Skip to main content

I recall a recent thread concerning circuit breakers between TIUs and other devices leading to track connections. I would like to add one to our club's Yard module for protection from the inevitable  shorts due to derailments. Can anyone steer me in the right direction?

I currently have a 8 amp SFE fuse installed, but this is unsustainable.

 

Conductor Earl

Anthracite Hi Railers

 

Original Post

Replies sorted oldest to newest

Earl,

Circuit breakers are useful but the problem is they take a while to trip.  We often think a 10 A circuit breaker will "break" at 10A.   Note this chart from a Mouser datasheet for a typical pushbutton type circuit breaker (when it pops, a button pops out and is reset by pushing it back in)

cb data

Looking at the chart, a 10 A circuit breaker will pop after one hour of a 14.5 A load!  That's OK for a direct short circuit (lots of amps more than 10), but even for a 40A load, this 10A breaker may take between 1.6 and 4.5 seconds to "pop".   Not much protection for delicate electronics.

Finding a high speed circuit breaker becomes the issue of $$$$. A cheap one may be OK for a real solid short circuit, but that's about all it's good for.

Ed

Attachments

Images (1)
  • cb data
Last edited by eddiem

I looked at the PM42, one thing that is odd.  They don't say what kind of track power.  I'm assuming they'll handle DCC, so it should handle AC, but it would be nice if they were specific in the specifications.  They went out of their way to tell you that you could use AC or DC for their separate logic power supply, why not specify the specifics of the track power?

gunrunnerjohn posted:

Susan, the PTC isn't all that fast unless it gets a significant overload, same issue as with the circuit breakers.

They're priced so reasonable I've been able to size mine using different values for specific uses and they trip fast enough to protect my electronics and postwar accessories.

Your correct that the 8A and 9A values I use on my main lines for multiple passenger cars don't trip fast enough with low power faults.  The 2 A values can be placed in the engines for max protection.

I use 1.1A hold PTC's in series with motors using the ERR Cruise Commander Lite to prevent damage due to a motor stall.  I cooked one of my beta test units when the K-Line Interurban snagged a switch and stalled the motor, that was my solution.  So far, no more cooked boards in the half dozen I've installed in my stuff and the others that have gone out the door.  Of course, any protection is better than none, and I like the PTC for inside units as you don't have to open them up to fix anything if the trip.

RJR,

   I have been aware that the 7 Amp and the 10 Amp Scott type breakers have been breaking pretty much at the same time for many years, probably the 10's were built to light from the very beginning, but they work great on my layout, and they do not sell the 7 Amp Breaker Banks at Train Electrics any more.  They were pretty much phased out when Scott got bought out.  I still recommend them they work perfectly with my DCS/Legacy Christmas layouts

PCRR/Dave

OK, just so you know.  They are thermal breakers, so their internal characteristics might change with time

5 & 7-amp breakers are still available from Defender Marine & other marine supply houses.  I use 5's on all circuits but one where I often run 3 or more locos; that's 7.

I recently replaced a KW with a Powerhouse 180; it appears the Powerhouse's breaker opens on derailment before an external 5 amp. 

Gun Runner John,

Thanks for posting the schematic of the overload control relay for the PowerHouse 180.  I had not realized that Cam had made a drawing for this device.  It answers a number of questions that I have had for a while (15 years?) and not had the time to get to  .  One I must post as a caution against the direct use of this diagram to build a breaker.  Early on (135wPH) I noticed on this this OGR board a couple of failures (written off as factory defect by the posters), where sufficient info was posted to infer the true cause of failure.  One failed closed (which will often go unnoticed), and the other failed open (no output), in common with several "failed" reports with no other info.  The most common feeling was "breaker" not working, IIRC.

From this you can infer that you are dealing not with a breaker, but an electronically supervised overload relay, and that the relay contacts have failed from trying to interrupt current over rating.  This failure can be either open (burned contacts) or closed (welded contacts).  I immediately opened one of mine to confirm the use of a relay (not recommended; you have to make your own tools to do it).  It had a Millionspot relay (10 cents at the time and about 1/2" cube) which appeared to have a 10-amp rating (less than in the 180 PH dwg), but I could not find complete specs, mainly its interrupting rating.  I did measure the short circuit current of the 135, and found it to be 77 amps (by extension, for the 180 PH, assuming similar design only larger, that would be 110 amps).

For the 180 PH, using this schematic, I determined that the switching current rating for its particular relay appeared to be 15 amps, or 150% of transformer rated current of 10 amps.  The short circuit current can be as high as 1100% (11 times) this rated current.  To fully protect the transformer, a specially formed output plug is integrally wired to the transformer.  This plug forced connection to several downstream Lionel devices which were rather quick acting and capable of interrupting 110 amps now (and probably then).  Any device containing a MOSFET voltage-regulating circuit falls into this class of necessity (to self-protect the MOSFET chip), and I believe the Lineside shack lockon has sufficiently large contacts even though it too is a relay.
This arrangement is called cascade protection, which has not always been looked on with favor.

The UL standard 697 for Toy Train Transformer did not then address anything within its 50 or so pages (price $600+) downstream of the transformer itself, but I saw (wangled a free look) that that year the standard had been revised to continue the breaker test, but to NOT continue the requirement that the test had to be passed for all of the possible currents the transformer could produce under all conditions. [The single-copy price of UL standards apparently became a national scandal, and a couple of years ago I noticed a $640 standard of a bit under 100 pages had been reduced to $40, still an exorbitant price.]  This change effectively permitted the cascade protection without mentioning it.  I am puzzled that the Lionel instructions did not point out the reason and need for the special plug.

Another of the questions I have is whether the control circuit provides a time delay in the operation of the relay (697 permits a 60-second delay then; MTH used a 24-second delay in the Z-4000 (I think for overload only, the short protection being in the externally mounted breakers)).  In the case of a heavy short this delay would permit a heavier relay (itself possibly sub-cycle if not delayed in the shed lockon) and between a half- and a full-cycle delay in the MOSFET protection (as some of the detectors were only on one polarity of the output wave, such as 7A PowerMaster).  The answers lie in the array of 4 comparators (U1a, -b, -c, and -d) at the bottom half of the schematics.  I will look at this, but could use some help... (or more years of delay?  ).

Meanwhile, at a glance I notice the following:

1. The relay is energized (to reverse the moving contact from the position shown) against a spring to interrupt the current.  Otherwise, I would not have been able to run the standard short circuit test.

2. L1 is a current transformer, ratio 100:1, with an output of 0.10 amps at 10 amps to the track.  The associated resister divider furnishes a voltage of 1 volt per each 1 amp output to the track.  This voltage is fed to the network of comparators.

3. The source of the +8v supplied to the circuit board and the relay coil (it is a sensitive 9v coil, pickup down to a voltage below 9v but not found) is the U1 regulator and its capacitors and rectifier.

4. The comparators are fed +8v both as power and perhaps as a reference voltage, and act to cause Q1 to conduct and energize the relay to interrupt the current as some point over 10 amps (I have heard 10.5 amps, but this should be determinable from the schematic.

5. The source of the +8v will not be interrupted by the action of the relay.  This relay control circuit will be working down to 9v on the transformer red output terminal, which will occur at maximum power input into a short, at 9v times 110/2 amps, or 495 watts.  At max possible short output, the red voltage will fall to zero, and at some point before this the relay will be unable to pick up.  This is a kind of rudimentary protection against failure of the relay contacts, but whether it is effective for long service is suspect, because you do hear about PH failures [I assume mostly among PH's connected directly to TIU's or postwar track lockons].

6. SW1 is a reset switch, which causes Q1 to cease conducting, releasing the energized relay.  This would be a trip-free action if the short remained.

7.  The heavy lines on the drawing (dwg) appear to indicate the changes involved in Rev A.  For example, SW2 and the pilot light (to turn the transformer on, and so indicate) appear to be modified in some way (added? altho "SW2" appears to have been there); and one change I do recognize, the current transformer L1, which is at least a change from the resistance wire I saw in the 135 PH (IIRC), and the changes in the associated sensing circuit at this point [a change suggesting that the resistance wire got hot, degrading accuracy too much].  It looks now like the labels are in a different color or layer in the original dwg and don't show difference if added or changed.

8.  I notice from the dwg that the standard test for short circuit current can be run without destroying the control circuit board.  Some may recall that I accidentally started this test before I finished thinking about this problem.  I would however not exceed 15 amps with the Lionel test track (not sure if anything below 2x rating is possible; if so 20 amps might damage the relay, which would operate).  The standard test uses only the rated current, but does not test the relay, and can be dangerous if the procedure is not well understood.  I am not sure the test track can measure short circuit current to its maximum, without risking calibration damage, even to real breakers.

So, does anyone have time to tackle the effect of the resistor values on the operation of the comparator array?  My eyes glaze over in these situations, not to mention a basic aversion to more effort  .  

The questions are operational level (amps), delay if any intended (cycles or seconds), whether current affects delay, and whether relay energization (opening) is inhibited or delayed differently above some current level, and if so what that current level might be.  Off hand, I think there are not enough comparators to do all this.  And memory comes back: The LM324 is a well-known chip-- I dimly remember that it may have all 4 comparators on it.  Yes, I see the little pin numbers all over the place.  I might even be able to find some example uses of it.  But not soon.

Yes, I think building one of these is not the solution of choice for the question in this thread.  At least, not for me.  But understanding its operation is worthwhile.

--Frank

Frank, you actually measured, rather than calculated, the short circuit amperage of the PH135 and found it to be 77 amps???  I'm surprised the internal resistance isn't such as to limit it to less.  I wonder whether the contacts of most breakers could handle more than one or two such breaks.

 

The internal breakers on a Z4000 open quicker than 24 seconds.  I have never had the Z4000 external 10-amp breakers open.

As I recall, there are two paths for over-current protection to kick in with the PH180 breaker.  I believe Dale Manquen did an analysis of the operation some time back.  One was a slow path that would allow a modest overload, the other was a quick reaction to a direct short.  I know that when I have a direct short, the PH180's trip on the spot, well under a second.  You'll see two paths to energize the relay, one has a much larger resistor and capacitor, the other at a higher output level of the 100:1 sensor has a much faster reaction time.

John, you remember correctly.  There is one circuit that probably keeps the transformer from cooking due to continuous overload, and a second circuit for shutdown for a bad short.

Although I started out with some fast magnetic breakers, I must ask the question "Just what are you protecting?"

A derailment usually involves wheels, pickup rollers and wire, and during the derailment short the voltage across any electronics downstream drops to near zero.  When the short clears, there can be an inductive voltage surge to the electronics if current is still flowing.  If, on the other hand, the power has been removed by the protection device(s) before the short clears, the surge would presumably be across the protection device since the track is still shorted when that device opens.

If the protection is for the wiring, the wiring inside a locomotive will probably survive a quick jolt of current while even a "slow" thermal breaker or fuse is acting.  The problem arises with a sustained partial short that doesn't trip the protection, cooking the wiring.

I use the rocker-style TE breakers (8 amp) for wire protection (and convenient circuit ON/OFF isolation) and TVS protection for voltage surges. 

I used Polyfuses for a while, but I find the rocker breakers to be more convenient, keeping the power off while I clear the problem.

Dale Manquen posted:

John, you remember correctly.  There is one circuit that probably keeps the transformer from cooking due to continuous overload, and a second circuit for shutdown for a bad short.

Although I started out with some fast magnetic breakers, I must ask the question "Just what are you protecting?"

A derailment usually involves wheels, pickup rollers and wire, and during the derailment short the voltage across any electronics downstream drops to near zero.  When the short clears, there can be an inductive voltage surge to the electronics if current is still flowing.  If, on the other hand, the power has been removed by the protection device(s) before the short clears, the surge would presumably be across the protection device since the track is still shorted when that device opens.

If the protection is for the wiring, the wiring inside a locomotive will probably survive a quick jolt of current while even a "slow" thermal breaker or fuse is acting.  The problem arises with a sustained partial short that doesn't trip the protection, cooking the wiring.

I use the rocker-style TE breakers (8 amp) for wire protection (and convenient circuit ON/OFF isolation) and TVS protection for voltage surges. 

I used Polyfuses for a while, but I find the rocker breakers to be more convenient, keeping the power off while I clear the problem.

Just for discussion, in some electrical power applications the inductive voltage problem is solved by actually short-circuiting the load until the circuit breaker has time to open. With the load short-circuited there is no voltage available to cause any trouble, it is all forced to the breaker. The GE Arc-Vault system does this on large 480 volt systems to reduce the let-through energy to an absolute minimum, as an example. 

No reason this could not be implemented on model railroad equipment to eliminate transients originating from the transformer during fault conditions.

 

John, thanks for the summary of Dale's work on the functioning of the PH relay controls.  I missed that; at times I don't look at the board very often.

RJR, yes I did measure the 77-amps rather than calculate it, for the 135w PH.  Altho it is the product of two measured voltages and one measured amps, so you need a clamp-on ammeter and a voltmeter.  I'm against posting the procedure on this board because any mistake can generate dangerous voltages, but I'll write it up and PM it to you.  --Frank

Add Reply

Post

OGR Publishing, Inc., 1310 Eastside Centre Ct, Suite 6, Mountain Home, AR 72653
800-980-OGRR (6477)
www.ogaugerr.com

×
×
×
×
Link copied to your clipboard.
×
×