Another capacitor discharge for O27 switch question

The only thing that is significant here is the resistance value of whatever combination you put together for the resistors.  50 ohms is 50 ohms whether it's one resistor or ten resistors.  Since we're talking about DC here, the coils will dissipate power based on their resistance, the resistance of the series resistor, and the total voltage applied to the circuit.

Maybe I'm just not following something. When I pair two resistors into this system, the smaller one  always gets hotter when a consist is on the switch. So, let's say I only use one resistor and it's very large....won't the COILS get hot just like the smaller resistor in the first example?

That's kind of where I was getting at with the triple 100 idea.  I'd think that it would take a larger item longer to heat up.  On the flip side, it'd take a larger item longer to cool.

Chris,

     I just went down and wired up a triple 100 for you. It's speed performance (as we would expect) is exactly the same as the 100/50 combo. Now that we know that the RS resistors can handle 525degrees, heat should not be an issue. I wasn't willing to take a chance with them and how they felt, but that high end temp is a diff. story. However, I'm awaiting the arrival later of the 75s. So, I'll try a 100/75 and we'll see how that does. And John doesn't think it will be a problem and that says alot. 

The smaller one has lower resistance, hence the greater power dissipation of the pair.

Let's take a 50 and 100 and wire them in parallel, and we'll put 12 volts across them in our little example.

Ohms Law: Voltage * Current = Power

With 12 volts across a 50 ohm resistor, we have .24 amps through that resistor.

12V * .24A = 2.88 watts power dissipation in the 50 ohm resistor.

With 12 volts across a 100 ohm resistor, we have .12 amps through that resistor.

12V * .12A = 1.44 watts power dissipation in the 100 ohm resistor.

Stands to reason the lower value resistor would get hotter, no?

That was great, John. Just what I needed.  My concern was that the coil was the smaller resistor in the system (especially when I measured it at 9.6ohms). So I was concerned that getting by with a small resistor(but larger than the impedance of the coil) in the cap system would result in more heating of the coil.  But the coils don't seem to heat no matter what. I'm hoping I don't get a surprise down the line when I actually start running trains over them and not just testing.

Roger, in a series circuit, all the currents are equal in the series elements.  Note that we're strictly talking DC circuits here, we'll cover AC in another class.

For example.  If you have a 25 ohm coil and a 50 ohm resistor in series with 12V across them, you'd have 75 ohms total circuit resistance and a current of .16 amps.  The 50 resistor would dissipate 1.28 watts and the coil would dissipate .64 watts.  The reason for that is you're dropping twice the voltage across the resistor, hence twice the power is being dissipated in that component.  Knowing that relationship, you can calculate the power dissipation for any of your combos.

That concludes Ohms Law 101 for today.

Perfect. Thanks, as always John. Do we have any homework on this?

You betcha', and I'm a harsh grader!

I'm not as worried about the resistors getting roached as I am about their temperature possibly roaching the train table.  That's why I was toying with using more.

The way to protect the table is to have them in free air with space between them and the table.  Truthfully, if the resistors get really hot, I'd be more worried about the switch coils, that's what we're trying to protect!

Ok men.....final results. Just put a 75 in combo with a 100 and had a 100/50 pair adjoining. The 100/50 is of course, faster. Not by much. It's almost as fast as you can toggle the switch. The first snap is truly a snap and the second and following ones are much slower, but the frog makes it across. With the 100/75, it takes about a second after a toggle and have it make it across. Both are way faster than you are going to need for your train to make it through and come back. You could also use your three 100s in place of the 100/50 for the same speed, but it's alot of extra parts.  I've got a consist on it now and will check it after lunch. If the 50ohm resistor is "holdable", I'm going with that option.

Roger

And John......I measured 4.3V across the resistors. For what that's worth

More importantly, what did you measure across the switch coils?

Tell me how to do that. When I put the leads on two of the terminals, I get about 17V, I think. I'll check that.

I see what's going on. With no consist over the switch, I get about 17V. If I put some cars on it (and the trick apparently is to trigger the anti-derail), I get about 1.5or so. It's easier to see the anti-derail engage on the switches with the bulbs as they light up when a train is on it. 

Probably reading an open circuit when there's nobody parked on the rails.  If you're reading 1.5V across the coils, and they read 9.6 ohms, you're dissipating around 1/4W in the coils, which is a non-issue.

All switches wired and operating. I was pleased with the bulb setup, but this is better. Much faster recharge. Chris....you're good to go.

John.....since we're discussing resistors.....I'm on to my next mod. I'm moving my K Line smoking diner to a new spot to make room for a new building. I always soup up my smoke units. Doesn't matter if it's an accessory, tank car or engine. But 2 years ago, when I did the diner, it was overkill. I have to cut the voltage to it to make it look prototypical. Otherwise it smokes like a banshee (too much). So, I thought, since I have to move it, I may as well use the opportunity to open it up and cut down the smoke (the opposite of what I always do). Would I hook up resistors to it and if so.....the smoke element , the fan or both?  I'm thinking the element as I've found most of the time, the fan runs off a VR.  Maybe I can use some resistors from the big pile I now have and didn't use.  

Roger

Roger, thanks for the doing the experimentation.  I know you've put several hours under the layout.  So, was your final configuration just putting in the paired 100/50, or did you just put in a single resistor given that they are made to take the heat?

Chris

Chris,

It's funny.....when you started this thread, I had no intention of messing with my own setup. I thought it was fine. And then I found I could make the caps recharge faster and now I'm very happy with it. Worth the time. For me, the best part of this hobby is the tinkering and modifying. Almost everything on my layout, or that runs on my layout is modified in some way. I'd rather do that than go down there and just run trains.

As for the final setup....I went with the 100/50 (10 watts for both). The single resistor idea produced way more heat than I was going to be comfortable with. In fact, the resistor pair part of the electronics is hanging from cable staples under the layout under each switch.   I can, if you want, post a short video clip of a switch toggling so you can get an idea of it's speed. Let me know.  Good luck with the fixed voltage mod. Try the "window" method to save yourself alot of hassle.

-Roger

First step is to measure the smoke unit resistance, and we can work from there.  I'd start by putting a resistor that's about 1/2 the resistance of the smoke unit in series and see how that works.

Ok....figured it had to go in with the element. I don't remember what the resistance was. When I open it up next week, I'll measure it and try one resistor. Right now, at 16 volts, it blows smoke out better than any of my engines. Not exactly prototypical for a grill smokestack. If it was on it's own power line, it wouldn't be an issue. But it's on one of my accessory circuits and they all run at 16.  Thanks, John.

So far I've converted 2 of the turnouts to fixed voltage.  Had to cut a window in the bottom to get enough wire to play with (these were Hong Kong mfd MPC era).  I have 2 new Chinese mfd MPC/LTI era to convert, but they look like the one's in Mike Reagan's video, so I might not have to cut a window.

My Digi-Key order arrived today.  Wow, these caps & resistors are big!  I was only looking at the electrical specs, not paying attention to the physical specs.  The caps are almost like C-cell batteries!  I hope to get everything switched over this weekend.

Hanging under the layout, you won't see that stuff.

John,

A question about reducing the smoke volume from the smoke unit in the smoking diner (above). I'll probably get to that this week. You suggested using a resistor on it....half of the resistance of the element. How many watts should that resistor be and....what is the advantage of using a resistor over diodes?

Roger

AC voltage rises from 0 to a peak in one direction, then falls back to 0 and rises to a peak in the opposite polarity; it does this 60 times per second. 

The reading on a voltmeter in AC mode does not indicate the peak voltage attained; it measures what would be an equivalent steady-state voltage that would deliver the same power as the rising and falling AC -- it measures the root-mean-square (rms) of the AC voltage changing across the cycle.

If you convert the AC into DC with a rectifier and then apply the rectified voltage to a capacitor, and if the capacitor is not "drained" rapidly, then the capacitor will charge up to the peak voltage of the AC cycle.  The peak voltage is 1.4 times higher than the root-mean-square voltage.  When a voltmeter is put into DC mode, its reading will be this peak voltage.

Interesting post, Ralph. Thanks!

You'd need a string of diodes I'd expect, but if you want to go to half, just use a single diode.  You'll get half-wave rectification and half the heat in the resistor.  That may be too low, you'll have to experiment.

John,

What size diode would you recommend for this and if I use more than one, would they be in series?  And I assume I would leave the fan connections as is.

Roger

For this task, the 1N4003 or similar 1A diode will be plenty.  If you want to drop AC voltage, you have to connect diodes in pairs back-to-back, each pair will drop .5-.6 volts.  If you have 16 volts AC going to the unit, and you want to drop six volts, for example, you'd need between 8 to 10 diode pairs.  This is why I suggested the resistor, a lot less parts.  For the low power of the smoke unit, a somewhat hefty power resistor should not get all that warm.

John,

Here's the electrical situation on the transformer that has been running the diner. It's a KW that mostly runs all my lighting (with the exception of this accessory). I usually have the throttle on 16V but it's actually putting out around 12 because of the lighting load (although it's only around 6amps). At that setting, the smoke unit is making way too much smoke. When I cut the voltage down to around 8 or 9, it's much better. So, if I use a resistor....start with a 10 ohm? Should whatever resistors I use be 10Watt?

Again, measure the resistance of the smoke unit and let us know what that is.

Will do, John. I'm actually waiting on it a bit. I ordered a "creeper" (after reading through another post on the Forum). I've been crawling around under there on a chaise lounge pad all this time. So, I ordered a good one and I'm waiting on it. I'll break it in with this project. LOL

Can your smoke unit operate with DC voltage?  It should.  In which case another option is to adjust the DC voltage going into the smoke unit using a DC regulator module.  These have been discussed in other OGR threads;  they are available on eBay for under $2 including shipping from Asia (search for "LM2596 module").  You would need to convert the AC from your transformer to DC with a bridge rectifier (or 4 individual diodes configured as a bridge rectifier).

The benefits of this approach are:

- more resolution via screwdriver adjustment vs. the coarse diode steps or finite choices of resistors

- regulation from changes in your AC supply which seems to vary based on your other accessory loads.  so when your AC supply voltage changes, the voltage to the smoke unit will not change as it would with the diode or resistor method

- more electrical efficiency.  these modules efficiently convert from one DC voltage to another whereas the diode or resistor dropping methods dissipate the "excess" voltage as heat which is wasteful (IMO).

More and more layout accessories now operate on DC - such as LEDs vs. traditional incandescent bulbs.  A single bridge rectifier can be used to supply the DC voltage to multiple adjustable DC regulator modules to efficiently power multiple DC accessories.

OK, I've got all my turnouts converted to fixed voltage & capacitor discharge.  Everything tests out OK so far, just have to do a re-test after I get done tidying up the wire underneath the table with my new cable tacker.  I wound up using triple 100s.  Everything runs cool, even after a 5-min jumper test.  I can get 3 or 4 fast successive snaps before things slow down.  Thank you all so much for the help!

Chris,

Glad you like the setup. New life to those 1122 switches.

Stan,

I'm intrigued by your suggestion. I'm going to look around on the Bay for one of those regulators. 

-Roger

Stan,

I was part of that older thread. I ended up (with John's help), building a VR for a small fan I installed in some Lionel Industrial smokestacks. Works great (I may start a thread about that one). In any case, we used a cap, diode and the fixed voltage VR. Adjustable seems like the way to go for this app. I'm wondering about the diode and cap.

How about this?

http://www.digikey.com/product...ADJ%2FNOPB-ND/363709

Then I wouldn't have to wait for ship from China.

Roger

Roger, those switching regulator chips require you to build a circuit around them, they're not a complete module like the eBay ones. The diagram below is basically what the eBay module consists of.

Here's the diagram from the datasheet for the typical application for the Digikey part.

If you want a couple of the ones from eBay without the China shipping delay, send me an email and I'll sell you a couple.  I have a bunch of them that I stocked for similar projects.

One suggestion, check the temperature of the switching regulator chip after running the smoke unit for a few minutes, you may want some thermal epoxy and a small heatsink to insure you don't overheat the module.  Loctite 383 Thermal Epoxy is a good choice.

For this module, it already comes with the input capacitor, so you just need the bridge rectifier.  I can toss one of those in as well if you don't have one.  You don't need a heavy duty bridge, I'm sure the smoke unit in a diner isn't that current hungry, not like a locomotive smoke unit.  I'd probably go with a 1 or 2 amp bridge, plenty for this task.

The reason I mention thermal epoxy is the top of the chip is simply flat, so you don't have anything to screw a headsink to.  OTOH, it's quite probable that the current required here will not be enough to require a heatsink.