F1 pump failure: Paging the electonicists

[CatmanV2]

Re-visiting an old idea of mine in light of Doris's unfortunate failure:

The F1 pump relay is a pretty well understood weak point, but there seems to be so much mythos around it. It's just a switch.

Here's my thought:

1) Replaced with a solid state version. This would eliminate the issue where arcing across the terminals causes them to stick. Using a higher rated relay *should* also make this rather less of an issue but. If this isn't a plug an play replacement, a simple wiring harness adapter would be totally trivial to construct.

2) Rather more sophisticated (and possibly overkill), a simple circuit that monitors the voltage across the relay output terminals. When they go 'on' a timer starts.
When they go 'off' the timer is re-set. Once the elapsed timer exceeds a certain threshold (no one's quite sure what that is right now, but again, simple to determine. Bit of a two man job though) sound a buzzer / beeper / LED / $random device). Again, massively simple to interface to the car. Simply plug into the relay socket and create a tiny loom to plug the relay into.

Of course to be really clever, combine the two

So who wants to help me out on this?

C

 

[mjheathcote]

I think a solid state relay with a custom loom plugged into the existing relay base is a good idea, with maybe simple led indication on the relay to indicate its state, open or closed, for diagnosis purposes. Incorporating a timer circuit with a solid state relay...the fact its a solid state relay makes this redundant surely, as a sticking relay is a mechanical relay problem not a solid state one?
Do Ferraris with the F1 box have the same issues? I assume, so the market for a plug in solid state replacement worldwide could make it viable?

 

[CatmanV2]

Well the LED is trivial, you just need to connect across the coil.
I think the fear around the use of SS on it's own is replacing this magic box with an unknown.
I don't think *I'd* be worried (at least not after the first few days) but I can imagine plenty of others would be.

No idea about the Ferraris. It would certainly be a nice little side line a bit like the contactless upgrades for the 3200.

Even a one off, I reckon £20 would cover it.
C

 

[CatmanV2]

Well the LED is trivial, you just need to connect across the coil.
I think the fear around the use of SS on it's own is replacing this magic box with an unknown.
I don't think *I'd* be worried (at least not after the first few days) but I can imagine plenty of others would be.

No idea about the Ferraris. It would certainly be a nice little side line a bit like the contactless upgrades for the 3200.

Even a one off, I reckon £20 would cover it.
C

Well, close

http://www.tmswebshop.co.uk/buy/low-cost-solid-state-relay-proswitch-lc.html#SID=85

Not *sure* that would be suitable, the specs are a bit light, but it only took 15 seconds of searching

C

 

[mjheathcote]

Number of duty cycles would be critical I guess?

 

[davy83]

Solid state relays are not like mechanical switches they are semiconductor switches and have an "on" resistance. So when they are in the on position there is a volt drop across them, so it will not be exactly the same as a mechanical relay. If you are having trouble with a relay welding closed then you have an inductive load and so you get big current transients, which is why the contacts spark making them stick together. The motor might not like having a resistive connection (it might be fine but...) A lot of motors will draw large current just briefly at start up and so your SSW would need to be rated for an inductive load so it can cope with this. So look for a solid state switch which is designed for an inductive load (the motor) and then look at the ON resistance. work out from the motor current what volt drop you will get with the SSW in place (V=IxR) Simples! its a 12 volt motor so you would want the volt drop to be less than 0.5 volt, preferably 0.1 or 0.2 volts.

 

[Grinzzz]

If you measure the resistance of the pump with a multimeter it'll give you at least an idea of the size of the current it will draw at start up. Power MOSFETS are the way I'd probably start looking.

 

[CatmanV2]

Agree totally about the inductive EMF on start up, but it's just a motor. I really doubt it'll worry too much about a couple of ohms worth of voltage drop. Simple enough to check.

If it is a problem, then the 'too long on' warning would still be useful.

C

 

[Grinzzz]

Shouldn't be a problem, it just has to be designed correctly. The main problem with semiconductors and inductive loads is back emf. Cant remember of the top of my head but I think a good rule of thumb is that it has to cope with around 2x the input voltage reversed.

Here you go, bound to be something here http://uk.rs-online.com/web/c/relays/solid-state-relays-accessories/solid-state-relays/

 

[CatmanV2]

Will look when I get sensible access


Sent from my iPad using Tapatalk
CatmanV2