Protection against voltage spikes

You have the right idea Paul.  The issue with the TVS parts is they can fail open and you'll never be the wiser.  I will say, they more often fail shorted, which is something you won't miss.

Voltage spikes are still a reality, and they can still take out electronics.

Heavens, John and Paul (not george and ringo), I thought I was safe with my TVS installations.   Now I find out they fail open.  Is there a VOM test to check them out?  Thanks

If you are using them at the transformer, I would advise putting another one at the track connection. That way you have double the protection. And then replace both of them yearly just to be safe. They're cheap enough....

Even if you had great stability everywhere, a static charge could wipe a component out. A TVS helps there too.

Remember, a large 1500W TVS has significant capacitance, so if you use too many, it will start to affect DCS performance.

As far as testing for open, it's a lot more complicated than an ohmmeter.  You need a current limited variable voltage supply to test the breakdown voltage of the TVS to see if it's still meeting it's specs. 

John and everyone,

Are you using anything special to protect your trains?

I last bought the bag of diodes from Mouser in 2000 , however I think I found the right part: https://www.mouser.com/Product...jLZ6j4EQ5nBpTQ%3D%3D

Using the part # I found this on digikeys site: https://www.digikey.com/produc...179ee36dbba38034f216

Given that the voltage protection circuits in the TIU don't last and neither do the add-on diodes and that manufacturer's warranties are only a year, I think this is a good idea. Are the ones above good ones to use or what do you recommend? If I find them, I guess I'll replace them annually like suggested above.

 

 

Those are just fine Paul.

Are TVS diodes necessary if you're still running all postwar, and nothing "modern"?

Paul, 

As a novice in electronics, when you say add the diode to the banana plug, is the stripe on the diode on red/power side and blank end on common?  A picture truly would be worth a thousand words for many of us.

Ron

The TVS diodes used in this application are bi-directional(no band).  Here’s a couple of pics of acceptable diodes and install locations.  At the track connections is also a good location.  Hook up straight across the hot and common terminals.

I also place TVS at the transformer outputs( old photo-now 180 PoHos). More important I place one on the jumpered railpower distribution terminal strips between the Hot (green) and Common(white) terminal screws on a 5 position power distribution direct to rails in a single power districtI

I have this TVS PN. Little fuse 1.5KE36CA, 1500W33.3V. Are these ok to use? Or is this a old PN? Is there some other product that would be better than a TVS?

Jon1443 posted:

Are TVS diodes necessary if you're still running all postwar, and nothing "modern"?

Nope. Some modern components are sensitive little buggers. Especially memory oriented chips.

Post war is more robust electro-mechanical for the most part.

That said, after GRJ pointed it out once, TMCC seems pretty robust to me too, but for a few pennies a TVS is additional insurance worth while for command folk imo. I only have two TMCC. I don't use sound, and don't care if the boards die, so have abused them without a hitch.

   MTH's approach early on didn't impress me at all in that way. I think my brother would have saved a whole lot of cash using TVS. A few never made a full loop, dying on the switches.

feet posted:

I have this TVS PN. Little fuse 1.5KE36CA, 1500W33.3V. Are these ok to use? Or is this a old PN? Is there some other product that would be better than a TVS?

That one is fine.

gunrunnerjohn posted:

feet posted:

I have this TVS PN. Little fuse 1.5KE36CA, 1500W33.3V. Are these ok to use? Or is this a old PN? Is there some other product that would be better than a TVS?

That one is fine.

Thanks John, I thought it would but better be safe than sorry.

Thanks for the photo's and info, very much appreciated.

You may find the attached surge suppressor article interesting.  Early experiments by Otto Schade, Jr. in 1999.   Hope this works, my first attempt at an attachment.  And my first post in a very long time.

Cam

Ron, others have posted good pictures...and like Ted said they are bidirectional so you don't have to orient them.

Cam, that is a great article - I have to admit it's over my head a bit, and seem to recall reading it years ago. I saved it to my computer just now.

fwiw: I've had a number of trains over the years become expensive doorstops...mostly PS1 stuff, but some PS2 and my latest failure was a Legacy engine, which as I mentioned was under warranty so I am taking precautions as I seem to have derailments and collisions quite often (operator error The transformer circuit breakers always trip but that is not adequate protection for our sensitive trains.

Static shocks can be damaging too as Adriatic pointed out.

 

I thought there were articles on the forum for actually putting these TVS' inside the locomotive, am I in error?

Ray

I don't know about articles, but the absolute best place for protection diodes is right across the power pickups of individual locomotives.  Obviously, it's also the place that's the biggest PITA to install them!  

In general ESD protection should be close to the circuit you're trying to protect.  If your TVS is at the transformer, when the train is 30 feet away and any circuit is interrupted that has inductance, the spike has to travel all the way to the TVS in order to be suppressed.  The inductance of the entire path between the TVS and locomotive affects the ability of the TVS to effectively suppress the spike.  OTOH, if the TVS is in the locomotive, any incoming transient is first met with the TVS protection diode.

Rayin"S" posted:

I thought there were articles on the forum for actually putting these TVS' inside the locomotive, am I in error?

Ray

If you run your trains on club layouts or a friends layout then in the engine is the place to put the TVS 

John explained well;  but a other version,... across the power feeds (+/-)(hot,com) as close to the board as possible. (cars with boards or acessories would benfit with one too.

And then the more tvs equiped items run the better.

(one more excuse for a long train, silly but true 😁)

John & Trainz, 

Thanks again for your input, I need now to input these into my locomotives.

Ray

TedW posted:

The TVS diodes used in this application are bi-directional(no band).  Here’s a couple of pics of acceptable diodes and install locations.  At the track connections is also a good location.  Hook up straight across the hot and common terminals.

Ted and others,

I noticed in Ted's picture above - the Mouser P/N differs slightly from the diode referenced in the article Cam posted, above...in Ted's pictured is Mouser P/N: 576-1.5KE36CA, where the one referenced in the article that Cam posted is P/N: 576-1.5KE33CA

For a non-electrical engineers like myself, is one preferred over another?

Thanks.

Paul

 

Either will do the job, the 576-1.5KE36CA has a 3V higher voltage rating before clamping.  Typically, the voltage spikes that do the damage are much higher in amplitude than either of these clamping voltages.

gunrunnerjohn posted:

I don't know about articles, but the absolute best place for protection diodes is right across the power pickups of individual locomotives.  Obviously, it's also the place that's the biggest PITA to install them!  

In general ESD protection should be close to the circuit you're trying to protect.  If your TVS is at the transformer, when the train is 30 feet away and any circuit is interrupted that has inductance, the spike has to travel all the way to the TVS in order to be suppressed.  The inductance of the entire path between the TVS and locomotive affects the ability of the TVS to effectively suppress the spike.  OTOH, if the TVS is in the locomotive, any incoming transient is first met with the TVS protection diode.

Which begs the question, “Why do the manufacturers not install the TVS diodes in the factory?”

For a few pennies, the company could save themselves a bundle in warranty cost. 

Can you send us a picture of what the diodes look like attached to the engine?  I want to know how to correctly install.

I don't have any pictures, but they simply go from the center roller connection to frame ground, pretty simple install.

gunrunnerjohn posted:

I don't have any pictures, but they simply go from the center roller connection to frame ground, pretty simple install.

Thanks John.  It does seem easy, but I have some steamers that have multiple pick-up rollers.  You only need to install one diode to one roller, correct?  Also, what about the tender?  If a diode is placed on the engine, would the tender need one as well or is that "overkill"?

 

You just need one, if you have multiple rollers, the ideal place for them is after the wires are all connected near the electronics.  For MTH stuff, one will do close to the board, for Lionel with wireless drawbar, you'll want one in the locomotive and tender as they are independent.

Thanks, John.

For you guys thinking of putting them in the engine...keep in mind that you'll likely need to replace the diodes after time, just like those of us who use them at the transformer or track connection. I understand that these diodes can only negate so many voltage spikes before they become in-effective, maybe its just one spike, I have no idea. I also understand MTH places them in the TIU's, but again, after spikes they become in-effective.

gunrunnerjohn posted:

...

As far as testing for open, it's a lot more complicated than an ohmmeter.  You need a current limited variable voltage supply to test the breakdown voltage of the TVS to see if it's still meeting it's specs. 

So what is the thinking on preventative-maintenance or testing a TVS?   

I think you could count on one hand the number of OGR'ers with a current-limited power supply that goes up to ~40V.

Soooo.  I was thinking maybe a 25 cent capacitive voltage-doubler circuit that would convert 18V AC that presumably every O gauge user has to higher voltage DC...then use any DC voltmeter with and without the TVS to see if the DC output clamps to ~33V, ~36V, or whatever.  The key point is the low-cost of this "tool".  If there is interest I can elaborate.

Or, to Paul's point, maybe everyone replaces all the TVS diodes (good or bad) on their layout at the same time they change batteries on their smoke detectors? 

 

Stan, I am always interested in your (and GRJ's) ideas, projects, etc. so I would be one that would like to hear more about TVS testing. I have a current limited DC power supply (from a kit), but as you point out I don't think it goes up to 40 volts (or even over 33?).

I have an HP 50V P/S with current limiting, so I could do a simple test.  OTOH, since you have to remove the TVS to test it, why not just replace it with a new one, they're cheap.  Why waste the time testing it?  Another point, they most often fail shorted.  Here's one of many references, this one in a Microsemi "MicroNote" about TVS devices.  In that case, it's easy to determine as you have a short circuit.

In the vast majority of cases, a TVS will fail in a shorted or severely degraded mode when overstressed.
This has often been the preferred failure mode for many applications, since the device would then
protect the remaining sensitive circuit from further severe transient threats due to its “electrically
shorted” shunt path to ground in front of the protected load until the device is replaced.

gunrunnerjohn posted:

I have an HP 50V P/S with current limiting, so I could do a simple test.  OTOH, since you have to remove the TVS to test it, why not just replace it with a new one, they're cheap.  Why waste the time testing it?  Another point, they most often fail shorted.  Here's one of many references, this one in a Microsemi "MicroNote" about TVS devices.  In that case, it's easy to determine as you have a short circuit.

In the vast majority of cases, a TVS will fail in a shorted or severely degraded mode when overstressed.
This has often been the preferred failure mode for many applications, since the device would then
protect the remaining sensitive circuit from further severe transient threats due to its “electrically
shorted” shunt path to ground in front of the protected load until the device is replaced.

Sure.  But what does severely degraded mode mean?  In the DCS-TIU signal degradation fiasco, IIRC the DCS signal level would slowly degrade over a period of months...presumably from repeated "strikes" of transients due to the harsh/severe operating conditions of the club layout.  Weren't there TVS protection in place that failed in the "open" condition or some "severely degraded mode"?

In any case, I'm thinking that anyone who takes it upon themselves to routinely replace the albeit "cheap" TVS protectors would find it interesting to see what the deal is...in the spirit of "inquiring minds want to know." 

rtr12 posted:

Stan, I am always interested in your (and GRJ's) ideas, projects, etc. so I would be one that would like to hear more about TVS testing. I have a current limited DC power supply (from a kit), but as you point out I don't think it goes up to 40 volts (or even over 33?).

Yup.  Without getting into the tedious nitty-gritty you run into certification limits (e.g., UL, CSA, etc.) when voltages reach 40V or so.  Let's get to the fun stuff.  So here's my 25 cent TVS "tester".

I figure everyone on OGR has at least 18V AC.   There are many ways to step-up an AC voltage to the 40V or so needed to mess with the 33V or 36V TVS clamps which seem to be popular amongst the OGR crowd.  I choose not to comment on the whole TVS concept and simply offer this with a positive pro-active mindset... take it or leave it. 

So with generic components that electrically inclined DIY'ers should already have lying around in their parts stash you can build a so-called voltage-doubler which converts AC to DC.  The capacitor is 0.1uF rated 100V.  The diode is generic - I show a 1N4003 as I suggest this as one that any self-respecting OGR DIY'er should have just because (and if you hunt around you can get them for 1 cent).

With 18V AC input, the voltage-doubler generates a ~40V DC output.  I show a "free with coupon" Harbor Freight DMM simply to show that you don't need some fancy dancy meter!

1) apply 18V AC

2) confirm ~40V DC output

3) place 33V or 36V TVS diode in question across the DC output of the voltage-doubler.  In above video I show a 30V and 22V clamp diode; I don't have the 33V or 36V TVS that everyone talks about.  If the TVS is failed in the "open" mode then the DC voltage will not change (stays at 40V or whatever).  If the TVS is failed in the "shorted" mode then the DC voltage measurement will drop to 0 V.

Note that there are many variations of AC-to-DC "doubler" or "multiplier" circuits.  I am suggesting this one because it will NOT draw excessive current if you place a failed-shorted TVS across the output.  That is, there are some versions of AC-to-DC multipliers that can draw Amps of current which would be untoward and annoying and probably burn your fingers.  For the technically curious, the "trick" here is to leverage the impedance of the 0.1uF capacitor to limit current.  That is, a 0.1uF capacitor at 60 Hz "looks like" a 25K Ohm resistor which is a good current limiter.

 

 

 

 

Stan, tell me about the Oscar Meyer Wiener song in the background!

However, since you have to have the TVS out of the circuit, it's easier to simply replace it.  Of course, you could test the new ones you're putting in...

Thanks Stan! Earlier I tried a sample circuit from a book I have and got 52 volts from the Lionel 18 volt accessory transformer on the bench. It used 100uf caps and mine were only 50 volt so I quit after seeing the higher voltage to be safe. Must have been one of the 'variations' you mentioned above and more than a doubler.  

I like your setup much better and will give it a try tomorrow. I think I have all the parts needed for it and I think the 0.1uf caps I have are like 1000 volt or something really high like that. I might even try a small PCB or something? Maybe like a battery tester or something where you just touch the TVS leads to some contacts on the PCB. Thanks again!

Yeah, I figured you'd get a chuckle with that.  Of course it's from that MP3 you posted for your O-gauge Wiener-mobile that you inspired me to do.  Now I know you actually watched the video! 

To each his own, but I figure some guys would be curious to know if the TVS is functional upon replacement.  Again, we're talking 25 cents!  As you may recall, my offer still stands:  at any time of your choosing I will send you a quarter for free as long as you pay shipping!