banging my head, 1 wire, 2 functions

The issue is obviously the fact that you basically have two resistors (the relay coils) wired in series.

One way to eliminate that is to add transistors.  You can add a PNP ((+) switches it on) and an NPN ((-) switches it on) transistor and then wire the relays to fire off the transistors.  The transistors will isolate one coil from the other and thus eliminate the feedback. 

2n3904 for the NPN and 2n3906 for the PNP

Obviously a bit more complicated then using diodes alone.

In the diagram above the 3906 will turn on when a - voltage is applied and the 3904 will turn on when a + voltage is applied to the base.  The transistors collector output is inversed from the base input - I'm guessing that is why you think it is backwards.

When you twist the wires together do both relays activate or just one?  Is this with or without the switch wire connected?  When I get home tonight I'll try to build the circuit and see what kind of results I get.

KPierson wrote:

---

Yep, I messed up, sorry!

---

I gave you an out.  

i am an idiot wrote:

---

KPierson wrote:

---

Yep, I messed up, sorry!

---

I gave you an out.  

---

haha, thanks.  I guess I'm just so used to messing things up it doesn't bother me at all anymore to admit it.

djfearny2 wrote:

---

and ok so the 3096 needs neg to activate so shall i have that one controlling the relay that uses a negative input to negative output? cause the signals dont change the shifter is neg so i have the switch set up to throw neg and the air horns are pos so again the other switch is setup to throw pos.

---

I'm not exactly sure what you are saying here.  I think what you are asking is if you should run a ground through the relay that uses the 3906 because it needs a ground to switch and the answer is it doesn't matter.  The coil and the contacts are completely isolate so you can run anything you want through each set of contacts.

I hooked it up here at home and I got the same results - it works great until you tie the bases together.

That got me looking at the diagram and we actually have the same problem - voltage is backfeeding through the top transistor out the base in to the 2nd transistor.  I appologize for not catching that when I drew it up.  I'll keep messing with it and see what I come up with.

I don't know how involved you are willing to get with this, but I have another idea that involves using an op amp as a two voltage followers.  The outputs of the op amps can then drive the transistors above.

Have you tried the relay diagram with the diodes?  If that works that would be quite a bit simpler.

djfearny2 wrote:

---

im into it until it is accomplished.

I have a dodge and they are notorious for multiplex systems. one wire with multible functions via ground resistance. I can wire the switches for anything i need or want. Do you or anyone know of a system or setup that can be done to create a multiplex system.

---

Multiplexing gets a bit more involved and can be done in several ways.  For a voltage multiplex you can use a microcontroller with analog to digital converter or you can use op amps set up as voltage comparators.  I've personally done it both ways and it worked both ways.  When I went the op amp route I used to op amps for each multiplex input and created a voltage "window".

Since you are only using battery voltage and ground I think it can be done with one op amp chip (with multiple op amps on it).  I'll try to draw something up later today.

I would give this a shot.  The op amp (available at Radioshack stores) should provide isolation between the transistor bases.  The op amp will be set up as a "voltage follower" on both channels and will output whatever  you put in to it.  So, Pin 1 and Pin 7 will mimic your wire coming from the steering wheel.  The seperate, isolated channels should stop any feedback.

Zener diodes are rated for voltage and wattage.  I have no idea what you need for what he is suggesting.  But following should help to get a number when you get a voltage and wattage.

https://www.fairchildsemi.com/sitesearch/fsc.jsp?command=eq&attr1=AAAFamily&attr2=Zener+Diode

dualsport wrote:

---

You can put your two 470 ohm resistors in parallel to maintain the same current to make sure you drive the transistors into saturation. If your relays require a higher current you should decrease the resistor value to get more drive current.
Measure the Vce voltage across the transistor that's turned on to make sure it's driven fully on, should be close to zero volts.

---

You should be able to leave the resistor right at 470 as long as it is a 12vdc system.  These transistors (from memory) have a gain of 35 so saturation really isn't an issue (theoretical output at 12vdc and 470 ohms is ~900mA).

Can you explain how the top PNP transistor will be biased through the zener?  I understand that the bottom zener will start conducting after the break down voltage is passed (6v in this case) but how is the top zener going to start conducting the (-) signal required to bias the PNP 3906?  Also, if you have 12vdc feeding from the collector of the 3906 to the base won't it avalanch both diodes anyway?  I'll have to add this to my list of circuits to build to see if it works out - I must say I'm skeptical at this point.

It's the same thing with the top zener when the input being supplied is a ground signal.  As long as the voltage across the zener is greater than the breakdown voltage, it'll start to conduct and pass current through.

For the top zener, the voltage being applied is 12V in the standby condition, and then when the ground signal is applied, the anode of the diode tries to go to zero through the biasing resistor, so it'll conduct and maintain the 6.2V across it.   Pretty much like ground switching a relay with the other side of the coil at 12V vs. switching it on by applying 12V to it with the other side of the coil grounded.

Because the zener drops the voltage by about 6V, then you would need to use 6V as the driving source for calculating the bias resistor requirement.  If the gain is 35, and the relay needs 200mA, then 6V would be ample for the base current.

Since you already built up the original circuit it should be easy enough to try out.  Give it a whirl and see- mouser has the diodes for pennies- :)

If you have a very high supply voltage, you may want to use a higher zener voltage to prevent conduction in the standby condtion (basically the two zener voltages added up should be greater than the supply voltage)

I built the op amp circuit and it worked!  I built it exactly as shown with no added components.

I also built the zener diode circuit and it worked as well.  However, you have to be VERY careful when picking your zener diodes.  I only had 5V zeners on hand so I used them.  At 12vdc the output of the NPN would latch on and stay on until I activated the PNP.  I slowly turned the voltage down and around 8.8vdc things started working as expected.  So, I would assume that a 6v zener would only work at 10.5vdc or below.  To get up to automotive range (14.8vdc) you would need around a 9v zener to work correctly.  The low side of the input voltage with the 5v zeners was 6.5vdc.  Below that the NPN wouldn't trigger.

Thank you dualsport for not only posting the diagram but explaining how it works.  I've learned quite a bit on this thread.

Jon to answer you question about multiplexing you can use the op amps to set up a comparator circuit.  You will need one op amp to be a low threshold and one to be a high threshold.  You can run the outputs of both comparators through a NAND gate - if both inputs are off the gate will turn on to drive a transistor to drive a relay.  I've done this before for a car and it worked, although it's far from elegent.  A better solution would include an ADC and a microcontroller - far less parts count and much more flexibility.

KPierson wrote:

---

---

The problem with putting pull up/down resistors on the input is that if you do, you again lose your isolation (as you would have both a pull up and a pull down resistor going from battery to ground, which would give you exactly 1/2 your battery voltage in the standby condition on both outputs.

One thing I did observe is that if the inputs of the op amp were disconnected the relays would turn on.  Installing resistors would obviously fix this situation.  However, when I connected both op amp inputs together both relays turned off and then the circuit operated as expected.  I should have measured the output voltage in the standby condition but I didn't. 

How did you end up wiring it?  Make sure you are hooking the PNP and NPN up correctly - if you reverse the collector and emitter it won't work correctly.

You shouldn't have to change any resistor values, the 470ohm should be more then enough to bias the transistor.

djfearny2 wrote:

---

ok so here is my status

It is working basically.

I had the Transitors wired opposite so switched that, and now here is what im getting

1- have to diode isolate inputs.
2- negative output not strong enough to control a relay.
3- the possitive output is resting at possitive which is ok seeing as i can just wire the relay differently,.

now as far as the negative goes, it is strong enough to control the shifter without a relay which is better as I prefer not to here the relay clicking when shifting since i use the feature often. plus I have to diode isolate the output to the shifter as well or it disables the whole manual shifting.

However if i change the resistance value can that increase the output power if needed?

---

How did you end up wiring it?  There are several different schematics posted.

When you asked about "both wires" resting what two wires are you talking about?  If you are talking about the wires comming from the collectors of the transistors to the relays then the answer is yes.  They will rest at nothing but the opposite polarity of the the output.  This is because you are reading voltage through the coil of the relay.  The actual transistor, with the relay disconnected, should rest at open.  With the relay hooked up the collector should show whatever is on the other side of the coil.