MTH Proto 2 Secondary A Unit constantly runs

There is a short somewhere damaging the slave board in the trailing A unit. You're best off checking the wiring in the units before installing a new board.

@Lou1985 posted:

There is a short somewhere damaging the slave board in the trailing A unit. You're best off checking the wiring in the units before installing a new board.

That was my first thought but finding the short is another story.  As I ran the unit for about 1/2 hour with no problems, it's obviously an intermittent short circuit.  I'm thinking the tether line may be the best starting point.  Was hoping someone here had already encountered the problem and knew the cause.

I'd check the wire harness for scuffs or broken insulation. Wiggle it around. Check the wiring to the trucks (pickup rollers and ground) as well. If there's nothing visible break out the multimeter and start probing wires for shorts.

Don't for get to check the pass through tether on the B unit as well. There could also be a short there.

Check the solder connections on the sockets of the tether.  They can bridge connections and short.  Also, in the B unit, the harness is crimped to the frame, which can pinch the wires.

The motor FET on the slave board shorted,  Pretty easy to replace.

Did you ever find a fix?

After reading multiple posts on O Gauge Forum and speaking with several of my fellow club members it seems there is an inherent problem with some early model Proto 2's; the ones that use the 9V type battery or BCR.  All it takes is a simple derailment to blow the board in the secondary unit.

Ultimately, I removed the motors from the secondary A unit and just run it as a single powered A-B-A set.  I'm not about to buy another circuit board for the secondary unit and have that one get shorted out too.  Replacing the motor FET on the board, as suggested in an earlier post, is outside my pay grade and almost seems senseless if another derailment will once again cause it to blow.

By the way, I should have clarified earlier that in my original post when I said that the circuit breaker on the transformer popped about an hour in to using the newly replaced circuit board, it happened when I had a derailment.  I never made the correlation until I read several more posts on the subject. 

Thanks for that info! That is precisely what has happened to both my ABA trailing units! I run O gauge Tubular track with O31 curves and all my other untethered units will run forever with out derailing. But not these two powered sets. I think I am headed in your direction. I just want to walk away and watch them run...Not work on them!! Thanks again for the info I was beginning to think it was just me!!! Bill

@TAGurit13 posted:

After reading multiple posts on O Gauge Forum and speaking with several of my fellow club members it seems there is an inherent problem with some early model Proto 2's; the ones that use the 9V type battery or BCR.  All it takes is a simple derailment to blow the board in the secondary unit....

In your research, did anyone suggest adding a protection device in the slave unit ... such as a 50-cent TVS (transient voltage suppressor), resettable fuse, or whatever?

Derailments are a fact-of-life so it makes one wonder what makes the $1 motor FET in the secondary so sensitive...and what can be done about it.

It's so sensitive for 2 reasons. These are just my opinion, so don't shoot the messenger.

#1 it's not heatsinked other than the board PCB trace itself. Heck the pad isn't even that large.

Photo by @Norton

#2 The circuit is basically an analog amplifier. Yes, you read that right. This reads a direct motor voltage from the lead engine, converts that to a logic level signal, and then drives effectively  the relay and 2 FET based dual direction motor control. So, should you spuriously send it AC in a derailment it's trying to slam this slow switching relay based motor drive circuit back and forth rapidly.  So yes, as a limitation of the basic circuit design you might blow it up with such fast cycling of forward and reverse. FETs can change (ON /OFF) fast, however the relay cannot. That could lead to inductive spikes as the relay opens and closes the contacts while current is flowing as another possible answer. Again, the slave board input is direct motor power from the lead engine. This takes that PWM DC signal and reduces it to logic level voltage. That said, that is analog and while normally should be a PWM DC voltage in one direction, pause, and then switch polarity for direction change and then smooth PWM mostly DC. However, in a derailment, a condition could exist where the slave board has power, and these 2 motor leads are swinging wildly or appearing as an AC track power signal rather than rectified and PWM'd DC.  Due to limitations of how fast this circuit can switch and other details- that is very bad for the slave motor driver board and often one or both FETs (also not heatsinked) fail.

@stan2004 this is your photo from this topic

On one hand, from an electronic engineering standpoint I do understand why this design was chosen. It eliminated the P channel MOSFET needed for an H bridge and then the gate drive problem. P channel MOSFETs are not as efficient (lowest ON resistance) as N channel, as well as N channel being very easy for gate drive circuitry. However, the tradeoff is there is a mechanical relay with contacts, there is a maximum speed which that can switch much slower than the FETS.

What i was trying to get at is, I have a few engines that use this slave board, and yes, I've had to replace one or repair one of them. So I do understand the frustration in a very personal way. Yes, TVS diodes can be a good safety for certain things and voltage spike are a constant battle with our electronics.

That said, I don't think (again, opinion, don't shoot me) that a TVS will solve this issue. One thing that seems easier (at least in my head) is rob some heatsinks from a dead PS2 5V board and add them 2 the 2 motor drive FETS. That would help the thermal side of this.

Image by @Ron045 in this topic

As for how to prevent a fast switching input signal that happens in a derailment from slamming this slave board, a TVS won't fix that. Not in any way I can think of.

Again, the basic motor drive circuit is a copy more or less of the same circuit (1 relay, 2 FETs) on the PS2 5V and 3V lead engine control boards. The idea is the response to copy the basic signal (direct motor drive voltage coming out of the lead engine, is copied and driving the slave motor. Down to the point the PWM is sensed and just replicated analog style. The thing is, all fine and dandy when that signal is nice and smooth and one direction, and then pause, and then the other direction like normal operation. However, in a derailment, that signal can just go nuts. It's those swings, and this board trying to copy that signal that then blows it up.

I had a brainstorm this morning, just a thought and believe me, I have no background in electronics as it appears you two gentlemen do.  After replacing the circuit board in the slave unit would it alleviate any electrical spiking problem that might damage it by simply programming the lead unit to travel in only one direction?   Just wondering, maybe if it's not trying to change direction due to the derailment,  it might be less likely for the board to blow?

Once again, no electronic knowledge here, just pulling at straws to alleviate a problem.

I am still curious if in your research if anyone has suggested protecting the FET(s) themselves.

I don't have a slave board.  I only know that this failure mode (slave engine immediately takes off when powered) seems to pop up on OGR like clockwork!

From what I can tell, the most common cause is a failed FET, and the solution is to send it to a skilled tech as it can be a tedious DIY exercise working with surface-mount components.  And even though the FET is a 50 cent part, the overhead of packing, two-way shipping, labor, etc. could turn this into an expensive habit.

As noted, there are no published schematics so it's just educated guesses as to why derailments are a problem.  And I don't think there's been enough OGR discussion to determine if this is more likely to occur when operating in Forward vs. Reverse, high vs. low track voltage, high vs. low speed, or whatever.

So I was thinking if any of the repair techs ruled out simply adding a 25 cent TVS or similar clamping component in case some FET-busting wild voltage excursion occurs during a derailment.  I was thinking something like the 33V (?) TVS commonly sprinkled around the layout.  In fact, the surface mount version might easily drop right next to the FET as illustrated below.  I suggest 33V because the IRLR024 is a 60V part, and even at command track-voltage it operates at only about 25V.

Again, this is taking a very narrow view into the problem given the lack of real-world field data.

You're right Stan, while it's not that common that I've seen them, I have fixed a few, and it was usually the FET.  The chips next to the relay occasionally are the problem as well.

I don't know if this helps but my 3 derailments that cooked the trailing units were at 9 volts except one that was just throttling up and all in forward. I only run conventional with Z1000 transformer and control and I just measured this voltage now as I have a unit running the layout at the slow speed I like. It sure would be fun to find a fix for this. I look at the box the premiers came in and it says 8 to adult and I will guarantee at 8 I crashed more trains  then I do now at 68! And the old ones are still running!!! LOL! I do like Stan's response but am unsure where a voltage spike may come from when the track voltage is metered? Sorry if I sound dumb!

@WEB posted:

...

I do like Stan's response but am unsure where a voltage spike may come from when the track voltage is metered? Sorry if I sound dumb!

Nope. You're asking the right question.

So here's the way I look at it.  And, again, I've never seen or touched a slave motor board and simply commenting in the spirit of OGR being a discussion forum to improve our collective hobby experience!

I say it's a dead-end trying to meticulously catalog the triggering event that destroys the FETs.  There could be many causes...one apparently being a derailment.  The matter-at-hand is we have damaged FETs. 

So strictly looking to protect the FET, I'm suggesting clamping voltage spikes which are a known to damage FETs in any circuit...not just toy train engine electronics.  The reason you would get a voltage spike in a motor drive is because of the coil windings in the motor.  The motor windings are directly connected to the two FETs.  Think of the spark coils in your fossil-fuel car.  You start with only 12V DC from your battery.  But spark plugs need some 10,000 Volts.  The way you go from 12V to 10,000 Volts is with a spark coil which stores energy in its windings and then "violently" releases the energy in a spike of many times the starting voltage.  That's the mechanism by which you might be getting voltage spikes from a "metered" track voltage.

So I am speculating that when the slave unit is moving, and the motors are spinning, there is stored energy in the momentum of the spinning motor(s) and their windings/coils.  Then when the engine derails or suddenly stops, this stored energy is "violently" released as a voltage spike that destroys the FET which is directly connected to the motor.  OK, there is a bit more techno-geek explanation that I'm glossing over but it's basically a simple value proposition:

So if you have a broken slave unit, and are going to send it in for repair, then why not additionally install 16 cents of additional parts - that is, two 8-cent TVS clamps.  I'd guess it would take a repair tech maybe 1 minute to install a TVS.   This is the same "reasoning" that is widely advised about adding TVS voltage clamps around the layout to clamp spikes on the track voltage.  It's cheap insurance.   It's a value proposition.

WEB...Another question...were all of your blown slave boards on the older Proto 2 units that used the 9V type rechargable battery?  These are the only units that I am aware of that have this problem.  I have other Proto 2 units that use the newer 2.4V battery and have never had (knock wood) problems.  If yours are all 9V style, therein may be part of the problem

Slave boards are the same between 3V and 5V PS/2, there was only one slave board part number.

@TAGurit13 posted:

WEB...Another question...were all of your blown slave boards on the older Proto 2 units that used the 9V type rechargable battery?  These are the only units that I am aware of that have this problem.  I have other Proto 2 units that use the newer 2.4V battery and have never had (knock wood) problems.  If yours are all 9V style, therein may be part of the problem

Yes they all 9v type batteries but I changed them to BCR's

Thanks Stan for taking time to explain things! I have to agree there is no harm in trying!!!

Okay, so you have had the same exact problem as me.  What I am getting at is in a derailment situation, the locomotive continues to receive power from the battery/BCR.  On the 9V system, the slave board blows. on the 2.4V system it doesn't.  Sorry, I'm no electronic mastermind or engineer but I'm also not a big fan of coincidence.  Seems much more than strange to me that the problem only occurs with the 9V battery/BCR.

@TAGurit13 posted:

WEB...Another question...were all of your blown slave boards on the older Proto 2 units that used the 9V type rechargable battery?  These are the only units that I am aware of that have this problem.  I have other Proto 2 units that use the newer 2.4V battery and have never had (knock wood) problems.  If yours are all 9V style, therein may be part of the problem

TAG, Do you know what years  Proto 2 start using 2.4 v batteries? I think I am going to fix my two using Stan's idea .... If they fail again they will be "dummies"! I unplugged the motors from the slave boards and tethered the set together and  the other functions (smoke, directional lighting) work just fine so they would still look good being dragged around!

I have no idea what year the switch over took place.  I'm done with replacing slave boards. Secondary "A" units are all dummies now for me.  As you said, without the motors they roll just fine and I rarely, if ever need the pulling power of 4 motors.

@TAGurit13 posted:

Okay, so you have had the same exact problem as me.  What I am getting at is in a derailment situation, the locomotive continues to receive power from the battery/BCR.  On the 9V system, the slave board blows. on the 2.4V system it doesn't.  Sorry, I'm no electronic mastermind or engineer but I'm also not a big fan of coincidence.  Seems much more than strange to me that the problem only occurs with the 9V battery/BCR.

Yes I agree we have the same problem... I am not convinced the trailing unit doesn't cause the derailments... My layout is a "U" 14x11x14 with reversing loops and just does not derail... unless I do something dumb! It is only these two ABA sets which are new to my collection.... Not a fan of coincidence either nor will I keep fixin the same thing expecting a different result.. as we all know that definition!

Insanity!!!   LOL    Not sure as to the cause of the derailments but, with me, the derailments I've had were not with the locomotives, they were derailments of the cars immediately behind the secondary A unit.

@stan2004 posted:

Nope. You're asking the right question.

So using the surface mount TVS, it is placed between grounding pad and 1st leg of FET?  I just want to make sure where to place these.

Thanks.

@DaveGG posted:

So using the surface mount TVS, it is placed between grounding pad and 1st leg of FET?  I just want to make sure where to place these.

Measure twice, cut once.  I see I scrambled the FET terminals between my original diagram motor bridge diagram and the slave board experiment diagram.

The idea is to clamp the Vds voltage to around 33V using a TVS.  As mentioned previously, the IRLR024 in an O-gauge motor drive will only see Vds voltage of around 25V DC+ in normal operation.

And I suggested a 33V TVS clamp simply because that seems to be commonly used by the guys to clamp track voltage transients; and I believe this is the part that MTH places on each channel of their TIU.  Anyway, the clamp should be some voltage above the normal operating voltage but less than the max handling voltage which is 60V in this case.

This is the correct diagram.  The TVS shown is bi-directional so can be installed either way.  It goes between the D(rain) terminal and the S(ource) terminal of the FET.  I have no idea if the SMD version of the TVS is "easier" to install than the leaded version.

Oh, and just to be clear - the "grounding pad" which is connected to the FET's Drain terminal is NOT electrical "ground."

So glad you asked for confirmation!

The early 5V systems where apparently more prone to this and the slave boards were modified to replace 2 diodes, 2 resistors in the signal path and one of the chips.  So the 3V engines should have the modified slave board. This is not a spike taking it out, but circuit wise with a lose of signal to motor drive, or noise that Forward motion Fet can burn up in a second.  I think MTH fix was twisting all motor lead wires in the Lead A to the connection PCB.  Any exposed lead in the Trail being twisted to help reduce motor noise.  Also making sure motors in trail unit respond quickly to movement.  I think the circuit mod helped with response of slave board also.  If the trail motors don't start turning, or hang up, the backlash locks them and you also wind up with a stalled motor.  Can be very frustrating working on them.  I have repaired a ton of Slave boards and units.  Most don't come back.  But I have had a few that kept coming back re damaging slave board.  So friction in truck, imbalanced motors, noise circuit, etc....   Plus many of these boards even direct from MTH are reworked.  I had a brand new board installed and it burned up soon as power was applied to the slave board.

It is not a heat sink issue either.  PS-2 3V or PS-3 don't have heat sink on fet, they are similar.  Only Added to the 1 scale PS-32 boards because of the bigger motors.  Even the PS-3 boiler board are small chip and used on Premier now.

So one solution is upgrade the Trail unit to a full PS-2 or PS-32 engine.  Run in a MU (lashup).  It is the only way to do upgrades of PS-1 ABA with PS-3 diesel, because there is no slave mod kit and even MTH did away with PS-3 slave after a short initial run.  G