I think I just fried a DCC loco chip - diagnosis assistance needed

[imt]

I have just started doing my own chip fitting, and am moving on to non-DCC ready locos. So this was an experiment and as such no big deal BUT..... I really need to try and find out what went wrong to prevent recurrence.

 

It's a Lima 37 chassis so nothing special, one of the two wires to the rear bogie sort  (some have only one - I have one of those too) ideal for a low cost experiment.  I cleaned the wheels (old and filthy) to make sure it ran on DC (on my programming track with a DC controller attached), it did.  I wired in an 8 pin female socket with a pre-attached loom. Put in a blanking plate to make sure it still worked on DC on the programming track, it did.  Put in cheap generic Hattons 8pin chip to see if that worked, attached NCE PowerCab to programming track and it worked fine.  Moved on to DCC Concepts Zen 360 with stay alive.  Result was wonderful, got a slow crawl, ran well on the full layout after fiddling with the back to backs.  Really amazed and really pleased.  Pride went before a fall!

 

There were two things left to do: cut off old large tension locks and replace with NEM pockets for Kadees and "ease" the chassis so that a Hornby Class 37 body would fit.  The NEM fitting required a lot of mucking around with the chassis upside down.  When I came to fit it all back together with the body on and put it on the track the motor ran at full speed and had to be taken off the track rapidly. This is a fixed performance, I cannot put the loco on the track it seems to be at full power.  I assume therefore that putting it on the track again shorted something in the chip resulting is this aberrant performance.  All wires are still soldered, all shrink wraps are still in place.  I can find no bare or worn wires.

 

So it went from a wizzo DCC chassis to a useless completed loco via being upside down for a while whilst bits were cut off and others filed so pockets could be stuck on the bogies.  I changed nothing electrically deliberately.  Maybe something in one or other of the bogies shifted?

 

Any ideas? Help in where to look would be appreciated.

Edited September 8, 2017 by imt

[Phil Bullock]

That doesn't immediately sound like a fried chip! What chip interrogation facilities do you have...?

 

Phil

Ps have you removed the suppressor?

 

Phil

[imt]

That doesn't immediately sound like a fried chip! What chip interrogation facilities do you have...?

 

Phil

Ps have you removed the suppressor?

 

Phil

 

Thanks Phil.  Yes the suppressor is out, has been from square 1. Do you mean something like an ESU 51900? I don't have anything but other locos, and I'm a bit scared of putting the chip in another loco!  As you will understand I have no way of interrogating the chip in it's current location.

[PaulCheffus]

Thanks Phil.  Yes the suppressor is out, has been from square 1. Do you mean something like an ESU 51900? I don't have anything but other locos, and I'm a bit scared of putting the chip in another loco!  As you will understand I have no way of interrogating the chip in it's current location.

Hi

 

Can you put the loco back on the programming track and read the address?

 

Cheers

 

Paul

[imt]

Unfortunately not.  The moment the wheels hit the rails it goes into flat out running.  I cannot do anything with it for fear of the motor overheating.  Both bogies need to be on the track as the left line feed only comes from the rear bogie - so no way to read anything!

Edited September 8, 2017 by imt

[Andrew F]

It sounds like you've removed it from the track whilst the chip is still responding to top speed (ie maybe loco stalled on bad track/frog). Maybe try putting the address into the controller with the throttle off and putting it back on the program track..

 

I've bought loco's off ebay before that have shot off like this and you don't know the address. Best to try it on the program track and if you can't stop it do a decoder reset.

 

Good luck

[Pete the Elaner]

It shouldn't be able to go flat out on the programming track.

Running the motor flat out is not so bad though.

Maximum load is worse. This will put a high current through the coils. Flat out creates a higher back emf so the current through the coils (which is what heats it) is low.

 

If the chip was dead, I doubt you would get any movement.

A decoder tester would be useful, but they are all so expensive, apart from the MERG kit.

[RAF96]

Was there any metal swarf generated by the chassis filing if so it may have bridged something in the decoder. This could be proven if you could put the chip in another loco and see if the snag follows the chip or stays with the chassis.

Rob

[Phil S]

Do you have a multimeter ?  If so - you can unplug the decoder, and check the impedances of each path:  Rail- Socket (should be close to zero to the appropriate pins (red and black wires)   Socket - Motor - Socket ( Orange-Grey ) expect something around 10 ohms (with a wide variation - but not a short, and not open circuit) etc -   and NO CONNECTION between either red or black and orange or grey  - assuming also at this stage there are no lights connected to blue and white/yellow/green/violet...   Ensure blue is not touching anything at its 'exposed' end.

 

If you don't have a multimeter - they can cost less than a single silent decoder.  Thoroughly recommended if hard wiring conversions.

[PaulCheffus]

It shouldn't be able to go flat out on the programming track.

.

Hi

 

I've had N gauge locos do exactly that on the programming track of my Lenz system. Traced to a connection between the motor and the chassis that was bypassing the decoder.

 

Cheers

 

Paul

[imt]

It shouldn't be able to go flat out on the programming track.

 

 

That's exactly what I thought too. BUT IT DOES hence my initial belief that this was a damaged chip - since I believe that the only power to the motor (orange and grey wires) gets there via the chip (unless I have something horrible wrong).

 

Maybe I'm confusing people by calling something a programming track.  Mine is not in a "use program track" mode unless selected by the PowerCab, its just a DCC (AC) powered piece of rail entirely separate from my layout.

Edited September 8, 2017 by imt

[imt]

It sounds like you've removed it from the track whilst the chip is still responding to top speed (ie maybe loco stalled on bad track/frog). Maybe try putting the address into the controller with the throttle off and putting it back on the program track..

 

 

This is the nub of my problem.  The program track is a piece of Peco ST-204 (about 650 mm?), I have a separate PCP in a box to feed it, the NCE PowerCab is physically disconnected from the layout and connected to this entirely separate track.  Other locos perform properly (i.e. sit there and do nowt until given a command).  This loco when put on the powered programming track - but without the NCE POwerCab having its loco number set or any accelleration set - goes wild.

[imt]

Well, just to prove one thing I have just removed the chip, inserted a blanking plate, attached my DC controller to the "programming track" and driven the chassis back and forward in DC mode.  It works fine.  If I put the chip back in and set up the DCC controller it flies off the end of the rail - just caught it! - before I can do anything (such as set the rail in "set program track" mode or even put the loco address in).  

[73c]

Have you tried the chip in another loco to see if it does the same thing?

[PaulCheffus]

Well, just to prove one thing I have just removed the chip, inserted a blanking plate, attached my DC controller to the "programming track" and driven the chassis back and forward in DC mode.  It works fine.  If I put the chip back in and set up the DCC controller it flies off the end of the rail - just caught it! - before I can do anything (such as set the rail in "set program track" mode or even put the loco address in).

 

Hi

 

This suggests there is a connection somewhere between the chassis and the motor.

 

Cheers

 

Paul

[Nigelcliffe]

This is the nub of my problem.  The program track is a piece of Peco ST-204 (about 650 mm?), I have a separate PCP in a box to feed it, the NCE PowerCab is physically disconnected from the layout and connected to this entirely separate track.  Other locos perform properly (i.e. sit there and do nowt until given a command).  This loco when put on the powered programming track - but without the NCE POwerCab having its loco number set or any accelleration set - goes wild.

 

There is a problem in understanding here.....

 

Programming track should really not have any power at all.   BUT, with the PowerCab, there is power on its outputs UNTIL you put it into programming mode, so its not really an ideal device for a programming track. 

Therefore, when testing a new installation, you should first put the PowerCab into programming mode, and only then, put the loco onto the rails.  Then try to read a value back from the decoder.   If the installation is "good", then the read should work and have a sensible value.  If not, then something is wrong, remove loco from rails and investigate.   

 

Or, better still, invest in the NCE Auto-Switch (costs less than a decent decoder) which ensures the programming track is without power at all times.

 

 

 

 

Its quite possible to find locos with a wiring short from pickups to one of the motor terminals.  This won't show on DC running with the blanking plate fitted (because the blanking plate connects the pickups to the motor terminals), but is a disaster for DCC chips if not spotted and corrected. 

[imt]

Well that's that then.  Having proved that it works in DC mode, I have taken the DCC Concepts Zen 360 out of the 8 pin socket and recovered a simple Hattons 8 pin chip from a Class 55.  Doesn't go like the splendid Zen did, but at least it performs properly and proves I hadn't done anything wrong anyway.  The conversion is correct and works in either DC (with blanking plate) or DCC with the Hattons chip.  I then put the Zen into the Class 55 and it showed the same symptoms - trying to rush of the track.  As a final test I cut off the stay alive and again tried the Zen in the Class 55 - and it still did it's flying rush. The Zen has failed - but why I don't know.  I have contacted DCC Concepts to see if they have any suggestions - but await a reply.  When I get one I'll let you know.

 

Great relief - I DIDN'T cock up the conversion after all!

[73c]

Some good news then imt. I've got a Bachmann 2-epb that wants beat the land speed record, tried different chip, same thing. Will send it to Bachmann for repairs.

[imt]

There is a problem in understanding here.....

 

Programming track should really not have any power at all.   BUT, with the PowerCab, there is power on its outputs UNTIL you put it into programming mode, so its not really an ideal device for a programming track. 

Therefore, when testing a new installation, you should first put the PowerCab into programming mode, and only then, put the loco onto the rails.  Then try to read a value back from the decoder.   If the installation is "good", then the read should work and have a sensible value.  If not, then something is wrong, remove loco from rails and investigate.   

 

 

 

Sure is a problem of understanding - mainly mine but I am trying to catch up.  I put the PowerCab into "set program track mode" and my trusty multimeter assures me that there is 13.5 volts AC there.  It's only a cheap Maplin one so it probably isn't very accurate, but there is AC power there. "Normally" behaving locos of course just sit there - the duff Zen chip likes the power and goes wild.  Sorry but you are possibly in error or my PowerCab doesn't work properly.  It is possible but it seems to do everything else right?

 

Anyway I think I have proved the chip is duff - see posts 13 and 17 above. Happy for any other comments from you - I NEED to learn these things.

[Butler Henderson]

Try CV8 = 8 on the Zen, that will reset it to its original settings.

[imt]

Try CV8 = 8 on the Zen, that will reset it to its original settings.

Yes I do know that. Believe me I'd really like to, and if you have any ideas on how to stop the loco shooting off so I can establish control for long enough to do anything I'd be most grateful.

[Pete the Elaner]

Yes I do know that. Believe me I'd really like to, and if you have any ideas on how to stop the loco shooting off so I can establish control for long enough to do anything I'd be most grateful.

Place it on a rolling road.

Place a heavy object in front of it (or hold it with your hand) & let the wheels spin.

If you have a continuous layout, remove all other locos & let it go round as fast as it wants. You can reset it on the main.

[John ks]

Have you tried disconnecting one wire from the motor & doing a reset.

 

Or build a simple test rig

Start with a 8 pin socket, connect 2 wires from the track to the track pins & connect a resistor(30 to 50 ohms should work) or a spare motor across the motor pins.

Plug your 8 pin decoder into this & program the decoder & or check CV's

 

John

[imt]

Have you tried disconnecting one wire from the motor & doing a reset.

 

......

 

John

Thank you. It's obvious when you stop panicking and think! Which is why I asked on here, cos I was doing the panicking. I'll try that when I get some time this weekend.

[RAF96]

Some decoders need a load like the motor to accept programming so disconnecting the motor may not guarantee being able to reset the decoder but worth a try.

Rob