Geronimo!!

Well the Parachute has turned up.. It is made out of a lovely 'thin mill' fabric and only weighs in at 11 gms so result. We went for the 24" as we worked out it should give us a decent rate of 5.61 meters/sec and should be a safe landing rate...but looking at it....we need to do some testing but "I think we need a bigger boat"

Flashing lights

Phil noticed that our LED was not flashing on receipt of data, but rather flashing on the absence of data. Whilst not exactly required, I decided to include an inverter circuit so that the LED would operate as expected. It sounds impressive, but it's a very simple circuit requiring no more than a couple of resistors and an NPN transistor.

This has been tested and it works fine. Whether it's a good idea to build it onto the production board, I'm not so sure. It adds complexity to a circuit that although should be simple, we've had enough problems on the stripboard to prove otherwise.

This will be the last you will hear from me until the 21st September as I'm off on a break. Hopefully Phil will be here to keep you upto date on the latest whilst I'm away. See you in a few weeks!

Back to basics...

...back to the breadboard. I've rebuilt the receiver on the breadboard and all appears well. Why the circuit never worked (properly) on stripboard is a mystery, but it confirms that we weren't doing anything very wrong - and that the components are working.

One of the things Andy wanted to work was a flashing LED to show that the device was in fact receiving data. This was lacking from the original design and so never made it to the stripboard version. Now that we are back on the breadboard I've had a chance to try a couple of things, the result of which is... a flashing LED when data is received.

With both circuits fully working, we need to make an effort to get these into their 'production' state. In the meantime we are waiting for our camera to be returned as it died shortly after arriving and before we could even load the new firmware - let alone test it.

Troubleshooting: Day 3

Another frustrating day. Everything that can be tested has been tested, yet we're still no closer to finding out why we get dropped packets when the eval board works perfectly (see yesterdays post).

It's looking more and more likely that the receiver will need to be on a proper PCB. We always knew the transmitter would have to be completely fabbed from scratch as the strip board would be unsuitable. To be honest, once the board is printed, its a lot easier to work with a proper PCB - its just getting to that stage is not completely straightforward.

To make things just that a little bit harder, I'm going to be away from the middle of next week for a fair while. Andy doesn't return until after I leave which means virtually all the electronics work will come to a halt, at least for a few days. Hopefully I'll be able to get enough prepared so that he'll be able to continue with the radio fabbing whilst I'm away!

Troubleshooting: Day 2

The eval kit from Radiometrix arrived today allowing us to work out which board was faulty. It turns out that it is a problem with the receiver which means a more difficult diagnostic process. So far we have tested the Rx module, the IC and the power regulator - all which appear to be fine. I now suspect that the transistor that manages the RS232 to TTL conversion could be faulty - possibly heat damaged in the soldering process. We have learnt that it might be wise to load the Rx module in a SIL carrier - as its a real pain to keep re-soldering.

One bit of good news though - another range test was conducted using our transmitter on the breadboard and the eval receiver board. The transmitter was still inside our building and the helical antenna pointing horizontally rather than vertically. We were also in a highly urbanised area, which we knew would dramatically reduce range. In spite of all this we still managed to transmit an impressive 4500 feet. With so many improvements yet to be implemented (fabbed board, half wave dipole at transmitter end, elevated antenna, etc) we expect to be further impressed in future tests.

Troubleshooting: Day 1

Last night I noticed that the NBEK IC did not have a decoupling capacitor right next to it. Whilst there is one on the Vcc rail, it really is located too far away to be effective at controlling ripple. It was a long shot, but I thought this might actually cure the problem. It didn't.

We've also taken the null modem cable out of the equation as well as the RS232 to TTL converter on the transmitter side.

More kit allowing diagnostics turns up tomorrow.

Missing data

We've hit a snag... What had been working, now seems to have gone a little wrong.

We are currently investigating why we are losing data packets over the radio link. As yet we cannot be sure if this is due to a problem with transmitter or receiver. The only real change has been that the receiver is now a soldered circuit. This is a little unexpected as the fabbed board should make the receiver better, not worse. Another theory is that the transmitter is causing an issue due to it being on a breadboard either because of interference or inductance. Luckily Radiometrix have offered a loan of some evaluation kit which will pin down exactly where the problem is. Once we know which board has the problem we will know how to proceed.

Phil has bet on the receiver being faulty, whereas I think its going to be the transmitter. I hope I'm right as the fix will be simply to fab the board. If Phil happens to be right, that means we need to examine our finished receiver... not something we would want to dismantle!

GPS update #2

The device turning up so soon caught me by surprise and so was completely unprepared for testing. However Phil and myself went up the road and picked up the required components and have now completed the following tests:

1. Plug in and check data is received -PASSED
2. Set to 1200 baud and check data is received - PASSED
3. Disconnect power, reconnect power and ensure data rate still 1200 baud - PASSED

The next test will be to check that we can send (and receive) this over our radio link.

GPS update

As you can see from the photo the GPS device has already arrived! US to UK in a weekend!

GPS ordered & end of week update

A Garmin GPS 15 L has been ordered for testing with our transmitter. We should be able to hook it up to a spare prototype board to program it before hooking it up to the test transmitter circuit. This device is yet another item not easily available in the UK (or the EU for that matter) and has had to be purchased from the United States. Not that I have any problem ordering from the States, but the long lead times as a result of international shipping can be a little frustrating.

Next week we will be testing the camera. Getting deja vu...? I did report that last Friday too, but due to Phil's holiday we have been unable to progress that any further this week. Phil hasn't been completely away from the project though - he has been working on the Zephyrus 1 capsule. Hopefully there will be more to report next week on this.

Andy has been doing further development on the back-end of our tracking solution (as well as adding and tidying the gadgets and layout of the blog). Although the main component is functionally complete, he has set some additional goals beyond what has already been achieved. The new target is a full dashboard - effectively a head-up display for Zephyrus 1. I can't tell you how detailed the HUD will be, but potentially it could allow us on-the-fly predictions of burst point and subsequently landing point. As Andy is away for two weeks, I'm afraid you'll have to sit tight for further information.

Tracking software completed

In a week which has seen a lot of team members off, progress has slowed down somewhat - this slower pace is to be expected for the following 5 weeks due to holidays. This had already been accounted for in the launch date, so no targets have slipped.

Despite this, I'm pleased to report that Andy has completed the tracking software. Anyone that has access to Google Earth and has been given the URL for our tracking page will get live updates of Zephyrus 1's position. The link also works in Google Maps on Windows Mobile devices, although a manual refresh is needed (Google Earth automatically updates).

Launch date

Very much like a NASA shuttle launch, we are unable to give a definate launch date as we are at the mercy of the weather. High winds and/or rain will force us to scrub any launch attempt as would any equipment failure on the day. What we can do, (again like a shuttle launch), is give the earliest date that a launch can take place.

The 'not before' date for Zephyrus 1 is 03 October 2009. Make a note in your diary.

Receiver update

Our new receiver has now been soldered up and tested - and it works! This is largely down to Andy's impressive soldering skills (the IC was pretty much machine quality!) and despite the fact I'd put the voltage regulator the wrong way round - though, in fairness, the Micrel regulator device was counter-intuitive.

Next week we will be begin working on the transmitter, which won't be on stripboard like our receiver, but will be a proper PCB. There are a quite a few reasons for this, but weight and size are a major factor. We will also be taking our first look at the Canon A560 camera and seeing what its capable of.

New board and an update on software

As I suspected, the stripboard could not take the pain of further re-soldering, so a new board had to be acquired. This time the whole board is being laid out first, (as you can see from the photo), and will be soldered in one go. I’m fairly confident that the major components (the IC’s) will have survived – but we won’t know for sure until tomorrow. Lets hope Friday is third time lucky.

Andy has been putting in the hours on the tracking software and the latest news is that it can now error check the data that is received from Zephyrus 1. During the flight it’s likely that the radio range will be close to its limit and, as a result, some information may be corrupted. In that instance, we don’t want to be chasing shadows – we want to know for sure that the data we get back is good. Thanks to Andy’s software, any points plotted on Google Maps will be pre-checked giving us full confidence in where our payload is.

The joys of resoldering...

As I predicted, a mistake had been made. After a resolder and retest it transpires my first fix had caused a short circuit. I've corrected this and checked over the board, which now seems okay. A test will be carried out tomorrow to see if I'll need to get a new bit of stripboard - as further resoldering is probably unwise.

Meanwhile, Phil has been investigating parachutes and Andy is developing tracking software that will, among many other benefits, allow multiple recovery teams access to live location information.

Receiver complete

The receiver is now functionally complete (although the power switch has not been wired) and will hopefully be tested tomorrow. I fully expect to have made a mistake somewhere!

In other news from today, Windows Mobile continues to impress. One of our major concerns is that GPS equipment will fail due to ITAR/COCOM restrictions. Thankfully a free (and powerful) scripting utility for WinMo allows us to restart the GPS whenever we like. I'll cover what we are using, and how we are using it, in a future post. No matter how good the radios are, Windows Mobile is booked on the flight.

Status update

I thought it might be appropriate to show exactly what we have and where we are with the project.

Balloon - FLIGHT READY
Capsule - DESIGN PHASE
Parachute - OUTSTANDING
Phone - FLIGHT READY
Camera - PURCHASED; AWAITING DELIVERY; FIRMWARE TO BE LOADED & TESTED
Transmitter - PROTOTYPE PHASE; FIRST TEST COMPLETE
Receiver - PROTOTYPE TESTED OK; BUILDING PRODUCTION VERSION
Rx Antenna - COMPLETED BUT NOT FULLY TESTED
Tx Antenna - OUTSTANDING

Voltage regulator

The voltage regulator is the first part to get soldered in place. Doing so allows me a chance to brush up on my precision soldering without risking costly components. It's also a good candidate as it can be tested quickly and easily. Everything went to plan and it checked out perfectly providing exactly 5.00 Volts.

A better receiver

Although the testing phase is incomplete, to actually conduct these tests more easily, we need to have a 'finished' receiver. Try holding a breadboard, power pack and laptop still on your lap whilst holding a metre long antenna out of a car window!

The prototypes work, but are fragile and cumbersome. Having the receiver in a single, easy to transport and handle unit with sockets for antenna (BNC) and serial (DB9) will make subsequent test runs much smoother. As you can see here, the work on such a unit is just beginning.