Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Hacking steering wheel controls to radio

Status
Not open for further replies.
In the LM3914's 'dot' mode there is a slight overlap of the respective voltage levels for the dots. Is this likely to be a problem?
 
For information, please see the attachment which is a scan of the schematic sketch which I used to build my converter.
Some aspects of it may be a bit cryptic.

Overlap on the dots was not a problem for me.
I did have to use two of the 3914 outputs for one switch position to ensure correct operation under all conditions.
Despite my best efforts at juggling the 390 ohm potential divider resistance, there was always on switch case which fell very near one of the 3914 switching thresholds, so I used both outputs above and below the threshold.

JimB

Mazda Radio controls converter.JPGMX5 Audio Controls.JPG
 
The chips arrived today and I have everything I think I will need. I calculated my resistor network based on the second page of the data sheet which uses a 6.8-18v input and a 0-5v signal. The couple things I noticed were that the 3.83k and 1.21k ohm resistors called for in the diagram don't really exist in the 100 count multipack I picked up from Radio Shack, secondly the formula provided results in 5.206v for the ref out. These seem like very specific resistances to end up with a formula that doesn't equal 5.000. That being said, on to the steering wheel switches.

I chose these values, voltages are based off 12-14v
R5 (to ground)- 150
R1 (switch)- 15k, 1.01-1.27v
Phantom R (pressing the following 2 simultaneously)- 800, 1.89-2.21v
R2 (switch)- 470, 2.90-3.39v
R3 (switch)- 330, 3.75-4.38v
R4 (switch)- 248 (390 and 680 in parallel), 4.52-5.28v

I figured I could use 2 outputs per switch to make 5 steps instead of 10 to ensure the correct command is registered. I will use three leds for R1 and one for R3 and R4(they overlap in the 7th band). Or I could use a ref out of 5 exactly and have steps of 0.5v which work nicely if 5.28v would still register and send a signal.
Any thoughts?
 
You could just use four 100 ohm resistors on the switches and a 220 ohm from Vin to Gnd. That gives a voltage of 0.69, 0.52, 0.42 and 0.35 (as a fraction of RHI), which will operate four contiguous outputs.

Alternately, the E12 standard values 120, 56, 91, 150 (and 220 from Vin to gnd) provide fractional voltages of 0.65, 0.55, 0.45, 0.35 (of RHI), which fall right in the middle of each comparator window.

The positive wire going to the switches should come from whatever you're feeding into the LM3914 RHI pin.
 
Last edited:
The positive wire going to the switches should come from whatever you're feeding into the LM3914 VHI pin.

Why? I understand that its convenient to do so but had I, I would not have been able to create a different function from pressing 2 buttons at once. If the consensus is that my method will not work then yes, I'm going straight to your solution, but I like my 5th function and I already did the maths.
 
I could rehash it using the Vhi as a source but I still want to keep the switches in parallel for the sake of the 5th function.
 
Ok, sorry, I didn't notice the extra function. The reason to use the same reference as the LM3914 is to make it irrelevant what the supply voltage is. If you have values that work, then go for it.
 
These are some values for parallel resistors that give you 7 outputs from 4 switches if you use combinations of two at a time.

Code:
Rgnd:    220    
      
SW       R      Vo
SW1      120    0.65
SW2      270    0.45
SW3      680    0.24
SW4      1200   0.15
      
SW1/2    83.07  0.73
SW2/3    193.26 0.53
SW3/4    434.04 0.34

The resistors are 120, 270, 680, 1200 ohms (220 RHI to gnd). This should activate outputs 6, 4, 2 and 1 for single switch presses, and outputs 7, 5 and 3 for combination switch presses.

You may have a problem using combinations of buttons because you won't be able to press or release them at exactly the same time, and therefore you will have unwanted outputs when you try.
 
These are some values for parallel resistors that give you 7 outputs from 4 switches if you use combinations of two at a time.

Code:
Rgnd:    220    
     
SW       R      Vo
SW1      120    0.65
SW2      270    0.45
SW3      680    0.24
SW4      1200   0.15
     
SW1/2    83.07  0.73
SW2/3    193.26 0.53
SW3/4    434.04 0.34

The resistors are 120, 270, 680, 1200 ohms (220 RHI to gnd). This should activate outputs 6, 4, 2 and 1 for single switch presses, and outputs 7, 5 and 3 for combination switch presses.

You may have a problem using combinations of buttons because you won't be able to press or release them at exactly the same time, and therefore you will have unwanted outputs when you try.
You, sir, are a god among men. Thank you!

Is the Vref a function of the Vin. That is, if Vin goes from 12v to 14 would the Vref adjust proportionally?
 
The Vref provides a constant voltage regardless of the supply voltage (provided V+ is large enough). So if you set Vref to 5V, it will be 5V if V+ is 8V and Vref will still be 5V when V+ is 15V.

I just read in the datasheet that RHI should be < V+ - 1.5V, so you'd want to use Vref or a voltage dropper. I've attached the schem. if using Vref. The transistor is just to allow sufficient current to go to the resistors without affecting the LED current too much (the the LED current is proportional to the current flowing out of Vref). You may wish to put a capacitor between RHI and GND if there are noise issues, however there may not be any...
lm3914.png
 
why did you use an 8k2 resistor for R2? I like using an Vref of 5V so I'll be able to use 1/4 watt resistors. If I followed the diagram on the data sheet and use an 3k83 resistor would I still need the transistor?
 
R2 can be changed to any value you want, but you should keep the transistor. Because the LED current is 10x the current drawn from the VREF-OUT pin, it is important to buffer the reference with the transistor, or your LED current will skyrocket when you press a switch.
 
I put in a 3k9 and hooked everything up on my breadboard, I get 4.9V for my Vhi. Now the problem is that the resistors are not tripping the correct step on the chip. I replaced the resistor for a potentiometer and while it can turn a little over a 180 degrees the lights only start illuminating when on the final 45 degrees of rotation. I double checked my Vlo and it is properly grounded.
 
If you have a linear pot between RHI and RLO, with the wiper to VIN, the LEDs should come on with roughly equal rotation of the pot. If you're using a log pot then the changes of the LEDs lighting will be bunched up toward one end of the pot.

This is assuming that you haven't connected the steering wheel resistors or the 220R resistor between VIN and GND.
 
I have everything hooked up and it works great. Thanks for all your help. If anyone wants more details on what I did please just post a reply to the thread. I will be subscribed to this until the end of time.
 
A success story, sounds good to me.

JimB
 
Sorry about resurrecting an old post, I came across this thread via a google search and I believe you've done exactly what I am looking to do with steering wheel control buttons.

I'm not the most educated when it comes to knowing what components to use to get my project accomplished, but I know how to run a DMM, soldering, and most of the simple steps. Not so much an engineer.

My car is a 2001 Buick Regal LS, It has 8 buttons on the steering wheel, I am planning on getting a bluetooth remote and having the steering wheel buttons operate it. I wont be running a standard headunit so I can't use a typical steering wheel adapter.

Did your steering wheel buttons already have the resistors attached to them so you are getting different voltage out of each button when pressing it, or did you have to install the resistors onto each button? When I checked my buttons they seemed to have the same voltage no matter which button I pressed. But this is when my battery was weak from 2 days of working on the vehicle, battery is currently charging so I can test at full capacity.
 
SyndromeOCZ said:
Did your steering wheel buttons already have the resistors attached to them so you are getting different voltage out of each button when pressing it, or did you have to install the resistors onto each button? When I checked my buttons they seemed to have the same voltage no matter which button I pressed. But this is when my battery was weak from 2 days of working on the vehicle, battery is currently charging so I can test at full capacity.


If you look at post #10, you will find 12 V somehow connected to the network. If you measured the "voltage" with this 12V connected, you would get nearly 12 V for all combinations of buttons. If there was a "device" connected, you should read different voltages.

The point is, the resistance should change when the network is potential free.

If your resistances turn out nearly the same, then each output from the bar graph IC could drive a FET optocoupler. This OPTOCOUPLER can be placed across pushbuttons on your remote.
 
Last edited:
hi
i used the same schematic for the Mazda, 1*LM3914 2* 7414 2*4066. my problem is everything its okay when the jack is not plug in the stereo, when i plug it i got just 2 buttons correct the others are completely wrong, any idea please, i used i diode still got the problem
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top