All posts by wyldbytes

Home Server – Install Linux from USB

Since I haven’t got a SATA CD/DVD drive to connect to the board to install ubuntu I decided I would install from the USB memory stick. Finding out how to do this when you haven’t got a Linux PC proved to be harder than I thought.

I came across many pages recommend various programmes for doing it, but when I downloaded the current versions they no longer had the option and was just for making LiveUSB installs of linux.

I found a guide here with a manual process for LiveUSB and also a guide here for using Linux to do what I want. So using those I came up with a windows method.

In brief

  • USB memory stick formatted as FAT32
  • Download the iso of the linux you want. I’m using 12.04 Server.
  • Download 7-zip for extracting the contents of the ISO and zip files, if you alrady have a method for doing this then you don’t need it.
  • Download latest syslinux in zip format and extract the contents
  • from the extracted win32 folder in a command prompt run syslinux.exe –mbr –active –install e: where e: is the USB stick
  • Open the iso using 7-zip or your method of choice. Extract the contents of the iso to the USB memory stick.
  • Rename the isolinux folder to syslinux.
  • In the syslinux folder rename the isolinux.cfg to syslinux.cfg

Home Server – case/psu

The selection of a case and power supply is ongoing. I ruling out cube cases as the location of the Home server doesn’t suit that shape. Most of the mini-itx case I’ve seen come with a 300W psu which way of what I need and also PSU may not be that efficient running at low percentage of there power. So I’ve decided for the PSU to use PicoPSU-80 which is 80W which should be more than plenty. They also claim a high efficiency at low load which is what I’m looking for. I’m putting of the selection of a case for the moment.

Home Server – Storage

I’ve decided to have two drives one for the system and one for the data.

For quietness and low power the system drive will be solid state drive. I could buy a SSD but since I don’t need that much storage or performance for the system drive I’ve decided to use a compact flash card with a sata adapter.

For quiteness and low power ideally the data drive should be 2.5″, but cost per GB is about twice that of 3.5″ drives.  Initial the choice has been made for me as I will use the HD from my dead NAS

I have yet to decide what format I should use for the data drive as I think it might be handy to have it in format that can be read by windows. Since this is a windows home and that would make it easy for recovery if needed. That choice might have a performance penalty so I think I will need to do tests to before I decide.



Home Server

After having my second home nas box die and not being impressed with the speed of them I’ve decided to do something a bit different. I could buy a more of an up market and more expensive, but instead I take the opportunity to build my own.

The requirements I’m setting for it are:-

  • To have good performance
  • Low power
  • Quite
  • Reasonably compact

My first thought was to use a Raspberry Pi as this would definitely meet three of the requirements also I want to have a reason to use a Pi. Performance wise from what I’ve read it looks like it can just about max out it’s network which is a 100Mb, but I would like the possibility of being a bit faster. A down side is that it would have to use an external drive.  It might also be but underpowered if I wanted it to do other stuff as well.

So I’ve decided to go the x86 root. Since I want it to be compact I’ve decided on the mini-itx form factor and for quietness and low power a fanless design. The Intel D2500HN seems to meets the requirements and should have enough performance. I will be using Linux to save costs as it can do everything I want or may want plus I want a chance to us Linux.

The board is ordered which I’m waiting for while I decided on the other componets.


Servo controlled launcher

First thing to find out was a RC servo is strong enough to pull the launcher release. Did some succesful dry tests up to 60psi as these were indoor tests I didn’t go above 60psi for safety reasons.

As I want the servo to be controlled by a microcontroller so a camera could also be servo controlled I needed to find out if the servo would work over a long cable. At 100 meters with the cable still on the reel there was to much loss for the power to be carried at that length. With the power on a shorter cable and just the signal on the 100 meters the servo worked but there was a lot of jitter.

Since I would not be needing anywhere near that length I repeated the test at 10 meters. Power loss was still a problem but there was no problem with the signal. So I have a additional power source at the launcher for the servo.

I have a microcontroller controlling the servo and also a second servo which will pan the high speed camera up at a predefined speed at launch.

Video processing

I use VirtualDub for processing my videos. I noticed it has this feature where you can export the video clip to a series of photos and import a series of photos to a video clip. Which gave me the idea I could do some of my own processing.

Now since I’m terrible at keeping the camera steady and also tracking the rocket. I though it would be a good idea if I could video the rocket more zoomed out then afterwards zoom in on the video clip once transfered to the computer. Also now you can buy HD video camera it makes it even more worthwhile With VirtualDub producing a series of photos I thought this would be possible.

Since I didn’t want to  individually open up each photo in a photo package and crop and then save. I thought a little program to automatically open the photo I would then click on the rocket and the program would then crop and save the photo and then open the next one.

I’ve successfully wrote such a program but it was still a bit boring clicking on the rocket for each frame. Now since the photos in question are usually of a dark object against a light background I thought I might have a chance of writing the program to actually find the rocket itself.

A few more days of programming and a few unexpected problems I’ve met with success. At least for the one bit of footage I’ve got and while only the sky and rocket is in few. I’ve managed to exclude the ground to a certain extant by making a few assumptions. Trees and Churches on the horizon do cause problems still.

I need to get some more footage to try it on and then I’ll be able to post the results. A HD camera would be handy as well. Christmas is coming 🙂

It’s also given me an idea for another project.

What went wrong with premature stager launch.

My stager is a gardena type connector with the spring removed and arranged so the spring pushs it open. It’s held closed by the pressure of the first stage inflating a bicycle inner tube against the inside of a pvc pipe connected to the ring of the connector.

On removing the pipe I found unsurprisingly the inner tube had got wet from the first launch. The water acted as a lubricant between the inner tube and smooth PVC pipe. I need to add a ridge to the inside of the PVC pipe which I was going to do originally but skipped it since it had such a good grip without (dry of course).

Launch Day 5

Planned launches

  • Tango Ia with green nozzle 60psi
  • Tango IV to test stager and parachute if above succesful

Pumped Tango Ia with the green nozzle up to 60 psi to test the nozzle would hold the pressure. Then launched which was difficult since the restricted nozzle launcher needs a lot of force to release.

Tango IV is a combination of Tango III with modifed Tango II as a second stage. Each stage is 700ml and was filled with 300ml of water each. Pressure was 60psi.

The nosecone departed to early so the launch didn’t test the deployment device. I was pumping the rocket up for a second flight when at about 60psi the first  stage released the second stage. The nose cone stayed on until the deployment device pushed it off. But since it was a premature launch without the first stage it hadn’t got very high and was about 2 meters before impact with the ground. So the parachute didn’t get a chance to open. The deployment device was wrecked.

I’ll be moving on to my next range of rockets with a higher volume before trying the parachute deployment device again.

Launch Day 4

Planned Launches

  • Tango Ia from stager with restricted black 9mm nozzle at pressure 35psi with 250ml of water
  • Tango II at 35psi with 300ml of water to test chute
  • Tango III at 35psi 1st stage 250ml, 2nd stage 300ml water
  • Tango Ia with restricted 9mm green nozzle at 50 psi with 250ml water.

I was pressed for time due to only having a bit of free time at the end of day and there isn’t much daylight this time of year. So I went and forgot to take anything to measure the water out so I think I under filled the rockets which might of been the cause of some failures.

First launch was to test that stager released the second stage when the first stage loses pressure. It was also to test that the second stage would pull out the parachute of the first stage. So the rocket was pressurised and then released the pressure from the first stage. This does pose a problem for launch aborts.

Second launch was to test parachute deployment of final stage. I think due to underfilling the rocket there was not enough flight to for the deployment.

Launch 3 was to test the stager in a actual launch.

Future parachute deployment tests will be done at higher pressures with enough water to give a chance for the parachute to deploy.