Halo Lights

I was just thinking about you - I hadn't seen your name pop up for a while on here.I hope you are well. Please could you email me your email address - I need some urgent advice but it's something that is and isn't to do with

4wd and you would be the only person to know. snipped-for-privacy@ntlworld.com many, many thanks.

Gaynor

I ordered the Crystal 7" lamps for the Defender (Look exactly the same as on the 2003 model) from the same chap with xenon H4 bulbs and ICE white sidelight bulbs. They took only minutes to fit and look really smart, the xenon light is very close to a daylight spectrum and with the perfectly clear glass (diffusuion & dip pattern is achieved on the backing chromed plate) the amount of light they put out on full beam is amazing, several times better than my old standard Halogen lamps but a lot of the chrome backing had rusted off those!

When I rang the chap up who works for a company called Mini Trailers in Ilkley he said that there was a wide range of lighting acessories he had in the works for the Defender. These included remoddled headlamp finish panels doing away with the small sidelight (sidelight is built into all their 7" lamps) and replacement modernised indicator units in a similar style to the new Range Rover and 2003 Discovery. He said a modernised front grille was also being developed which might incorporate additional 5 or 6" driving lights. Replacement rear lamps were also on their way apparently.

I know a lot of you will be thinking these sound like Max Power mods for a Defender and it won't be long before we get tea tray spoilers and chrome exhaust extensions, but with the amount of problems I have had over the years with loose connections, bad contacts and water ingress on the standard lighting I'd be very keen on trying a new modern design, as long as they dont look too hideous anyway. The existing lamps are only generic/trailer board units after all.

The chap at Mini Trailers recons all these items will be the must have accessories in 6 months time so I suppose we just wait and see what develops.

Fergus

Whats the weather like on Mars?

Julian

--------- = Pretentious Sig required =

Don't forget to wire your front fogs so they come on with the ignition, then you can wear your baseball cap backwards with pride

:D

Alex

Eh? Do you really mean me?

The email address in the headers is valid...

Martyn

Rewd Dewd!

Martyn

I just want to be able to see where I'm going. 101 lights are not renowned for their rabbit stopping properties...

Martyn

Just sent you an email Martyn.

Gaynor

An equally useful upgrade would be to attack the wiring. Since changing all my wiring and fitting relays, the light output in vastly better. I'm now getting >13V at the bulbs with the engine running which is around 90% efficiency. There is a Hella report somewhere that states that the following voltages at the bulbs result in the stated light efficiency....

13.5V 100% 12.825V 83% 12.15V 67% 11.475V 53%

I was getting around 11V at the bulbs before the exercise!! I'm about to carry out the same exercise on the 130.

Malcolm

On or around Thu, 8 Jan 2004 14:32:56 +0000 (UTC), snipped-for-privacy@cix.compulink.co.uk enlightened us thusly:

interesting point. The implication is that their nominal 12V bulbs are actually designed for 13.5V. You'd need a *very* sound system, bearing in mind that most alternators seems to run at about 14V, to get 13.5V at the bulb.

One of the most relevant points for a relay is doing the main lighting supply switching, IME. Often the switch is the weak point.

Cue my idea for rewiring a car at some point (possibly skippy).

Rather than the current wiring look you put intelligence at each corner - I'm planning on a PIC based unit with multiple relays. The dashboard talks to the remote units over an RS485 bus and just tells it "Light 1 on, light

1 off" for the indicator with a preset series of light settings for each possiblity.

Powere would be delivered via a common power bus which would run right round the car. Because you're only doing one power run to each system rather than

5 or 6 in the case of tail light clusters you can use much heavier cable to deliver the 'leccy and get major improvements in efficiency. Add a single loop for the RS485 and you have a complete wiring setup for your car that uses two distinct cables. The RS485 would probably be run on Cat5 STP in this scheme to give it shielding as well as the built in noise cancelling inherent in 485. Running a loop would make it self healing to some extent.

I figure that the parts for the remote units shouldn't run to more than about 25quid a corner and should more than pay for themselves in the simplicity of the vehicle wiring.

Of course, I first need to do my in-dash computer to drive all of this and to turn button pushes into RS485 signals, but that's well under way anyway.

I just have this vision of me retrofitting this to Skippy and having a 20 year old RRC with a dashboard that wouldn't be out of place in KITT.

I'd probably run one additional cable around the car hard wired to the relays for the indicators to allow a purely manual hazard warning system.

P.

Oh dear. Remind me of the staggering lack of progress on my Digital Dashboard Project then why don't you...

It is a legal requirement for there to be a visible indication that an indicator buld is broken - which is why cars still use the shitty unit which clicks and flashes fast if a bulb is out. Even using a phase generator is classed as illegal (been down this road when I put my LED indicators on Grumble).

This is still a fecking good idea and leads to a lot of other possibilities. I had previously thought about USB / Network camera for reversing, ultrasonics, IR and an array of other useless essentials.

Touch screen backlit TFT monitor and a basic housing is about as far as I've got so far. Oh, and a knackered old laptop with a basic RedHat install. The D2A2D stuff is what making my hair migrate at present ;-)

I've lost count of the times I've thought about this (aloud actually) and the times the likes of Hobbs and Davies have given helpful words of encouragement such as "it can't be done" or "can you build a GPS mapping system in?"

This may be the workaround for my initial comment about indicators.

One final thought. A GPS is legally fine for indicating speed, however you need to have a means of measuring total milage. This is near fecking impossible with a homebrew digital system.

Martyn

[SNIP]

So you have an active test that triggers back to the dash. My Rover 75 has this as did my XJ-S. It's just a case of periodically polling the bulbs at too low a voltage to actually illuminate them and then see if you get anything back. I already have a 2 way bus available, why not use it to report the status of bulbs. Poll them once every 5 seconds when they're and everything's hunky dory.

Again, the 2 way bus is there and the data volumes are reasonably low. I wouldn't put a camera on this bus, but you could also run a separate high speed bus with the A bus being low speed but guaranteed latency stuff and the B bus being high speed but non-guaranteed delivery. Ethernet is fine for this.

Which D2A2D stuff in particular? I can't see many places where this would be necessary in this sort of rig, but I'm sure I've missed something. I've got a couple of PC104 boards and some interface cards for this sort of thing (VGA->Composite convertor, 48 channel digital I/O card, CF drive card, PCMCIA carrier). The monitor is a 4.6" unit I was planning on using along with a couple of 2x40 LCDs to replace the conventional dash. I was planning on keeping the rest of the controls as vanilla as possible, just attaching them to different things. I was also considering voice feedback.

"Yes, you can. It's far easier to just make the GPS output common and let something better than you can write yourself do it".

I'm planning on teeing the output from the GPS into the onboard 'puter for tracking and using the PocketPC system I have to do actual complicated things like map display and direction giving.

It's a legal requirement to have an odometer?

Piece of piss to do really.

You modularise.

All you do is build a unit which is fixed purpose. It counts things. Specifically it counts pulses read from a magnet strapped to the propshaft, probably read by a hall effect sensor. You find out how many turns of the propshaft equate to one mile travelled and have this dedicated purpose module emit a pulse every one mile.

Your onboard computer can read the pulse and this go "one mile covered", however you also have the pulse going to one of these:

which can be mounted somewhere you can get easy access to it, say in the instrument binnacle under the cover.

Have the sender unit for this have its own backup batteries - a couple of NiMH AA cells should be enough - and you have a perfectly good digital odometer that keeps a running tally up to just shy of ten million miles whether or not your computer is working.

Can't see the parts for this costing more than about £15 at absolute worst case, and Tandy don't exist any more.

P.

On my previous RR (with VERY dodgy wiring) I fitted a pair of relays close to each headlamp unit with separate fused feeds to the nearest battery. The original lamp feeds then simply switched the relays. Nice!

Richard

Don't see this as difficult (but convince me otherwise)?

If you didn't insist on using that bloody Linux stuff I could be of more practical help (and probably up for a collaboration).

Wot t'other chap said I guess. How do electronic systems in cars work?

CANBus is surely the way to go?

I really do like that idea - saves all the aggro of a complete rewire I suppose.

Showing my ignorance, let's postulate that I have 9volts at the headlamps of the 101. Am I right in saying that the low current side of the relay isn't going to be too worried about voltage I put through it, and a nominal 12v relay would still be switched by the 9v? Or do I have to measure it and choose a relay accordingly?

[SNIP]

Which is why you do a network device. You let various systems play to their competencies. Dedicated purpose systems where they make sense, Linux where it makes sense and other things elsewhere. You separate the critical stuff from the presentation and have the simple stuff done by simple kit.

Probably pretty much the way I suggested except using an RRPROM in the ECU rather than a mechanical counter.

Damn expensive way to go. RS485 has most of the same benefits and doesn't cost as much. CANBus is (I believe) still patent protected. RS485 isn't.

And there's several guys at work who know RS232 and RS485 inside out..

Given that this discussion has strayed *wayy* off list topic, and that there is a definite "all talk, no trousers" thing going on here, shall we take it off to mail.

I've set up a mailing list just for this.

send a mail to:

snipped-for-privacy@geekstuff.co.uk

with

subscribe landietronics

as the first line of the body. You'll be asked to confirm in a followup mail.

When you get this mail strip everything out of it except the "auth whatever " line and make sure you don't have any quote characters before the auth

Send it back to snipped-for-privacy@geekstuff.co.uk and you'll be subscribed.

P.