3 aspect model RR signalling help plse

[Raiway Pete]

Reply to Angie1199

Hi Angie,

I did not see a jumper across the output of the bridge rectifier. Maybe I'm old and my eyes are dim but I just saw two diodes "pushing" up against each other and another two "pulling", the current flow being zero. I'll take a closer look.

I am assuming that you have parked cars on a block without an engine.

The circuit should work OK, but the maximum value of resistance should be around 1.5K in the load circuit. Higher than that and the there won't be enough current available to "light" the photocoupler's LEDs. You have to perform a balancing act between the number of cars per block and the max current your DCC power supply can hande. If you want to detect a single car then to make sure that at the detector fires cleanly you should put a 1.5K resistor across all four of its axles. If three fall out then you still have enough OOmph from the last conducting one to fire the detector. When all four axles are conducting then the total resistance per car will be 375 ohms.
Now when you have ten cars like this on your layout the total resistance will drop to 37.5 ohms. Make it twenty and your 2 amp booster is supplying half its max load. AND you havent moved an engine yet. Yikes!

Lastly the 1.4 volt drop is constant and applies to each block on your layout. the 1.4 volt drop is not accumulative. If you use this detector then all your blocks will have 14.6 volts across the track. No Problem.

 

[Raiway Pete]

Jumper

Angie

There is no jumper across the output of the bridge rectifier in the diagram you sent me.

Go to the junction oF D1 and D2. Mark the spot as A.
Go to the junction of D3 and D4. Mark that spot as B.

Connect a jumper between A and B.

Before doing that ask your simulator to determine the voltage across your photocoupler's LEDs. I bet it will be around 14 volts. That will smoke them. After you add the jumper it should be around 1 volt.

 

[angie1199]

curious, how did you get the diodes at 45degrees? angles??

It's a dodgy story

I did the squarewave thing but it still looks like DC when I simulate in AC. Is this right?

 

[angie1199]

The circuit should work OK, but the maximum value of resistance should be around 1.5K in the load circuit. Higher than that and the there won't be enough current available to "light" the photocoupler's LEDs.

Aha, ok, so this particular circuit is ok if you only want to detect a loco. Add wagons and it's useless. So we're back to the Twin-T which I have changed to PNP power transistors, 35v 4A, and it's still not switching the relay

 

[angie1199]

Angie
There is no jumper across the output of the bridge rectifier in the diagram you sent me.

There is! I see it, it's the horizontal line. This is why it says -1.79v at VF3. Or am I looking at the wrong place?

 

[angie1199]

The circuit I posted was designed with 2N4401's. Hfe on the order of 100. Also have used 3904's, 2222's, a few others over the years.

Darlingtons have extraordinary gain compared to bipolars. But there is some loss of speed. I had a couple of applications where I used MPS-A27 darlingtons. They work well enough to drive a relay. They didn't work reliably driving the logic transistors. Base-Emitter voltage is higher than a bi-polar.....

Artificer,

I think the schematics you put up are a little too deep for what I need. I only want to switch on an led in a photocoupler, the signalling circuit I already have will do the rest.

Also, the reason I haven't commented on them is I don't understand them, I'm so new to this stuff I understand transistor, just

 

[Raiway Pete]

Aha

I was looking at the wrong (earlier) diagram. But the same operational conditions apply.

I checked the twin-t circuit. It looka fine but I have some questions about the symbology and the relay.

The relays I use are 24v DC and have a coil resistance of 2.5Kohms. They work well at 18 volts. Your simulator might not like that though. So I suggest you tell it these facts (somehow).

Your simulator also uses a 16volt power supply (I think - 'cause VG2 is 16).
Make it 24 volts and try it.

Make the relay coil 650 ohms too. The cheap relays with multiple contacts have these coils. I used them too and they work well.

I don't know the symbol you are using for the track. Is it open all the time or what? I suggest you remove it and put in a 100 ohm resistor and a push button instead. (push to test). Right now. if the symbol has continuity, you are shorting the DCC power across the Twin-T transistors.

On your old original Photo coupler diag your DCC input frequency is 50 Hertz. Er Um No. DCC is 9600 baud which is around 10,000 Hertz (10 kilo cycles). Make your simulator assume that too.

Apart from a lower (16 v) Twin-T power source, an unknown relay coil and a maybe open connection across the track I see nothing wrong with the circuit.

 

[angie1199]

I was looking at the wrong (earlier) diagram. But the same operational conditions apply.

I checked the twin-t circuit. It looka fine but I have some questions about the symbology and the relay.

I'm trying to keep to 16VAC for the DCC side of things and 12VDC for everything else. That's why I liked the comparator one I originally mentioned. It can run on both the DCC 16vac and 12vdc.

To be honest I haven't been able to get one circuit running so far. It seems like this is a non-runner of a project. I may have to bite the bullet and by the detectors then try to work out how to connect the things to my signalling circuits.

I didn't think it would be that difficult to light an led from a 16vac circuit, but I feel useless as I have no idea what I'm actually doing. I don't know how the different components interact, like the diode in your circuit which as far as I can see simply shorts or bypasses the relay.

I can throw components in a schmatic and simulation app but have no idea why they don't work when they don't work. I can change components but I have no idea if what I try will make it work.

This was the circuit I wanted to use. **broken link removed**

Nice 'n' basic, and the creator has tested it. Now I have no idea why it doesn't work in simulation. Maybe I should just build the thing and pray. Might be reality makes it work.

 

[angie1199]

Ok, I got your circuit working but the VF readings are way off your indicated values on the schematic.

A couple of things I'd like to change.

Firstly I'd really prefer using a photocoupler instead of relay.

And second, does it have to be a huge 16VAC secondary supply? I have 12Vdc for the signalling circuit and hoped to use that as it's already going to have that supply to the signalling circuit.

 

[MrDEB]

I agree wth you Angi
I can't get the Twin T to work either.
Going to try the ORGINAL circuit (your last post) using LM741 op amps as comparators.
will get back to you

 

[angie1199]

Ok, so here's a modified twin-t.

I've retained the twin-t and inserted a led and phototransistor (to replcate photocoupler).

Changed power supply to 12vdc.

Forgot to change the R8 resistor to 470R but it still works either way.

Put 100R resistor as load.
Switch open - No loco (v4e jpg)
Switch closed - loco on block (v4f jpg)

All works in sim. So with all that in mind, are the VF readings ok or does something need to be 'hacked'?

As I say, I just throw in a few components and hope

 

[Artificer]

Lets try this again

The attached file is the basic Twin-T circuit. The previous files were intended to show the circuit in place and illustrate how the Twin-T integrates into a system.

This circuit will detect a dry finger on the rails, in most conditions. It has been known to give false positives in high humidity. The 5K6 resistor is a standard wheelset sold in the US for the purpose.

Metal trucks and wheels will work just as well to put the resistance across the rails. At a considerably lower price. Some users of DCC need a resistor across the locomotive pick-ups. If the loco is stopped, some decoders will go high impedence.

Pete uses PNPs and a negative supply. I use NPNs and a positive supply. The circuits are otherwise the same. Preference is a matter of what is available and at least cost. Power transistors vs small transistors and diodes...... The theory is the same in either case.

The block "Your Circuit" is the actual signalling logic. There is a choice of using an opto-coupler, relay, or direct drive, as I do. So, three options for connection. Whatever is appropriate for "your circuit".

There are some circuits a simulator can't wrap around. The Twin-T is one of the bumble bees of electronics. According to aero engineers, a bumble bee can't fly. Make up one and try it, it does work.

 

[Raiway Pete]

Congrats you got it working.

Ok, I got your circuit working but the VF readings are way off your indicated values on the schematic.

A couple of things I'd like to change.

Firstly I'd really prefer using a photocoupler instead of relay.

And second, does it have to be a huge 16VAC secondary supply? I have 12Vdc for the signalling circuit and hoped to use that as it's already going to have that supply to the signalling circuit.

After looking at Artificer's schematics and your actual requirements I would go with his circuit. You just want a panel light. His circuit has fewer components to do this more than adequately well. My extra components merely debounce the detector output preventing relay chatter and flickering signal aspects. That circuit you sent me that uses the 555 chip as a debouncer could be used in you future expansion to drive a two aspect signal. That would probably do it for you. As for the VF readings, I should not worry too much about them being way off. If you build the circuit and use a meter you may find that they are quite close. There are some things simulators are not good at. One final note. The twin-t is very sensitive as Artificer noted. DCC square waves are very noisy. Should your system be a noisy one then we need to divert this noise with a 0.1 micro farad capacacitor across the twin-t transistors. We can get into details later but it is a very minor 'maybe' problem.

The diode across the relay is not shorting it out. If you notice its installed backwards across the the power supply. Should the relay be replaced by a resitor and LED (Like you suggested) this diode would have no function at all and should be omitted. HOWEVER!!!!! a coil is a device that hates change. When voltage is applied it initially resists current flow until the ever increasing magnetic field overwhelms this resistance. Now comes the kicker. When the current source is cut off suddenly by the transistor the coil tries to keep current flowing (remember it does not like change). The stored energy in the magnetic field generates a 'back' electro motive force (EMF) that's the reverse of the original voltage across the coil. If you have an old door bell ringer put your fingers across the coil while it's ringing. You will feel a tickle. That is the back EMF. Unfortunately this back EMF can destroy a transistor. So we insert a diode that shorts the back EMF out. The voltage cannot rise to dangerous levels because the diode conducts the current. That is about it in layman's language.

Now go build a twin-t. You will love it.

 

[MrDEB]

WE have a winner!!

it works great!!
the LED is the opti couplier or ??
the transistor base shows 35ma

 

[angie1199]

There are some circuits a simulator can't wrap around. The Twin-T is one of the bumble bees of electronics. According to aero engineers, a bumble bee can't fly. Make up one and try it, it does work.

As MrDeb says after your message, TINA does simulate this circuit well. It does work.

Thank you so much all of you. Hope this helps others out too. What a nightmare!

I added a phototransistor as the other half of a photocoupler and a second led and resistor as an activated led.

Angie's happy again.

One question though. This is with a 5k6 resistor, one wagon? I simulated 10 wagons (5k6 is parallel) and the simulator says it still works. What problems or failures etc could I experience, if any, as I don't know what readings to take from the circuit to select the correct components.?

Should the diodes be able to handle 16v at 3A as that's the current each block will have (max)?

 

[MrDEB]

WHERE DID YOU FIND the opti coupler in TINA??
and how did you get the diodes the other day at an angle??
inquaring mind wants to know

 

[MrDEB]

I would use the 3 amp diodes

I simulated with as little as 10 ohms and it comes up 1.3 amps.
would lose the ground you have in your sim post as it could (not likely but?) cause a ground loop or ??
ground not needed as your connecting directly to separate power source.
why add potential problems conflicting with the AC from the DCC?
Murphys law states that if something could go wrong it might go wrong.
now about thatopti coupler in TINA?? where did you dig that up??

 

[Raiway Pete]

Twin-T

As MrDeb says after your message, TINA does simulate this circuit well. It does work.

Thank you so much all of you. Hope this helps others out too. What a nightmare!

I added a phototransistor as the other half of a photocoupler and a second led and resistor as an activated led.

Angie's happy again.

One question though. This is with a 5k6 resistor, one wagon? I simulated 10 wagons (5k6 is parallel) and the simulator says it still works. What problems or failures etc could I experience, if any, as I don't know what readings to take from the circuit to select the correct components.?

Should the diodes be able to handle 16v at 3A as that's the current each block will have (max)?

Hi Angie,

Glad you got something working.

It's not how much current is flowing inthe track that counts. That part of te circuit can take care of itself. The diodes should be rated above the DCC boosters trip out current. If you have a 2 amp DCC booster connected to the track then a set of 3 amp (or more) diodes should be used.

The number of wagons to use? I suggest that you try and keep on increasing the resistance across the track until the twin-t LED goes out. The resistance value is the MAX resistance at which the circuit will trigger. Subtract 5% of that value to give you wiggle space and use that value as the resistance to install across your axles.

The resitance should be quite high. Mine here trips at around 65K ohms. During experimenting I tried ten passenger cars with 100K ohm resistors across each truck (4 wheels). Thats 50K ohms per car time ten is 5K ohms. I dont think my DCC booster even noticed that. Then I took out the resitors and put lighting in the cars. This takes 20 Ma each. Again times ten and I added 200 Ma to my booster load. No problems again. With the extra components on my detectors I have adequate debouncing so that my signals don't flicker. Overall I'm quite pleased with flexibility and stability of the system.

Now I have a silly question for yer. You are using an opto-isolator LED to fire a photo-transistor to light another LED. Whaffor? I mean why not just drop the opto-isolator and use a nice super bright LED in the twin-t common collector circuit? You mentioned something about panel lights and then also hooking up to some form of signalling system. I guess I really don't know what you are up to.

Can you fill me in?

pete

 

[angie1199]

During experimenting I tried ten passenger cars with 100K ohm resistors across each truck (4 wheels). Thats 50K ohms per car time ten is 5K ohms. I dont think my DCC booster even noticed that. Then I took out the resitors and put lighting in the cars. This takes 20 Ma each. Again times ten and I added 200 Ma to my booster load. No problems again. With the extra components on my detectors I have adequate debouncing so that my signals don't flicker. Overall I'm quite pleased with flexibility and stability of the system.

Now I have a silly question for yer. You are using an opto-isolator LED to fire a photo-transistor to light another LED. Whaffor? I mean why not just drop the opto-isolator and use a nice super bright LED in the twin-t common collector circuit? You mentioned something about panel lights and then also hooking up to some form of signalling system. I guess I really don't know what you are up to.

I forgot the resistance get's less the more you add.... There's me thinking it get's bigger and eventually will not allow enough current to power the led let alone a loco!

As for the led being lit by an optoisolator, the led shown won't actually be there, the collector of the photocouple connects to the detect output on the signalling circuit. When the photocoupler lights, the detect circuit is made and signal turns red (block previous to one loco is on. This in turn set's the previous block to that to yellow. If wagons stretch to that previous block, their resistors across the wheels turns that block red, and the previous one yellow.

Ta-Da

And here's the full circuit, for each block and signal.

J3 goes to block ahead of loco to switch yellow when block ahead switches red.

J2 goes to block behind loco to switch the previous block to yellow when J4 is switch by the block detector.

I'm so happy, you've all had reputation added. What a team...

Now to work out the PCB for the detector bit. That way if the detector goes wrong, I can swap just that section, likewise with the signalling part.

 

[MrDEB]

up to 89.1 K for resistance

went wild and simed up o 89.1K at 89.2 the LED goes out (off, no light, not on etc.)
he DCC voltage would have effect on this for sure.
tried varing the battery voltage from 12 - 24, didn't make any difference.