TVS Sacrificial? - Shorted out

If your transformer has enough power you can blow them up. I have vaporized a few with my 500 watt Ampzilla. Most arcing and sparking will not destroy them though.

Pete

Please provide more details of the TVS used and the transformer used. 

Dale H

They have two failure modes, shorted and open.  It's not uncommon for them to short when they have failed, we see that pretty often with the MTH TIU, it has a TVS on each channel.

Absolutely.  Any diode can fail shorted depending on the fault.  Some times they fail open.

The TVS is for protection and it did it's job.  So replace it.  That would have been the first item to check after removing the train from the track and still seeing a short.  G

So there is a "TVS" inside the TIU?  Reason I ask if that my channel 1 variable no longer will operate.  A few years ago I think I had a short and that channel no longer worked and I just never did more than open up the TIU and check the fuses (which I believe were all ok).  

But if this is some component I didnt' know to check/replace then maybe I can save that channel myself?  

If the channel is shorted, you can replace the TVS and maybe it'll restore it.  If the channel isn't shorted, while the TVS could be open, that won't cause the channel to fail.

There is a TVS on each channel of the TIU, they're the 1500W models, very similar to the ones we've recommended from time to time here.

My TVSs are rated for 1000W - my postwar ZWs are 250/275W. I've had my fair share of shorts on the track prior to this. Replacing a TVS every so often is not expensive or time intensive - I just had an incorrect idea of their how fragile they are and that they would be something I would want to have replacements on hand for.

The power rating is for a few milliseconds, maybe less. Inductive spikes can exceed the continuous power rating of the transformer and last longer than the TVS is rated for.

Pete

I am currently using 1.5KE36CA, which actually has a peak of 1500Watts, not the 1000 I mentioned earlier. Would a higher rated TVS, like a 5KP36CA be better suited?

I think you just have to expect that they'll occasionally fail.  At least they're doing the job.

I have always wondered why you can't put a resistor in series with those devices to make them intentionally soft.

Well, since the point is to damp transients in the nanosecond ranges, I suspect the effect of the resistor would pretty much kill any effect that the TVS would offer.

Tackindy, if you don't have a short, I doubt that a bad TVS would kill a channel.

A resistor in line means more power is going to the Engines you are trying to protect. Let the TVS die and replace it, It's cheap, engine electronics are not.

And Yes, A TVS is like a fuse, expendable to protect more expensive parts.

I design Industrial Products, I use TVS' like popcorn.

One or more on every connection off the board. and then I put Series 1/2Watt wire wound dropping resistors in line into the board inside the TVS.

The contacts are all rated 10VDC.

You can hook up anything up to and including 24VDC to any of my connections and it will just pop your fuse. It will not hurt my product.

Sorry, No, they are not meant for Model Trains. They are Industrial Controls.

Then you are defeating their purpose.  They need to react fast to protect your trains.  They only should fail when you exceed their design limits.  Most spikes caused by derailments or what have you should not blow the Transzorb(unless the one picked in a TIU or by you isn't rated high enough for the environment you attend to use it in).  It's only those really big spikes that should the Tranzorb out.

I wouldn't call them fuses either. They are two diodes connected end to end to clamp voltage/current spikes.

Wow! Sure is a bit of knowledge around here! So a TVS has two failure modes:

1) Short

2) Open

If it fails as a short, I know what that is like now.

What happens if it fails OPEN so I know to replace it?

To check for an Open Fail you need a power supply rated higher than the rated voltage of the TVS with a limited current flow.

If current does not flow, it's Failed open. If current flows, the limit will keep you from smoking the TVS.

OPEN is a much lower occurring fail than shorted.

I read that as being silently unprotected in the event of an open failure 

Russell, before trying that, I would think all output should be removed.

Come to think of it, the test could destroy it also.  Correct?????

Gentlemen,

    Sounds like the TVS did it's job as required.  bmoran4 has just learned how a TVS safe guards his layout.  It is too bad that TVS's are not resettable, but they must be replaced as far as I am aware.

PCRR/Dave

You read that correctly, nothing's perfect.

well if they open up you have NO protection , i would rather one shorts at least it save the channel,where as if it opens you have lost the TVs diodes protection and the channel circuit now can be damaged!

Alan

 A new tvs out of the bag would give what for resistance readings and could you pull one leg out of your TIU and compare?

There is no law that says you can not put multiple TVS units on a circuit. At work sometimes we put 3 or 4 on one circuit if it was troublesome. Early instruments with computers were very sensitive to spikes and engineers then often did not consider the problem in circuit design. Usually they fail open. I imagine you could test the device with a 28VAc transformer which would peak around 39v.Put it in series to a small load. Should clip the top and bottom of the wave and maybe light a bulb. 

Dale H

Can some one post a picture of where the TVSs are located on the TIU.  Have some tiu's that have channels not working and want to check the TVS.

thanks 

rad400

GRJ posted some pix in last few days in another thread.  If channel isn't shorted, tvs wouldn't cause failure.

A new TVS looks like an open circuit until you reach it's voltage rating, then it's a short circuit.  The only way to test them is to apply in excess of the rated voltage with some current limiting.  The most common failure mode is shorted, and before the total short, the trigger voltage starts to be reduced, which of course hastens the demise of the TVS.

Yes, Do the test on the TVS all by itself, removed from circuit.

Also, destruction in test is why I specified a Current limited power supply.

Set the current limit very low, then apply low voltage.

With Voltage below the threshold of the part, no current should flow, if it does, it's shorted, Replace it.

Then raise the Voltage slightly above the threshold.

If no current flows, the unit is open - Replace it.

If current flows, it's good

Discontinue the test quickly to maintain the part.

And as I mentioned before but may have been misunderstood.

TVS devices are cheap, So treat them like a fuse. (but not AS a fuse)

IE, Keep spares and toss out failed ones.

Repair of one Engine will buy hundreds of TVS'

  So if shorted would you get a meter low ohm reading for example my test leads when touched read .4 ohms?

I had a K-line F3 where the TV shorted and killed my track power too (not sure of the reason). It was across the motor circuit. I replaced it and it started running again and no problems since that happened a few months ago. In my case Long story "short" 

If you're reading that directly across the TVS, it's shorted.