voltage modification 2

[bigTomHanks]

I've been trying to make a drive by wire throttle body work with a different model controller and am having an issue with developing a circuit as a work around. Please forgive me if I am way off base here as I am a mech. eng. and have very little electrical experience.


From what I gather when the throttle blades move the throttle positions sensors voltage should look like this chart:

http://chevythunder.com/TP sensor chart.BMP

When I move the throttle blades on the one I have both TPS 1 and TPS 2 follow a path similar to TPS 1. I think this is due to an internal redesign of the TPS. To work around this I figured I could make a differential op. amp. that would convert the TPS 2 output to equal the following equation so it would send the computer the signal it is expecting.

TPS 2 (to computer) = 5V - TPS 2 (coming out of throttle body)

I went to radio shack and bought the following:

(1) 742 op amp
(4) 1k ohm. resistors.

I connected them same as the circuit shown here:

http://en.wikipedia.org/wiki/File:Op-Amp_Differential_Amplifier.svg

Pin 1 - not connected to anything
Pin 2 - TPS 2 signal voltage
Pin 3 - 5 Volt constant from PCM
Pin 4 - - 6 volt
Pin 5 - not connected to anything
Pin 6 - should equal (5V - TPS 2 signal voltage)
Pin 7 - + 6 volt
Pin 8 - not connected to anything

http://www.st-andrews.ac.uk/~

http://www_pa/Scots_Guide/datasheets/Opamps/741.html

So very simply I am trying to subtract one variable voltage from a constant one to give the expected reading to the computer. The circuit works but it is not giving the expected output voltage. Any help that can be provided will be greatly appreciated.

 

[bigTomHanks]

Nevermind. I noticed that the tolerance on the resistors was 5% so I checked them and one was significantly different than the others. I swapped that one out and the circuit now works as it should.

Also, edit on the OP amp type it's a 741 not a 742.

 

[geekEE]

Glad to hear you got it working. I had a couple comments though. First of all, make sure that the rated temperature range of your op amp is not exceeded. If you have this circuit under the hood, the temperatures can be very high.

Secondly, you said that your supply rails are +6V and -6V. How are you getting these voltages? If you are generating them from the car's 12V power, make sure that your regulator circuits are able to handle the horrendous voltages swings that can be seen there. A load dump can put a high voltage on the line.

Be careful with this, because it could be really bad if this circuit fails in a full throttle manner.

Glenn

 

[bradrs]

Can you simply swap the 5V and ground wires on the TPS 2? That should get you the same result without needing to add an opamp.

As posted, the +6V, -6V power has some risks associated with it. You could use an opamp powered just by 5V, but then you need to make sure it is capable of rail-rail operation

 

[ScottyUK]

The 741 is not really a device of choice these days for anything remotely accurate. Plenty of good op-amps to choose from, many more capable than the 741.

 

[itsmoked]

Pretty much ALL of them more capable than the first one ever designed..

Keith Cress
kcress -

http://www.flaminsystems.com

 

[IRstuff]

Yeah, nice rhetoric, but not even close ;-)

The 741 was preceeded by, at least, the 702 and the 709; and that's just the monolithic ones. The 702 and 709 were designed by Widlar, himself, and the 702 is supposedly the first actual monolithic op amp. Philbrick made the first solid state op amp.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[bigTomHanks]

Hi Scotty - I'm not familiar with the readily available OP AMPs on the market. What would be a better one to use?

 

[MacGyverS2000]

Determine what parameters are most important to you, then use your favorite part search engine (Digi-Key, Mouser, et. al.) to find a matching part.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[bigTomHanks]

Ok changed direction on this one. I ditched the OP AMP circuit due to the complexity in getting both TPS voltages to what I needed them to be. I'm now using an Arduino Uno to take the input values from both TPS 1&2 and then convert them to the correct output voltages that the computer expects using an equation. Before doing this I would only get a quick on/off from the throttle body. With my new setup I'm getting a more sustained attempt by the throttle body to cycle itself fully open and closed. It'll open very slightly and cycle over and over like it's trying to do what it's supposed to but then times out after about 5 sec. I'm thinking that my problem is due to the output of the Arduino being PWM. Would application of a RC filter after the Arduino turn the signal into analog? Also I'm not familiar with how RC circuits work. Would they cause a voltage drop or does it just clean up the signal and covert it analog?

 

[IRstuff]

Yes, an RC configured correctly will filter, but there will probably be some signal loss as well. That's inevitable going from a sequence of square waves to a semi-analog signal

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[bigTomHanks]

Thanks to all who have provided input into my problem. I am going to try adding the rc circuits but don't know what good capacitor and resistor values would be for my application. Can someone help me determine what would be good values for 0-5 volts with very fast movement to steady state?

 

[OperaHouse]

You may not be through with that pesky op-amp yet. When I first saw this post it occured to me that this might be a good
application for the UNO. The RC filter will work best with a very light load so an op-amp may be needed to buffer the output
of the RC filter. Of course this is old analog thinking. The UNO has pretty good drive characteristics and I would suggest
using a resistor of 470-620 ohms and a capacitor of 4.7uF to start for this 490Hz PWM. You have six A/D inputs and one of them
could monitor the output and adjust the PWM drive count to correct for losses. In worst case when a step function happens, a
calculated one cycle overdrive count could be added/subtracted to the PWM drive to give a better response. This will all take
some playing around if you are going for some extreme control. It takes a different way of thinking but it is far easier than
twisting a bunch of pots.

I would suggest that you take a running sample on the input signal to reduce noise and to create larger numbers to work with.
It is quite easy to work with a count of 5,000 to represent 5V. Makes it much easier to wrap your head around numbers like that.
Avoid floating point at all costs. Binary number limits like 255 and multiples will easily match up to 5V if you reduce the voltage
slightly to the TPS with a resistor drop.

5V is not 5V. Every regulator is different. One of the best features of the UNO is the serial USB output to a laptop. With that
you can display real time data on the screen. I built a fridge controller for my solar powered camp. Temp was calibrated dead on.
Worked great until I actually hooked it up to the fridge. The laptop powered the UNO at a slightly higher 5V than the on board
regulator of the UNO. That will change all your A/D slightly but noticeably. Had to open up the USB cable and cut the +5V wire
in order not to change readings.

 

[bigTomHanks]

I use the internet to solve a lot of problems so I wanted to report the final result.

I set up the Arduino to take the TPS 1&2 sensor readings and convert them to the values the computer was expecting to see for a given TPS open percentage. I also added RC circuits after the Arduino and before the DBW computer to ensure the signal was analog or very close to it. Even though the output voltage was correct the DBW system did not work. Given this I'm going to ditch the DBW and go with a regular cable operated throttle body.

Thanks again to all who provided input.