Model Train Forum banner

Need Transistor Circuit

13K views 109 replies 6 participants last post by  Hutch 
#1 ·
I have a signal bridge. I want to use a 3 wire red/green
Led target signal to indicate clear route thru 3 turnouts.

The signal would be green unless any one of the turnouts
was set against the main in which case it would go red.

I have done complex circuits using relays...but I have only the
1.5 to 2 vdc LED current available...so only a transistor would
accomplish this switching.

Using a relay with double throw blades, the green would be
connected to the one making contact when relay was not
energized. The red to the contact made when energized.
The relay coil would be connected thru a diode from each
turnout indicator on my panel...thus be energized when any
or all would be thrown...but I don't have the relay voltage.

I don't know transistors well enuf to select one that can
do this...either as one unit or two...

So, I beg of you guys with solid state heads...help...

Don
 
See less See more
#3 ·
It's not big deal to diode-OR the three LED drives into a transistor and have that switch a relay to accomplish this. I'd add a 1K resistor and assuming the on state of the diodes is positive in respect to circuit common, a 2N2222A transistor should do the trick.

Connect the diodes with the anodes to the LED's, the cathode connected to one end of the 1K resistor. The other end of the resistor goes to the base of the transistor, the emitter goes to circuit common, and the collector goes to the relay coil. The other end of the relay coil goes to your positive power supply suitable for driving the relay.
 
#4 ·
5 simple options:
(1) Use a good dual coil switch machine with extra contacts.
(2) Use a slow motion switch machine with extra contacts.
(3) Use a atlas snap relay, or a snap switch machine with the extra contacts.
(4) Use the switching circuit for a slow motion switch machine to control the LED's (this is how I have done it)
(5) Digitrax sells a SE8C signal detection control board.

2 not so simple options:
(1) Add a cable or throw bar attached to the points to sense the direction of the points. and use that to control the contacts.
(2) Build a circuit to sense the point direction. The reason it's not so simple option is that the signal that's sent to the dual coil switch machines are only Powered for a split second, that doesn't leave much time for the circuit to sense that there was a change.
 
#6 ·
Need Transistor circuit

...sure do appreciate the input from you guys...but does look like
I have not made clear why a switching solid state device is
what I need...

The turnouts are Pecos with Peco undertable motors...no
added contacts...

They are close together and 'permanently' down....so can't
pull 'em...

I'm an odd ball when it comes to panels...Because I wanted the
panel to indicate turnout position...and also 'throw' the turnouts
I'm using DPDT switches for each...one side controls Straight or diversion
on the turnout...the other controls the red/green LEDs in the panel...
The turnout side also has a momentary button in the circuit for each to actually pass the capacitor discharge current to the turnout motor.
(I have since thot of a better ideas but alas...too late...and true, if I had used the tortoise motors the problem would be solved)

There are a number of LEDs on the panel...I use one higher wattage
resister for the whole panel rather than try to crowd even more
resistors behind the panel...thus the available DC voltage on the switches
is at the LED level, appx 2v or so...(source is an old DC power pack
'throttle' set to appx 4 v dc and sent out on a buss to 3 panels...may
not be graceful but it works without protest)

So...that's what I have to work with...I'd like to avoid
any relays...which is why I am hoping for a pure solid
state device solution...in effect a transistor that would
be an equivalent of an STST relay controlled by
the appx 2v dc fed to the three turnout panel LEDs...but isolated by
diodes so as not to have back feed errors.

If only I had not stopped my electronic education with Tubes...
loved those 6L6s...that, I read, some higher fi gurus are still
using.

A very Merry rest of Christmas to y'all from Jacksonville, Fl.

Don
 
#7 ·
So, you have multiple LED level voltages coming in, and you want a single voltage coming out to indicate if any of the switches are in the "out" position, right?

Take all the collectors of the transistors I described previously and tie them together. Note I'm using a 2-lead bi-polar red/green LED.

Try this circuit.

 

Attachments

#8 ·
Gunrunnerjohn

You are a man of many talents...sure glad to have your help.

Actually...the control panel LED circuits are supplied by a buss running appx 4 vdc
input thru a resister and the various switches resulting appx 2 vdc at the LEDs.
It is this voltage I would use to input the solid state circuit.

I am using a door bell transformer to supply the capacitor discharge voltage
for the turnout motors...think it is running 16 or 18 vac (rectified in cap dis ckt)
but could rectify for 12 vdc for your circuit...

I'm seeing the 12 vdc gnd for common on transitors and using
the 2 vdc from LED circuits from separate power supply...do we have
a problem with that?

What transistor are you specifying?

Don

Don
 
#9 ·
I use the 2N2222A as my "go to" transistor, it's cheap and robust. The 12VDC common does indeed have to be the same common as supplies your switch LED's. Can you just connect the common's together if they're independent power supplies?

There could be an opto-isolator added pretty easily if you wanted to have them totally isolated, the TIL-113 is one that I use regularly.
 
#10 ·
Physically, yes, the common of the LED 2 vdc can
be easily connected to the common of the turnout
supply voltage...each has its own ac transformer...
is there a sneak electrical 'short' with this connection?

My electronics knowledge has faded with the years...
but as I recollect you can do this with rectified dc voltages...
is that correct? If so, it looks like you have solved my problem.
The new circuit would physically hang on the back of the
yard control panel...thus only 2 wires to the Signal LED.

Sure sounds like a winner.

D
 
#11 ·
As long as it's only the common, there should be no "sneak" path. I'm assuming that the rest of the circuits are separate.

Remember, those transistors are connected, one for each of the switch LED's, so there's more than two wires. The actual LED only requires 2 wires, but it's getting to that state...
 
#12 · (Edited)
Gunrunnerjohn...

Just wanted to make sure that I am understanding
the circuit you have drawn...

I see that an LED voltage from to any one of
the transistors would cause it to pass ground to the LED thus light
the Red...which gets + thru the 1k resistor.

I see there is a positive voltage thru the 220 ohm resistor
that would light the Green...which then gets - thru the 1k
resistor to ground.

Does the grounding of that LED terminal by the transistors
negate the positive voltage from the 220 ohm?

I assume the 1 k resistors are dropping the voltage from 12 vdc
to LED voltage. But at 1 volt drop per 50 ohms...would it not
be negative voltage?

When I say 2 wires to the signal LED I assumed the rest of
the circuit to be at the panel...thus the 2 LED leads are
simply longer than normal...about 4 feet.

Do apologize for my density but like I said...it's been decades
since doing any serious electronic thinking...

And do appreciate your efforts to guide my shaky hand.

Don
 
#13 ·
When the transistor conducts, the voltage on the collector will indeed be very low. Also, remember that the LED doesn't need 20MA to light nice and bright, I was thinking less for this task. If it's not bright enough, you can reduce the value of the two 1K resistors, say to 470.

Yep, you can run the LED leads as long as you like, at least within reason. :)
 
#14 ·
John

Hope your memory hasn't faded on this project.

I've finally got around to wiring it.

Have some questions.

The Red/Green/yellow LEDs. I have the 3 wire type.
How do I connect them to this circuit for Red or Green?

The panel LEDs run on 2vdc. That will be the trigger
voltage to the base of the transistors. Is a 1k resistor
a bit high for the circuit?

Don
 
#15 ·
You have the three-lead LED's, the circuit was made for bi-polar LED's, i.e. two-lead.

1K resistors work fine, that will give you about 5ma at 12V, plenty to light the LED. If you want them brighter, you can cut the values of all the resistors, but remember that the current will go up all around.
 
#16 ·
That's what I thot, not a problem, I'll just switch to the 2 lead LEDs.

By the way, I added an additional transistor because there are now
four turnouts that are on that track run. I copied your circuit and
made it identical to your three. Don't see why that should affect
anything. If I'm wrong let me know.

Thanks again for your help.

Don
 
#20 ·
John

More tests:

First I tried a direct feed from panel LED
to the Base bypassing the resistor. That
shorted out the Panel LED circuit, all went out
but came back when I lifted the wire from the Base.

I then bypassed the 1k with a 220 ohm to the
base. No effect.

Next, I tried the 12 vdc thru a 1k resistor to
a base (after disconnecting the panel LED feed).
That did cause the LED to burn a tad brighter. It
did not change the color.

Then the 12 vdc thru a 220 ohm That also caused
the LED to burn brighter yet, but no color change.

Finally, I hit the base with the full 12 vdc, a momentary
brightness then back to normal. Think that fried the
transistor because further tests with the 220 did not cause any
change in brightness.

I'm at a total loss.

What say thee?

Don
 
#22 ·
John

More tests:

First I tried a direct feed from panel LED
to the Base bypassing the resistor. That
shorted out the Panel LED circuit, all went out
but came back when I lifted the wire from the Base.

I then bypassed the 1k with a 220 ohm to the
base. No effect.

Next, I tried the 12 vdc thru a 1k resistor to
a base (after disconnecting the panel LED feed).
That did cause the LED to burn a tad brighter. It
did not change the color.

Then the 12 vdc thru a 220 ohm That also caused
the LED to burn brighter yet, but no color change.

Finally, I hit the base with the full 12 vdc, a momentary
brightness then back to normal. Think that fried the
transistor because further tests with the 220 did not cause any
change in brightness.

I'm at a total loss.

What say thee?

Don
 
#25 ·
Don,

Well, I was curious what happens, as I can't see anything wrong with the logic.

I build the little circuit on a proto board and tested it. I'm using a 12VDC bench supply to power it, and another supply has 2VDC out for triggering the color change. This circuit is exactly what my schematic shows.

GUESS WHAT! It works exactly as I expected. I only used one transistor, but additional ones that aren't turned on appear as an open circuit, so they'd have no effect on the operation. If you want red when it's not triggered, just turn the LED around.

I knew brain cells are dying at my advanced age, but I didn't think that many had died. ;)

The threshold voltage for the LED light change is around .7 volts, that's the junction drop of the transistor. Below that voltage, you get no change, above you get the alternative color.


Below are the pictures of the test.

Transistor input open (tried it grounded as well)

Electronics Wire Technology Electronic device Plant



2VDC input to 1K resistor on base of transistor

Technology Electronic device Electronics Circuit component Breadboard
 
#26 ·
John

If I had any, I'd pull out my hair.

I have no idea why my lashup failed. I'm going to try
just what you did with only one transistor and see if I
can get the same result. I couldn't make out the color
codes, but I assume you used the same 1K resistors
and the 220 ohm from the 12 vdc +.

I note that your transistor s a round metal device. The transistors
I used are a smaller black round with one flat side style. Could
there be a difference in the 2 devices? It's numbers were
what you specirfied.

My first thought is that I had reversed the emitter and
collector leads but that is not the case. It's such a simple
circuit, it's easy to trace and I've done that a hunnert times,
it seems, but obviously something is amiss.

In the meantime, I found a 1.5 vdc DPDT relay for $3.83
at Mouser Electronics. I'm going to try using it with a diode
feed from each of the 4 panel LEDs. I think that will work also.

Don
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top