Slow Window Fix for less than $30.00

That's the one!

 

Is no one buying the commercially available (e.g. from Superformance for £60) 3-wire "window booster" combo relays anymore? I'm curious because I've stumbled across a fairly straightforward modification I can make to an existing OEM component to create essentially the same thing, potentially for less money, but I don't want to either 1) undermine somebody's livelihood (something made me think Superformance was sourcing them from a mom-and-pop sort of outfit), or 2) make the effort to produce such a thing in quantity if there's no demand. But I also kind of hate to see people paying $70–100 for what is essentially a pair of $5 relays sealed into a little box...

Certainly the pair of standard off-the-shelf relays (as described earlier in this thread) works just as well as what I'm proposing, but if people are interested in a "plug and play" alternative, I could offer such a thing, in either of two flavors: 1) would be a 6-wire device similar to the Oono F-1020 linked above (in that it switches both power and ground to the motor, bypassing all the existing voltage drops on both sides of the circuit, if installed 'correctly' in the door), but superior in that it's switched by +12V (rather than switching to ground or to the 7th wire of the Oono device) and is thus compatible with the stock switch/wiring configuration; or 2) would be a 3-wire device equivalent to what Superformance et al have been selling for years (i.e. only improving the ground side of the circuit, but offering super easy installation without having to run a new source of +12V into your door [which is what's needed for a 'correct' installation of the Oono device or my 6-wire option #1]).

Any interest in such a thing? I would probably want to charge ~$20 for option #1 and ~$40 for option #2 (which may seem backwards, given the greater functionality of #1, but #2 requires more of my time—probably a lot more than $20 worth, frankly, if I think about it too hard). My windows aren't slow enough to want to open up a door for the sake of experiment, so if there's interest in #2, I'd probably be willing to give someone my prototype for free, or at least on pay-me-later-only-if-you're-really-satisfied terms. (#1 can be tested sufficiently thoroughly on the bench, if anyone's wondering why I'm not making a similar offer for that flavor of the device.)

-Joe

 

It may be possible to use two 10A diodes to complete the ground circuit at the motor instead of grounding through the switch.

 

I'm curious what you're envisioning (it's not intuitively obvious to me how that might work) but, in general, the voltage drop across a diode (~0.7 V) makes them undesirable for applications like this where the whole point is to bypass the various voltage drops that already occur in the wiring of an old car. (I forget if anyone shared any actual measurements earlier in this thread but, for example, the window motor is designed with an assumption that it will see ~13 V; with the length of wiring involved in these circuits and a bunch of marginal connections in an aging car, it ends up seeing only ~9 V. The improved grounding provided by one of these "window booster" solutions improves that to perhaps 11 V, so adding a diode to the circuit takes you back closer to 10 V, i.e. potentially undermining ~half the improvement.)

 

The switch damages easily by the inductive kick back on the negative terminal of the switch, and causes the circuit to become open. Yes, the diode has a .7V junction loss, but it is an easy fix that I have used many times with good results. Voltage drop in the positive lead is another cause of this problem, the diode only works when the switch becomes damaged and the feed voltage is correct.

 

You could also use diodes with 0 voltage drop (need to use a few in //) , but frankly the voltage drop is not an issue: in any case when the engine runs, the voltage raises to 13.6v

The switches are old and non rebuildable, so leveraging the load is a good idea. Despite what was said earlier about my crude relay hack, the windows have been working strongly eversince. So basically any solution that takes the burden away from the switch is a vast improvement.

While you are there, securing the electric mirror switch with a diode, cap, and a ferrite is also a good idea: my amp was going nuts each time the mirror solenoid was triggered. When I opened the switch (the dual mirror switch are rebuildable), I could see the sort of sparks that were comming from the mirrors and destroying whatever electronic was in my car!

 

One other point, charge voltage with the engine running should be 14.2V or so. This will also help the windows run a little better. Voltage drops in these older Ferraris get worse as time goes by, and they could be from either the positive path or ground losses.

 

Oh, as a flyback diode? So you weren't suggesting that adding diodes will speed up slow windows, but rather as a preventative measure that will protect the switches from the wear that eventually produces a big voltage drop across the switch and thus slow windows?

 

Well, the voltage regulator isn't magic--it'll maintain that voltage (actually hopefully 14.3 V) only at the location of the regulator (which is to say, at the alternator output for modern cars, but I'm not sure if your 400i may have an older, remote-mounted regulator). Anywhere 'downstream' is subject to voltage drops across corroded connections, worn switches, and of course the wires themselves; for example, if you put 4 A through a 1 mm^2 wire, you'll drop roughly 0.07 V per meter of wire length. Even without considering any deterioration in the connectors, that can add up to half a volt lost in the wiring for a circuit like the windows, where the current has to flow to a console-mounted switch, to the motor inside the door, and then back to the console-mounted switch before it's grounded!

I haven't followed this whole thread, but the likely criticism of the relay solution in your earlier post is that it just bypasses the switches, whereas installing the relays in the doors (with a local ground and dedicated power) bypasses all the wiring (and associated connections) in addition to the switches. So if your solution works for you, that suggests that your biggest voltage drop was across the switches, and that your wiring and connectors are all pretty healthy. What's interesting is that a lot of people (including the commercially-available "window boosters") implement the relays such that they improve the ground side of the circuit only, with the +12 V still traveling through the switch and all the associated wiring. So while relays in the door with dedicated power and local ground is the ideal solution, it wouldn't surprise me if your implementation is superior to the popular ground-only implementation in many cases; you're bypassing the switch on both sides of the circuit, but none of the wiring, whereas the popular fix bypasses half the wiring and bypasses the switch on only one side of the circuit. Of course, in practice, it all depends how deteriorated the switch is vs. how deteriorated the wiring and connectors are, so different people will have different results with different cars.

 

A fly back diode could be a good idea, but the 2 diodes solution was a fix for burnt switch contacts. Both the relay solution or the diode solution are a good fixes for systems that have damaged switches. It is the return ground through the switch that fails due to arcing when the switch is released. Since these switches are not easily sourced these days an alternate solution may be the only fix.

 

I guess I'm still not visualizing what you're describing—from the switch's ground terminal to each of the switch's output terminals (to the motor)?

 

No need to gain access to the switches, just the motor in the door. Connect the cathode (stripe) end of one MR751 diode to each of the motor leads, and connect the anodes of both of these diodes to a good ground in the door. The diode does nothing when positive is applied, and then conducts the ground for the motor. When reversed the same happens for the motor.

 

Maybe there's a typo in there somewhere, but what you're describing would not conduct the ground for the motor, as current can only flow through the diode from anode to cathode, and grounding of the motor needs to conduct current from the motor to ground.

 

A diode blocks with positive on the cathode, and conducts with negative on the anode.

 

Well, half of that is true.

 

Paralleling diodes will not change the forward voltage drop (which is about 0.7 volts for silicon type diodes). Maybe you are thinking about ohm's law and what happens when you put resistors in parallel.

 

To be more circuit specific, the diode blocks the positive applied to the motor and thus does nothing to that leg of the circuit. Then the diode completes the motors path to ground in the circuit. This is to overcome the fault in the switch that does not complete the motors path to ground. This is what I was indicating, and was not electron specific in that sentence.

 

No my concern is more about the amp-draw. It's easy to source regular diodes that can sustain high amps (eg: alternator diodes), but low voltage drop diodes that can sustain high amps are harder to come by. That's the reason I would use a few in parrallels.

I am not sure it is neccessary to bief up both motor brushes: when lowering the window the amp-draw the amp-draw remains within the reasonable limits for the switch.

 

MR751 diodes are more than capable of doing the job, and are easily available on eBay.

 

I think if you draw the circuit diagram you’ll see why the diode, installed as you’ve described, cannot complete the motor’s path to ground—that would require current to travel ‘backwards’ (from cathode to anode) in the diode, which is impossible (except for Zener diodes).

 

+1 -- I'm not getting what jimmyr is suggesting.

jimmyr - Can you post a schematic showing how the diodes are connected to the motor leads?

 

Not able to do a diagram, but my previous note described the connection process. To give the electron path: the positive from the switch flows to the motor, and the connected diode at this point does nothing (blocked), but the second diode on the other motor lead goes into conduction as the positive voltage exists the motor at this point and that connection through the diode conducts to ground. The opposite is true when the switch reverses the motor. This system has worked on many installations with damaged switches; Daytona, 330, Boxer, etc.

 

This is where I am not understanding. A diode placed to allow current to flow from the negative side of the motor (call this side A) to ground when the other side of the motor (side B) is at +12V, would also allow (huge) current to flow from side A to ground when side A is at +12V to reverse the motor direction.

 

This is what I get from the discussion. Windows up, V1 > V2. Windows down, V2>V1. Regardless of the polarity, if, due to losses in the wiring, V1 and V2 are greater than the diode forward voltage then both diodes will conduct. For example say V1 = 11v, V2 = 1V. Or let V1 = 1v and V2 = 11v.

Now if you flip the diodes, neither conducts ever because V1 and V2 are both more + than ground.

Image Unavailable, Please Login

 

Except if the diodes are positioned the same, to block +, then they will never conduct because if the switch ground side had high resistance, the ground side of the motor will still be at a higher potential than ground. It's not the voltage on each side of the motor that matters, it's the voltage between the ground side of the motor and actual "ground".