Pick up problem with DCC?

[dmsmith]

I have built a Craftsman 02 0-4-0 diesel shunter with a High Level 108:1 gear box and Mashima motor. The pick ups are phosphor bronze wire bearing on each of the wheel flanges. The loco runs well on DC with a blanking plug. However, I can’t get it to work with DCC. The motor turns a few revolutions and then stops. Given a nudge, the motor turns again, but for a few more revolutions. After each nudge, the motor runs for longer until eventually it turns continuously.  
 

I installed an NEM651 6-pin socket, so that I could swap decoders and blanking plugs. I’ve tried several different types of decoder, but to no avail. My most recent experiments have been with a Zen Blue+ (DCD-ZN6D.2).  To eliminate one variable,  I hard wired in the track feed to the decoder socket. The loco then runs perfectly on DCC … better than DC. That makes me think that it can’t be a problem of excessive current draw and makes me suspect the pick ups. I therefore invested in a Zen small stay alive, (DCD-SA3-SM.1). That seems to help a bit, but the loco is still unusable.

 

I've tried the logical things like deactivating DC and back EMF. There is no suggestion of a short across the pickups.  I’ve cleaned the track and wheels meticulously. Does anyone have any ideas? I’d be very grateful for your help.

 

Best wishes

David

 

 

 

[Ray H]

Have a look at the start voltage CV3 (I think). Increase the value a fair bit (think the default is 0) and see if that does anything, then slowly lower the value back down until the problem re-appears then increase the value again slowly until the problem goes.

 

We had a loco at our club that was a lumpy runner at low speed. Doing as indicated above all but erased the problem.

 

The good thing is (?) once you know (and record) the first value of any CV you wish to alter you can alter that CV using Programming On the Main - but don't forget to keep a note of any changes. Saves a lot of faffing about using the Programming Track.

[RFS]

9 hours ago, dmsmith said:

 

To eliminate one variable,  I hard wired in the track feed to the decoder socket. The loco then runs perfectly on DCC … better than DC.

 

 

Not sure I understand what you mean here. Surely both track feeds (and motor feeds) are hard-wired to the DCC socket in the first place?

[dmsmith]

When I built the loco, I installed a 6 pin socket with connections to the pickups and the motor. To try to diagnose the problem, I’ve made a short harness with a 6-pin plug at one end and another socket at the other that I can insert between the old socket and the decoder. The wires to the motor run straight through, but I’ve attained some fly leads so that the power now runs straight from the new decoder socket and can be connected directly to the track to bypass the pick ups.

Thanks for your thoughts!

David

[dmsmith]

38 minutes ago, Ray H said:

Have a look at the start voltage CV3 (I think). Increase the value a fair bit (think the default is 0) and see if that does anything, then slowly lower the value back down until the problem re-appears then increase the value again slowly until the problem goes.

 

We had a loco at our club that was a lumpy runner at low speed. Doing as indicated above all but erased the problem.

 

The good thing is (?) once you know (and record) the first value of any CV you wish to alter you can alter that CV using Programming On the Main - but don't forget to keep a note of any changes. Saves a lot of faffing about using the Programming Track.

Thanks for the suggestion, Ray!
 

Unfortunately, that was one of the first things that I tried. It didn’t seem to make any difference at all! It was after that test that I installed the direct connection from the decoder socket to the track feed to bypass the pickups. The loco then works faultlessly, even with a CV3 setting of 0. 
 

It’s really strange!
David

[Oldddudders]

CV3 is the acceleration rate. CV2 is the starting voltage. 

[dmsmith]

Yes, Oldddudders! It was CV2 that I adjusted. I didn’t tinker with CV3.

[Izzy]

Under DCC poor current collection can become an issue when it isn’t under DC. The make of decoder and the firmware it uses can also be a contributing factor. Using DC the motor just runs at whatever voltage has been set and interruptions to the supplied current won’t alter this. If these pauses in supply are fractional then often the overall performance can be quite good enough not to be noticed.

 

But…with DCC even the smallest break in current can cause a decoder to reset, to go back in it’s commands and restart. So lights go out and need re-setting and the motor starts off again from zero speed. Multiply this a few times over a short period, a second or two, and problematic running is the result. And of course intermittent current can also perhaps be seen by the decoder as commands it’s been sent to add to the problem.

 

However, much also hangs on the decoder make and it’s firmware. Some - the better makes as regards firmware - try and make allowance for these problems by remembering settings so the current supply issues aren’t so bad. Think Zimo etc. Of course fitting some amount of stay-alive will also help get over the problem, and a Zimo + SA (even a minimal amount) - my default fitment these days -  will generally prove helpful in producing a decent and reliable answer.

[Nigelcliffe]

As per Izzy,  fit a decent decoder (Zimo), and a decent stay alive.  On the stay-alive, if there is room I'm now using the Lais units available from Digitrains (three different package shapes).  If tighter space there are other approaches, right down to DIY with surface mount parts.    

 

I'm assuming the pickups are working properly, and the wheels are smooth & clean. Quite a few of the wheels sold into the kit market have terrible tyre quality: pitted and scarred, which leads to poor pickup.    They can be improved by careful polishing with fine wet&dry papers.  

 

For ease of wiring, drop 6-pin and adopt Next18 instead.  Next18 sockets are available (Lais again) with large solder pads on the back for wiring.  Those include the decoder positive and negative needed for stay-alive units, and speakers (if fitting sound).  Which then makes decoder install a doddle - clip it in and job done.  

 

 

- Nigel

[kev]

If you have a multi meter do a check from the wheels to the socket you should show a short from the wheel set to the socket  if not clear that first i had a half ohm reading on one loco and it caused poor running

[dmsmith]

Thank you all for your suggestions. My assumption had been that the decoder is resetting if there is a momentary interruption in power from the pick ups. I was just surprised that there isn’t any hesitancy on DC. I did check the resistance between track, via wheels and pickups, to the decoder socket, and did get a short.
 

It is possible that the wheel flanges are pitted. I’ll try turning them smooth, but I am nervous about removing them from the axle too many times. they are Alan Gibsons, and they don’t take kindly to being taken on and off. I’ll let you all know how I get on.
 

Perhaps I am expecting too much of the Zen small stay alive. The DCC Concepts website suggests a run-on distance of 150 mm, but that must depend on speed, and with a 108:1 gearbox, that’s just a crawl. Interestingly, I can’t find the capacitance listed on the website. There’s not much space in the cab, but I’ll look at the Zimo range and a bigger socket.

 

Best wishes

David

[SamThomas]

7 hours ago, Nigelcliffe said:

For ease of wiring, drop 6-pin and adopt Next18 instead. 
 

Why ? 6 (or even 8) pin sockets are more than ample for many applications & of course the 6-pin tiny & ideal for small locomotives.

[Nigelcliffe]

26 minutes ago, SamThomas said:

Why ? 6 (or even 8) pin sockets are more than ample for many applications & of course the 6-pin tiny & ideal for small locomotives.

 

Explained in the rest of the paragraph you snipped out.   Because there are solder pads for stay-alive, so don't have to wire things to decoder's PCB.   It really is a easier.   

 

 

[Nigelcliffe]

2 hours ago, dmsmith said:

It is possible that the wheel flanges are pitted. I’ll try turning them smooth, but I am nervous about removing them from the axle too many times. they are Alan Gibsons, and they don’t take kindly to being taken on and off. I’ll let you all know how I get on.
 

Perhaps I am expecting too much of the Zen small stay alive. The DCC Concepts website suggests a run-on distance of 150 mm, but that must depend on speed, and with a 108:1 gearbox, that’s just a crawl. Interestingly, I can’t find the capacitance listed on the website. There’s not much space in the cab, but I’ll look at the Zimo range and a bigger socket.

 

 

Wheels - try running them in the loco, with power direct to motor (from DC, no decoder fitted), and use some fine wet&dry stuck to a wooden stick.  Move through grit sizes, ending up at polishing grades (might get nice sticks quite cheaply in some places for work on fingernails).    Don't forget to do the place where the pickups rub.  

 

Not much space ?  I've got stay-alive in my 2mm scale class 02!     It may require a bit of thinking about where the space can be found, and components may have to be distributed around the loco to get as much in as possible.   There's often lots of space in the chassis of 4mm locos, between the frames.     

 

Zimo will offer you tiny decoders, all support stay-alive, but most of them require the stay-alive to be soldered to pads on the decoder.   

 

- Nigel

[dmsmith]

On 02/10/2021 at 12:54, Nigelcliffe said:

As per Izzy,  fit a decent decoder (Zimo), and a decent stay alive.  On the stay-alive, if there is room I'm now using the Lais units available from Digitrains (three different package shapes).  If tighter space there are other approaches, right down to DIY with surface mount parts.    

 

I'm assuming the pickups are working properly, and the wheels are smooth & clean. Quite a few of the wheels sold into the kit market have terrible tyre quality: pitted and scarred, which leads to poor pickup.    They can be improved by careful polishing with fine wet&dry papers.  

 

For ease of wiring, drop 6-pin and adopt Next18 instead.  Next18 sockets are available (Lais again) with large solder pads on the back for wiring.  Those include the decoder positive and negative needed for stay-alive units, and speakers (if fitting sound).  Which then makes decoder install a doddle - clip it in and job done.  

 

 

- Nigel

Thank you all once again for your suggestions. I polished the wheel flanges carefully, as suggested, which made a small improvement, but the Zen decoder still kept shutting down. I’m more convinced than ever that the contact between wire pickup and flange is sending spurious signals to the decoder which interprets as commands. So, I followed the next suggestion and invested in a Zimo decoder and Lais stay alive from Digitrains, which came 24 hours after ordering. Problem solved! The running is now exceptional and the stay alive does everything that it should.

 

I opted to stick with a 6-pin decoder because I’d already fitted the socket on the loco and fed it through a small hole in the cab floor. I was reasonably confident of my ability to solder the capacitor wires to the chip, and managed it without any problems. However, I can certainly see the appeal of the much bigger solder pads on the 18-pin socket.

 

Now  to see if the same solution works with my Judith Edge NBL 0-4-0 …

 

Best wishes

David

[Nigelcliffe]

9 hours ago, dmsmith said:

Thank you all once again for your suggestions. I polished the wheel flanges carefully, as suggested, which made a small improvement, but the Zen decoder still kept shutting down. I’m more convinced than ever that the contact between wire pickup and flange is sending spurious signals to the decoder which interprets as commands. So, I followed the next suggestion and invested in a Zimo decoder and Lais stay alive from Digitrains, which came 24 hours after ordering. Problem solved! The running is now exceptional and the stay alive does everything that it should.

 

I opted to stick with a 6-pin decoder because I’d already fitted the socket on the loco and fed it through a small hole in the cab floor. I was reasonably confident of my ability to solder the capacitor wires to the chip, and managed it without any problems. However, I can certainly see the appeal of the much bigger solder pads on the 18-pin socket.

 

Now  to see if the same solution works with my Judith Edge NBL 0-4-0 …

 

Best wishes

David

 

Glad it worked, 

I can't see why the NBL 0-4-0 wouldn't be similarly improved by the same collection of parts.   If a £20 Zimo won't fit, then there are smaller ones at higher prices (MX616, and the even smaller MX615 may be in the shops before Christmas).   Similarly, there are smaller stay-alive packages possible, but not with the same quantity of energy stored.   

 

There's still a need to get wheels/pickups/track to a high standard.  Then fit a good decoder and stay-alive to get even higher quality running. 

 

 

- Nigel