Sorry to learn of this. I’ve been using 2 CW80s for power on my layout for years with no problems. However, I took this information down from a forum thread (don’t recall which forum), just for reference in case any problems arose.

A TA-27 triangle head screwdriver (which you may already have)

Use 8 amp fuses

Use 40mm Evercool ball-bearing fans

You should be able toGoogle all of the above and order them off the Internet.

Hope this helps and good luck.

"Why is the unit putting out voltage when the throttle is down and is there any way to correct that?"

Easy, you blew the transistor. When you do that, they blow shorted so it's full track power.

What EXACT model of CW80 do you have because that matters greatly.

Pictures from the latest new model

https://www.onsemi.com/pdf/datasheet/fdp047n08-d.pdf

Again, if this model of new CW80, then you most likely blew the matched pair of fdp047n08 (the bottom two as the picture is oriented or the rightmost 2 when level.

On the older models there is only one TRIAC per output so one is accessory and one is for track variable. Again, entirely different in the old model in that there is a heatsink and fan. The new models have no heatsink or fan.

That said, if it is a new transformer- send that thing to Lionel. http://www.lionel.com/contact

Make their warranty department aware of the failures, make them support you the customer. Their short circuit protection (or lack thereof) failed. Again, those CW80s just came out, all new design, all "new and improved".

On your old CW80, that is a run of the mill 40mm X 40mm X 10mm thick 12V brushless common fan. Because 3D printers and all kinds of DIY electronics use them it is very easy to find. https://ogrforum.com/topic/cw-80-fan-noise

One wonders why a short across the output would blow up the transformer, that certainly doesn't inspire confidence!

@gunrunnerjohn posted:

One wonders why a short across the output would blow up the transformer, that certainly doesn't inspire confidence!

Agreed John. What's the purpose of the fuse then?

@RSJB18 posted:

Agreed John. What's the purpose of the fuse then?

Besides, it's not like you never have a short across the rails if you do any amount of actual running!   I would have thought the new CW-80 would have improved the circuit protection, and that includes not killing the transformer!

@RSJB18 posted:

Agreed John. What's the purpose of the fuse then?

Same as with the old CW80. The fuse is there so when the output gets stuck on full and so now your failure is loading the transformer, you blow the fuse hopefully before a fire starts at the transformer. It is a fuse of last resort. It does NOTHING to protect the electronics, it does nothing to protect your train. It is there to protect Lionel from a lawsuit.

Same comment, a simple track short should not destroy the transformer!

My general opinion of CW-80's.

They make good door stops for the train room......

The CW-80 was the latest ("improved") model as I ordered it three weeks ago and got it a week later. And, it was not a simple short, I mistakenly turned on a PH 180 that was also connected to the track so there was a lot of current.  I don't think I am up to replacing transistors so maybe I'll have to write this off as a learning experience. Is it safe to use the accessories terminals so I can at least get some use out of it? And, does anyone have a link to how to replace the old CW-80's fan/bearings? Maybe I could do that.

Odd that it does not have over electronic current protection other than a fuse, like the much less expensive power supplies that are available for various voltages.  Probably another dollar in components in mass production to have that circuit in place and chances are the main regulator chip has an input for over current that is not used.  If you open it up and it has good component part numbers, not ones that are Lionel for common transistors or IC's, it can be fixed if you know how to troubleshoot the circuit.  But if it is all surface mount components, forget it.  I have a friend who installed one of those fancy WIFI thermostats.  It gave him so much trouble, he went back to a basic mercury stat.  You can complicate your life with modern stuff thinking it will make things simpler, but most times not, while an old reliable 65 year old 1033 powers on.

The CW-80 is not overly complicated.  It's definitely not anywhere near as complex as a WiFi thermostat.

Robert's "short" by his own admission wasn't a simple issue either.

The situation he described was not likely one that the designers anticipated with a PH-180 backfeeding the CW-80 (and apparently doing so out of phase).

Should they have thought of this as a possibility?  Definitely.

Did they?  Probably not.  It's tempting to blame a few pennies but it's rarely that simple.  Instead it's about thinking each and every possibility through carefully.

Having said this it should be obvious to us that transformers get hammered often, because so many of us are amateur electrical engineers, and just learning our craft.  As a result incorrect connections happen very frequently and must be expected.

Those who designed the 1033 knew this about their customers.  Did the folks who designed the CW-80 know it?

Mike

After reading through this, I’m glad mine are the older models.😳

After reading this I'm glad that I don't use the CW-80 for anything.

I still have some 50+ YO motors running equipment on our campus. We've replaced newer ones more than once.....

@gunrunnerjohn posted:

After reading this I'm glad that I don't use the CW-80 for anything.

HA HA HA!

@CALNNC posted:

Odd that it does not have over electronic current protection other than a fuse,...

It does have "fold-back" circuitry to limit the output to 5 amps.

(glances over at vintage ZW) (raises eyebrow)

I use 2 MTH Z-1000 Transformers on my layout and have had very good luck with them. Before that I had two CW-80's that came with starter sets that both failed about 15 months after first use. They had both cooked themselves internally. I only use the CW-80 for a AC power supply on my workbench since I can adjust it for both track and accessory power O-18 Volts AC. I use an old PC power supply for DC power, -12, +3.3, +5, & +12 Volts DC. I use them both for testing anything Before it goes onto the layout. Fuse protection on each voltage output. GFIR, Fuse, and Surge Protection on the 120 Volt Input for the entire workbench and layout.

@Robert Earl posted:

The CW-80 was the latest ("improved") model as I ordered it three weeks ago and got it a week later. And, it was not a simple short, I mistakenly turned on a PH 180 that was also connected to the track so there was a lot of current.  I don't think I am up to replacing transistors so maybe I'll have to write this off as a learning experience. Is it safe to use the accessories terminals so I can at least get some use out of it? And, does anyone have a link to how to replace the old CW-80's fan/bearings? Maybe I could do that.

It may be safe to use the transformer for accessory operation. I would test the output with a DVM to make sure it is putting out the voltage you want before hooking it up to anything and keep an eye out for any signs of overheating. Also, since you successfully replaced the fuse, depending on the track voltage you need (not sure if you're running command needing 18 vac), there's nothing that I know of that says you couldn't also use the accessory terminals for your track voltage.   

Replacing the fan is pretty straightforward once you have it opened up again. You could replace the fan bushing with a ball bearing, but with the price of replacement fans being so low, it's probably not worth the time and effort.

I would counter that and say put the best fan you can find (longest lasting high quality- reasonable airflow and noise) given that no, it's not that easy to replace. Just my opinion, but between the security screws, the fact to get to the one board mount screw you have to also remove the control potentiometer, and angling and working the board around the terminal screw posts- no it's not what I consider user friendly serviceable design. Again, just my 2 cents, but something like a Noctua NF-A4x10 FLX, Premium Quiet Fan, 3-Pin (40x10mm, Brown) would be worth it to do it once for a good long time before having to open it up again.

Just out of curiosity has anyone scoped a new CW-80 and if so is it still the same ugly chopped sign wave output?

@zhubl posted:

Just out of curiosity has anyone scoped a new CW-80 and if so is it still the same ugly chopped sign wave output?

Would you expect it to be different?  I'm not sure that I would.

It's hard to find miracles nowadays, especially at that price point.

I would expect that the budget went into fixing other problems that the original version had, and perhaps into better current limiting and voltage spike protection.

And the ugly waveform isn't actually a unique problem if all of it's similar competitors also have a chopped waveform.

Mike

@Mellow Hudson Mike posted:

Would you expect it to be different?  I'm not sure that I would.

Actually, since the old one was triac based and this one uses FET's for voltage control, I'd be very surprised if it's not different.  Maybe not better, but different.

@Mellow Hudson Mike posted:

And the ugly waveform isn't actually a unique problem if all of it's similar competitors also have a chopped waveform.

Au Contraire!  In truth, the CW-80 has by far the ugliest waveform of any of the chopped waveform units I've seen.

CW-80 half throttle

MTH Z-1000 half throttle

Lionel ZW-C half throttle

MTH TIU Variable Channel half throttle

• Channel 1 Transformer output full throttle (MRC AH-501)
• Channel 2 TIU Variable Output (fixed mode, half throttle)
• Channel 3 TIU Fixed Output

GRJ,

Ugly is ugly.

On the other hand you have a good point.  These differences are likely to be a matter of filtering and that may indeed be a matter of a few cents, or perhaps a dollar or two, worth of components here or there.  Not much in the broader scheme of things.

Mike

@zhubl posted:

Just out of curiosity has anyone scoped a new CW-80 and if so is it still the same ugly chopped sign wave output?

Yep, did that the first day I could my hands on one.

Full Throttle screenshot

Partial power

For for back to back comparison the old fan based CW80

Full power screen capture

Partial

Also, specific short circuit and overload testing I performed on the new version of the CW80.

Testing using a common 3D printer heated bed that has either 12V or 24V paths. The 24V path is 6 Ohms cold resistance, and the 12V path is 3 Ohms resistance.

Some quick tests using a Fluke 101+ meter (doesn't have AC current measurement so going on voltage over known resistance)
With nothing more than a small light bulb for load output is 18.98V AC RMS max throttle
With the heated bed 6 Ohm load and the lightbulb output dropped to 18.0V and did not trip overload. 18V into 6 Ohm load should be about 3A = roughly 54 Watts
With the heated bed 3 Ohm load, raising the throttle to about 50% on the lever results in 7.5V and not quite tripping the red blinking overload light. Raising it higher to just above 8V triggers the red light overload blink but power is still being provided for 5 seconds. I ran this test several times, a slow load based overload trips in 5 seconds. Just did another reading by removing the load at this setpoint to measure the unloaded voltage VS loaded voltage at the trip point. Test condition throttle at 50% mark, results in 10V unloaded, and then adding the 3Ohm load drops to 8.5V with the light blinking. In theory this is about 3.33A, I didn't have my good ammeter on hand for actual measurements.
The bad part IMO, the overload light blinking starts a 5 second sequence and then power is dropped.

Further short circuit testing:
Raise throttle to 100% with 6 Ohm load plus lightbulb indicator.
Heated bed dual load is approximately 2 Ohms.
Simulate overload with 3 Ohm in parallel to existing 6 Ohm Load= Light dims and voltage is about 9V output, red light blinks, holds load for 5 seconds, then output is off, red light solid. Must turn handle to 0 to reset.
Simulate intermittent short using 3 Ohm load on top of 6 Ohm load. Counting seconds how long the intermittent short is loaded.
This is basically a 2 Ohm load.
1 Second, light blinks, returns green after overload is removed. Power is not interrupted.
2 Seconds, light blinks, returns green after overload is removed. Power is not interrupted.
3 Seconds, light blinks and even if load is removed, continues blinking for 5 Seconds and then power shuts down and red light becomes solid.
Again, the 3 second test, the power continues for 5 seconds and then is shut down.

Using 30 feet of 14 gauge solid house wire as a short load- touching it across the terminals instantly blinks overload switching to solid red and output is shut down.
Again, a very low resistance short results in an instant shutdown.

Further testing accessory side on the New CW80 Fanless transformer:
Works basically identical to the track power side from a safety or overload perspective.
Same set operation at previous CW80. Hold the 3 buttons DIR, HORN, BELL, light blinks green, raise handle to desired accessory voltage, release buttons and lower handle back to 0.
Max voltage with only a lightbulb was 19V.
Set to 12V for testing
6 Ohm load no issue.
3 Ohm load red led blinking and then shutdown in 5 seconds.
Once accessory output is tripped you must power cycle the transformer to reset. -Different than the track power as lowering the throttle to zero will reset a fault on that channel.

Again, my impression is, the new CW80 is an improved design over previous CW models. The new control board design that is fanless is a big bonus and it's not just about removing the fan, the overall circuit is just a lot better. That said, this is still a microprocessor controlled device. The firmware and design decisions put into that ultimately control how this works.
The choice of how to handle intermittent overload and or shorts, and when they trip off VS allowing the load could probably be a multi page debate.
Because the bell and whistle as well as the direction button tell the microprocessor the user input, and then the processor controls the output- there is a delay between user and output. I don't have good numbers on what that is, but it is there- just like the previous CW. That matters to MTH users trying to perform specific bell and whistle sequences.
My gut says that overload works, however, I could also see how you might be able to smoke some thin Lionel lighted car wiring in a derailment. They made and sold both parts of the system- the transformer and the rolling stock, and absolutely, between the thin wiring used for the Fastrack power entry section, to the track joints, to a derailed lighted modern passenger car like a polar express, I'm not 100% sure that the 5 second overload timeout will kill power in time to prevent smoke.

Again, another question asked to me by friends what was the voltage output and as measured, I saw a max of almost 19V, under load dropping to 18.29V.
In theory, overload is kicking in somewhere above 3A and without my good meter handy with an ammeter function, I had to go on measuring voltage across the load and that's not the most accurate method to measure.

My thought is that it's maybe possible with too much accessory + track load to blow the internal fuse that is on the output lead of the transformer before the new control board. I did not try this destructive testing.

I did load the accessory output at 12V setting with the 6 Ohm load, and then applied the 3 Ohm load to the track output at 50% power and tripped the overload red blinking LED and did not blow the fuse and the accessory output and track power both were operating.
I have no concern of it not meeting the labeled 80 watt output. That rating might be combined power and so you may or may not get exactly 80 Watts out of one output or be very close to tripping overload.
I didn't have a fine variable load to achieve that kind of testing.

Also, temperature rise testing since the new one has no fan.

Just for reference, light load testing for temperature of the transformer rise.
I used the accessory output set to 12V RMS unloaded measured with the Fluke 101+ and a small 20ma light bulb just to visually see what the power did or changed.
I used the 6 Ohm load. Ran now for 30 minutes. Transformer is cool to the touch, has not risen in temperature. This was an approximate 24 Watt load. 12v, 6 Ohms, 2A

Pretty impressed that going from the old fan cooled type to this with no vents of any kind, no heatsinks on the FETs, this appears to be a nice cool running transformer.
Again, given I am using the transistors to reduce the output to 12V from the internal 19V, this shows that their switching losses under reasonable loads is very low.
Modern MOSFET tech at work.

Is the Fluke 101+ a true-RMS meter? I can find lot of info on the 101, almost nothing on the 101+.

Just wondering.

@PLCProf posted:

Is the Fluke 101+ a true-RMS meter? I can find lot of info on the 101, almost nothing on the 101+.

Just wondering.

The 101+ is just a 101 with extra accessories. Same basic 101 meter

Based on a quick search, "true-RMS" advertising is first seen on the 115 model- about a $200+ meter

@PLCProf posted:

Is the Fluke 101+ a true-RMS meter? I can find lot of info on the 101, almost nothing on the 101+.

Just wondering.

Not a true RMS meter.

I want to thank everyone who responded to my original questions - it developed into a detailed and informative thread about CW-80's that I am working hard to fully comprehend. Based on the information I received and following a few of the links provided I was able to replace the fan on the old CW-80. And, it still works after putting it back together! Much quieter but I can hear a little whine which tells me the new fan is operating. Buoyed by that small success I am thinking about trying to replace the transistors I burned out on the new CW-80. I watched a short video on how to desolder components and it may be within my reach. But I need to gather a littler more information to feel comfortable.

1. Does it matter which transistors I get to replace them as long as they have the same number (fdp047n08)? Any advice on brands or sources?

2. Does anyone have experience with this that would help me prevent rookie mistakes? Anything I should be careful to do and/or not to do?

3. Is there any way to check the transistors to make sure I am only replacing ones that are fried?

4. Anything else I should check/do while I have the unit dismantled?

Thanks again to all of you, your knowledge is greatly appreciated.

@Vernon Barry posted:

Yep, did that the first day I could my hands on one.

Awesome, thanks that does look better than the older one. And a big thumbs up for all the other testing you did.

too bad it doesn’t trip immediately with a obvious overload or at least like the ZW-L a thermal breaker (granted a thermal breaker isn’t necessarily the best it’d be an improvement)

before you go down that rabbit hole of repair- did you call Lionel? Might be worth the trouble to see since it is their new product.

Checking is simple- you know it's the ones for the track powered side. There are 2 per channel, so your accessory is good, your track side is bad- then it's that pair. If you want to be extra sure use a meter and look for the 2 rightmost pins to be shorted on the bad ones.

They are https://www.onsemi.com/pdf/datasheet/fdp047n08-d.pdf

Your choice of who you order from, Mouser, Digikey, there is only one brand https://www.digikey.com/en/pro...mi/FDP047N08/1853911

FETs are static sensitive so ideally you want an ESD safe soldering workbench and proper temp controlled soldering iron that is also ESD grounded.

FWiW, the easy way to understand the orientation of the board is the 3 pin connector for the potentiometer is the same side the handle is on, and that is the same side the track power channel is.

Remember, when you open the transformer, that board is facing the buttons up into the housing.

@Robert Earl posted:

I want to thank everyone who responded to my original questions - it developed into a detailed and informative thread about CW-80's that I am working hard to fully comprehend. Based on the information I received and following a few of the links provided I was able to replace the fan on the old CW-80. And, it still works after putting it back together! Much quieter but I can hear a little whine which tells me the new fan is operating. Buoyed by that small success I am thinking about trying to replace the transistors I burned out on the new CW-80. I watched a short video on how to desolder components and it may be within my reach. But I need to gather a littler more information to feel comfortable.

1. Does it matter which transistors I get to replace them as long as they have the same number (fdp047n08)? Any advice on brands or sources?

2. Does anyone have experience with this that would help me prevent rookie mistakes? Anything I should be careful to do and/or not to do?

3. Is there any way to check the transistors to make sure I am only replacing ones that are fried?

4. Anything else I should check/do while I have the unit dismantled?

Thanks again to all of you, your knowledge is greatly appreciated.

You may want to invest in a "solder sucker" for the desoldering process to remove old solder. It's hard to juggle everything with only two hands, so a helper works well.

As recommended I checked with Lionel to see if they could help with the CW-80 but since I had replaced the fuse the warranty was void. That is fair because it was my own bonehead move that fried the transistors and fuse but, it was worth a phone call to make sure. So I am left with two options: 1) Use the accessory posts which I can set to any output, or 2) try to replace the transistors to get the unit fully operational again. I would like to try the latter as the parts are only $6 for the 2 transistors and $10 for a solder sucker. But I do not have a ESD safe solder station and my iron (Hakko FX888D-23BY) is not grounded to my knowledge. Does this mean I should give up and just use the accessory posts or is there a reasonable chance I could succeed even without the ESD items.  The soldering doesn't look too difficult but if I am likely to cause more harm than good I'll probably give up trying to fix it. Thanks.

It's warming up and the ESD risk is reduced compared to the dead of winter where it becomes extremely dry.

That Hakko is a decent soldering iron.

On the desoldering of the holes, I personally use the Hakko FR-301 desoldering tool but that is not cheap and I did this with much cheaper tools for a very long time. Again, this is a $270 tool, not expecting you to use that level, but if you ever get into a full time repair shop. The tool slowly pays for itself in that scenario.

As a rule, I always buy a few spares of FETs since the shipping of any one order, if you only order the 2 you need, Murphy tends to kick in.

The challenge for you is going to be first removing the old ones and then clearing the holes so you can insert the new ones. After that, it gets pretty simple just soldering the new ones in.