Mechanical engineer here. My job is heavy with electrical work but I don't have a ton of training in EE. So looking for help on this one. As the title says I am looking for a way to create a circuit that outputs a linearly changing voltage from 2.3V to 2.6V upon receiving an input signal.

The input signal will be either 12V or a short to ground.

With no input signal, output voltage should read 2.3V.

Once that input signal is received I need to start ramping up the output voltage to 2.6V at a specified rate. Still working on what that rate should be. It will then sit at 2.6V until the input signal is removed.

When input signal is removed the voltage can go right back to 2.3V. Rate of decrease does not matter.

Any help is greatly appreciated.

 

What happens when the voltage reaches 2.6 V? Does it remain constant at that value?

 

What happens when the voltage reaches 2.6 V? Does it remain constant at that value?

Yes, it will remain at that constant value until the input is removed.

 

This can easily be done.
Do You have experience with building a simple Op-Amp Circuit ?

What causes the Control-Voltage to be at a static-level of ~2.3-Volts to start with ?
Do You have a Schematic, or at least a Block-Diagram, of the existing Circuit ?, Please post it.

What does the "Control-Voltage" control ?
Why does it need to ramp at a specific, but "unknown", rate ?
Is the Linearity of the "Ramp" rate critical ?, or approximate, and why.
.
.
.

 

what is the environment (temperature changes), as this can affect drift. how often (if ever) would rate and limits need to be adapted? if they need to change, who would be in charge of making such adjustments? maybe adding digital part makes sense....

 

I am looking for a way to create a circuit that outputs a linearly changing voltage from 2.3V to 2.6V upon receiving an input signal.

Does this NEED to be "Linear"? Or do you just want a timed change from 2.3V to 2.6V?

Also, is this a DC circuit?

 

I have some experience with Op-Amp circuits.

This is an automotive application where sensors are missing. The ECU is trying to run an operational test and since those sensors are missing, an engine code is thrown.

Basically what happens is:
1. ECU closes the tank drain valve. This is the 12V or short to ground signal.
2. Once that tank drain valve is closed. The ECU looks at the tank pressure sensor and expects it to increase from a low pressure reading of 2.3V to a high pressure reading of 2.7V.
3. Once 2.7V is reached, the tank drain valve closes. Losing the 12V or short to ground signal. The ECU then expects to see the tank pressure sensor voltage drop back down to 2.3V.

If the ECU doesn't see this happen it will assume the driver has forgotten to put the gas cap on and throw and trouble code.

The rate at which the ECU expects to see this rate increase is not known or readily available information. So I plan to take a trial and error approach with this.

The circuit will be inside a vehicle but will see major temp swings. However, fluctuating +/-.5V is fine.

 

Does this NEED to be "Linear"? Or do you just want a timed change from 2.3V to 2.6V?

Also, is this a DC circuit?

It does not need to be linear. It just needs to go from 2.3V to 2.6V over a certain time. It is a 12V DC voltage. I have a 12V regulated supply I can pull from.

 

It does not need to be linear. It just needs to go from 2.3V to 2.6V over a certain time. It is a 12V DC voltage. I have a 12V regulated supply I can pull from.

And that certain time is...?

 

Depending upon the manufacturer of your Car,
You may be able to get a local Hot-Rod-Tuner-Shop to simply turn-off the test procedure.

I use "HP-Tuners" Software to tune my GMC-Truck.
Evaporative-Emissions-Testing was one of the first things that I turned-off.
So far, I've picked-up roughly ~25-Horsepower, and a 6-MPG Gas-Mileage increase.
( these are not realistic numbers for people new to Tweaking Engines )

HP-Tuners covers a fairly wide array of PCMs,
there are other various Softwares as well, with different specialties.
.
.
.

 

And that certain time is...?

Lets say 5 seconds. But I am not sure what the ECU is looking for and I am going to need to test multiple rates.

 

Depending upon the manufacturer of your Car,
You may be able to get a local Hot-Rod-Tuner-Shop to simply turn-off the test procedure.

I use "HP-Tuners" Software to tune my GMC-Truck.
Evaporative-Emissions-Testing was one of the first things that I turned-off.
So far, I've picked-up roughly ~25-Horsepower, and a 6-MPG Gas-Mileage increase.
( these are not realistic numbers for people new to Tweaking Engines )

HP-Tuners covers a fairly wide array of PCMs,
there are other various Softwares as well, with different specialties.
.
.
.

I'm actually the product design engineer at a tuner shop. I would just delete codes from the ECUs in most cases. However, for this project the ECUs can not be tampered with. I've got circuits to get rid of every other evap, cooling, and CAN comm codes. This is the last remaining.

Congrats at that extra HP and MPG!

 

OK,
I will "assume" that the sensor is fed the standard 5-Volts Sensor-Voltage and that
the resulting Sensor-Output-Voltage going back to the PCM is,
at Atmospheric-Pressure,
~2.3-Volts.

I would estimate that the 2-Voltages may need to be adjustable to insure reliable operation.
Are You absolutely sure of the precise Voltages that You quoted ?
What is the Sensor-Voltage after sitting overnight with the Gas-Cap off, and Ignition-On ?
What is the normal Voltage of the correctly operating EVAP System with the Car running and warmed-up ?

This means that no Sensor needs to be installed on the Vehicle,
and that, 2 preset Voltages need to be supplied to the Sensor-Signal-Wire going back to the PCM.

This also means that the branch of the Wiring-Harness that normally goes to
the general area where the EVAP-Valves are normally located may be removed if desired.

This also means that the Standard 5-Volt Sensor-Supply-Wire can be used to power the
new Circuitry that satisfies the "expected" behavior coming from the original Sensor.

I will further make a guess that the PCM does not continuously monitor the Voltage from the Sensor,
meaning that, it is not important that the Voltage changes "smoothly" over a specific period of time,
but simply "waits" for "X" period of time and then samples the Voltage again, repeatedly.
I would estimate that the Voltage is sampled approximately once per minute,
while the Engine is running, regardless of any other circumstances,
therefore, I'm going to guess that a crude "Timer" that is adjustable from
~30-Seconds to ~120-Seconds should be a reasonable range to start out with.

Does all of this seem correct to You ?
.
.
.

 

I'm going to guess that a crude "Timer" that is adjustable from
~30-Seconds to ~120-Seconds should be a reasonable range to start out with.

If that is the case, and bearing in mind the temperature range a vehicle experiences, the ubiquitous 555-based timer is probably not the best choice. Anything more than a few seconds is a l-o-n-g time in 555-land, if reliable timing is required.

 

I'd be tempted to use a small processor board like the Seeeduino XIAO. It has an output pin with a 10 bit DAC (Digital to Analog Converter) which means it can output 0V to 3.3V in equal steps of around 0.0032 volts. Power it with 5.0V with a 7805 linear regulator connected to your 12V supply. A DAC output value of 713 should give a voltage of around 3.3 x 713/1023 = 2.3V. It's not exactly linear so you may want to tweak that value up or down a bit.

When the input signal is connected to a digital read pin (via a potential divider to drop the voltage from 12V to around 3V) the code can increment the output with a delay between each step up to a value of 806 which should give you 2.6V.

Each loop the code would check if the input is high or grounded - when grounded it could step back down to 713 again. If the 93 steps are not as smoothe a climb as you want you could filter a bit via a resistor and capacitor but this may not be important. The advantage is a simple circuit and the ability to change voltages and ramp speed with simple code changes

 

Most likely,
all that is required is an LDO Voltage-Regulator set-up for 2 Voltages,
and a TL431 setup as a Timer to select the Voltage.

This setup will act as a 2-Stage-Timed-Switch, with no "ramp" required.

This will also provide a very Low-Impedance-Drive to eliminate any Electrical-Noise getting into the system.
.
.
.

 

Here is a circuit using a LM317 regulator, with the signal input at 12V in , the output voltage is 2.6V and with the input at ground the output voltage is 2.3V.

If you want 2.7V output then use 92 ohms instead of 84 ohms.

 

the output voltage is 2.6V and with the input at ground the output voltage is 2.3V

Neat. Can that be made to ramp up and down with a resistor and capacitor?

 

I might have missed this above . Ramp characteristic - linear of exponential?

The TS said "Once that input signal is received I need to start ramping up the output voltage to 2.6V at a specified rate. Still working on what that rate should be." I'd guess that a bit of non-linearity would be okay - it's not like its moving to and from zero

 

If an exponential ramp still is acceptable, then the circuit in #17 just needs an added capacitor across the 84 ohm resistor (reference designators - !).

If my read of #7 is correct, one of the input transistors can be deleted for the correct logic polarity:

0 V = ramp up
12 V = snap down

Note that the LM317 output can source a lot of current, but is not very good at sinking. If this is an issue, add a resistor from the 317 output to GND to set a minimum current sinking capability.

ak