R2LC Smoke Unit Output Voltage

First, kudos to GRJ for his thorough documentation!  So now, in addition to GRJ's measurements, we have Mike Reagan's comment that:

"If you measure the voltage with a digital or analog meter (specifically pins 5/6 for AC hot and pins 3/4 for ground) you will read 12VAC. While holding down 9 you will see 14VAC (this is providing you have used auxiliary code 8 in programming and have 18VAC applied to the track)."

And Norm's data suggesting the voltage (with 18VAC applied to the track) is:

9.84VAC (normal) and 16.69VAC (boost).

I can see why this can be confusing to the casual observer.

One conclusion to draw from all this is to be mindful of the limitations of your test equipment.  Back in the day, AC meant a pure sinewave and DC meant a stable battery-like voltage.  Now, AC can be chopped sinewaves (from triacs) and high-frequency track AC (DCC squarewaves) and your father's Buick DMM can give unexpected results.  Likewise, DC can be pulsed and at remarkably high frequencies (such as for DC motor drive).  A typical DMM's DC measurement will not meter short high voltage pulse in a way that tells you what's really going on.

I do have one quibble with  Norm's chart, the normal voltage output of the R2LC is not AC, but rather DC.  That's what started this whole thread in the first place!

I probably should have dragged out an 18V transformer for all the tests so it would match with all the other stuff, but I was just testing on the bench, and my old 1033 is what powers the bench.

If I had to describe the output voltage, I would have said half wave rectified AC and boost is AC. The final answer was in the first post in this thread...use an oscilloscope if you want to know the output voltage. "A scope picture is worth a thousand digital meters".

I agree Chuck.  That was my point too, more precise description of the wave form.

One thing about Norms data is we don't know the transformer (chopped) used and what device measured the voltages, though they are similar to Johns.

Would be interesting to see if a pure AC 18V source like the PH Brick approaches the 12V output Mike states.  On Norms chart the incremental increase in smoke voltage was going up .4-.5V per 1V AC, until 19VAC.

John, One last question did you use a smoke unit resistor or was it just a light bulb?  G

It was just a light bulb, I started the "project" just to demonstrate how the triac reacted in normal mode and boosted mode.  I wasn't going to absolute voltages.  However, there's no way under load that the voltage would be more, so I'm guessing the voltages I measured can only go down in the real world.

First of all - Bill where did you find that? I don't think I have any of that stuff around anymore. I would say that dated back to around 2000-2001. At the time everybody was trying to hop up their Lionel smoke units. These readings were taken with my trusty Fluke 87 III, a bulletproof true RMS meter, made around 1997.* This was in support of many posts back around that time on how the R2LCs of that time worked with the early version (8057) smokers. The readings are of the 'good enough for who it's for' variety and geared toward the task at hand - which was to show how the R2LC controlled the smoke units of the time when taking into consideration operating mode (command/conventional), speed steps and CAB-1 aux-1, 9 buttoning.

There was a little known speed-related 'auto boost' that was programmed into the pre-Odyssey 1 R2LCs. For some reason, it disappeared by the time the first RS11s came out. Manual boost still worked though (I think?).

This early experimentation pre-dated the ACRG controlled second-gen smokers with the heavier wicking and lower resistance elements. By that time boost was gone in all forms.

I wouldn't get too hung up on scopes vs. meters. These things are like calculators - they only prove what you already know, right?

Other things to think of - RMS voltage of half-wave rectified AC - is it really .707 times the peak voltage? I can't remember but Googling the thought may help.

*These older Fluke DMMs can be found on eBay for around $100 most of the time. I recommend this model for a good, accurate, repeatable and reliable tool for your workbench and layout. Always choose a true RMS meter if you can.

Half wave rectified sine wave RMS is peak/2.

I have an 87 at home too...bought it for $100 as a matter of fact at a hamfest in Ft. Wayne, IN. The vender had a pile of them...picked out a like new one. I think they were near $400 new.

That makes sense cjack and that would help explain that boost is not exactly X2 normal. I'd say the old 87 didn't too bad showing the relationship there.

I have my old 87 for home/hobby use and I use a 189 out in the field for troubleshooting and calibration.

One of the meters I was using was a Fluke 8012A True RMS meter, it got very similar readings to my Fluke 77, I've had both of them for a long time.

"I wouldn't get too hung up on scopes vs. meters. These things are like calculators - they only prove what you already know, right? "

Norm, This is true, but what is funny is that we should be seeing about 12+ Vrms for a half sign wave of 18VAC input.  Which is the value Mike gave.  18VAC x 1.414 is a peak of 25.5V.  If a half sine wave 1/2 the peak is about 12.5VAC.  Yet we seem to read consistently a lower True RMS value of about 10 Volts.

But this has been health experiment, and I think it helps explain why Lionel smoke units working on an AC input and single resistor value, don't seem to work as well as the MTH PS-1 equivalents that uses DC and a changing resistor value to compensate for high voltage operations.

Lionel has to pick a higher initial resistor value to handle the higher voltages to prevent burning out the unit, yet this seems to under perform when lower voltages are present.  G

Actually, I'd expect the true RMS value of a half-wave rectified signal to be half of the true RMS value of the full signal, minus whatever the diode drop incurred in the rectification process adds.  I'm going to ignore the diode drop for now, but it is a factor in the real world.

The RMS value of an alternating current is also known as its heating value, as it is a voltage which is equivalent to the direct current value that would be required to get the same heating effect. For example, if we applied 18V AC RMS to a resistive heating element it would heat up by exactly the same amount as if we had applied 18V DC.  By rectifying it with a single diode (or triac), I'm cutting the heating value in half, as I'm losing half of the energy.  Logically, I'd expect my 16 volts RMS to end up to be around 8 volts RMS after being rectified.  I actually get a bit more in my measurements, I attribute that to the fact that my older true RMS meter is probably not "state of the art" as far as accuracy, and the DC bias may be throwing it off a bit.  I probably should have taken the true RMS reading with the digital scope as it probably would have done a much better and more accurate job.  Next time I do such a test, I'll think of that.

I always like the 'effective heating' explanation for RMS voltages. It would be interesting to compare my older 87 against my 189 someday. I hope to get a bit more layout construction behind me before I can get back to my workbench though.

I started poking around in the smoke units with my meter before experimenting with different elements. You could get the old 8057 units to work well electrically once you understood how they were controlled. Then the other trick was getting them to work well thermally.

I will poke around my old hard drives to see if I have anything that might help. I may even have the dual element readings vs input (track) voltage for the old MTH PS1 smokers.

I thought the Vrms was V peak divided by 2 for a half sine wave.  The peak is 1.414 X Vrms.  So a 18V rms half sine wave should read a peak of about 25.45V.  So Vrms of this half sine should be about 12.7Volts less diode losses.

DC offsets inside a AC wave form is a slightly more complicated rms calculation.  Average power is easy though:-)  G

Actually, RMS is not peak voltage, it really is the heating value or actual power the waveform can provide.  You need to look at my previous definition.

Here's a better definition that I'll be able to type: http://digital.ni.com/public.n...CBEC86256CD10063FC0B

G is accurate. He's just using Vpeak to compute Vrms for tha half wave sine. As to the power, remember that it's Vrms squared over R. If you use G's values, the power is half of the 18 Vrms power.

John, We are saying the same thing.  Each time I have given a formula you have restated in a different format.

Basically, the conversions all assume a pure sine wave signal.

Because we are given a transformer that we believe is rated in RMS voltage, I just use the 18VAC RMS to calculate what V peak is.  Since the TRIAC pass half the signal, we now have a formula that works for converting Vpeak to V RMS for a half sine wave signal.  G

Way back in this thread someone (GGG?) asked where I got the 18 ohm heaters from for converting older Lionel 8057 smokers.

I got several from BoxcarBill I believe, and a couple from Chuck Sartor, or perhaps Jim Sandman.

I found in command mode using 18 VAC track power, the 16 ohm MTH heaters run a little too hot and create a rather acrid smell. The 18 ohm run a little cooler and a charge of smoke fluid lasts longer and it smells much better.

Regarding the whole RMS thing, I just want to check that my understanding is not out to lunch, if I may.

I thought that the RMS (root-mean-squared) voltage is the equivalent heating value of the same DC voltage (as stated above), and that it represents the area under the curve as far as pulsating or AC voltages are concerned.

So if we use the example of 18 VAC RMS, the peak voltage is 18/.707=25.45 volts. And the peak-peak voltage is 50.9 volts.

But it only has the heating value equivalent to a steady 18 VDC source.

And since a half wave rectified voltage has only one sine curve peak for the positive part of the cycle, and none for the negative part, it should have a true RMS equivalent value (area under the curve) of only 9 volts, even though the voltage peaks are still 25.45 volts? Does this sound right?

And a heating value equivalent to a steady 9 VDC source.

The .707 factor is correct for only a true sine curve, as was stated above. But for any other wave form it is some different factor, which is equivalent to the integrated area under the curve of that wave form, right?

Thanks,

Rod

Thanks Rod.  Yes, except there is a formula that for half rectified sine AC that V rms is V peak divided by 2.  Since the peak Voltage is the same, the V would be about 12 not 9.  G

No. A 9 VDC source has a quarter (not half) the heating value of 18 VDC.  Half the heating value would be a steady 12.7 VDC source.  12.7=0.707 x 18. As stated earlier heating into a resistor is proportional to voltage-squared.

To repeat, you need to confirm if your DMM can really measure the RMS voltage of a pulsed DC signal. Try measuring a steady DC battery voltage in the AC mode of your so-called true-RMS DMM. Many will read 0. Additionally, many (most?) DMMs measure average (not RMS) voltage in DC mode. In such cases neither the DCV or ACV mode accurately measures the RMS "heating" voltage of a half-wave rectified smoke signal.

Very good point Stan, and as it turns out, my somewhat aging Fluke can't measure the RMS of a DC voltage, so any readings I have are suspect.

I will have to check and see if the digital 'scope is smarter about RMS.

Of course, you are correct about the heating value, half the voltage would indeed be 1/4 the power dissipated, I forgot my Ohm's Law for a moment.

This all came from considering the voltage only and not the current. Power is VI and that is where the Vsquared/R comes from.

Anyway, I was thinking about how to derive the RMS of a half of a full wave sine. I think it is intuitive, but you have to go back to the original graphical derivation of the RMS value. The Root of the Mean of the Squares. We derived Vpeak/sqrt2 for the full wave. Now the Mean of the Vsquared values are just half of what we had for the full waveform. But we still have to re-take the sq Root. So the RMS value of the half wave is 1/sqrt2 of the RMS value of the full wave. So...Vpeak/sqrt2 times 1/sqrt2 equals Vpeak/2 for the half wave rectified sine.

Well, you guys with your bench scopes, how come you're not checking A1 vs. A2 (or M1/M2?) on the triac? I also think you need to simulate the load along with 5V regulator circuit hanging off there. Do you guys remember how these fans wouldn't even run if the element opened up?

The lamp load would be more suited to the R2LC programming for cab lights, etc. I wonder how the output would look for the different behavior modes (reset codes).

Norm, Which is why I asked if it was the heating element or a bulb.

Attached is a theoretical performance of a MTH PS-1.  I know I have seen a chart that measured actuals, but I can't find it.  Again this gets back to optimizing a Lionel smoke unit for use in conventional for me, but I am not sure it can be easily done though I am thinking of using a DC regulator to drive it.  G

Guys, I wasn't trying to do a full analysis when I started measuring stuff, I was just looking at the waveform to determine if it was indeed half-wave rectified DC, that contention was challenged in a previous thread.   This discussion has morphed into something much more, however lots of interesting stuff has come out of it.

George, no reason you can run a smoke unit on DC, the standard Lionel smoke unit will run on DC, and the fan regulator will even work properly.

Can you even turn a triac on in a steady state? Or does it need to be pulsed in some manner? Lionel liked to flip flop the firing of those things. If you really want to see some weird stuff, play with their old LCRX boards!

They turn off when the current drops below the holding current. That would be about at the zero crossing. So you need to turn them on for each half cycle with a properly phase shifted trigger of the voltage waveform you are trying to control.

I have copies of your LCRX work, good stuff!  G

John,  Yes I know.  What I am trying to come up with is one that will smoke well regardless of TMCC mode or conventional, yet smoke low/medium when in neutral and yet increase to the sweet spot at normal train speed in conventional.  I have the 8057 models, but with 8 ohm resistors.

I ran one off direct AC using the Z-4000 and had about 1 amp draw at 8 VAC (Z-4000 meter).  Smoke was about what I wanted for idle condition.  I had repacked with new batting.  Just doesn't seem to smoke as well as the PS-1.  Maybe Rod is correct and I need to use the 18 ohm resistor Lionel has.  G

I'd like to get some of the 18 ohm resistors and see how they work.

I've been tempted to marry my motor sensing circuit with the smoke unit to allow it to "idle down" when stopped, but run full-power when running.  That would be more desirable operation for me.  Most of my stuff is targeted at command operation.

I take it the new Legacy engines do this. Idle smoke at idle and really smoke when running. They change the heat and the fan...or just the heat or just the fan?

If I recall that circuit correctly, could be a candidate for a 555 PWM circuit driving a power FET.  That way you can set the running PWM % to something less than 100% (as an alternative to switching to 18 ohm heaters).  And then drop to an even lower idle PWM %. If starting with half-wave rectified AC as the power source, it would be an interesting waveform - PWM on top of half-wave rectification. Warning: your DMM might get a migraine measuring it!

They vary the heat, but I "think" the fan remains at a constant speed, at least for diesels.  I don't hear a pitch change for my noisy U-boat fans when they stop, but the smoke volume does taper off.

If I were doing this, I'd be tempted to run off track power and just rectify and filter the smoke output for a control signal.  The rectification would just be for the boost full-wave signal or if it happened to start up in conventional mode.  Then I could more or less depend on having plenty of power available for the smoke function.