Piko 218 and Massoth XL problems

[jimmielx]

Having read some great reviews of the Piko BR218 particularly regarding reliability and having decided that my wallet does not stretch to a new Countess, I jumped at the chance to acquire a second hand 218 in Maroon. I have added red LEDs for the rear lights and some little GOW bulbs in the cabs for some cab illumination. All of this is wired to a Massoth XL decoder. This is by far the largest and heaviest engine I own, the others are an LGB Track cleaner with MTS chip (I guess), a Baguely with an ESU LokPilot XL (lots of additional weight and skates - now an excellent runner), a GRS 0-4-2 on an LGB Toytrain chassis with a tiny CT Elektronic DCX51 decoder and a Scratchbuilt GW railcar on two USA trains power bogies with a Massoth XL decoder and power buffer.

Initial results with the 218 were very promising all the lights work well, no flickering at all as the loco goes round so presumably there's good pickup, however the loco variously stutters and jerks during running. I put this down initially to needing to set up the Back-EMF correctly, however I've now tried pretty much ever combination of back-EMF settings I can think of including turning it off, but I still get stuttering at some speed. I have tried taking the power buffer from my railcar and putting this into the 218 which certainly helped a little, but has not cured the problem, it now sometimes slow right down to a stop, pauses (maybe while the capacitor is recharging?) and then speeds up again. Interestingly the Railcar has similar problems when run without the power buffer, but runs perfectly with it. My railway is pretty modest in sizing as it squeezes into a London garden, however I am aware that there are some voltage variations along the line as I have current detection sections for computer controlled automation (which is how it runs most of the time) which are independently fed from the shed (probably on slightly under size cable) and I wonder if this may be the problem. The main reason I'm not sure is that my other four locos would often run for up to eight hours without a hint of a problem and with lesser quality decoders. I suppose what's different about this loco is it's weight (my railcar which also has two motor blocks is very light - probably too light, but since it doesn't pull anything and has the two motors it has no problems with traction), it is much heavier than anything else I own. I can't directly measure the current pulled on the track, but I can measure the current to my Lenz command centre. The track voltage is nominally 18v and the supply voltage is to the command centre is 21v DC. The 218 draws around 1A from the supply (so I guess a tiny bit more on the track) which is about the same as my Baguley draws.

So I am a bit stumped as to why I am experiencing these problems, my next action is to take the ESU decoder out of the Baguley and try it in the 218 instead. I am wondering if the Massoth decoders perhaps don't deal with the voltage drops well, although I'd have thought the cheapo CT Elektronic HO decoder would be the one to show problems. The ESU lokpilot does have a small power buffer built in to it and in the Baguley has been faultless so I'll give that a go in the 218 next. However if anyone does have any bright ideas or suggestions I'd be very grateful to hear them! I hope I've explained it all properly...

Cheers

 

[mike]

possably its the loco? have you checked that both motor blocks are picking up power? did you link the pickups form both sets off motor boggies? use a mulitimeter to test they are picking up power, try cleaning the wheels, ect

 

[Gizzy]

Fitted an XL to a 2095 Wizzy Cranks recently, but only connected one set of bogie wires from the decoder to the motor pcb on the loco. Only one bogie was driving.

Realised after having traced the copper tracks on the motor pcb that I needed to connect the other set of bogie wires to a parallel set of terminals on the motor pcb, as these were not commoned!

I think your problem though may be down to a combination of dirty wheels and track. The Piko wheels seem to have a reputation for this?

Your suggestion of cable feeds could also be valid. You could try to double up on the cable if you don't have any heavier duty replacement....

 

[don9GLC]

. . .
My railway is pretty modest in sizing as it squeezes into a London garden, however I am aware that there are some voltage variations along the line as I have current detection sections for computer controlled automation (which is how it runs most of the time) which are independently fed from the shed (probably on slightly under size cable) and I wonder if this may be the problem. The main reason I'm not sure is that my other four locos would often run for up to eight hours without a hint of a problem and with lesser quality decoders. I suppose what's different about this loco is it's weight (my railcar which also has two motor blocks is very light - probably too light, but since it doesn't pull anything and has the two motors it has no problems with traction), it is much heavier than anything else I own. I can't directly measure the current pulled on the track, but I can measure the current to my Lenz command centre. The track voltage is nominally 18v and the supply voltage is to the command centre is 21v DC. The 218 draws around 1A from the supply (so I guess a tiny bit more on the track) which is about the same as my Baguley draws.
. . .

Despite your detailed description there are too many aspects that could be contributing to this problem to offer definitive advice.

However, I would be suspicious of the track occupancy detection modules. So far as I am aware they work by measuring the current to the track and do this by inserting resistance in the feed to the track. You are right that undersized cables will not help but they should not be causing the erratic behavior you describe.

My suggestion is to disconnect one module from the power and track wiring and connect the wires to bypass the module. It may be possible to link out the module without disconnecting anything but without knowing the internal circuit I cannot be sure that no damage would be caused.

If this cures the erratic behavior in the affected section of track, repeat for other sections to determine if you have a few faulty modules or if they are all incompatible with your loco.

On the other hand if this erratic behavior is taking place on sections of track that do not have detection modules you have another problem! But no one can guarantee you have only one problem, so provided the effort is modest, I would try bypassing one module to see if there is any difference.

There could be all sorts of other reasons for your problems of course.

Hope this helps,

Don

 

[jimmielx]

Thanks Mike and Gizzy for the replies.
Yup, both sets of bogies are picking up power, one wheel on one side is not, but I don't think that should make any odds. I did attempt to clean the wheels with my peco HO wheel cleaning brush and scraper which is how I clean the Baguley's wheels - I must say it didn't seem to make much visible difference as the wheels remained pretty black, but were (apart from the one mentioned) picking up ok. So the way these Piko wheels look and work is yet another new one on me. LGB and USA trains wheels seem to remain pretty spotless and never need cleaning, Baguley wheels a quite well documented on here and I find need a clean every 10 hours or so and the Piko wheels seem to me to perform better than it would appear they might.
As I said I did swap out the Massoth XL decoder for the ESU XL decoder from my Baguley, just connecting the motors and pickups. And the difference is night and day. With the ESU decoder not only is all stuttering gone, but the general running characteristics of the loco are world better - ignoring any questionable track sections the loco runs much smoother with the ESU decoder generally, it achieves slow speed crawls which no amount of tinkering with the start voltage on the Massoth could get and at all speeds just looks smoother. It also sounds a little different, with the Massoth decoder there is a bit of a growl, whereas the ESU decoder seems to purr a little more. Both were set to the highest DC motor frequency which on the Massoth is 16kHz, and on the ESU is 32kHz, whether that makes a difference I don't know. I didn't tinker with CVs so the ESU just had default Back EMF settings and certainly didn't slow down on my radius 1 bends, but did pull another half amp from the supply.
So for me the simple conclusion is to go with the ESU decoder - everything about them just seems better... I may actually try the ESU decoder in my railcar which has never been as smooth as the others (I had put that down mostly to one motor block being ever so slightly faster than the other) and does also growl a bit, although certainly was never jerky once the power buffer was installed.
I also measured my track voltages. The main voltage on sections with a direct connection to the the command station is 17.5v which remains basically constant. On the furthest section from the shed (and therefore the current detection modules!) the track voltage is 16.6v, which dropped further to 15.6v when the 218 was on that section of track. So Gizzy you're right I could certainly do with using some thicker cable to these sections. Although in practice pre 218 I'd only ever noticed a very slight level drop on the headlights of incandescent lamps on the locos (and then only if looking closely), the locos have never visibly slowed on these sections, perhaps the Back EMF is taking care of this.
My feeling is that the Massoth decoder either does not like the voltage variations in my line, or perhaps that it's not wild on the lower voltages (G scale being usually 18-24v) that I'm running at (for me though this is more than enough and all my decoders have a top speed setting of around half what they are able to give) or perhaps a heady mix of the two. I say that because I now have two of them which don't appear to be the best performers for me. In the railcar the Massoth XL has very little work to do, it does run two motors, but has a relatively tiny load to pull, although it does require a power buffer to have acceptable running and I now suspect may run smoother with something else. In the 218 which is a much heavier beast the XL just doesn't seem to have coped well with the challenges at all, even with a power buffer which does improve things. I certainly don't intend to bash the Massoth decoders, apart from anything else plenty of people on here use them very sucessfully, but just want to report my findings. For me the biggest surprise was how much smoother the loco runs, particularly at slow speed, but noticeably at all speeds with the ESU XL. I will report back on how the ESU XL fares in my railcar in comparison with the Massoth too.

Thanks

 

[jimmielx]

Don
I just saw your reply too. Thank you for that. Not all my track sections have occupancy detectors, but it felt like the places where I was seeing most of the problems was just after the loco passed into or out of a section which was detected, I'm not certain of that though.
Aside from I think having found the solution for me above, you pose an interesting question which is whether the voltage drop is due to the occupancy detectors themselves or the wires from them to the track. Without taking too much apart I can bypass some of the occupancy detectors and see what difference that makes. My suspicion is that I am losing approximately one volt to each detector so I am measuring 16.6v in the detection sections compared to 17.5v in the undetected sections with no load, but that possibly the drop of a further volt under load in the detection sections to 15.6v is due to the cable. I'll have a look at this later too.

Cheers

 

[bunnyrabbit03]

Hi James,

Your problem is most likely caused by the Massoth decoder itself, not your line, installation etc. Massoth decoders are known for having problems with (very) slow running, in particular in combination with the 5-pole Piko motor in comparison to the 7-pole LGB motor. Following a recent test in the German railway magazine Gartenbahn Profi, try the following CV settings for the Massoth-Piko combination recommended by the magazine and Massoth:

CV 2: 3
CV 60: 2
CV 61: 150
CV 62: 255
CV 63: 63

Maybe these settings will help you

Marc

 

[jimmielx]

Hi Marc

Thank you for that, I've tried those settings and I'd say that it is running as well as it has under them with the Massoth decoder, particularly at the slowest speeds. Unfortunately I am still seeing some stalling and restarting.

 

[jimmielx]

I've also tried the ESU decoder in the railcar and again the performance is better. Slow speed is most noticeably much smoother, but at all speeds it seems to glide along more effortlessly.

 

[bunnyrabbit03]

Hi James,

I am glad the settings were of some use. The ESU decoder has a power buffer built in, so there is already an advantage. Just for your information, the recommended settings for the ESU-Piko combination from the same magazine article are:

CV 2: 2
CV 53: 50
CV 54: 32
CV 55: 25

Best,

Marc

 

[jimmielx]

Hi Marc

Thanks for the ESU Piko settings. They will be most useful!

 

[bigjack]

I have a 218 with a Massoth Sound decoder fitted to it, I have found simular problems with it as you are experiencing. The fault with my loco appeared to be dirty track and wheels. The decoder looks for a constant 24v supply and if the track/wheels are dirty, then the decoder won't pickup the signal correctly.
Personally I prefer the ESU decoder as they look for less voltage (18v) so run more smoothly as the track is supplying more volts than needed. I also believe that the LGB MTS decoders run on 18v as well, but I may be wrong on this point.
So, make sure your track and pickups are clean.

 

[jimmielx]

Hi BigJack

That certainly rings true. The Massoth literature states that the decoder needs 0-24v (presumably it won't work on 0v!), my suspicion is that it's optimised for 24v and will work best at higher voltages. My lgb track cleaner which presumably has an lgb decoder, runs very well on my track, so it certainly works well with lower voltages.

 

[dunnyrail]

I have had the odd jerks and stalls with the Massoth decoders as you appear to have. However on making sure that the track is scrupulously clean I find that the problems went away. However my track is all securely clamped up with Hillman Clamps, could you be getting voltage drop with possibly a less than secure feed all round your line.
JonD

 

[ntpntpntp]

my suspicion is [the Massoth decoder] is optimised for 24v and will work best at higher voltages. My lgb track cleaner which presumably has an lgb decoder, runs very well on my track, so it certainly works well with lower voltages.

No experience with Piko locos yet, but I do have at least one Massoth XL decoder in service with no problems at all (think it's in a Bachmann Mallet but I'd have to check my install notes). My line is only running on around 18v.

... presumably it won't work on 0v!

All decoders need a certain voltage level to "wake up" the circuitry. It varies a little by brand, but on average around 7v from my observations of the different brands I've installed.

 

[Cliff George]

Hello James,

Although this probably isn't going to help much I can confirm some of your observations about Massoth decoders in a similar setup to yours.

I have recently fitted my first Massoth decoders to two of my LGB locos. I haven't seen any stuttering as you report maybe because both also have power buffers fitted. They both slow down, stop and start again when they shouldn't. I put this down to them invoking the 'brake on DC' feature when they shouldn't. The decoders don't seem to have a way of turning this feature off. I emailed Massoth about the problem but got no response from them.

I also have occupancy detection but am using the type with current transformers that have no direct connection to the track and so should in theory have no effect on the track voltage at all. I do have several power districts and the slow down and stop problem does seem to occur more regularly at power district boundaries.

I also have a Lenz system set at 18V, I use 4mm house wire to all of my detection sections to avoid as much voltage drop as possible. I measure about 17.8V on the track.

I too have tried tinkering with a lot of the decoder settings, but to no avail.

I do not have any similar problems with my other brands of decoders (Lenz, ESU, LGB) if they are fitted with power buffers or not.

I indend to replace these Massoth decoders and have crosssed Massoth off of my possible supplier list, which is a pity really.

Sorry this is not a solution.

 

[jimmielx]

Thanks for all the replies chaps,
So I've looked at the connections to the motor, I ended up bending the connectors on the motors themselves slightly in to make a better contact with the pins on the motor blocks, I'm not sure if this had been a problem, but the connection didn't look very satisfactory there.
I've also cleaned the track and wheels and am basically getting good contact - the lights certainly never flicker which I'm taking as a good sign - as I'd expect from such a heavy loco with so many pickups.
Having ensured all of this I am still getting the slowing down, stopping and then accelerating away again which you describe Cliff, so the brake on DC engaging seems very possible. In addition ESU decoder is giving a smoother ride at all speeds.
Cliff, I use LDT occupancy detectors, is that what you are using?
Basically I've reached the same conclusion as you which is that the Massoth decoders seem to be incompatible with my setup for some reason. I could spend a lot of time rewiring the system with heavier gauge wire which following your note above might or might not work, but having seen how well the ESU decoder works and given that the rest of my stock runs so reliably I am just going to swap the Massoth decoder out for an ESU one, and will probably stick with those in future.

Cheers

 

[Cliff George]

Hello James,

Cliff, I use LDT occupancy detectors, is that what you are using?

I'm using the Trak-DT from Dallee Electronics: http://www.dallee.com/

I first put some of these in about 5 years ago and I've stuck with them as they work well. There are more options available now and if I were starting now I would certainly evaluate the new options.

I have some LDT equipment but not the occupancy detectors. I use the SA-DEC-4-DC to control my air activated points.

Good luck with solving your problems. If you find a way of making the Massoth decoders work for you I'd be interested in the solution.

 

[whatlep]

my suspicion is [the Massoth decoder] is optimised for 24v and will work best at higher voltages. My lgb track cleaner which presumably has an lgb decoder, runs very well on my track, so it certainly works well with lower voltages.

No experience with Piko locos yet, but I do have at least one Massoth XL decoder in service with no problems at all (think it's in a Bachmann Mallet but I'd have to check my install notes). My line is only running on around 18v.

... presumably it won't work on 0v!

All decoders need a certain voltage level to "wake up" the circuitry. It varies a little by brand, but on average around 7v from my observations of the different brands I've installed.

Chaps - all the decoders have to abide by the NMRA standards for voltages to be deemed NMRA-compatible. Both Massoth and ESU are. For G scale, the decoder standards are:
minimum voltage - 7 volts
maximum voltage - 27 volts
expected voltage - 18 volts (i.e. the expected output from the DCC central station)

For what it's worth, the only decoders that have ever given me problems have been from ESU. Two supplied, two failures within a month in 2009. I've never used them since. Massoth and their LGB twins have given no problems at all over the past 11 years.

 

[jimmielx]

Thanks for that clarification Whatlep, that makes sense. I don't wish to bash the Massoth decoders generally, apart from anything else a vast majority on here find them highly reliable, just reporting my experience under some quite specific circumstances. It's certainly very useful to hear a different experience to keep things in perspective.
I think I may have leapt to a few too many assumptions in later posts. I've just been running the loco for a few hours outside. I feel confident that the problem I have is that same as Cliff has experienced, namely slowing to a stop and then restarting again, in my case almost always where the loco has just gone into an area with a separate supply (via an occupancy detector). When I was seeing the jerky action reported in earlier posts I suspect this was trying to fix that having much too much Back-EMF dialed in. I have now used the Back-EMF settings recommended by Marc and although I am in no doubt that the ESU decoder I tried in it produced smoother running at slow speeds, I would be happy enough with the Massoth decoder.
So I believe that I have one problem - the loco decides to stop and then restart again occasionally. That would appear to be the same problem that Cliff has experienced. Apart from both running automated railways with current detection (and I do use a quite different system of current detection to Cliff) we both run a Lenz system, I have no idea if there's anything in that as I do find mine very reliable and everything should conform to NMRA standards and be cross compatible.
I still have to try bypassing the current detection modules and see if that makes a difference.
I am pretty sure that I am going to put an ESU decoder in the 218 as my experience with them has been good so far and I'm hopeful that will keep my problems at bay.
This thread has produced a wealth of information though so thank you all for that, I'll let you know how I get on. In the mean time of course if anyone has any further thoughts do add them here!

Cheers