Aristides,
“I mounted the relays on the battery support, very convenient as it’s really close.”
The XJ-S battery is in the boot, right? How do you get that bundle of wires up to the cowl?
Ed Sowell
'76 XJ-S coupe, red
Aristides has an XJ12 with the battery under the bonnet/hood.
Andy.
Yes, I forgot you guys have the battery at the rear end of the world…
Aristides
If you measured 12+V at the motor, we’re clearly not talking about the same thing. As a reminder, that voltage needs to be checked while the motor is running and trying to move the wipers across the windshield. If you get more than 8V with the stock wiring, I’d be very surprised. 12V just simply isn’t possible, and might not even be possible with the relays added to the system!
I mounted the relays on the
battery support, very convenient as it’s really close.
I’m guessing this is a saloon? Because the battery support for the XJ-S is at
the other end of the car!
– Kirbert
Let me restate my years-old recollection as getting readings that were not very much below normal system voltage. To be perfectly honest, I don’t even remember if it was with the engine running, or just using battery voltage. So, I’m not going to quibble on absolute values. But I do take issue with your “8 volts” speculation.
Here’s the way I look at the issue.
It so happens that I have measured the resistance from the ground pin of the Electrolux motor to various pins in the motor bulkhead connector, and made notes of the resistances. Therefore I can say that the resistance from ground to the motor Slow pin Rs = 1.2 Ohms, and to the Fast pin Rf = 1.5 Ohms (I may have these two reversed but that isn’t important here). Now, about the harness wiring. Let’s assume 10 feet of 18 AWG wire, for which the resistance Rw is about 0.064 Ohms (from Google… I assume it was for stranded copper). Focusing on the Slow mode, the wire and the motor windings/brushes are basically two resistances in series, wit 12 volts at the Slow pin and0 volts at the ground pin. We’ll call the voltage at the Slow pin Vs. From elementary circuit theory Vs calculated as below, where I is the current:
Vs = 12 – I*Rw
= 12 – (12/(Rw +Rs))*Rw
= 12 *Rs/(Rw + Rs)
= 12*1.2/(0.065 + 1.2)
= 11.38 v
Similarly, Vf = 11.5 v.
This ignores resistance of the switch contacts and connector contacts, but I don’t think those factors would bring it down to 8 volts. Moreover, your earlier justification for the relays was the high resistance of the wiring.
Ed Sowell
'76 XJ-S coupe, red
That’s what I think too, but the OP is asking about an XJ-S.
Ed Sowell
'76 XJ-S coupe, red
I haven’t measured resistances before and after the relays installation, but 1.5 Ohms total resistance would not explain the difference in the wipers speed, and it’s is quite substantial, before they were crawling, now they are actually moving fast ! Same thing with the electric windows… before I installed the relays the windows would barely go up, very unJaguar like, now they go up with almost twice the speed…
And if space is a problem you can use these relays (no affiliation):
They are cheap, they work very well and they are really really small, 16 x 12.5 x 14mm, rated at 20A, I was able to squeeze eight of them under the center console !
Best,
Aristides
Therefore I can say that
the resistance from ground to the motor Slow pin Rs = 1.2 Ohms, and to
the Fast pin Rf = 1.5 Ohms (I may have these two reversed but that
isn´t important here).
I didn’t run such calculations, I just measured the voltages at the motor when
running fully loaded. I suspect the problem may be that the motor pulls more
current when loaded, but whatever, adding relays to get full power to it
GREATLY improves its performance. At least it did on my '83, and others
who have made similar mods have reported similar experiences.
Relays added to the headlight circuits and the window circuits reaped similar
benefits. The headlights weren’t as far down on voltage without the relays,
but light output is HIGHLY responsive to voltage and gaining even a half a
volt at the headlight terminals yields dramatically better illumination.
It’s also perhaps worth noting that apparently Jaguar agrees, and with the
facelift started using relays to control several of these high-load accessories
from the factory.
– Kirbert
Kirby,
“I didn’t run such calculations, I just measured the voltages at the motor when
running fully loaded. I suspect the problem may be that the motor pulls more
current when loaded, but whatever, adding relays to get full power to it
GREATLY improves its performance. At least it did on my '83, and others
who have made similar mods have reported similar experiences.”
Interesting, Kirby. Do you remember the details of how you did the test? E.g., where did you measure the voltages? I did my measurements at the wiper motor harness using insulation piercing test probes. To do this the grille/wiper assembly had to be removed from the car, so the motor was just working against friction in the linkage mechanism which had just been thoroughly cleaned and lubricated. However, it necessarily did not see the load due to wiper blades against the windscreen.
It occurs to me that you might have found some place to measure voltages while the wiper blades were in action, i.e., fully loaded conditions. Perhaps that explains your lower readings. Check this out:
The upshot is a DC motor is very sensitive to voltage, so high resistance in the wiring or switch could cause slow running. Moreover, over time low voltage can damage the motor windings.
I notice that in the early XJ-S the wiper is fed through a 35A fuse (#11), much more than indicated by the static resistance based calculation of 10A. In addition to the wiper, this fuse carries only minor loads, i.e., washer motor and a few relay windings. That suggests the design wiper motor current must be, perhaps, 17A. But, even with 17A the wire resistance would cause a voltage drop of only 1.1 volt, and with the full (fuse-blowing) 35A, 2.2 volts.
If my previous calculations have any value at all it is that loss in the wiring is probably not very important. Poor contacts in the switch could be a problem, but I just measured one I have on the shelf and found the resistance between pin 1 (to which power is supplied) and the fast and slow pins with the lever in the respective positions to be lower than my DVM can measure.
I also want to say that the ’89 Electrolux system I recently retrofitted, including a new stalk switch, moves those blades smartly across the windscreen. So far I have had only one opportunity to drive it in the rain, but they seem to be as fast as those on my ’05 Avalon. In the retrofit I have only one relay in a very simple configuration: ign +12 goes to the relay coil pin 86, bat +12 goes to pin 30, pin 85 to ground, and pin 87 goes to the wiper stalk switch pin 1. Thus the switch gets as close to system voltage as can be achieved. However, all paths through the stalk switch carry the full motor current. I have not measured voltages, but based on wiper performance I judge that neither wire resistance or switch contact resistance seem to be a problem.
One final thing. The ’89-’92 Electrolux system draws wiper current through a 15A fuse, suggesting far less required motor torque. Since the wiper-on-windscreen load is the same as earlier models, the gearbox/linkage system has far less friction.
Ed Sowell
'76 XJ-S coupe, red
Do you remember the details of how you did the
test?
Sure. Stuck pins into the wires close to the motor and attached alligator clips
to the pins. Reassembled the wiper grille so the motor would be running
under load. I think I did it on a dry windshield, just to make it a worst-case
scenario.
The upshot is a DC motor is very sensitive to voltage, so high
resistance in the wiring or switch could cause slow running.
It’s not as sensitive as incandescent lighting, but it is pretty sensitive. And as
voltage drops off, not only does torque drop off but efficiency drops off pretty
quickly, too. As a consequence, a higher percentage of the energy is
converted to heat and less and less is turned into useful power.
Moreover,
over time low voltage can damage the motor windings.
That’s well-known among electric motor experts. They will always tell you
that undervoltage is a bigger motor killer than overvoltage. Still, I don’t know
of anyone who has actually burnt windings or the like on a Jaguar starter or
motor by running it on low voltage. The Lucas motors seem remarkably
tough in that respect.
I installed relays on both my wipers and my windows, and in both cases the
motors immediately ran dramatically better, zipping with authority rather than
looking as though they are grossly underpowered. But perhaps worth noting,
they continued to run better and better over the following couple of months,
as though operating on full power was cleaning out the cobwebs or
something.
I notice that in the early XJ-S the wiper is fed through a 35A fuse
(#11), much more than indicated by the static resistance based
calculation of 10A. In addition to the wiper, this fuse carries only
minor loads, i.e., washer motor and a few relay windings. That
suggests the design wiper motor current must be, perhaps, 17A.
There’s no telling what Lucas’ theories are on fuse sizing. I know that, in
general, the rule of thumb is that a motor will pull 5x as much current at
startup (“locked rotor” current) as it will at running load. That gets into
whether the fuses are fast-blow or slo-blow, whether they can take the spike
in current for the instant of startup without blowing.
But,
even with 17A the wire resistance would cause a voltage drop of only
1.1 volt, and with the full (fuse-blowing) 35A, 2.2 volts.
Thinking more about this – and now about 15 years since I sold my XJ-S – I
now recall that, after doing those relay mods, I later discovered that there
were wiring issues throughout that car. It wasn’t the wires or the switches so
much as the connectors. For example, some of the connections underneath
the fusebox in the engine compartment became so corroded that it started
smoking at an intersection (!) and started to melt down the fusebox. It wasn’t
the glass tube fuses or the clips that held them, it was the crimped
connections between the wires and the brass terminals on the underside of
the fusebox. My Jaguar did a fine job demonstrating that the studies that say
crimped connections are just dandy are hogwash. I could have fixed those
particular connectors by soldering, but with the fusebox already partially
melted I opted to replace the entire contraption with a modern generic
fusebox using spade fuses.
We’ve also heard from a couple of members here that the rubber block
connectors used all over this car develop issues inside the rubber block,
where you’d think they’d be protected from corrosion and flexing.
Add the rationale that, despite them being the Prince of Darkness, it’s difficult
to imagine that the engineers at Lucas didn’t go through exactly the same
calculations you did. All together, I’d say it’s quite probable that the voltage
issues in my window, wiper, and headlight systems were more due to aging
and corroded connectors than the length and gauges of the wires
themselves. Too late for me to check, of course, but the fact that others
have reported similar benefits from installing relays would seem to indicate
that I’m not the only one who experienced such issues. Going to relays fixes
the problems, of course, because even tired and corroded connections carry
enough current to engage a relay.
– Kirbert
“Thinking more about this – and now about 15 years since I sold my XJ-S – I
now recall that, after doing those relay mods, I later discovered that there
were wiring issues throughout that car. It wasn’t the wires or the switches so
much as the connectors. For example, some of the connections underneath
the fusebox in the engine compartment became so corroded that it started
smoking at an intersection (!) and started to melt down the fusebox.”
In this connection (no pun intended), I really don’t know what caused my Lucas system to stop, mid stroke, but it was probably the wiring. Since I had thoroughly refurbished it a few years earlier, and still found it to be painfully slow, I wasn’t interested in working on it anymore and set about the Electrolux retrofit. Before installing on the car, I successfully bench tested the motor/gearbox connected to the new relays (power & delay) and stalk switch. Then I installed it on the car using the OE wiring that connects the left side stalk switch with the right side bulkhead connector, plus a couple necessary new wires pulled through the center console. It didn’t work. After much testing I found that the LG wire that provides ign +12v to the switch had no voltage. Since that wire disappears into a bundle behind the instrument cluster I was never able to trace it to the point where it connects to ignition voltage. At that point I simply abandoned the OE wiring and built an entire new wiper harness all the way across. Works perfectly, thank goodness.
I don’t intend to pursue it further, but can’t help but think it’s a failure at the connection point for the LG wire. I might at some point test the Lucas motor. If it turns, that will confirm the problem was in the wiring.
Ed Sowell
'76 XJ-S coupe, red
After
much testing I found that the LG wire that provides ign +12v to the
switch had no voltage.I don´t intend to pursue it further, but can´t help but think
it´s a failure at the connection point for the LG wire.
Hah! That’s gotta be some sort of jag-lovers record in going the long way
around to find a simple wiring fault!
Was your Lucas wiper motor the early solenoid-park system or the later
reverse-park? Either way, the Electrolux is reportedly a considerable
improvement, so you did a good thing even if a quick wiring repair would
have gotten the original Lucas up and running.
It would be interesting to see what happens when people with slow dragging
Lucas wipers revise the power supply alone rather than installing the
complete relay system I devised. Heck, there might be one particular
connection buried in there somewhere that’s causing ALL of these dragging
wipers. I have no doubt that a better 12V power supply – whether by relay,
as you did it, or just a new wire run from somewhere – would make the
wipers run better. The question is whether it would attain 20% or 80% of the
improvement seen with the entire relay scheme.
– Kirbert
“Hah! That’s gotta be some sort of jag-lovers record in going the long way
around to find a simple wiring fault!”
I will bow and humbly accept whatever award is due for this achievement. I am unable to take an easy path if a more difficult one is available J
“Was your Lucas wiper motor the early solenoid-park system or the later
reverse-park?”
Solenoid park. I was so exasperated when it failed so soon after the refurbishment I decided I’d rather put effort into retrofitting rather than trying to diagnose the actual cause. I’m happy with the results.
Ed Sowell
'76 XJ-S coupe, red
Kirbert1 Kirbert
January 5
It would be interesting to see what happens when people with slow dragging
Lucas wipers revise the power supply alone rather than installing the
complete relay system I devised.
Well now,
That first reply did not work.
I think because I did it direct from Gmail, and removed formatting to implement plain text which the old system demanded.
So I will cut and paste from the Gmail “Sent Mail” folder and see if that works.
Still have the Lucas wiper system in the coupe.
The big problem of not parking was eventually fixed by replacing the
Lucas switch on the gearbox with a generic microswitch driving a relay
coil, no longer switching the total motor current.
At the same time cleaned up the motor, linkage and then applied grease
to moving parts.
With this simple fix the wiper motor will probably go on way beyond
the life of the car.
I dare say if you drive in tropical areas with high rainfall the above
statement might not apply.
I do not drive the coupe often, and rarely in rain, but I think the
wipers work a bit better than before.
Richard Dowling, Melbourne, Australia. 1979 coupe + HE V12 + manual;
1989 convertible; 2003 XJ350.
That is a very clean piece of work you did. One less thing to worry about for sure!
Now for my next question. Is there a voltage regulator, reducer or something British that will make some of the gauges read low when it isn’t doing its job? I filled my tank up today and installed a new sender unit. It only read 3/4 full.
Russ
Unfortunately, no. I’ve also replaced the sender and get low readings. I’ve fiddled with bending the arm of the float but no joy. There have been postings here over the years about adding a resistor or something but I never explored further. If the Archive search is working on this new site, do a search.
Ed Sowell
'76 XJ-S coupe, red
To check the gauge ground the sender cable, it must read full.
Even if the float is new check if it’s pierced and that it has no fuel in it.
With the sender out try to see if the gauge will read full when the arm is in it’s uppermost position.
If all is OK, bending the arm must do the job.
If not, the sender resistance might be too high, or you have a problem with the wiring / connections.
Aristides
Well I guessed wrong then. I just put it in and was expecting it to work. The gauge will show full when the wire is shorted and the old one did have fuel in the float. The new one doesn’t have a hollow float. I guess now I should have checked that first before installing it. Oh well, it won’t take too long to burn this tank.
Thanks
Russ