[Diy_efi] Sending unit confusion + schematic suggestions

Thu Feb 20 23:51:15 GMT 2003

J,

Guessing by the design of the OEM EWD you have a JZA80? (I just went
through all my MA70 notes to find them irrelevant ;p).

The problem here is that you don't really know where the wires connect, as
the EWD doesn't give you enough information.
I'd speculate FE means Fuel Earth, but that's not consistent with the Br-B
(assuming that's Br-W on EWD) having 15k to ground.

Chances are there's a resistor going to 12V on the FR pin, hence when the
SUB fuel sender is not hooked up you get 12V on the Y wire (no current
draw means no voltage drop across the fuel sender and possible resistor).

You could work out what the resistances are likely to be (if you can't
measure them - you can also pull apart the BODY ECU and trace the pcb)
from what you know, but in any case you don't really need to to simulate
the 0-3.5V range. I'd do as someone else suggested and just use a pot (10k
or 55k if you want, doesn't really matter) with each end connected to
ground and 8V, and then use the tap for your testing voltage; you'll get a
0-8V range, but for testing that's fine.

The other issue you might run into is that your ciruit may draw some
non-negligible current from the fuel sender resistor ladder and cause your
fuel gauge to show an incorrect amount. The data sheets for the LM339
shows less than 1uA so it wont affect it, but something to remember if it
cause weird problems.

Some notes on your schematic[7] (finally posting some to you after having
composed a few mails :)).

Typical NPN transistors are used in open collector applications pulling to
ground. So the typical way of hooking it up is as follows. +ve of the
relay goes to 8 or 12V, negative goes to the collector, emmitter goes to
ground, and the base goes to a +ve siwtch-on signal through a resistor. A
diode accross the relay is needed, Anode (arrow part) to negative, Cathode
(bar part) to positive.

In other words you should interchange the relay and transistor. The reason
for this is that you need to reliably develop over 0.65V across the BE
(Base-Emmitter) junction. If you hook up E to ground, and pin2 is at 8V
you will always have over 0.65V at B. The current going into B will be
(8-0.65)/1k=7.35mA which will turn it on without doubt.

If you hook it up the way you have, E will sit somewhere nearer 8V due to
the resistance of the relay and the transistor may not even turn on.

Another poster mentioned the LM339 having only open collector outputs,
meaning that it can only reliably pull something to ground - it wont pull
pin2 to 8V - you will need a "pull-up resistor" to pull pin2 to 8V. A 5k
pull up at pin 2 should turn on a transistor and is within the 339 limits
(but a better way is to use a PNP transistor and reverse the logic).

For PNP, E to 8V, C to relay +ve, relay -ve to ground, diode across relay
(cathose to +ve, anode to -ve), 10k pullup between 8V and B (to make sure
it turns off) and 1K from B to pin2 should be fine. In this way, when the
-ve 339 i/p is over the +ver 339 i/p, the o/p will be low turning on the
transistor, and the relay.

In Schematic 5 the reference 3.2V (#2) is blocked by C3, so no DC
component will reach pin 4 (fixed now but..) Dave's suggestion was to hook
it up in parallel with the IC, ie positive of cap goes to pin4, neg goes
to ground (to help with possible noise problems).

Note: this has not been tested, just theoretical rambling.

Mos.

-- 
84 AE86, 90 ST185GrpA, 91 MX83Gr, Sydney, Oz.

On Thu, 20 Feb 2003, Toyota Supra wrote:

> hi all :)
>
> in my never ending quest to conquer this circuit, i have encountered
> something very strange. when i removed the fuel level sending unit to try
> and test my circuit, i found that i am not able to simulate its electrical
> function.
> the sending unit has 2 wires, a yellow and a brown w/ black tracer. when the
> sending unit is not plugged into the body harness, the yellow wire has
> +12vDC (or whatever battery voltage may be) on the body harness side of the
> connector, and the brown w/ black tracer has approximately 15k of resistance
> to ground, or 35mV of positive voltage.
>
> when the connectors are plugged in, the yellow wire will have anywhere from
> 0 to 3.5vDC, dependent on the position of the "float". the variable resistor
> will have resistance ranging from 0k to 55k.
>
> now, my problem is, i can't simulate the variable voltage outside of the
> vehicle, to test my circuit. if i put 12vDC into the yellow wire, the brown
> w/ black tracer will also have 12vDC, so what could that wire possibly be
> doing, or connected to, that would make the circuit and unit work properly?
>
> i have enclosed some scans of what the factory manual diagrams it out to be,
> and some pictures of the sending unit, with descriptions
>
> here is a description of the sending unit, and pics of its resistor
> http://68.7.7.84:8080/schematics/sendingunit1.jpg
>
> http://68.7.7.84:8080/schematics/sendingunit3.jpg
>
>
> here is the oem wiring diagram
> http://68.7.7.84:8080/schematics/oem-schematic1.jpg
>
> here is the way they describe the sending unit in the manual
> http://68.7.7.84:8080/schematics/sendingunitdescription1.jpg
>
> and here is the latest evolution of the overall circuit
>
> http://68.7.7.84:8080/schematics/schematic7.jpg
>
>
> once i can simulate the sensor working, i can make sure the schematic works,
> and implement it. thanks to all that have helped me through this!

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi