Sinclair C5 Indicators

A while back, I bought a Sinclair C5.


I love it. It’s an amazing piece of history and something really unique. Sure, it’s not that practical so I don’t drive it every day, but it’s great fun to take out for a spin every now and then.

I’ve got various projects on the go to improve it, and one of them was making indicators for it.

The original indicators looked like this:

The front ones are fairly standard but the rear ones are specially made for the C5 – as you can see they fit flush against the curved edge. The problem is, these original indicators are very rare and expensive nowadays — they can easily fetch over 100GBP for a set.

So I set about making my own. My friend Eddie Tindall, very talented designer, recommended that Renault Megane indicators would go quite well with the style of the C5. So I bought some of those and pryed the plastic diffusers free from the backings. The glue was quite brittle, so they came off ok with a little bit of care.

The reason for doing this was that I wanted to use LEDs rather than the original incandescent lamps. I wanted to avoid retrofits though because I thought they probably wouldn’t be bright enough.

I hacked up a quick PCB in EAGLE CAD. It was really rough and dirty — the polygons were all hand drawn and had funny shaped gaps between them, etc. But in the spirit of my previous post, I didn’t want to bother spending ages getting it perfect. Once I was fairly happy the design was OK, I sent it off to Smart Prototyping in HK for manufacture. 18 days later, the 5 copies of the PCB arrived.


Each board is for two indicators — one left and one right, since the two indicators are mirror images of each other. You’re looking at the front of one indicator PCB and the back of the other here. They needed to be guillotined in half along the silk screen diagonal line. The circles on the top right of that image (also present on the back at the bottom left) are my idea for thermal management: I calculated that if I soldered coins to these pads (1p coin for the circular ones, 2p coin cut in half for the semicircular ones), it would give the indicators enough thermal mass to keep cool in normal use. This is pretty essential, since the LEDs I was using would be dissipating quite a lot of power.


Here’s one of the boards soldered up, from both sides:

I couldn’t find a hacksaw, so I just blobbed a load of solder where the 2p halves should have been for now. It’s not quite as effective, but it might be good enough. It will run for quite a few seconds before the 1p coin becomes noticeably warm, but the solder blobs heat up a bit quicker. Soldering this was fairly easy since I have a ~90W iron, but would probably be impossible with a cheap maplin iron.

Here’s a picture of one working, hooked up to a 700mA constant current supply.
I’m not entirely sure how to measure the brightness, but they were bright enough to give me spots in my vision. For comparison, the backdrop of that picture is in mid-evening british summer sunlight and the exposure was apparently f/2.8, ISO 50, 1/250 seconds.

I’m reasonably happy with how these turned out. I still need to work out how I’m going to fix the PCB to the diffuser (probably some sort of high temperature epoxy), and then fix them to the C5. But before I can do that, I need to get the rest of the electronics done — and that’s a job for another day.


Perfect is the Enemy of Done

I’ve finally gotten round to re-hosting my blog — I was previously using Gandi‘s Dotclear installation which came free with my domain name but I got fed up of it because the comment spam filtering was useless and I found the markup difficult to use. The old blog is currently still hosted at but may move in the near future. This blog is now hosted by the lovely people at the SRCF (full disclosure: I’m on the committee of the SRCF).

I’ve been intending to re-host it for a while but I spent a long time deciding what platform to use. I ruled out wordpress early on because it was too bulky and slow — PHP and MySQL, really!? For just a blog!? I tried a few static content generators but ruled those out for some reason or another too, and I eventually decided that this was a job for later, when I could find time to do it perfectly.

And that’s the problem. Recently, I realised that a lot of my projects end up unfinished because I can’t do them perfectly. My bench frequency reference, my reprap, my Sinclair C5 indicators (I’ve bought a Sinclair C5 since my last post — more on that another time) and various programming projects — all ended up coming to a stall. Or if for some reason I absolutely have to finish something, I tell myself that there’s no point in doing it just alright — if it’s not going to be perfect anyway, I may as well put the absolute bare minimum effort in and do a horrific bodge of it.

But enough is enough. As they say, perfect is the enemy of done. Over the last week or two I’ve been trying to get myself into that mindset. I’m going to start finishing some projects which have dragged on for too long. I may not be able to do them perfectly but that’s fine – I’ll at least get them done. First up was migrating my blog. WordPress may not be perfect, but it will get the job done. So I’ve used that and now the job is done.

To help me keep this in mind, I made a motivational poster to go on my wall — I took a piece of scrap paper, scrawled “PERFECT IS THE ENEMY OF DONE” on it and taped it up in front of me with parcel tape.


It’s not perfect, but hey — it’s done.