Diagnosing a short in a BLI Reading T1 4-8-4 (N scale)

[thomasjmdavis]

Hello all.

Looking for the wisdom of the group, here, especially anyone who has encountered a similar issue, or has a lot of experience with BLI steamers, especially this little 4-8-4.

I am looking for some guidance on my recently acquired BLI 4-8-4.  I bought it used.  I think the fellow who sold it to me is an upstanding citizen, and I don't think he sold me something he thought was a lemon.  I got a good deal, and so if it takes some time, effort or money, well... so be it.  This is both my first BLI product, and my first sound equipped loco- so... I have lots of experience taking locos apart and reassembling them, and as a rule, I can fix whatever is busted inside one. BUT... this is the first time I have encountered this particular brand and loco, and I don't want to find myself doing one of those things that all of you know to NEVER do.

After receiving the loco (package via USPS, no apparent damage or rough handling, but you never know), I put it on my DCC++/JMRI programming track. Initially, all was ok- was able to read the DCC address, and looked at some CVs. Played around with a couple settings... loco appeared to accept a couple of changes (one of which was to turn off DC mode and opt for DCC only). And... circuit breaker on the Arduino kicked out.  Tried resetting- immediately blew again. Removed loco from the track, reset- all good to go.  Programmed a different loco- no issues.

Put the BLI back on the track- circuit breaker trips as soon as power is applied. Let it sit overnight (didn't want to perform surgery through the middle of the night). Today, opened the tender.  Disconnected the front plug (loco to decoder), still got the short.  Tried with the front plug connected, rear plug (decoder- speaker & rear headlight) disconnect- still trips the breaker.  Disconnected both plugs and reset the breaker- no short.

Now, this seems to indicate to me that the decoder is where the fault is. And the Paragon decoders don't seem to have a very good reputation here on the Railwire.  However, I don't want to make the assumption and buy a $100+ replacement and THEN discover that the short is somewhere else- some connection or loose wire that feeds into those harnesses.

The only modification I am aware of is that the smoke unit has been removed (or, the seller said he removed it, I haven't removed the boiler to look- it has crossed my mind that this might be the source of the short, even with the smoke switch in the off position.

So, 2 questions.... (1) can I remove the boiler without drive wheels, gears, and miscellaneous draft gear all over the tabletop? (it looks like just 4 screws, not unlike the tender- is it really that easy?)
(2) Are there any common short possibilities I should be checking out?

[jwaldo]

You might want to check the gauge on the wheelsets. I've encountered a couple locos where if the wheels were gauged too close together the half axles ended up close enough together to cause intermittent shorts. Correcting the wheel gauge was all it took to fix them.

[thomasjmdavis]

The rear driver (with the traction tire) was just a fraction of a fraction tight, but when checked with a meter, showed no short.  Everything is now in gauge.

[peteski]

I assume that tender picks up power.  Is there a separate tender track pickup connector in that decoder or tender pickups are wired together with the harness connected to the locos pickups?

[thomasjmdavis]

I assume that tender picks up power.  Is there a separate tender track pickup connector in that decoder or tender pickups are wired together with the harness connected to the locos pickups?

Yes, the tender picks up power. 
Here is the discussion of the loco with several pictures-
https://www.spookshow.net/loco/bli484t1.html

There is a second board underneath the decoder, I haven't pulled the decoder yet to get a look at that and how the power from trucks and from the drivers is put together.  There is no separate plug for the tender trucks on the decoder.  However, as I noted in the first post, unplugging the decoder resolves the short. That leads me to think that the power is not the problem - since any short between power wires or one of those wires and the chassis would still cause the circuit breaker to trip.  But... I will take a look when I get deeper in.

[peteski]

You are most likely correct.
Likely one or more rectifier diodes are shorted.  Assuming you own a multimeter, have you tried to check the resistance and or continuity between the wheels trying both polarities of the test leads?  Do you see zero ohms in ope or bit polarities?

[thomasjmdavis]

You are most likely correct.
Likely one or more rectifier diodes are shorted.  Assuming you own a multimeter, have you tried to check the resistance and or continuity between the wheels trying both polarities of the test leads?  Do you see zero ohms in ope or bit polarities?

My meter's ohm range is relatively limited, but I am getting about 17k ohms across the wheelsets, or between any wheel on the left and any wheel on the right. Split frame is doing it's job.

[peteski]

If 17k ohm is a true reading then it is far from a short.  Zero to few ohms would be a short.  But if only one out of 4 diodes in a bridge rectifier shorted out then an ohmmeter would not accurately show the problem.  Since you already determined that the decoder seems to be the problem, I would not expect a short through the split frame.

I wonder of BLI has a stealth PC board available for this model? If they do, and are willing to sell it, you could replace the Paragon decoder with the stealth  board and install a decoder of your choice.

[thomasjmdavis]

If 17k ohm is a true reading then it is far from a short.  Zero to few ohms would be a short.  But if only one out of 4 diodes in a bridge rectifier shorted out then an ohmmeter would not accurately show the problem.  Since you already determined that the decoder seems to be the problem, I would not expect a short through the split frame.

Sorry if I did not make it clear, yes, I know that a short is zero (or a few) ohms.
That was my point ( in reporting 17k ohms)-- the wheels/frame are NOT shorting.  I did not check under power with the DCC engaged.  (Never tried that with a digital meter) But in any case, since the circuit breaker trips when the loco is powered up, I can't check with the decoder powered (and in any case, it obviously is shorted when the decoder is powered, hence the circuit breaker tripping).

What I was hoping for were suggestions on ways to eliminate other possibilities like accessories (motor, smoke, lighting or sound wiring) that are are the output side of the decoder.  The wheels and frame and power wiring to the decoder all seem clean. 

[peteski]

You cannot test for shots if the circuit is energized!  That can damage your meter, and will not give you correct reading. I was asking to check for shorts between wheels while the decoder is installed and connected while no power is applied.  Or if you can probe the decoder's connector, check the resistance between the decoders left and right rail connections.

I guess I'm still confused.  If the meter shows 17k between the wheels while the decoder is disconnected that is odd.  It should show open circuit (infinity reading) not 17k.

[thomasjmdavis]

You cannot test for shots if the circuit is energized!  That can damage your meter, and will not give you correct reading. I was asking to check for shorts between wheels while the decoder is installed and connected while no power is applied.  Or if you can probe the decoder's connector, check the resistance between the decoders left and right rail connections.

I guess I'm still confused.  If the meter shows 17k between the wheels while the decoder is disconnected that is odd.  It should show open circuit (infinity reading) not 17k.

It showed 17k while the decoder was engaged. 

With the decoder disconnected there is no short (as per previous posts).

[Maletrain]

Because I have a shorting problem with a BLI USRA 2-8-2, I am following this thread to see if I can learn something.

In my case, the loco shorts when it is unplugged and detached from its tender.

However, the short was initially very intermittent, so I spent some time with false assumptions that I had "found the cause" when the short went away after I did something, only to have it return and NOT respond the same way from doing the same thing again.

So, in your case, did you already unplug the tender from the loco and detach its drawbar to see which piece of equipment still shorts when on the tracks and powered by DCC?

[peteski]

It showed 17k while the decoder was engaged. 
With the decoder disconnected there is no short (as per previous posts).

That does make more sense to me. Have you tried measuring the resistance in both polarities of the meter?  Was it 17k in both polarities? It would be interesting to find this out.

To measure resistance there is actually a voltage sent to the circuit under test and the meter then measures the current, internally calculating the measured resistance.  The voltage is fairly low, usually not high enough to get any silicon diodes in the measured circuit to start conducting. But sometimes the voltage is just high enough for the diode to start conducting and that can show up as a medium value resistance. 

[thomasjmdavis]

Because I have a shorting problem with a BLI USRA 2-8-2, I am following this thread to see if I can learn something.

In my case, the loco shorts when it is unplugged and detached from its tender.

However, the short was initially very intermittent, so I spent some time with false assumptions that I had "found the cause" when the short went away after I did something, only to have it return and NOT respond the same way from doing the same thing again.

So, in your case, did you already unplug the tender from the loco and detach its drawbar to see which piece of equipment still shorts when on the tracks and powered by DCC?

When the loco is disconnected, it does not short, but the tender does, unless the rear plug on the decoder is also disconnected. (Not sure if all Paragon 4s are the same).  In any case, essentially the same as observed when everything is assembled- if the decoder is disconnected at both ends, there are no shorts. If it is connected at either end, it does short.  In my case, not intermittent- it shorts every time power is applied if the decoder is plugged in.

[thomasjmdavis]

That does make more sense to me. Have you tried measuring the resistance in both polarities of the meter?  Was it 17k in both polarities? It would be interesting to find this out.

To measure resistance there is actually a voltage sent to the circuit under test and the meter then measures the current, internally calculating the measured resistance.  The voltage is fairly low, usually not high enough to get any silicon diodes in the measured circuit to start conducting. But sometimes the voltage is just high enough for the diode to start conducting and that can show up as a medium value resistance.

The caveat here is that my meter is not very precise (a meter for electrical work, not electronics).  But, the resistance with polarity reversed is approximately the same.  It could be that it is 16.5 one way and 17 in the other,  I get some variance in readings as I go along various wheelsets. 

And just for the record -  when tested, the motor runs fine when decoder is disconnected and  DC is applied from my old Tech II powerpack.