Where’s the best place to take a live feed from?

[Paulfc]

Hi,

Some guidance please. I’m thinking of installing relays to ease the load on the headlamp circuit/switch with, hopefully, an improvement in brightness. Never used relays before and would appreciate any advice as the where best to locate them, where to take the switched current from etc. I’ve seen the battery positive terminal and the alternator mentioned as suitable source points. Also, what cable rating should I use? Plus, of course, any associated do’s/don’ts, product recommendations, advice on fusing and so on.

Many thanks.

[Clive]

Each car is likely to be different. But on my spitfire, I located the relays behind the ns headlight, and ran a 30A cable back to teh battery positive, well, I have an extra fusebox, so it ran to that and is used at 30A, and that is powered direct from the battery.

I then used the bullet connectors on the bonnet as teh connection points to "trigger" the relays, and then the original looms fro teh bullets as teh feeds to the lights.

For reference, a 55W filament bulb draws a little under 5A. so 2x dipped or main will be 10A, but if you have dipped lights on and use the flash, that will be 20A, so a 30A cable made good sense to me. No doubt 17A cable (seems to be a common size) would be fine, as the flash is not used for more than a second or so.

[NonMember]

On my Vitesse I put the relay (it's for the main beam only at the moment) behind the N/S inner headlight, where the bonnet structure has a convenient threaded hole. I ran the power feed (30A cable) to the alternator as that's a bit closer than the battery. I figure the main beam (in fact, any headlights) should really only be in use when driving, so the alternator is the true source of electrical power. The other connections were made by cutting the loom because that part of it (where it passes the relay) was all falling apart anyway.

My Spitfire was treated to a brand new loom, which had the relays included from AutoSparks. They're mounted on the radiator support frame.

[dougbgt6]

I got one of these, there are tags in the bonnet nose that hold the wiring in place, it fits neatly in there. I extended the power wire and ran it back to the fuse box.

I know Clive had one where the contacts rusted, but I've ever had a problem in 5 years. 

Car Truck H4 Headlight Relay Wiring Wire Harness Headlamp Light Bulb Socket Plug 8051628548476 | eBay

Doug

[PeteH]

So long as the relay is not in a "Wet" site. The cable should be sized as capable of coping with the Amp draw, I would always go one size larger to allow for voltage drop in the cable. (No of bulbs  x Watts /12) Most relays sold for general auto use are 30amp rated, some are fused.

Pete

[johny]

I think the selection of components needs to be in this order:

work out the maximum likely load amps,

the fuse (mounted as close as possible to the supply) needs to be rated a bit higher than the calculated load,

finally the wire should be of a higher current capacity than the fuse (as Pete says you can as large as you like to reduce volt drop).

This might seem a bit ott but if an earth fault occurs to the wire the fuse must blow before the wire overheats...   

[Paulfc]

Hi All,

Many thanks for your advice and guidance, I’m much clearer now on what and how to tackle things.

Paul

[68vitesse]

Don't forget the earth return cables also needs to be upgraded to get the best result.

Regards

Paul.

[Clive]

4 minutes ago, 68vitesse said:

Don't forget the earth return cables also needs to be upgraded to get the best result.

Regards

Paul.

Good point. The bonnet loom/bullet connectors are not great, and often cause issues. Overcome by careful cleaning, and wise to put prtective spray wax/grease over the connectors to keep them corrosion free. 

[Ian Foster]

Paul

A search on previous posts (search relays) shows previous discussions.

This is one thread 

 

Ian

[Paulfc]

Thanks Ian.

[Wagger]

Sorry to chip in so late on. You may find that your bulbs blow more often as there is less series resistance. The filaments have a very low cold resistance, therefore quite a high current flows at switch on but for a very short time. This can blow a fuse.  My Vitesse main beams glow very dim on dip to pre-warm the filaments ready for full current. Somebody before me did a mod.

Way back I had a 6v BMW car with 60watt headlights. That used relays and 6v at 60w is 10 amps. It managed 6v at the bulbs when running and a mere 5v when not.. My pal fiited these to his 6v beetle and only had 3.5v at the bulbs.

I added relays for him.

I would run feed and return in 20 amp cable and use the existing wires to operate the relays. Put a 20 amp fuse in the feed. Remember, fuses are to protect wire, not the load. They have inherent resistance, depending upon size and type.

[NonMember]

2 hours ago, Wagger said:

Funny stuff is Tungsten. It is much brighter at 14 volts than at 12v.

Nothing to do with tungsten, that. Simple physics. First off, it's the power that matters, so you need to square the voltage - and 14V gives a good 35% increase over 12V. Then there's the "black body radiation" formula (a bulb filament is a black body for the purposes of this) which makes the peak energy "colour" depend heavily on temperature. That 35% increase in power shifts the temperature up from where it's nearly all infra-red light to a more visible-light-centric spectrum (though still very much at the red end - you need to be as hot as the surface of the sun to get your peak emission in the middle of the visible octave).

[PeteH]

Using lower voltage requires heavier cable too. US 110/115V wiring is far heavier that UK 230/240 . Most HGV`s are 24V. Which would indicate less voltage drop in the cable over long runs.

Pete

[Wagger]

If you have all switched on (heater lights accessories etc) then the voltage at the point at which you have chosen to take your feed from may be low. (12v). When just using the lights, it could be 14v.

If any of you have the kit to measure volts and current, at the same time. Switch your lights on and measure the current drawn by just one 50 watt bulb and the voltage across it, without the engine running.

Now, start the engine and measure those again. You will find that, if the voltage increases by 10%, the current will increase by less percentage.

 

[Paulfc]

Fascinating stuff this physics! Thanks to all. Genuinely, I now have several evening’s research and study to digest and properly understand those latest posts.
 

Unrelated I know, but am I better off, should I opt to, installing modern sealed beam units or replaceable lamp type lights?

Thanks.

[PeteH]

2 minutes ago, Paulfc said:

Unrelated I know, but am I better off, should I opt to, installing modern sealed beam units or replaceable lamp type lights?

For and against both. But the replaceable units (Should) have the facility to take Halogen, should you want to?.

Pete

[NonMember]

2 hours ago, Wagger said:

I have replied to 'Nonmember' above separately.

And I've replied saying Wagger is mistaken in his physics. Tungsten has the same resistance characteristics as all metals. The reason it's chosen for light bulbs is its very high melting point. I do agree that its resistance rises with temperature, so the 35% gain is, in fact, slightly less than the simple square. But not by a huge margin.

[Chris A]

2 hours ago, Paulfc said:

Fascinating stuff this physics! Thanks to all. Genuinely, I now have several evening’s research and study to digest and properly understand those latest posts.
 

Me too, in fact this is the only physical exercise I like 

[Wagger]

I measured the resistance of three coiled filament bulbs.

A 12v 55w car headlanp bulb. This measured 1 ohm.

If it had this resistance with 12 volts applied then it would draw 12 amps. 

Next was 230 volt ac 15 watt lamp. This measured 343 ohms, that would result in a current of 0.67 amps.

Lastly, a 230 volt 42w lamp gave a reading of 93 ohms. That would result in a current of 2.47 amps.

 

OK, I measured all of these at room temperature, so we need to know the operating temperature of the filament, and I should repeat this at a range of increasing ambient temperatures.

When choosing switchgear, we must consider this phenomena, especially if there are many lamps connected in parallel. Larege currents do flow for a short time and could trip protective devices such as MCB's.

[NonMember]

7 minutes ago, Wagger said:

I measured the resistance of three coiled filament bulbs.

<snip>

So, was I wrong about Tungsten?

Yes, and you were wrong in your experimental method. The cold resistance of bulbs is irrelevant because ALL metals exhibit a positive temperature coefficient of resistance. As a simple "idiot's explanation" (i.e. not the actual truth but close enough to get a feel) - when they get hot, the atoms jiggle around lots and keep knocking the electrons off course, so it's a lot harder for the current to flow, which is more resistance. That's true of ALL metals. Copper wire does it, too, it's just that you don't usually heat your copper wire until it glows white hot (because it would melt).

Also, stop talking about inductance, it's practically non-existent. The only reason filaments are coiled is to make them longer, so their resistance isn't too small.

[GrahamB]

Quoted from the BBC Bytesize revision site.

If the temperature of a metal conductor increases, the ions of the metal vibrate more vigorously.

This increases the number of collisions between the free electrons and the ions.

Hence, for a metal, resistance increases with increasing temperature.

Often the increase in temperature is caused by an increase in current.

An example is a filament lamp.

As current increases, the filament gets hotter and the resistance of the bulb increases.

This applies to most (all?) metals so the only "unique" property of tungsten is its high melting point which makes it suitable for filaments.

The opposite applies to semiconductors where the resistance decreases with increasing temperature resulting in blown output transistors on your audio amp if there is an insufficient heatsink.

[chrishawley]

Halogen sealed beam v. H4 units (halogen):

You might find the deciding factor is aesthetic. On 'our' cars OE fitment would have been Lucas with a very pronounced dome to the lens. It's quite hard to find H4 7inch units which have enough curvature, and suitable lens pattern, to look right (or right enough). But with halogen sealed beam there's a decent chance of finding NOS or pre-owned Lucas, Osram or Wipac units (for examples) which have a more authentic appearance. Also be wary of cheapie/bargain price H4 units with claims to be 'universal' - which can amount 'universally doesn't fit anything' by virtue of the casting of the three lugs on the rear the being incorrect. But if you go sealed beam then it has to be halogen SB as standard pre-halogen SBs will more than nullify the effects of your wiring upgrades.

As an aside I'd mention not everything about headlights is watts and lumens. It also depends what direction those watts are pointing in (if a watt can 'point'!). My current project has some pretty average H4 units - a couple of hours spent adjusting then on the fly on unlit roads produced a far better, useable, result than was the case with the nominally correct alignement.  Condition of alternator and battery makes a difference as well.

Have you decided where to take the main power feed for the revised circuit?

 

 

[NonMember]

38 minutes ago, chrishawley said:

standard pre-halogen SBs will more than nullify the effects of your wiring upgrades.

I'd disagree on that (though everything else you say is good). Improving the wiring to reduce the voltage drop will definitely give a marked improvement even if you stay with the original sealed beam units, because they'll be working properly. In fact for a car that's not had any attention to its electrics in fifty years, the wiring improvement with standard headlights will usually give more benefit than changing to halogens with the original wiring.

[Wagger]

I do like graphs. Should have looked it up.

In 1995, I was tasked with making kit to measure anything that was powered from a 230v supply. (An advanced version of a Portable Appliance Tester). I was asked to measure the earth continuity, the load resistance, the insulation resistance at 500v DC and Leakage at 2500v ac. Then the power consumption in VA and Watts with Power factor also.

Luminaires (Lighing units) were a pain. Bulbs look like the load is short circuit, as do transformers. The fluorencents appear open circuit as do the first energy saving bulbs. LED's were not around then.

I had to add methods of measuring the mains source impedance and integrity. Ended up having to control every source in order to get repeatable results withing + & - 4%. It cost a fortune and took three of us two years to produce. We only ever marketed a cheap crappy version.

Learned a lot and remembered some of it wrongly. Oh Well!