you make things, you’re really just piecing together other things…and that means you need to purchase those other things.
- The “putting it all together” — The tough part in any field is associating the theoretical with the practical. Oftentimes, this isn’t perceived to be a separate field of study, but I think it is. If you assume theoretical things will “just work”, you’re going to get burned. If you think practical things will “always work”, you’re going to be iterating without any direction and will also get burned. The method of taking assumptions from the theoretical predictions and using experiments to verify these claims in a real world context (there’s that word again!) is something that needs more focus. In fact, even in this first session of Contextual Electronics, it can use a lot of work.
- The “just do it” mantra — Again, a nebulous topic, but something I wanted to emphasize in the course. Having guidance is nice and will be provided to students. But you need to jump in and make mistakes. So that’s another part of this course, to the point where the finished PCB likely will have some mistakes. Instead, we can call them “learning opportunities”. Session 1B, where we’ll be building and debugging hardware, will be all about this trial-error-analysis cycle.
- Openness — Another misgiving of the academic path is built into the structure of the institution; there are large competitive edges in keeping your work quiet or secret in an age of openness and sharing. So the course will use open tools (KiCad, SPICE), will produce open tools (all products will be licensed as Open Source Hardware) and much–though not all–work will be shared online (progress of the project is shared on GitHub).
I’m not saying that Contextual Electronics is perfect or finished. Just that it’s starting and that it’s an outgrowth of my frustration with the state of electronics education. I hope you consider taking the course. If you do, there’s more info about it here: http://contextualelectronics.com/session-1a-registration-is-now-open/
Thanks to Nick Thompson for the image of Aristotle