Is an additional flyback diode required for Fuel Pump Relay?

[Danbob]

Hi, I did a search on google and on the forum and found different opinions on whether an aditional "flyback" diode is required on the Fuel pump relay output of the main board?

I'm driving my stock fuel pump relay from the main board, I have no idea if the stock relay has any internal diode, but I expect it does not.

Is it necessary to wire a diode into the loom for the FPR?

If it is necessary how do I wire it in? In series? Or in parallel across the relay coil pins?

What type/spec of diode is best/suitable for this purpose?

I'm quite new to electronics so this is all new to me.

Thanks again
Dan

[WestfieldMX5]

As a rule of thumb, every relay needs a flyback diode.
Install a 1N4001 or 1N4002 with the banded side to a 12V source and the other side to the fuel pump output.
By doing this, high voltage flyback currents (thus voltages) will get diverted back to the 12V source instead of into the PCB.

[jsmcortina]

Yes, but there is already a voltage clamp in place.

James

[Danbob]

Yes, but there is already a voltage clamp in place.

James

OK, so does that mean an additional diode is not required?

Is Diode D4 providing this?

[Matt Cramer]

Exactly. That's what D4 does.

[Danbob]

Oh OK, starting to make sense then.

I thought that it would be strange if the mainboard circuits did not already include everything necessary to safely drive the fuel pump relay, when the output is labelled as FPR and it is shown in the hardware manual etc.

OK, so I accept that it is wise to install a flyback diode whenever driving a relay as a rule of thumb, but in the case of the Fuel pump relay drive on MS mainboard, this diode is already installed?


Thanks all

[Matt Cramer]

Correct, it's already there.

[dnz]

To avoid a new thread, I take it this is the case for V3 Microsquirt also meaning nothing additional is required?

[dontz125]

The uS uses a power MOSFET that is far more rugged and able to deal with flyback than the TO-92 NPN transistor used on the V3.0 MS board. Note also that the N-Mosfet in question has a body diode that will deal with much of the flyback pulse.

[scatt]

hi,
did read this post I must say I was thinking about adding an extra diode like 1n4001 or something on my Fuelpump relay.
I did think about the extra Diode because I had two broken ZTX450 transistors I had to replace.
When I switch on/off a couple of times the transistor gets broken.
I did measure my fuelpump relay current : 160mA, normally that should not be a problem ?
What could be the cause of that, do you have any ideas ?

thx

[jsmcortina]

What could be the cause of that, do you have any ideas ?

Are all of the other components in place? e.g. D4.

James

[DaveEFI]

hi,
did read this post I must say I was thinking about adding an extra diode like 1n4001 or something on my Fuelpump relay.
I did think about the extra Diode because I had two broken ZTX450 transistors I had to replace.
When I switch on/off a couple of times the transistor gets broken.
I did measure my fuelpump relay current : 160mA, normally that should not be a problem ?
What could be the cause of that, do you have any ideas ?

thx

I've replaced several ZTX 450 transistors on boards I've repaired. Of course since I only ever see the board, I've no way of knowing what caused the failure, if anything. I tend to replace them with a 2N2222, which was the original transistor on early boards. It can handle 800 mA which is less than the ZTX450 but I've not had any returns for it failing. It is more than adequate for a normal auto relay.

[scatt]

hi,
Thanks for the quick reply !
@jmscortina : yes, all is in place also D4 (so what I thought is confirmed in this thread, because of D4 no need for an extra Diode). I measured other components and seem to be ok.
the second time I replaced the transistor I did startup about 5 times and all was ok. The day after it was broke again, without giving the 1 second startup signal for the fuelpump, so I think it was broken when I switched off the last time the day befor (just a guess). When writing I start thinking maybe I need to check the soldering of the D4 ? If that gives bad contact there is no protection...

@DaveEFI : ok I will try with the 2N2222. If that goes down as well it must be sth. different as the transistor.

Just a question in general : for such (slow) signals, wouldn't it be safer to use a optocoupler (gives some more protection to the other parts of the board) ?
I think I will use one to drive the cooling fan ( also because I used the ZTX450 from the Fidle circuit to repair , I did not install because I have Idle stepper valve)
which I will put on the fidle or one of the leds.

I studied electronics (long time ago, about 25 years ago So I try to remember : what is the purpose of the 2N3904BU (Q19) together with the 1 Ohm resistor ? A kind of current stabilisor ?
as the current through R gets bigger Vbe Q19 gets bigger and Q19 starts conducting and recuding Ib from Q2 ?

thx !

[DaveEFI]

Reading your posts again, I'd be looking for another reason for the failure of the driver transistor. If one ZTX failed, might just be bad luck. More than that, likely a cause.
Where are you getting the +12v feed for the pump relay from? Is it from the same circuit as the ignition coils?

I'd add that although I've replaced quite a few ZTX on other's boards, they've been just fine on my personal MS2.

[scatt]

for the Relais feed current (I hope that is understandable english...) : Battery -> Fuse -> Contactswitch -> Relais.
Then from Relais -> 3 fuses to lambdaheater / ignition Coil / Fuelpump
Switching the relais : Same 12V point as via contact , ground via signal megasquirt
(now I switch by hand with a switch.... because that transistor is so difficult to solder...)
I think too there must be a reason for that failure but I cannot imagine what it could be if the relais only takes 160 mA.
Maybe I need to try with replacing Q19 also and resolder that whole area who knows.
After 2nd thought maybe better replacing with ZTX again untill I found the problem.
I will need to order somewhere online, because no ZTX in the local shop.
I will do it later on and let you know the result.
Now concentrating a bit on making a good tuning, I did not drive the car yet. First registering my TS
Thanks for thinking with me !

[scatt]

Just a small remark : that transistor says : 1 Ampere, but that is with very low Vce, there is only the max disipation of 1W...
I'll try later on and try to find the cause

[billr]

Have you actually measured 160mA going through the MS fuel-pump output pin (DB37 pin 37)?

[DaveEFI]

Since the coils are on the same 12v feed as the pump, are there caps at the coil between +12v and ground?

[scatt]

no, I just measured the current through the relais coil after things broke down, to have an idea of the current.
There is the standard cap between the 12V on coil and ground yes.

[Six_Shooter]

I studied electronics (long time ago, about 25 years ago So I try to remember : what is the purpose of the 2N3904BU (Q19) together with the 1 Ohm resistor ? A kind of current stabilisor ?
as the current through R gets bigger Vbe Q19 gets bigger and Q19 starts conducting and recuding Ib from Q2 ?

thx !

I asked this same question recently, and it turns out is is for over current protection. http://www.msextra.com/forums/viewtopic ... 23&t=60042