Cobalt Turnout Motor Problems

[newbryford]

4 hours ago, Andymsa said:

Some time ago it was pointed out to me that tortoises are outdated as they have never been updated from there design introduction. My thoughts are why reinvent the wheel, all the fancy intergrated dcc electronics, the lack of throwing the point by hand, the various redesigns have put me off colbalts. And judging by the issues users are getting a well founded product choice by the looks of it going with tortoises. How many reports of tortoise issues do you see, not saying they don't get faults, but if any of my 220 tortoises do get a fault just I take it out and put new one in, No redoing the dcc address. And one final point colbalts are not true stall motors as like the tortoise. I'm very surprised by the comment that you can run the colbalts off the track bus by the manufacturers although technically possible it's not very recommended especially when you run a point. Yes the colbalt is smaller but not by much but the size of the tortoises has never been an issue for me, So for me I'll stay with what works. I don't like to bash the colbalt but as a famous advert says does it do what's on the tin 100% of the time or even 99%? Judging by the posts I would say no.

 

Not quite sure where you get the information that they are not stall type motors from.

If you take apart a Tortoise and Cobalt - they have the same motors inside them.

 

A Digital Ip and Analog Ip will switch down the power to about 5mA to the motor after it has reached end of travel - i.e. stalled...….

The Omega will still take about 20mA when stalled - very similar to the Tortoise.

 

You are comparing apples with oranges when you put a Tortoise next to a Digital IP

The Tortoise is a purely analog motor that runs on a DC supply.

The Digital IP has an integrated DCC decoder that requires an address to work from DCC. Whether the feed is from the track bus or accessory bus, ultimately, it's the same DCC feed. (Digital IP can be powered from "straight" DC and use external pushbuttons to control them).

 

Ultimately, it is shorts that are creating a problem - and the usual cause of a short is either bad wiring or bad operation - running through points set against the train.

 

If you do want a separate track bus to feed the frog, then simple - switch it via the auxiliary contact on the Digital IP - in exactly the same way that this is done on Tortoises and micro switches mounted on solenoids.

 

Analog IP and Omega Cobalts are more similar to Tortoise in that they require a DC supply to operate - and like a Digital IP - are stall type motors.

 

 

 

 

 

 

 

 

 

Edited August 7, 2019 by newbryford

[Junctionmad]

To my knowledge there are two known faults across the cobalt range 

(a) a clicking from the gearbox which I understand is a manufacturing fault in a gear 

 

(b) the digital Ip versions have been reported to loose their dcc address from time to time as a result of a short in the  dcc supply typical of track shorts. 

 

Both of these need to addressed by DCC concepts as they are blighting a very good product 

 

having used all three variants and tortoise , I would contend the cobalt analog ip variant is the best of all four options 

Edited August 7, 2019 by Junctionmad

[WIMorrison]

The use of a filter across the track bus and optionally accessory bus has resolved the lost address issue for everyone that has used it - and it has generally improved DCC running.

[DCCconcepts]

1 hour ago, Junctionmad said:

To my knowledge there are two known faults across the cobalt range 

(a) a clicking from the gearbox which I understand is a manufacturing fault in a gear 

 

(b) the digital Ip versions have been reported to loose their dcc address from time to time as a result of a short in the  dcc supply typical of track shorts. 

 

Both of these need to addressed by DCC concepts as they are blighting a very good product 

 

having used all three variants and tortoise , I would contend the cobalt analog ip variant is the best of all four options 

 

Hi JM.

Thank you for your post, but we have long felt a need clarify some myths and misunderstanding regarding Cobalt motors.

 

In reply to an earlier couple of posts within this thread - Cobalt motors are true stall motors.

 

To address your points:

It is not a manufacturing fault with the gears.

There was one early batch (2012?) that had a case moulding issue due to incorrect temperature regulation, causing minute uneven shrinkage of the case when it was removed from the injection moulder. This could cause a gear mesh issue. Replacement cases were manufactured and supplied.

 

The current gear issue is caused primarily by using an unregulated supply - typically those on the side of H&M controllers with a "nominal" 12vDC output that is far from a smooth 12v.

It is relatively uncommon in the later batches of Omega/Analog Ip and Cobalt IP for gear issues to appear as the gear train was improved as a result of experiences with the first generation of Cobalt motors.

 

To make the Cobalt about 20% smaller than the Tortoise, the gear train carries a bit more stress through having smaller gears than the Tortoise. When used with regulated supplies, there are no issues, but unregulated/unsmoothed supplies can eventually cause the motor to "over-drive" and jump gears - once the damage is done, it can become worse. 

There is a lifetime warranty on Cobalt motors and we will exchange gears and advise customers with regard to suitable power supplies.

Initial gear issues with original Digital Cobalts were similarly caused by higher DCC track voltages - typically used by European systems.

A well documented "fix" was to wire a resistor in series with the power feed to reduce the voltage to the Cobalt.

 

The Digital IP and Analog IP have better voltage regulation within the internal electronics to overcome these issues. The Omega has a 6-12/12-18v selector switch.

 

There does seem to be a recent issue with some Cobalts losing addresses and we are currently trying to compile a database of users and systems that are having more than their fare share of problems.

It does appear to be those with higher current rated systems, inadequate short circuit protection and/or those with less than ideal wiring.

It seems that there was minor fault within the microprocessors that were fitted to a small batch made in late 2018 only. and combined with tolerance variations with other components on the circuit board, this has caused this issue of memory loss.

 

This has now been identified and these particular microprocessors will never be used again and the circuit board made more robust to prevent this happening on future batches.

 

As above with the original Cobalts, there is a simple fix for this small batch - if required. Albeit an old-fashioned one - that applies to this one and only batch of Digital IPs.

The addition of an inductive load in either the feed to the motor or the frog switch will cure the memory loss problem.

The simplest form of inductive load is a 12v/21W (or similar) car light bulb that will reduce the immediate impact of a short to the Digital IP.

We have already supplied them to a few customers that were having memory issues and this has immediately resolved them.

In fact, we will supply these bulbs free of charge if anyone requires them - please contact us via our website.

 

 

We are constantly striving to improve our full product range and welcome all customer feedback through the appropriate channels.

Please use the contact details on our website.

 

The Dccconcepts Team

 

Edited August 7, 2019 by DCCconcepts
typo

[Anthony566]

5 hours ago, blueeighties said:

Of course. Any DCC compatible point motor can be operated via a seperate accessory bus. But the major DCC Concepts selling point, is the fact that Alpha control systems use just the one bus for track and accessory control. That's a major selling point for customers in terms of saving time and money during installation. Less wiring, switching etc. Unfortunatley, this has not proved to be the case by a long stretch.

 

 

Hi Lee,

 

From my understanding of Alpha systems, they can plug directly into NCE systems and then the output from the NCE can be split into track and accessories. i.e. It does not have to be a combined bus as you seem to think it must be ?

It is the Digital IP that has the capability of being connected directly to the DCC bus and only one other wire powering the frog, negating the need for auxiliary switching of the track power to the frog.

Yet - it can also be wired from an accessory bus and track power switched through the auxiliary switch - as is the normal case for Tortoises and solenoids and servos with external micro switches 

 

I always find pictures speak a thousand words when problem solving ... do you have any photos of the layout wiring and maybe the power of the forum can help with this ?? 

[Andymsa]

7 hours ago, newbryford said:

 

 

The Digital IP has an integrated DCC decoder that requires an address to work from DCC. Whether the feed is from the track bus or accessory bus, ultimately, it's the same DCC feed.

 

 

This is not strictly true, if you have a separate booster for the accessory bus then would this not be a different DCC feed.

 

so if you do run a point regardless of the reason with a separate bus with its own dcc feed its a simple fix just throw the point. With a combined track/ accessory feed you have to shut down the layout and push the train away from the point and power up again..

[newbryford]

16 minutes ago, Andymsa said:

 

This is not strictly true, if you have a separate booster for the accessory bus then would this not be a different DCC feed.

 

so if you do run a point regardless of the reason with a separate bus with its own dcc feed its a simple fix just throw the point. With a combined track/ accessory feed you have to shut down the layout and push the train away from the point and power up again..

 

Maybe I should have said "ultimately it's the same DCC signal"? (Unless it's a completely independent DCC system - such as the Alpha system I use on Deadmans' Lane.

 

The point I was trying to make was that a Digital IP can quite easily work direct from DCC, whereas the Tortoise does not. as it requires a separate decoder.

 

It is well understood that having separate track/accessory  buses will allow the accessories to operate when the track is shorted - provided of course, that there is separation of them via a breaker or booster.

 

Yes - one of the main selling points of the Digital IP is that you can connect the frog directly to the motor if it is powered from the track bus.  This does simplify the wiring.

Great for small layouts when you can simply push the offending train away from the point and then carry on - not so great for a large layout if the train isn't easily accessible and then two buses becomes a much better proposition.

Edited August 7, 2019 by newbryford

[St Enodoc]

6 hours ago, WIMorrison said:

The use of a filter across the track bus and optionally accessory bus has resolved the lost address issue for everyone that has used it - and it has generally improved DCC running.

Which is good practice regardless of what type of point motors or other equipment you are using.

[St Enodoc]

7 hours ago, Junctionmad said:

To my knowledge there are two known faults across the cobalt range 

(a) a clicking from the gearbox which I understand is a manufacturing fault in a gear 

 

(b) the digital Ip versions have been reported to loose their dcc address from time to time as a result of a short in the  dcc supply typical of track shorts. 

 

Both of these need to addressed by DCC concepts as they are blighting a very good product 

 

having used all three variants and tortoise , I would contend the cobalt analog ip variant is the best of all four options 

I am using several dozen Cobalt iPs - a mixture of analogue and digital, for different applications. So far I have had no trouble with them whatsoever, with one minor exception.

 

If all the 50+ digitals are in circuit when I power up the DCC system (NCE Power Pro with a separate SB5 booster in dumb mode for the accessory bus), there appears to be an initial current surge that trips the 5A circuit breaker in the SB5. If I connect them progressively in groups of a dozen or so, all is fine. If Richard is still in the conversation, perhaps he can comment on this and let us know the value of the initial current surge (not the steady state stall current or the operating current - I have these from the manuals and online documentation).

 

Thanks.

[Andymsa]

3 hours ago, newbryford said:

 

Maybe I should have said "ultimately it's the same DCC signal"? (Unless it's a completely independent DCC system - such as the Alpha system I use on Deadmans' Lane.

 

The point I was trying to make was that a Digital IP can quite easily work direct from DCC, whereas the Tortoise does not. as it requires a separate decoder.

 

It is well understood that having separate track/accessory  buses will allow the accessories to operate when the track is shorted - provided of course, that there is separation of them via a breaker or booster.

 

Yes - one of the main selling points of the Digital IP is that you can connect the frog directly to the motor if it is powered from the track bus.  This does simplify the wiring.

Great for small layouts when you can simply push the offending train away from the point and then carry on - not so great for a large layout if the train isn't easily accessible and then two buses becomes a much better proposition.

 

Yes I fully  agree with size of layout and resolving running a point, and using a breaker or booster for the second bus. The real issue is design, manufacturing, and component issues seem to be a common factor with colbalts, although they seem to being resolved. 

[newbryford]

3 minutes ago, St Enodoc said:

I am using several dozen Cobalt iPs - a mixture of analogue and digital, for different applications. So far I have had no trouble with them whatsoever, with one minor exception.

 

If all the 50+ digitals are in circuit when I power up the DCC system (NCE Power Pro with a separate SB5 booster in dumb mode for the accessory bus), there appears to be an initial current surge that trips the 5A circuit breaker in the SB5. If I connect them progressively in groups of a dozen or so, all is fine. If Richard is still in the conversation, perhaps he can comment on this and let us know the value of the initial current surge (not the steady state stall current or the operating current - I have these from the manuals and online documentation).

 

Thanks.

 

With due respect, that's not a Cobalt issue.

 

No matter what the electrical system, if there is a large number of items connected at power up, there is the possibility of a cutout.

 

 

 

 

 

I can sometimes have a similar problem when a PSX breaker tries to reset on a section that has multiple sound fitted locos on it. Solution is to remove a couple of locos and all is well.

 

It's a bit like starting a four engined airplane - there isn't enough power to start all four at once, they are started progressively.

 

Simple solution is a switch - or switches - as you have found out - to progressively start up sections of the layout.

 

 

[St Enodoc]

13 minutes ago, newbryford said:

 

With due respect, that's not a Cobalt issue.

 

No matter what the electrical system, if there is a large number of items connected at power up, there is the possibility of a cutout.

 

 

 

 

 

I can sometimes have a similar problem when a PSX breaker tries to reset on a section that has multiple sound fitted locos on it. Solution is to remove a couple of locos and all is well.

 

It's a bit like starting a four engined airplane - there isn't enough power to start all four at once, they are started progressively.

 

Simple solution is a switch - or switches - as you have found out - to progressively start up sections of the layout.

 

 

Yes I realise that! For various reasons I already had the switches in place, so the problem isn't really a problem - but it would be nice to know the value of the surge current, so that I can work out the maximum number of motors per circuit other than by trial and error.

[Mooresby]

On 07/08/2019 at 16:04, WIMorrison said:

The use of a filter across the track bus and optionally accessory bus has resolved the lost address issue for everyone that has used it - and it has generally improved DCC running.

Is there a particular product that you would recommend to achieve this - i've experienced the lost address issue with my Cobalt IPs and am keen to remedy the problem.

Many thanks

[WIMorrison]

IIRC the values of the resistor was 150 ohms (2 watt?) and the capacitor was 0.1uF that I used, and they definitely weren’t branded - they were the cheapest I could find on eBay  

 

I have 2, one on the accessory bus and one on the track bus

[dj_crisp]

Hmm

Of my two colbalts fitted I now have one with a loud clicking sound on one throw. It had been ok for a year or so.  I guess it's faulty? is there a fix or is it for the bin like everything else I've owned from DCC Concepts?

 

cheers

Will

[DCCconcepts]

1 hour ago, dj_crisp said:

Hmm

Of my two colbalts fitted I now have one with a loud clicking sound on one throw. It had been ok for a year or so.  I guess it's faulty? is there a fix or is it for the bin like everything else I've owned from DCC Concepts?

 

cheers

Will

 

Hi Will,

 

We notice from your Bisley thread - you were using the original analog classic motor - 8 terminals -  powered with a DR4018/DR4101 combination and it was "buzzing". It is possible that there may be an issue with the output from the DR4018 and it may have damaged the Cobalt. (overdriving the gears)

It is noted that you now use Digitrax DS64s instead to operate your analog motors.

 

We are officially closed over this weekend, but still keeping a lookout!

Contact us via our email next week and we'll see what we can do to help - gear issues are usually fixable.

 

Best Regards,

The DCC Concepts Team.

[dj_crisp]

12 minutes ago, DCCconcepts said:

 

Hi Will,

 

We notice from your Bisley thread - you were using the original analog classic motor - 8 terminals -  powered with a DR4018/DR4101 combination and it was "buzzing". It is possible that there may be an issue with the output from the DR4018 and it may have damaged the Cobalt. (overdriving the gears)

It is noted that you now use Digitrax DS64s instead to operate your analog motors.

 

We are officially closed over this weekend, but still keeping a lookout!

Contact us via our email next week and we'll see what we can do to help - gear issues are usually fixable.

 

Best Regards,

The DCC Concepts Team.

 

 

Thanks for replying - you need to have a rest from work!

 

You're right that I've switched to using the DS64, as i got buzzing from both the colbalts and tortoises when i used the DR4018. As i did I also didn't leave the layout on for long until i could resolve. I'll take it apart and see if there is a gear problem

[dj_crisp]

Swapped out my clicking motor... now the replacement is very slow at throwing. I haven't timed it but roughly 30-40 seconds. Amazingly it does make it.

 

Has anyone else had this issue....Does this improve with time and use or one for removing?

 

Thanks

Will

 

[St Enodoc]

1 minute ago, dj_crisp said:

Swapped out my clicking motor... now the replacement is very slow at throwing. I haven't timed it but roughly 30-40 seconds. Amazingly it does make it.

 

Has anyone else had this issue....Does this improve with time and use or one for removing?

 

Thanks

Will

 

Does it take that long with no load as well as under load? If so, it sounds as though there's something not right.

[two tone green]

Is there a little switch on the point motor by the terminals. If yes what is it set for. It may be set for the wrong supply voltage you are using. 

[WIMorrison]

6 hours ago, two tone green said:

Is there a little switch on the point motor by the terminals. If yes what is it set for. It may be set for the wrong supply voltage you are using. 

 

the switch you refer to is for setting the address of the motor, it has nothing to do with voltage

[two tone green]

On the cobalt omega which is dc powered there is a switch with selects the operating voltage range. 
 

I was not sure if he had a DC or DCC version. 

[dj_crisp]

No switch on mine... I think i have some early ones. I'm definetly going to try it with no load tomorrow. If that fails I'll switch it out as I have a few more as I originally intended using them for the whole layout. Now down to just two where their size helps.

 

cheers

Will

[two tone green]

I replaced my original DC ones due to the clicking you mention. They were very voltage sensitive and I gave up with them with some over driving and not stalling. There were badly made with poor QC

 

But as I had all my point decoders I decided to go for the new DC versions so I bought the Omega DC ones with the voltage selector switch. I have a very good stabilised smooth DC power supply so I knew that the voltage on my DC power bus would be always whatever I set it at and set the little switch on the point motors accordingly. 
 

I have 25 of the Omega versions on my O gauge layout and have had them for a good few years now with no problems. I run them though of ESU switch pilots using three to two wire converters and cut the power off at the end of travel so the motor is not stalled as such and relies on the gearbox to prevent travel when not powered. Again never had any problems with this arrangement with the motors getting lots of use. 

[St Enodoc]

8 hours ago, dj_crisp said:

No switch on mine... I think i have some early ones. I'm definetly going to try it with no load tomorrow. If that fails I'll switch it out as I have a few more as I originally intended using them for the whole layout. Now down to just two where their size helps.

 

cheers

Will

If they're the original type you should be able to send them back to DCC Concepts for repair, replacement or refund. The later Omega or iP versions are fine in my experience.

Edited June 27, 2020 by St Enodoc