The picture above is of the safety circuit, it is mandatory according to the regulations of the Shell Ecomarathon. In the vehicle it will be plugged into safety buttons and a Hydrogen sensor. Should any of these trip then three things will happen:
Power to the fuel cell will be cut
The motor controller will be turned off
The safety valve connected to the hydrogen tank will be shut off
The circuit is relatively simple: three switches (hydrogen sensor functions as a switch) are connected in series to each other. This then connects to a relay which provides power to the solenoid, fuel cell and MMS systems.
The eight pronged chip (inverter) and MOSFET on the top left are to regulate current and power the hydrogen sensor.
When I started on the project this year the safety circuit was not working. After doing some measurements with my trusty multimeter I found out that one of the diodes was cracked. We did not have the part in our stocks so I ordered several of them for the fix and as future backups. I soldered out the old part and soldered a new one in.
After mucking around with which ports were actual switches (poor documentation) I finally got it functioning again for use with the fuel cell.
A friend at school came to me saying that his laptop was overheating and that no air was coming out the vent on the side. I figured the culprit was dust so I pulled together some tools and got to work.
I started by removing the components that were easily accessible: Hard drive, Modem, RAM, DVD drive and the battery. I then removed all the screws from the bottom of the laptop and carefully placed them aside. Laptop screws care extremely fiddly and tiny: one size does not fit all. If they get mixed up you could spend hours trying to figure out which screws go where. I was then able to remove the keyboard and disconnect all the cables and top case screws.
I then carefully un-clicked the plastic snaps and was able to remove the top case to get in the laptop
With the top case removed I was able to remove the motherboard and get to the fan. There we found the culprit:
Up against the heat sink and heat pipe was a carpet of dust that was obstructing airflow. With that removed it was just a question of reversing the previous steps to get the laptop in one piece. The laptop started up without issue and was running much cooler afterward.
One of the biggest issues with laptops is that they are designed to be as compact as possible with not much regard to maintenance and access. While Laptops have been getting better in this aspect in respect to hard drives and RAM, the processor is often forgotten. This is harmful to the average consumer that can not diagnose such issues, they end up paying hundreds for a job that requires few tools and parts but a lot of experience.
It is a major design flaw which could be engineered around, yet very few manufacturers bother.
You may have noticed a lot of posts on the same day. These are all experiences from my past year in automotive engineering. So the dates might not be accurate. This post however is only a week off.
Last week we were finally able to get a past student back to show us proper procedure for starting up our fuel cell.
As I have just gotten started on this project and my fellow fuel cell teammate had not worked much with the fuel cell before, it was necessary to have him show us what and what not to do when running the fuel cell. We are using a PEM fuel cell, although efficient, they are expensive (20,000€+) and easy to ruin. Due to the potential consequences of mistake in start up procedure, it was necessary to have an expert.
A quick background: The Shell Ecomarathon is a worldwide competition where teams compete to creat the most energy efficient vehicle. There are different categories, the cell is used to power the Triga, our vehicle riding in the prototype category. In the past the Triga has been able to ride 1,600 km to one liter of euro 95 petrol (equivalent).
The fuel cell had not been started in over three months meaning that the PEM was most likely dried out. This is bad for the membrane so we took extra care to make sure it was damp before we started it. After assembling the necessary tools (batteries, H2 meter, power supply, Hydrogen, safety goggles, etc) we got to work.
I was told that the Fuel cell management system needed to have it's programming tweaked to increase efficiency and that it needed shielding from electromagnetic interference. Further more the safety circuit was not working so by now we had a healthy list of tasks for the coming year. Another potential boost for efficiency was looking for a more efficient pump but that was less important then getting it running properly.
After hooking up all the bits and pieces we put our safety goggles on and fired it up.
Luckily our expert knew his stuff and had it running in no time. I had heard in years prior that it had taken hours and even days to get it fired up after the summer vacation, we were elated with the results . The fuel cell is still missing about 1.5V of output compared to years prior.
Unfortunately this can come from a multitude of places. Our next step is to hook up flow meters and get exact readings to pinpoint the issue. This will be a year full of learning.
A hobby of mine is broken things. Well not broken things, but fixing broken things. I recently discovered the multitude of cheap parts floating around on ebay and decided to start helping people out with broken smartphones, laptops and more. I recently had an iPhone come in that had been dropped in hydraulic fluid. It was quite a mess to clean up.
After taking it apart I had found the reason the screen was not working: Oil had gotten in between the different layers.
There was no saving this screen so I ordered a new one on ebay. After the part arrived I took the phone apart again
After inserting the new screen and reassembling it worked!
Now to finish the rest of the repairs(software and hardware):
My school has given me amazing oppurtunities to see new things and learn. History is not something to be forgotten and this also goes for the automotive sector. The Louwman museum was recently opened in The Hague and a group had been invited to take a tour after hours.
Taken From the Louwman website: "The Louwman Museum is home to the world’s oldest private collection of motor cars, compiled by two generations of the Louwman family. The museum dates back to 1934 and now comprises over two hundred and thirty antique and classic motor cars. Experts regard the collection as one of the most beautiful in the World. The museum exudes passion. Each car has its own story to tell, its own contribution to history. The motor car is a mirror of culture."
I learned that electric vehicles had been around since before the 1900's! Around that time 30% of the cars on the road were electric and some even employed primitive regenerative braking systems. Fossil fuel burning cars won out the race at the time and electric vehicles were a niche product until very recently.
Of course the Louwman Museum had many impressive cars on display from all corners of the world. Here is a sampling of a just a few:
A New York ladder-wagon so long it needed and extra driver for the back wheels
As a member of the PhidippidesShell Ecomarathon team I was asked to stand at the Ecomobile convention in Rotterdam. The convention is centered around green technology in the transport sector. Everything from electric bikes to sport cars were showcased.
My task was task was to inform people about our project, answer any questions and to promote our team and vehicles. We put our prototype vehicle (the Triga) on display and went to work. Our sponsor Rabobank was kind enough to let us use their "Young Innovators" stand to showcase our vehicle.
Of course it would be a wasted opportunity to not go out and see the other exhibits and learn something. Luckily I was there with a fellow student and we were able to switch off and check the convention floor.
There was a lot to see and drive! I took a spin on a segway (very cool) and a Renault Twizy.
I think the vehicle most eye-catching is the electric Mercedes SLS AMG pictured above. If Mercedes is willing to make an electric version of their famous sports car and sell it for 400k then there must be a future in electric vehicles!
Joking aside it was amazing to see the innovation and jumps in technology in the electric car sector. Having visited the year before I was astonished to see how much the convention and vehicle selection had grown.
One of my courses last year involved choosing a topic, designing a system around that topic and then building it.The goal of this course was not to have a working product but to follow the protocol of working in a group. This included having team roles (chairman, minutes secretary), creating a list of system requirements and creating and executing a plan of attack.
My group of five decide to tackle CANBUS and create our very own CANBUS system to control car headlights. This project alone could have used up half a years worth of blog posts but as I am doing this retroactively it will all have to fit into one.
In combination with developing the system we had to conduct group meetings, maintain hour logs and carefully document everything we had done along the way. I have a rather hefty binder full of information here at home and an empty toner cartridge for this reason.
It did teach me a lot about the process of working in projects. I was fortunate enough to be project leader of our small team and it was my task to steer the team into the right direction. Our team only had 5 members so everyone participated in the development on a technical level. One of the major issues of our team was a member who did not learn or add to anything and refused to do any of the work assigned. We followed our established protocols (written in our bylaws) as far as disciplinary action but nothing helped. After conversing with our teachers he was put out of the group. This was hard for us and him but if he could not contribute he was only going to slow the team down.
But enough of that, on to the technical bits!
We started with matrix multimedia. They have a system called eblocks that are pre-made chips ready for programing. Using a program called flowcode we were able to very easily put down basic programs to make buttons operate lights. When adding a CANBUS controller between the input and output it got a bit trickier.
After many hours of messing with flowcode I finally got a working example:
Of course our goal was to build a system that could be placed into a vehicle so we made the decision to build our own chips. By looking at the schematics of the eblocks chips we were able to find out the bare essentials of what we needed to get the system working. We started with a single chip without CANBUS. Using a breadboard we were able to create this:
We stumbled into issues of pull down resistors while searching for a way to get our lights to turn off after being turned on. After figuring this out we moved on to making a full system on breadboards and got this:
The photo doesn't show it but this system worked like a charm. The input was translated to and from a CANBUS signal and fed to the chip to create our output in LED's.
This system of course could not be fitted into a vehicle but we knew what worked so we could get on to creating a chip. We all had limited soldering skills but with a bit of practice we were able to create this beauty:
This is the chip with sockets for the PIC, CANBUS and transceiver chips. This was the input board.
Then disaster struck, in the rush to get everything done by our deadline a team member accidently hooked up a board without a voltage stepdown. The chips were designed for 5 volts, they did not fare well when hooked up to a 12v source. We had a limited number of chips and with the end of the school year only a few days away it did not seem we would be able to deliver a finished product.
This was a chance to use my black belt in google and I was able to find a supplier in Poland who had the parts we needed. We negotiated with our teachers to come in the week after school and present our finished product as soon as the chips were delivered.
We had the second issue of not being able to get our output CANBUS board to function. This was a also a major setback but we decided to present a hybrid system to show that we had grasped the technicalities. Our hybrid system had an input board that we had built and soldered ourselves but the output board was the Eblock system we had used in the early stages of development.
We did away with the LED's and push buttons and hooked up our system to a car switch and headlights. The headlights were controlled with MOSFETs to avoid ruining a second set of chips. Our finished product:
On the foreground is our own chip, behind it are the eblocks and in the rear is the MOSFET board. To the right are the headlights.
It worked! To prove to our teachers we had indeed created a CANBUS signal we hooked up a scope
It showed the block signal particular to a CANBUS signal including message ID, filler and message body.
As a team we learned a lot about the different setbacks that can happen during a project and it gave us a healthy respect of Murphy's law. Sorry for the wall of text, if you have made it this far here is a video of the product in action.
A quick disclaimer first: During the production of the system I had decided to add a feature that could showcase the versatility of a CANBUS system: the MAGIC button. When this button was pressed the lights would blink in a sequence, similar to a light show. The intention was to leave it out of the final product but by then we had become too attached to remove it. The video starts by showcasing that feature.
