Smoke unit burn out?

Unhook the wires from the unit and read the voltage.  Take the unit out of the mix.  What you want to see is 12 volts.

I should have mentioned. This is not from a locomotive. It's from a k line building.

Do you mean to measure the voltage at the power plug that would plug into that empty terminal in the picture? That reading is 15 volts. The directions say it can run on 12 to 14 volts optimal. Not sure if an extra volt might have done this.

The way this is wired is straight accessory power from the terminal clips to the Smoke unit plug.  There is no resistor or Rectifier in line.

Just out of curiosity, did you measure the resistance of the smoke resistor?  If the fan was running, it shouldn't get that hot if it's the normal resistor.  My guess from looking at the resistor is it's a lower value and it pumping out a lot of heat.

The "transistor" is an LM78L05 TO-92 5V regulator.

 I burned out two of them. The one pictured registered .03  The other one that doesn't look as bad but got hot enough it was melting the solder,  read .29.  Forgive my ignorance. Not real Savvy in electronics. Can you advise what that might mean? They both melted, yet they have drastically different ohm readings in my opinion.

Also can you explain the transistor info you gave?

Are you really saying .03 and .29 ohms?  I would suspect they should actually read around 27 ohms, give or take a few ohms.

Like I said, the little three terminal part is a voltage regulator for 5V DC.  The diode, cap, and regulator supply the fan power from the AC input power.

I told you I was "electronically" challenged! LOL. Ironically I can wire an entire house including the service box!

Okay I scraped the crud off and got better readings and used the correct scale. Pretty sure the readings are 32.2 ohms. And 2.3k ohms.   Any idea how to interpret this and what might have made them fry?  As I said, I was thinking of replacing them but don't want the same thing to happen.

I was thinking of using this part number - 6108057200 from Lionel to replace them with. Anyone aware if this would work?

That smoke unit would certainly work on 15 volts, but it may run a bit hot as well on 15 volts.  You'd like to be supplying this smoke unit with 10-12 VAC max, especially in a building.

Train Nut posted:

. Pretty sure the readings are 32.2 ohms. And 2.3k ohms.   Any idea how to interpret this and what might have made them fry?  

The math is fairly straightforward.  The power = Volts x Volts / Resistance.  So 15V x 15V / 27 ohms = 8.3 Watts.  That seems excessive power to me for a resistor of that size.  I'd think half that much power (4 Watts) would be more than sufficient for building smoke.  You can fiddle with the numbers yourself but, for example, 10.5V would generate 4 Watts using a 27 ohm resistor.

Also, what is your power supply type and what kind of meter are you using to measure AC voltage?   When measuring AC voltage using certain hobby-grade meters, power supplies like the popular Lionel CW-80 or MTH Z-1000 will under-report the voltage so you actually provide more heating power than you think.  This could be another factor in play here.

The part number you gave appears to be for an entire 27 ohm smoke unit - $32 at the first online supplier I saw.  You should be able to get just the 27 ohm resistor for $1 or so.

Remember that the resistance can change with temperature, so you need to measure resistance when the unit is at operating temperature.

 

Also, what is your power supply type and what kind of meter are you using to measure AC voltage?   When measuring AC voltage using certain hobby-grade meters, power supplies like the popular Lionel CW-80 or MTH Z-1000 will under-report the voltage so you actually provide more heating power than you think.  This could be another factor in play 

I'm using a spare cw-80 Transformer. Not using the fixed voltage terminal. I'm using the variable outputs.

Reading the voltage with a digital multimeter. The reason I had at 15 volts is because it seems like some of the other accessories do better at that voltage than at a lower voltage. Maybe a separate transformer set lower for this building and others like it is the answer??

I don't have a CW-80 but I just hooked up a MTH Z-controller which generates a similar type of track voltage.  I set the controller to read 15.0V AC on a low-cost digital multimeter.  Measuring the same track voltage with a so-called True-RMS digital multimeter, the reading was 18.5V AC.  True-RMS is the technically correct measure to determine power going into the smoke resistor.

So using the previous example of driving a 27 ohm smoke resistor with 15V AC (or what you think is 15V AC), you might actually be driving it with 18.5V AC which is now a whopping 12.7 Watts.  That amount of power will overwhelm that 27 ohm resistor!

Train Nut posted:

Also can you explain the transistor info you gave?

Using the DC voltage mode of your meter, pull the connector at "CON2" and measure the voltage of the 2-pins.  The "transistor" generates a 5V fixed DC voltage to drive the motor.  If you don't get 5V, then change the "transistor".  These are inexpensive components....if you "do" eBay, 10 cents each free shipping albeit from Asia.  Any US mail-order electronic parts store (DigiKey, Mouser, etc.) will carry these but for a bit more plus shipping. 

If you do measure 5V DC at CON2, re-connect the motor at CON2.  Either it runs or it doesn't.  I think replacement motors run about $5.  

That makes a lot of sense, if this was a regulated supply (i.e. the "transistor" is actually a 3 terminal fixed 5 volt regulator) and it failed, then it could fail in a manner to stop regulating and put the entire input voltage to the unit. That would also explain why the fan also failed.

That package is not good for a lot of amps, I'm surprised they did not use a TO-220 case, but I would guess the fan and the heater run about 1/2 amp total. Wouldn't think it would make a lot of smoke.

Greg

Greg Elmassian posted:

Remember that the resistance can change with temperature, so you need to measure resistance when the unit is at operating temperature.

That's going totally over the top I think.  The 27 ohm resistor is the standard for Lionel smoke units, and works well at 10-12 volts.  In addition, I just put one on my bench meter, it measured 27.1 ohms.  I then used my hot air rework tool set to 300C to heat it up for a couple of minutes, it got to 28.4 ohms max.  That's pretty insignificant, let's not complicate this more than it needs to be.

Just suggesting it could change, not anything over the top. The smoke units I use run an amp, and they do change resistance.

Pardon my ignorance on this specific unit, I was trying to be helpful.

Greg

An amp, you must do some serious smoking!

You betcha!

I think my little Hogwarts command upgrade smokes pretty well.

In my "fancier" locos (I'm in large scale) I like using the MTH Hudson unit. I've never gotten my hands on anything with a bigger capacity that is built as well. In LS we have plastic ones from Aristo (not great quality), the metal one in the video from USAT (reservoir too small) and some European stuff (overpriced and still plastic).

Greg

Greg Elmassian posted:

Just suggesting it could change, not anything over the top. The smoke units I use run an amp, and they do change resistance.

Greg Elmassian posted:

That makes a lot of sense, if this was a regulated supply (i.e. the "transistor" is actually a 3 terminal fixed 5 volt regulator) and it failed, then it could fail in a manner to stop regulating and put the entire input voltage to the unit. That would also explain why the fan also failed.

If you look at the pic in the original post, the entire input voltage goes directly to smoke resistor.   The 5V regulator only carries the fan motor current.   But to your point about package choice, when you do the math, the package is marginal in my opinion. 

The TO-92 regulator is marginal in this application, and I've seen quite a few that bit the dust when repairs come in.  Running 50ma through that regulator with an input/output voltage offset of around 8 volts results in a power dissipation of .4 watts.  At 180°C/W, that gets you a 72°C temperature rise and thermal shutdown!

You guys are good, but it's too much crap to think about for me.

Yeah, did not check the circuit, so input voltage goes to the heating element, and separately (sort of) the fan burned out.

This would make you guess over voltage hurt something. Of course the advice to check the regulator makes sense.

Looking at the components on the board, most 3 terminal linear regulators have anti-oscillation caps on the output when you check the app notes.

Without knowing the complete schematic, input voltage, etc. I'm just guessing. Clearly there are people here who are much more familiar with this specific component.

If the input voltage was 15 volts to regulate down to 5, well the math says you will have dissipation issues, as John states.

Greg

stan2004 posted:

Greg Elmassian posted:

Just suggesting it could change, not anything over the top. The smoke units I use run an amp, and they do change resistance.

Out of curiosity, how did you come upon this observation?

 

Are you making custom heater driver electronics?  And are you planning to dynamically brake the fan motor to get more distinct puffs?  It sounds and looks like the motor just spins down after each puff.

 

Stan, take a deep breath and turn off attack mode. I'm new here and having everyone show me who king of the jungle is, well that's wasted effort. I just showed an example of a smoke unit.

Now you have all these pointed questions. I may be new, but I'm not stupid. I'll answer the questions out of courtesy, but if all exchanges are attacks, then no one will benefit.

This example was running from a DCC/DC decoder that does not do dynamic fan braking. I was running the heater at a constant voltage because I was testing the long term temperature stability to see what average power I could run without overheating the unit.

I use many different types of decoders and electronics, and actually the decoder pictured can vary the output used for the heater element.

By the way, when you say dynamically do you mean short the motor leads, apply a resistance, or apply reverse voltage? I think dynamic is too broad of a term if you really want to get into the nuts and bolts. There are some really wild things you can do with a fan, think about using a motor driver output with BEMF and playing with the PID parameters, you could do some wild stuff.

Greg

stan2004 posted:

Train Nut posted:

Also can you explain the transistor info you gave?

Using the DC voltage mode of your meter, pull the connector at "CON2" and measure the voltage of the 2-pins.  The "transistor" generates a 5V fixed DC voltage to drive the motor.  If you don't get 5V, then change the "transistor".  These are inexpensive components....if you "do" eBay, 10 cents each free shipping albeit from Asia.  Any US mail-order electronic parts store (DigiKey, Mouser, etc.) will carry these but for a bit more plus shipping. 

If you do measure 5V DC at CON2, re-connect the motor at CON2.  Either it runs or it doesn't.  I think replacement motors run about $5.  

Checked and it measured 5.9 dc.  Dont know if the extra  .9 v or the heat killed it. My guess is the heat.

Going to try the new Lionel unit and run voltage lower.  

Well, when you initially introduced the concept of tempco of a heater resistor and knowledge of a TO-220 package, this is not beginner level!  I thought maybe you were working on a fluid-out or fan-failure detector for a smoke-unit which can be done by sensing the temperature rise in the heater resistor.  Hence I was asking what kind of electronics you were using since it is an interesting topic.  But apparently I was in error.

As for the fan braking, again it seemed you had advanced DIY experience with smoke units so I was going to discuss/suggest ways you can brake a DC motor for the purpose of smoke puffing for, say, 50 cents in parts...depending on what kind of electronics you have.  To your question, the typical way to do this is by shorting the windings electronically with a solid-state switch and let the speed collapse as it will...nothing fancy like attempting to control the speed profile of the fan...though again that's a curious topic especially for accelerating a fan when starting a puff to make even more realistic puffs.   

It appears I offended you so I apologize.  I will now cease and desist directing questions at you!

Train Nut posted:

 

Going to try the new Lionel unit and run voltage lower.  

If you're reading 5.9V DC then you definitely have a problem with the 78L05.

A smoke resistor is about $1, 78L05 about 10 cents, diode and capacitor are another 25 cents or so, and fan motor is maybe $5.  If messing with and soldering small components is not in your comfort zone then I get it.  So curious why you're going for a new smoke unit which is north of $25...especially since, as I understand it, you already tried 2 "new" units and they both failed.   I thought the rules say "3 strikes and you're out"

I understand the idea of using a lower voltage...but based on my comments previously about inaccurate AC voltage readings by hobby digital multimeters (when measuring a CW-80 type voltage) how will you know that you are applying a suitably low voltage?

Stan, thank you and let's talk, learning and presenting ideas is fun. I don't want to derail this thread more though (apologies to the OP). Simple dynamic braking just short out the motor between puffs, but doing this with a relay will work at slow speeds but not faster. There are some fast reed relays, and a DPDT one would let you short the fan out between puffs. That should slow the fan pretty darn fast and inexpensively. I have worked with some manufacturers on some advanced features, so building this capability into a decoder is probably the fastest way to an inexpensive solution, i.e. no extra cost in a decoder.

Standalone, I'd try the dpdt reed relay and see if high speed performance is good enough.

Greg

(oh, electrical engineer and tinker-er by profession and hobby).

Greg Elmassian posted:

Stan, take a deep breath and turn off attack mode. I'm new here and having everyone show me who king of the jungle is, well that's wasted effort. I just showed an example of a smoke unit.

Stan is one of the nicest, most knowledgeable and most helpful folks here on the forum. He goes way out of his way to help others try and learn about electronics for their trains or other electronic related projects. He has helped many, many others here and never typed an unkind word to anyone in the 4 or 5 years I have been around here and I try to read all of his posts as they are always very educational. I'm afraid you have this all wrong here.

I see we were both typing at the same time RTR12...

I think you see Stan's and my responses to each other, there's no issue (I guess I'll learn people's names as we go).

Thanks, Greg

stan2004 posted:

Train Nut posted:

 

Going to try the new Lionel unit and run voltage lower.  

If you're reading 5.9V DC then you definitely have a problem with the 78L05.

A smoke resistor is about $1, 78L05 about 10 cents, diode and capacitor are another 25 cents or so, and fan motor is maybe $5.  If messing with and soldering small components is not in your comfort zone then I get it.  So curious why you're going for a new smoke unit which is north of $25...especially since, as I understand it, you already tried 2 "new" units and they both failed.   I thought the rules say "3 strikes and you're out"

I understand the idea of using a lower voltage...but based on my comments previously about inaccurate AC voltage readings by hobby digital multimeters (when measuring a CW-80 type voltage) how will you know that you are applying a suitably low voltage?

One unit was the original, and the other i took from another k line building i had. Both were used for a few years no problems and then sat in basement storage for 10+ years. Neither were currently new.    Im comfortable soldering, but seeing the electrinics are cooked, the fan died  and the circuit board got somewhat cooked. About the only thing salvageble is the housing.  Figured given that, new unit is the way to go..

Train Nut posted:

One unit was the original, and the other i took from another k line building i had. Both were used for a few years no problems and then sat in basement storage for 10+ years. Neither were currently new.    Im comfortable soldering, but seeing the electrinics are cooked, the fan died  and the circuit board got somewhat cooked. About the only thing salvageble is the housing.  Figured given that, new unit is the way to go..

Understood on starting anew.  But I'm still concerned that you supply a suitable voltage.  It bothers me a bit that, as you point out, the manual says to use 12-14V, yet Marty says you want to see 12V and GRJ suggests 10-12V.  These are guys with plenty of experience. 

In fact, as a general question to anyone, if using a CW-80 variable output (or similar "chopped" sine controller) to drive an AC accessory like this, how are you supposed to set the voltage without an expensive True-RMS meter? 

You have 2 problems, first is measuring, and the second is appreciating the nuances of PWM.

PWM gives you variable width spikes of full voltage. This can have weird affects on electronics.

Greg

is that diode a zenner?,  voltage reg.  that unit looks like a piece of cake to fix, than an mth.

stan2004 posted:

Train Nut posted:

One unit was the original, and the other i took from another k line building i had. Both were used for a few years no problems and then sat in basement storage for 10+ years. Neither were currently new.    Im comfortable soldering, but seeing the electrinics are cooked, the fan died  and the circuit board got somewhat cooked. About the only thing salvageble is the housing.  Figured given that, new unit is the way to go..

Understood on starting anew.  But I'm still concerned that you supply a suitable voltage.  It bothers me a bit that, as you point out, the manual says to use 12-14V, yet Marty says you want to see 12V and GRJ suggests 10-12V.  These are guys with plenty of experience. 

In fact, as a general question to anyone, if using a CW-80 variable output (or similar "chopped" sine controller) to drive an AC accessory like this, how are you supposed to set the voltage without an expensive True-RMS meter? 

 

Well Stan, in your bag of tricks (modules,) do you have a switching power supply module with a metered output? No reason you can't run the smoke unit on DC (of the correct polarity); it draws less than an amp. 

Might be the most economical way to protect a new unit.

Greg Elmassian posted:

Standalone, I'd try the dpdt reed relay and see if high speed performance is good enough.

A poor man's dynamic brake is a resistor.  As long as you don't mind supplying more power to the fan motor, very effective dynamic braking can be had using a resistor in the 25-50 ohm range across the motor.

I do it the fancier way with my Super-Chuffer, I use a FET to clamp the motor after turning off the fan.  However, I also have a processor handy that's generating the chuffs, so applying the braking is a pretty small increment in the costs.

riki posted:

is that diode a zenner?,  voltage reg.  that unit looks like a piece of cake to fix, than an mth.

I assume you're talking about this diode.  You can tell by the location of the polarity "band" on the diode that it is on the output side so in this case it's undoubtedly a generic silicon diode and probably of the 1N400x type.  If it was a zener diode, the diode band would be on the input side of the circuit.  But indeed using the larger voltage drop of a zener, or even a resistor, is another economical way to distribute heat dissipation when the task is to reduce a voltage.

But to your point, I agree that this particular board looks salvage-able though I'm reminded that Nothing is so easy as the job you imagine someone else doing.

PLCProf posted:

Well Stan, in your bag of tricks (modules,) do you have a switching power supply module with a metered output? No reason you can't run the smoke unit on DC (of the correct polarity); it draws less than an amp. 

Might be the most economical way to protect a new unit.

Right.  It's so much easier to accurately measure DC voltages than the squirrel-ly AC voltages coming from some train transformers.  So here are a couple options - these photos are re-cycled from other threads but I think illustrate the concepts:

This option has minimal soldering.  The left module perform the AC-to-DC conversion, the right module provides the regulation to a settable DC voltage and includes the metering display (remarkable price when you think about it).  The meter shows 5.0V DC in this case, but has on-board adjustment to set it to 12V DC or whatever.  The point of course is the 12V DC provides the same heating power as true 12V AC.

Or if soldering and working at the component-level is in one's comfort zone, the following can save a bit on cost.

/

This option also illustrates the idea of performing the AC-to-DC conversion once using a bridge rectifier and capacitor.  Then this DC voltage can be distributed to multiple DC-to-DC regulator modules with each module set to whatever voltage.  This diagram was from a thread discussing how to individually control the voltage to a variety of layout accessories.  So one widget needed 6V DC to drive a motor, some Lemax/Miller lighting devices needed 4.5V DC, etc.  A hobby-grade multimeter can make the DC voltage measurements for initially setting the voltages - in this case the "free with coupon" meter you get at Harbor Freight Tools.

The two diagrams show AC coming from 14-16V Accessory Voltage because that was the application in the original threads.  The AC can be from any train transformer, CW-80, Z-4000, PH180, etc.