Hello all,

I've been building a device that helps firm up the shifts on my car. I have two controller designs built and assembled on one single Perf Board. They both are working well...

One design is based on the LM3914. The IC is being used a gate controller that closes PNP transistor switches. Each switch then feeds current to a slide POT, which adjusts the current to a Power Transistor...

The other controller design is a simple non-inverting op amp that uses a 10K POT as a variable voltage divider used to manipulate the gain. The output of the op amp feeds current to the Power Transistor.

Both designs are being driven by a .5 to 4.5V signal coming from my accelerator. This signal wire is fed through a 50K resistor that drops the current so low that if I either short the wire to ground, or I back feed 14V to the wire (the two worst case senerios), there is no effect to the car's throttle control at all... The current is way to low. I only mention this to insure the good people here that the signal tap is safe...

All this said... the op amp design is hugely more simple... BUT... The gain is linear... And although it works well. It does not work as well as the gate controller (which has a dedicated adjustment slide per gate).

When I look at ow the slides are adjusted on the gate controller, I would say that it's an increaseing curve...

See attached Thumbnail (red line).... (it's an increasing gain value)



Can something like this be done with an op amp? a no-linear increasing output?

In plain english I want to amplify a .5 to 4.5V signal to a 1V to 12V output, with more of the gain increase in the last half of the span...

Hope this makes sense...

I added a chart that I made up that sort of shows what's going on...

The left axis is Vout @ 400ma. 100 = 12V

The bottom Axis is the gas pedal signal shown as 0 to 100% (.5 to 4.5V)

The red line is what the gate controller is doing...
The blue line is the gain controller (op amp) with a gain = 2
The green line is the gain controller (op amp) with a gain at something > 2
The black line is what I'd like my gain controller to do. (I was thinking to try and introduce a step change somehow...

I'm trying to build on what I already have and not start over...

Here are a couple viedos showing what the device looks like...

All help will be appreciated. I have found a lot on non-linear op mp circuits on the web but I'm having trouble understanding what I'm seeing. I'm VERY GREEN when it comes to all this...

I've been toying with the thought of using zener diodes to manipulate the signal from the gas pedal etc... I decided it's best if I get some input before doing anything more...

Thank You...

---

 

First thing is confirm that this is for a track car and not for a car that will be driven on the streets.

 

First thing is confirm that this is for a track car and not for a car that will be driven on the streets.

And your concern is?

 

You will crash and people will die. That seems like a possible consequence of a malfunction.

 

If you are only driving on a racetrack then experimentation would be fine - risks of crashing are an expected consideration there.

 

You will crash and people will die. That seems like a possible consequence of a malfunction.

Your response is based on no knowledge of the system at all? Cocern and caution are waranted without a doubt. I've even run this design through a formal Hazop analysis... And your opinion (fear) is based on what?

Don't bother answering... I already know...

Thank You for your opinion... I feel the possible negative side effects have been duely coverd and no threat is lurking in the design...

All constructive help will be appreciated. Thank You.

--

 

Have you looking into non-linear gain circuits used in compandors?

 

Have you looking into non-linear gain circuits used in compandors?

I think I've seen that... But to be honest it's a bit over my head... I think I have an idea, but I'm sure it's not the best way to do it. And I'm not even sure it will work... I'll google " non-linear comparitors " and see what I can find.

thanks for the lead...

--

 

It breaks my heart to say this, but such a function is best done with a digital system where the many variables are easily handled and you could store as many different curves in memory as you want and call the one you want for various conditions.

It's possible to make non-linear analog gain blocks but not at all simple.

 

Umm, obviously you have checked with your road safety traffic authority about whether your car is considered as 'modified' and would need a engineers report to make it road worthy. What was the outcome?

Your mod doesn't come into the normal assortment of mods made by the public! Is it the same as remapping/chipping the transmission gear change control, or are you controlling other aspects of the car? I think people would be interested to understand what aspect of the transmission is actually being electrically controlled, or whther you are using an electromechanical interface.

Ciao, Tim

 

Hello,

As the safety is not garanteed, we do not want to support it anymore.

2. Appropriate conduct. Forum members are responsible for the content of their posts, including information they provide, validity of this information, and any extra content they may append to the post.

Illegal or inappropriate discussion or information will be removed. When there is ambiguity in the legal status of a request or reply, the Moderators and Admin will come to a consensus on whether the discussion is appropriate to remain in the public forum.

Forum members must not knowingly provide any information that may adversely affect another member; this includes, but is not restricted to, information that may potentially result in injury, death, damage or destruction of property and possessions. The administration takes this matter seriously and has the right to remove any contentious content and deal with the offending member as it deems is necessary.

We do not want others to be hurt or even killed by modifying a car.

Bertus