I'm a bit disappointed and now on the verge of ditching N in favour of OO.
I've just run in my Dapol Class 22 and J39 (from the Eastern Freight kit) on DC.
I needed to try out my Digitrax Zephyr so I put a DZ126IN decoded into the Class 22.
As I've not got much space for a layout I'm building a of shunting layout so I need the engine to move very slowly, so after a bit of messing around I tried it on the slowest setting possible. The engine moved in a very slow jerky / stepped movement, I speeded it up and to a point visually it just moved in quicker steps and was the same in either direction.
I went to Digitrains in Lincoln and asked them to show me the difference in slow speed performance between N and OO. Their N gauge demonstrator (I think a class 66) showed the same stepped movement as mine. They then showed me a OO gauge, it had the same stepped movement though not as obvious but it did move about 5 times slower than the N.
Problem for me now is that N doesn't seem to move smoothly enough but I've not really got enough room for OO.
Could I please ask what peoples experience with N is at very slow speeds.
Quote from: Dag on December 06, 2015, 08:39:33 PM
I'm a bit disappointed and now on the verge of ditching N in favour of OO.
I've just run in my Dapol Class 22 and J39 (from the Eastern Freight kit) on DC.
I needed to try out my Digitrax Zephyr so I put a DZ126IN decoded into the Class 22.
As I've not got much space for a layout I'm building a of shunting layout so I need the engine to move very slowly, so after a bit of messing around I tried it on the slowest setting possible. The engine moved in a very slow jerky / stepped movement, I speeded it up and to a point visually it just moved in quicker steps and was the same in either direction.
I went to Digitrains in Lincoln and asked them to show me the difference in slow speed performance between N and OO. Their N gauge demonstrator (I think a class 66) showed the same stepped movement as mine. They then showed me a OO gauge, it had the same stepped movement though not as obvious but it did move about 5 times slower than the N.
Problem for me now is that N doesn't seem to move smoothly enough but I've not really got enough room for OO.
Could I please ask what peoples experience with N is at very slow speeds.
Gooduns and baddans I'm afraid. Some are really good straight from the box others need some running in and they get slowly better. Some are just rubbish and will never work well. Out of 30 locos I only have a couple in the rubbish category. I used to model EM 4mm gauge but prefer N gauge cos of the sheer amount of railway you can get in a small space. :beers:
Quote from: Dag on December 06, 2015, 08:39:33 PM
Could I please ask what peoples experience with N is at very slow speeds.
Assuming the loco isn't faulty and has been properly run in and lubricated, modern design locos generally run very well at low speed.
A good test is this: stick the loco on a length of spotlessly clean flexitrack (maybe polish it up with some IPA prior to use, letting it dry thoroughly before use of course). Apply some current. If it crawls along nicely there, where conditions are optimal, but doesn't on your layout, then time to cast a critical eye at the layout! Wiring, tracklaying (and any weathering or ballasting) as well as your cleaning regimen will all affect the track-to-loco part of the circuit.
Don't for a second imagine 00 scale modellers never have problems with low speed running! "Grass is always greener" comes to mind...
Cheers, NeMo
My two class 22s positively creep along when needed. Sounds like the cvs need playing with to get smooth running. I used to use Digitrax but now wherever possible use CT or preferably Zimo as their motor control is superb.
The most important question is how smoothly did the same locos run on DC before you fitted the decoder??
(you say you did test run them first as recommended...)
If they ran poorly on test on DC why did you proceed to fit the decoder and not sort them or return to the retailer unmodified?
If they ran acceptably on DC, then its not the loco that's at fault - it's either the decoder or it's settings...
Thanks everyone.
Class 22 was run in on a rolling road for best part of an hour in each direction. I then put it on an oval of track and on DC it crept along really well.
I then set up a brand new piece of flexitrack to try out DCC. It worked great apart from slow speed. The model in the shop shows the same "stepping" nature and I've seen other posts which sort of mention it. According to the manuals everything comes from the factory set to 128 speed increments, I'm not sure if there is any way to adjust for smoothness at speed. I've had play with settings but only for acceleration.
I'll try again tomorrow with the piece of track cleaned as NeMo has said, Layout hasn't been started yet that's why I've still got the option to jump to OO.
Sorry, Dag, but if it ran nice and slowly on DC it isn't the loco as PLD points out.
Quote from: newportnobby on December 06, 2015, 10:06:39 PM
Sorry, Dag, but if it ran nice and slowly on DC it isn't the loco as PLD points out.
I thought as much, I'm just reading about the world of back EMF, torque compensation and other such delights :confused2:.
The other thought is that I'm expecting too much.
On my DC layout I can get the best locos crawling nicely over paintwork including the 3-ways for shunting. I'm led to believe it'd be possible to improve this if I went with DCC. Keep persevering mate and good luck!
Quote from: Dag on December 06, 2015, 10:24:15 PM
Quote from: newportnobby on December 06, 2015, 10:06:39 PM
Sorry, Dag, but if it ran nice and slowly on DC it isn't the loco as PLD points out.
I thought as much, I'm just reading about the world of back EMF, torque compensation and other such delights :confused2:.
The other thought is that I'm expecting too much.
Not at all, DCC driven loco's should run as well if not better than DC. I had a class 66 using a Zimo decoder take 30 mins to complete its own length, it was barely moving to the eye and would crawl most realistically at slightly higher speeds. It is certainly possible but you may need to investigate the tuning of your Digitrax decoder. They are a good make and offer many great decoders although I don't have much experience of them.
Try with a better decoder eg a Zimo or CT decoder, though it might be the settings that your DCC controller is creating as it programmes your decoders.
If it runs well on DC then it should run at least as well on DCC.
Some good news, I think I'm getting there. I've been messing around with CV09 (quiet operation) and have managed to smooth things out quite a bit without introducing too much buzz or whistling.
On the downside I've noticed that the controller doesn't do anything until it is turned half way to increment 1 and I don't think I've got speed step 01, I'm going to take it back to the shop tomorrow to see what they say.
Keep us posted on success or otherwise please, Dag.
Dag, it seems like you are getting it sorted. Divn't to be in too much hurry to go to 00. :thumbsup:
With the same chip as yours, I've found that the stepping effect can be reduced by applying an acceleration and deceleration, e.g. CV3=24 and CV4=20. The speed can be slowed by reducing the top voltage, e.g. set CV5=100. I find an convenient way to get slow speed running is to set the mid voltage to about 40% of the top voltage, in this case set CV6=40.
With luck this will improve things immensely.
If you want quality slow sped running in N, then switch to CT decoders.
I have found in both gauges that only some locos are good for shunting.
A lot comes down to track cleaning, wheel cleaning, engine maintenance (eg pick ups) and a good controller (talking about DC here though I think it all applies to DCC too).
Only thing is in N, a real life piece of dust or grime may have more effect as the contact areas are smaller.
Thanks everyone, took the engine to Digitrains, they put it on their track and the first comment was how well it ran, we tried low speed and they seemed impressed, I checked out their Zephyr and it worked the same as mine.
Over the past couple of nights I've tried everything that's been suggested and there's been no improvement, I've confirmed that I have speed step 001 by setting a totally flat throttle curve. The thing that changed everything for the better was CV09. Best I can manage is 15 seconds to move over its body length with no stepping, though that's well short of the 30minutes reported by Cutting42 above.