Fuel warning light delay

[Patrick]

Can anyone help with a delay device to be added to the circuit in the
attached web link.

The delay is required so that the warning light is not illuminated
until the sender signal at the set low level threshold has been solid
for say 2 minutes. So that bumps bends or acceleration doesnt flash the
light as the fuel sloshes around.

I am pretty inexperienced with electronics-I can understand and build a
basic 555 circuit,to give an indication of my level!

Hope someone can help.

Patrick

http://users.pandora.be/miata/english/misc/lowfuel.htm

 

[John Popelish]

Can anyone help with a delay device to be added to the circuit in the
attached web link.

The delay is required so that the warning light is not illuminated
until the sender signal at the set low level threshold has been solid
for say 2 minutes. So that bumps bends or acceleration doesnt flash the
light as the fuel sloshes around.

I am pretty inexperienced with electronics-I can understand and build a
basic 555 circuit,to give an indication of my level!

Hope someone can help.

Patrick

http://users.pandora.be/miata/english/misc/lowfuel.htm

I think you can use a 555 timer with the timing capacitor connected to
both threshold and trigger inputs (don't connect discharge). Thew
timing resistor is connected between the low fuel signal and the
timing capacitor. This is basically an RC low pass filter connected
to a high hysteresis Schmitt trigger. You could get better filtering
by putting a second RC time constant upstream of this resistor, using
about twice as large a capacitor and half as large a resistor.

You will have to protect the 555 from high voltage spikes from the
electrical system (18 volt zener across the 555's supply terminals, in
parallel with the bypass capacitor and a series resistor to +12).

The timing network might be something like this: (set the reader to a
fixed with font, like Courier to see this as a picture)


Input---47k--+--100k--+----+--Trig
+| | |
1000uF 470uF +-Thresh
| |
+---+----+
|
GND

When the voltage on the threshold pin exceeds 2/3 of the supply, the
output of the 555 output latches low.

When the voltage on trigger is less than 1/3 of the supply, the output
latches high.

See:
http://www.uoguelph.ca/~antoon/gadgets/555/555.html

 

[John Fields]

Can anyone help with a delay device to be added to the circuit in the
attached web link.

The delay is required so that the warning light is not illuminated
until the sender signal at the set low level threshold has been solid
for say 2 minutes. So that bumps bends or acceleration doesnt flash the
light as the fuel sloshes around.

I am pretty inexperienced with electronics-I can understand and build a
basic 555 circuit,to give an indication of my level!

---
View in courier:

<-----+
VBAT |
|A |
[1N4148] REPLACE |
| LAMP VBAT |
| WITH | |
[1N4148] / | |
| +12A / [SENDER] |
|K | / | |
[1K5] [1K] | |
| | E |
+----+-----|-----+-------B PNP YOUR CIRCUIT
| | | | C |
[2K]<--+ | [470K] | |
| | | [3K3] |
GND +------+-----+--C | |
| NPN B----+ |
| E | |
| | [3K3] |
| [1N4148] | |
| |K GND |
| GND |
| <-----+
|
|
+----|----+-----+------+
| | | | |
[510k] | +-->[100k] [0.1µF] +--A
| | | | |NOR Y
| | +-------+------+--+ +--B
+----|--|Rs Rtc Ctc | |
| | | +--A
+--|MR Q13 | |NOR Y
| +--------------+--+ +--B
| 4060 | |
| +---A GND
| NOR Y--+
| +--B |
| | |
| | A--+
| +--Y NOR +12A
| | B--+ |
| | | [LAMP]
+------------------|--------+ |
| C
+--[1K]-------B 2N4401
E
|
GND


VBAT>---[1N4003>]--+-->+12A
|
[1N5353]
|
GND


+12A is to be used for Vcc for the 4060 and the 4001, and as shown
above.


The way this thing works is that as long as the collector of your
NPN is high, which will be when the tank has over a certain amount
of gas in it, the 4060 and the RS latch made from the NORs connected
to Q13 of the 4060 will be RESET, the 2N4401 will be cut off, and
the lamp won't be lit. When the level of gas in the tank drops to
the point where your NPN turns on, the RESET will be taken off of
the 4060 and it'll start counting. If the gas doesn't bounce around
for a minute or so, the counters MSB will go high, SETting the
latch, which will turn on the 2N4401 and light the lamp. Any
disturbance causing your NPN's collector to go high will result in
another system RESET, and at least another minute before the lamp
can turn on again.

The two NORs shown with their inputs grounded are spares, and their
inputs _must_ be grounded as show.

 

[Ken Smith]

Did I miss something or:

[.. same as before ....]

| 1N914 or equiv
+----|----+-----+------+--->|---
| | | | | | <-- Diode kills
[510k] | +-->[100k] [0.1µF] | oscillation
| | | | | when Q13
| | +-------+------+--+ | high
+----|--|Rs Rtc Ctc | |
| | | |
+--|MR Q13 | |
+--------------+--+ |
4060 | | | |
| | +12A
| | |
| | [LAMP]
| | |
| -------+
| |
| C
---[1K]-------B 2N4401
R4 E
|
GND

[...]
Also:
I would put a resistor in the Vcc going to the CMOS parts and perhaps a
15V zener across them. This is going in a car and voltage spikes are
common in cars.

I'm make R4 10K or use a power MOSFET as the output transistor or
something. A CD4060 won't like putting out the current it is being asked
for.

I started off with the diode facing the other way and running from Q13 but
decided that if Q13 was really loaded with a 1K, I'd be better taking the
collector of the transistor as the input to the clock stopper.

 

[John Fields]

Did I miss something or:

[.. same as before ....]

| 1N914 or equiv
+----|----+-----+------+--->|---
| | | | | | <-- Diode kills
[510k] | +-->[100k] [0.1µF] | oscillation
| | | | | when Q13
| | +-------+------+--+ | high
+----|--|Rs Rtc Ctc | |
| | | |
+--|MR Q13 | |
+--------------+--+ |
4060 | | | |
| | +12A
| | |
| | [LAMP]
| | |
| -------+
| |
| C
---[1K]-------B 2N4401
R4 E
|
GND

---
I think you missed something....

Here's the whole thing with an added reference designator:


<-----+
VBAT |
|A |
[1N4148] REPLACE |
| LAMP VBAT |
| WITH | |
[1N4148] / | |
| +12A / [SENDER] |
|K | / | |
[1K5] [1K] | |
| | E |
+----+-----|-----+-------B PNP YOUR CIRCUIT
| | | | C |
[2K]<--+ | [470K] | |
| | | [3K3] |
GND +------+-----+--C | |
| Q1 NPN B----+ |
| E | |
| | [3K3] |
| [1N4148] | |
| |K GND |
| GND |
| <-----+
|
|
+----|----+-----+------+
| | | | |
[510k] | +-->[100k] [0.1µF] +--A
| | | | |NOR Y
| | +-------+------+--+ +--B
+----|--|Rs Rtc Ctc | |
| | | +--A
+--|MR Q13 | |NOR Y
| +--------------+--+ +--B
| 4060 | |
| +---A GND
| NOR Y--+
| +--B |
| | |
| | A--+
| +--Y NOR +12A
| | B--+ |
| | | [LAMP]
+------------------|--------+ |
| C
+--[1K]-------B 2N4401
E
|
GND


VBAT>---[1N4003>]--+-->+12A
|
[1N5353]
|
GND


Notice that if we start Off with a full tank the collector of Q1
will be high, which will RESET the 4060 and kill its oscillator, so
the 1N914 you inserted isn't needed.

---

Also:
I would put a resistor in the Vcc going to the CMOS parts and perhaps a
15V zener across them. This is going in a car and voltage spikes are
common in cars.

 

[Patrick]

Thanks ,guys for your help. I think I am probably a bit out of my
depth!
I can just about understand Kens circuit, which looks slightly
simpler. I am not familiar with what the NORs are in John's !!!
Couple of queries
1 Is it a 4401(referred to in cicuit diag) or 4001 (in note at bottom)
2 I am not familiar with the pin out on the 4060, from a diagram of a
4060B I can see
Ctc =9,Rtc=10,Rs=12,Q13=3,MR=16?,is that right?
Do I need to do anything with all the unused pins?
3 In my application the tank sender unit is grounded in the tank and
supplied off the back of the gauge,I'm not quite clear how that is
connected to the Emitter of the PNP,where your diagram shows VBAT
4 I will probably be using a bright LED as the lamp,protected by a
suitable resistor,hope that doesnt change the rest of your circuit.
5 Finally do you agree that if I protect the sender with something
like a .25A fuse, I will save it, if I cock something up?!

I am sure this is all very elementary stuff and I appreciate your help.

Patrick

 

[Patrick]

I missed the latest replies,while I was putting those questions
together!
Patrick

 

[Ken Smith]

[....]
I think I was right:

<-----+
VBAT |
|A |
[1N4148] REPLACE |
| LAMP VBAT |
| WITH | |
[1N4148] / | |
| +12A / [SENDER] |
|K | / | |
[1K5] [1K] | |
| | E |
+----+-----|-----+-------B PNP YOUR CIRCUIT
| | | | C |
[2K]<--+ | [470K] | |
| | | [3K3] |
GND +------+-----+--C | |
| Q1 NPN B----+ |
| E | |
| | [3K3] |
| [1N4148] | |
| |K GND |
| GND |
| <-----+
|
|
+----|----+-----+------+
| | | | |
[510k] | +-->[100k] [0.1µF] +--A
| | | | |NOR Y
| | +-------+------+--+ +--B
+----|--|Rs Rtc Ctc | |
| | | +--A
+--|MR Q13 | |NOR Y
| +--------------+--+ +--B
| 4060 | |
| +---A GND
| NOR Y--+
| +--B |
| | |
| | A--+
| +--Y NOR +12A
| | B--+ |
| | | [LAMP]
+------------------|--------+ |
| C
+--[1K]-------B 2N4401
E
^ |
! GND

This point
goes high and
stays that way

VBAT>---[1N4003>]--+-->+12A
|
[1N5353]
|
GND


Notice that if we start Off with a full tank the collector of Q1
will be high, which will RESET the 4060 and kill its oscillator, so
the 1N914 you inserted isn't needed.

The 1N914 I put in kills the oscillator when Q13 goes high and stops the
4060 in that state so the flip-flop isn't needed.



[...]

I have a 1N4003 in there to block the negative spikes, and a 1N5353
15V 5W Zener to clamp Vcc to 15V for positive spikes.

I'd expect to be picking bits to the 15V zener out of the surounding
plastic. In a "load dump" case there are many amps available.

The output transistor can be a high enough voltage one to take the voltage
but the CMOS need protection, hence the resistor in just its power line.


See above. Remember, the diode goes to the transistor's collector and
thus pulls down on the oscillator only when Q13 is high.

 

[Ken Smith]

, which looks slightly
simpler. I am not familiar with what the NORs are in John's !!!

IIRC: The CD4001 is the NOR gate he is refering to.

As my circuit shows, you don't need it but what John is making is an
RS-latch. Try it as sort of equations like this:


X = not(A or Y)
Y = not(B or X)

Start by assuming A, B, and Y are all low. Now observe what happens if A
goes high. Y will be high. At that point taking A low, will leave Y
high. You can only get Y low by taking B high.

Couple of queries
1 Is it a 4401(referred to in cicuit diag)

He has a 2N4401 transistor, over at the bottom right. The is a modest NPN
transistor. A 2N2222 would also work in his circuit.

4060B I can see
Ctc =9,Rtc=10,Rs=12,Q13=3,MR=16?,is that right?
http://www.fairchildsemi.com/ds/CD/CD4060BC.pdf

Do I need to do anything with all the unused pins?

All the unused outputs should be left disconnected.

 

[Richard Henry]

IIRC: The CD4001 is the NOR gate he is refering to.

As my circuit shows, you don't need it but what John is making is an
RS-latch. Try it as sort of equations like this:


X = not(A or Y)
Y = not(B or X)

Start by assuming A, B, and Y are all low. Now observe what happens if A
goes high. Y will be high. At that point taking A low, will leave Y
high. You can only get Y low by taking B high.


He has a 2N4401 transistor, over at the bottom right. The is a modest NPN
transistor. A 2N2222 would also work in his circuit.

All the unused outputs should be left disconnected.

I guess I would just do the whole thing with a small PIC.

PIC - the 555 of the 21st Century.

 

[Ken Smith]

I guess I would just do the whole thing with a small PIC.

PIC - the 555 of the 21st Century.

You'd need an LM7805.

 

[Patrick]

Here is a response from the gauge manufacturer about the sender
specification. How does this effect anything?
Patrick,
The resistance range of the telemetrix fuel sender is 10 ohms at empty,
to 180 ohms at fuel.
Hence, as the fuel level decreases, the voltage at the top of the
sender decreases, opposite to that required by your circuit.
Apart from sorting that out, don't see too many probelms as your
impedances are high relative to that of sender and gauge.
Regards,
Ian

Ian John
Design/Application Engineer
Caerbont Automotive Instruments Ltd

 

[Patrick]

Here is a response from the gauge manufacturer about the sender
specification. How does this effect anything?
Patrick,
The resistance range of the telemetrix fuel sender is 10 ohms at empty,
to 180 ohms at fuel.
Hence, as the fuel level decreases, the voltage at the top of the
sender decreases, opposite to that required by your circuit.
Apart from sorting that out, don't see too many probelms as your
impedances are high relative to that of sender and gauge.
Regards,
Ian

Ian John
Design/Application Engineer
Caerbont Automotive Instruments Ltd

 

[John Fields]

[....]

I think I was right:

The 1N914 I put in kills the oscillator when Q13 goes high and stops the
4060 in that state so the flip-flop isn't needed.
---
Yup.
---

[...]
I have a 1N4003 in there to block the negative spikes, and a 1N5353
15V 5W Zener to clamp Vcc to 15V for positive spikes.

I'd expect to be picking bits to the 15V zener out of the surounding
plastic. In a "load dump" case there are many amps available.


The output transistor can be a high enough voltage one to take the voltage
but the CMOS need protection, hence the resistor in just its power line.

---
Yup, and it only has to be big enough to let enough current out of
Q13 to turn on the output transistor.

There's another issue, though, with the same kind of problem and
that's with the reset line to the counter coming from the OP's
circuit:



VBAT
|A
[1N4148] REPLACE
| LAMP VBAT
| WITH |
[1N4148] / |
| +12A / [SENDER]
|K | / |
[1K5] [1K] |
| | E
+----+-----|-----+-------B PNP
| | | | C
[2K]<--+ | [470K] |
| | | [3K3]
GND +------+-----+--C |
| Q1 NPN B----+
| E |
| | [3K3]
| [1N4148] |
| |K GND
| GND
|


Which could be fixed like this, I think:



VBAT
|A
[1N4148]
| VBAT VBAT
| | |
[1N4148] |A |
| [1N4003] [SENDER]
|K | |
[1K5] [1K] |
| | E
+----+-----|-----+-------B PNP
| | | | C
[2K]<--+ | [470K] |
| | | [3K3]
GND +----+-+-----+--C |
| |K Q1 NPN B----+
| [1N4744] E |
| | | [3K3]
| GND [1N4148] |
| |K GND
| GND
|
+-> TO 4060 RESET

---

See above. Remember, the diode goes to the transistor's collector and
thus pulls down on the oscillator only when Q13 is high.

---
Yup.

First time around I read it like you were _adding_ the diode to the
existing circuitry, not substituting the diode for the latch. Good
save, a whole chip!

 

[John Fields]

You'd need an LM7805.

 

[John Fields]

3 In my application the tank sender unit is grounded in the tank and
supplied off the back of the gauge,I'm not quite clear how that is
connected to the Emitter of the PNP,where your diagram shows VBAT

 

[John Fields]

Here is a response from the gauge manufacturer about the sender
specification. How does this effect anything?
Patrick,
The resistance range of the telemetrix fuel sender is 10 ohms at empty,
to 180 ohms at fuel.
Hence, as the fuel level decreases, the voltage at the top of the
sender decreases, opposite to that required by your circuit.
Apart from sorting that out, don't see too many probelms as your
impedances are high relative to that of sender and gauge.

---
So your gauge circuit looks like this:?

+12V
|
[GAUGE]
|
+----E2
|
[SENDER]
|
GND

Where E2 is the voltage going into the emitter of your PNP?

 

[Patrick]

So your gauge circuit looks like this:?

+12V
|
[GAUGE]
|
+----E2
|
[SENDER]
|
GND

Where E2 is the voltage going into the emitter of your PNP?

John
Yes,gauge is also grounded ,if that makes any difference

See http://www.caigauge.com/minorgaugefit.htm.

I'm amazed at the interest that this has generated!! Even though most
of it is way over my head.
Any one fancy,drawing up the final circuit ,so a newbie can build it?
Why is it I cant seem to view the circuits in Courier on home PC, but
on work one ,I can?
Many Thanks
Patrick

 

[Ken Smith]

There's another issue, though, with the same kind of problem and
that's with the reset line to the counter coming from the OP's
circuit:

Since the frequency is low, we can just put a 100K resistor and a couple
of 1N914s in there to protect it. I really didn't look at that part of
the circuit.

I assume that the purpose was to make a comparitor with a large backlash.
The suggested circuit looks more complex than using a comparitor IC. A
quick count has it having about 20 solder joints vs about 18 for using an
LM211.

VBAT
|A
[1N4148]
| VBAT VBAT
| | |
[1N4148] |A |
| [1N4003] [SENDER]
|K | |
[1K5] [1K] |
| | E
+----+-----|-----+-------B PNP
| | | | C
[2K]<--+ | [470K] |
| | | [3K3]
GND +----+-+-----+--C |
| |K Q1 NPN B----+
| [1N4744] E |
| | | [3K3]
| GND [1N4148] |
| |K GND
| GND
|
+-> TO 4060 RESET

 

[Ken Smith]

Why is it I cant seem to view the circuits in Courier on home PC, but
on work one ,I can?

Theory #1: Bill Gates has it in for you personally and is messing with
your mind.


Theory #2: You aren't actually getting it set to Courier at home either
bause you don't have that font installed or you clicked something in some
settings box that overrides the font.