Pacific Stay alive fitment without major surgery

[Too Tall]

I was having an little issue with the pacific loco's randomly stopping when crawling.

I had a clean track I had checked all the wheel contacts, but it was still happening, I guessed the DCC chip was being very fussy and the tiniest loss of connection was upsetting it and causing a reset. This is what it was doing: 

The obvious fix was to fit a stay alive even though one would not expect too on a loco with pickup on 7 axles ! But despite all the electronics being in the tender, there is little room unless you want to undertake major surgery. I didn't so I looked for a smaller solution, I knew I didn't need lots of run time, just a reserve for a micro interruption in power. I found a board with dimensions that looked like it would fit, so I gave it a whirl.

There is no discernible run time when lifted from track, but it does exactly as I hoped and provides enough to keep it moving.

But ... I didn't have the sound running, so had to check that didn't push the power draw that little bit to much and cause the stopping to return, thankfully all was still good.

Of course, someone had to come and find out what I was up to, causing an outake:

The solution for me was a Zimo SACC16 board, these come with one 220 mf tantalum capacitor fitted, with room to fit up to 4 more. To fit without too much surgery, we can only use 2 !

The videos show it blu-tacked in place (but unplugged on the first one).

This will not work for fitting the tender body back on, I insulated and secured it in place with Kapton tape. I did have to slice out the 2 little tabs at the inside top rear of the tender, apologies I thought I had taken a before pic but I hadn't. Here is where the tabs need removing:

And it just fits on the pcb behind the decoder:

Disclaimer: This will invalidate your decoder warranty and anyone copying this does so at there own risk!

[96RAF]

You are ill advised to use an alien stay alive with a TXS decoder and any subsequent damage will void the decoder warranty.

The reason you see little or no run on is likely because the stay alive is not charging.

[Too Tall]

Yes I know it voids the warranty, but I needed a solution and Hornby don't have one stuck_out_tongue_closed_eyes

I've added a disclaimer laughing

If its not charging, its doing a good job of hiding it ! I Repeatedly tried it plugged and unplugged and every time unplugged it would stall/stutter at some point. with it plugged in I have had it crawl at the slowest speed it will run for 2.5 Meters taking well over 4 minutes without a hiccup.

I didn't expect, or TBH need a lot of run on time. I only have a total of 440mf capacitance, if the decoder only supplies 12v charge instead of the 16v the caps are rated for I suspect its less efficient at storage too.

The Hornby power bank provides over 3 times that.

[96RAF]

If it works it works, maybe under-volted but obviously adequate to task.

Hornny PB is 3x 1F in series so only a third of a farad storage.

[Too Tall]

Now you have confused me 96RAF - (admittedly that doesn't take much) So adding the caps together divides their storage capacity ? is that not the opposite of what is required (i.e. most storage in smallest space) ?

So With 2x220mf this board is effectively 110mf ?

The board also comes with a 1000mf electrolytic cap if you have room for it, but I assumed mixing them isn't wise.

[ntpntpntp]

Capacitors need to be connected in parallel to combine their stored charge. Connecting capacitors in series reduces the total capacitance due to the total plate spacing increasing (google it!)

[Rallymatt]

TT, I would go back to basics (many overlook this) why are you loosing electrical connectivity with 7 axles picking up power? The answer is they are not picking up power. If power supply to rails is proven then it must be wheel sets or pick ups . Re gauge the wheel sets and check the pick ups. Don’t fall into trap of using tech to solve non tech problems 😁

[LTSR_NSE]

@Too Tall - don’t think of capacitors as batteries, think of them as simply (storing charge &) supplying voltage.

Wired in series they supply their rated voltages multiplied (so 3x 3v will supply 9v) but their available capacity is proportionally divided (so 3x 300mf will effectively only store 100mf).

Wired in parallel they will only supply their rated voltage (so 3x 3v will only supply 3v) but their available capacity is multiplied (so 3x 300mf will effectively store 900mf).

*edited with correction

[Daedalus]

Capacitors store charge not voltage.

[Peachy]

TT, I would go back to basics (many overlook this) why are you loosing electrical connectivity with 7 axles picking up power? The answer is they are not picking up power. If power supply to rails is proven then it must be wheel sets or pick ups . Re gauge the wheel sets and check the pick ups. Don’t fall into trap of using tech to solve non tech problems 😁

 

 

Rallymatt, hope you don’t mind me asking, but have you actually run a Hornby TT120 loco with a HM7000 decoder using solely Bluetooth? (ie not using any other DCC system)

Because many, many, many, many, many of us have and it’s a simple truth that the Bluetooth connectivity is an Achilles heel in the system. One that is eradicated the moment a power bank/stay alive is fitted.

[Peachy]

I fail to understand why Bluetooth is considered a problem.

 

 

we all must be imagining it then, silly us

[Too Tall]

To be fair Peachy, that was under DCC control so it may be the decoders are quite sensitive themselves. I did not notice an issue when I had the Gaugemaster standard decoders fitted or when I was testing in DC before the scenery went down.

RM - The pick ups have all been checked and adjusted, I have not gauged the wheels as I am unsure how to adjust them if required. But as the first video shows the first stoppage without the stay alive connected was on a straight piece of track, and the second one only one leading wheel was on a frog. Both times the loco (decoder?) restarted itself with no intervention.

Thank you NTP,LT&SR and Daedalus for the explanations, from my limited knowledge the charging circuit board I am using looks to have the tantalums in parallel which would fit with what you are saying. Am I correct in thinking it unwise to use both electro and tants in the same circuit even if they have the same voltage rating ?

[Too Tall]

I know 96RAF knows more about the decoder than he can share with us, hence why I am a bit concerned about his thoughts the Zimo charging circuit isn't being charged.

I can understand it not being fully utilised if its only being fed 12v and the caps are rated for 16v, but if they weren't being charged it would indicate some form handshake is implemented between the decoder and Hornby power bank.

But I ran with and without the Zimo board plugged in 3 times, on every occasion it stalled without and ran fine with, so I think something is getting through or its a very freaky coincidence.

There is no visible run on, so I imagine not even enough to traverse a dead frog at a crawl if no other pick up was working, so the power interruptions I am having must be very brief, maybe this is best described as a brown out protection, not stay alive blush

[Peachy]

Sorry if I’m sounding a little frustrated chaps, but I’m just getting a little tired of being told I must be the problem and not Bluetooth connectivity with HM7000.

TT you quite clearly are having exactly the same experiences as I & many others have had.

Namely we have perfectly performing track and locos and have a reasonable expectation that they should work at a reasonable level of consistency when being controlled by the HM7000 app.

Absolute fact of the matter is, every single issue I’ve had that couldn’t be resolved by checking & double checking all the basics such as cleaning track and wheels etc have been 100% resolved by fitting a stay alive.

[Rallymatt]

TT, it’s sounding like you have a ‘micro short’ issue. I found this on a few locos (less so in recent ones). Peco track always seems to throw it up as it’s finer scale. It’s caused by the wheel back to backs being under size. This allows too much movement usually in front truck but have had a tender and some wagons do it, this in turn allows contact between rails of both polarity at points causing a short. Sometimes the momentum pushes through and the decoder reboots sometimes it stops the layout, sometimes the loco stops but sound continues then loco starts after a few seconds and sometimes sound restarts but layout needs restart to get loco moving and sound catches up.

Most TT:120 models are nominally set to 10.2mm but can be delivered under. The NMRA Target dimension is 10.5mm. I set my stock to 10.45mm (because I have suitable tool that size 😁) and it’s cured the issue completely

I think the decoders are very sensitive to shorts, for good reason, and my Dynamis DCC system is too.

Note DCC Concepts have released a TT:120 back to back tool but this is undersize and will not cure the problem, it’s 10.2mm. My special tool is a pair of pliers (new not old gnarly things) the jaws are slightly bevelled and widen to 10.45mm, I pop the axle in and gently squeeze, the wheels are pushed out to a consistent width. The jaws an are polished so the wheel backs are not scratched.

I have only ever had to do front truck wheels, tender wheels, wheels on the 10t tankers (all of them) brake vans and my Piko diesel. The drive wheels on Hornby steam locos have never been any issue.

[Too Tall]

@ RM - I do have said DCC Concepts B2B gauge,and the Mallard suffers the same as Blinky, but has not yet got the "brown out" circuit fitted., Ill check the pony truck and tender wheels and adjust if required, and see where that gets me. I am guessing they are both just push fit to adjust ?

@ Peachy, I know exactly where you are coming from, its not unreasonable to expect a loco to run fine out of the box, but many years in slot cars has taught me its often not the case !

[Rallymatt]

TT, the wheels are pushed onto the axle, it’s easy to do, The DCC Concepts B2B gauge isn’t wide enough so don’t rely on that, it’s within the NMRA/NEM tolerance but to cure this issue you need to be closer to the Target setting. A good quality vernier calliper is very helpful.

My feeling is it would be better to illuminate the cause.

I appreciate it’s frustrating and we should expect things to work however, as I have always said there is always a bit of fettling needed. The cause of the problem is not BT or even DCC, it’s loss of power/short circuits and then decoders response, it does it on other non BT decoders. There is another cause, fluctuations in power, if loco is bouncing over poorly laid track that also appears to upset decoders. The Roco turntables for example are warped, that sends the decoders into a real tizzy. Continental TT locos tend to have pivoting wheel sets to cope with poor track and sharp incline transitions. A spec of dirt on a wheel can also cause the short detection to kick in. I know when my Piko BR130 has picked up dirt on a wheel, it triggers a short when going into the reverse loop, as soon as it clears the isolation gaps in the rail. the RLM switches the polarity but the dirty wheel (we are talking a pin head) cannot pass power so the decoder sees a ‘short’ and shuts down. Took me a while to figure out exactly was going wrong!

[96RAF]

@TT

The Hornby PB has an onboard transistor that is supplied in two ways by the decoder. 1 - is to provide a charging potential and 2 - is to switch that charging on.

At present most TXS sound profiles only support charging on the move, with a couple having static charging. The intention is to enable all sound profiles with static charging, albeit with a start up lag to avoid multiple inrush currents from many locos overloading a controller or power supply.

It has been noted during team testing of ABC asymmetric DCC control that the TXS decoders seem to be overly sensitive to track voltage changes resulting in false reversals on a shuttle setup or false stopping on a drive thru’.

[Too Tall]

Thank you for your patience and explanations 96RAF blush

@RM - I bought a dud gauge checker ? ! oh well it wasn't expensive, where might I obtain a correct one ?

As an aside I the tender and bogie wheels were tight on Mallard, even with my apparently undersized gauge. I have adjusted them as best I can with the tools at my disposal, the Vernier reports between 10.32-10.39 over all of them. My driving wheels had 2 sets tight too, I was worried about getting all the conrods unaligned, so I carefully did them in situ with the baseplate removed .. and it has certainly improved it, but not cured it.

I also mopped up the lake of oil sitting in there !

And the Vernier's do report the tool to be 10.21mm.

[Daedalus]

Sorry, I appear to have upset people by posting my decoders work as expected.

[Fishmanoz]

Just to clarify the capacitance of the power bank v the Zimo as configured by TT:

• the power bank has 3 x 1 Farad caps in series, or 330 milliFarads
• TT has configured the Zito with 2 x 220 microFarad tantalums (they are electrolytic by the way) giving 440 microFarads or just under half a milliFarad

So if the power bank can run the loco for 10 seconds, the Zimo can run it for just 15 milliseconds, which is enough to overcome TT’s transient disconnections given his experience.

And could we please accept that individuals are having different experiences here, it doesn’t make anyone right or wrong.

PS. When putting in an original post and you want to include a lot of photos/videos etc, could you please put these in a separate second post so they don’t take up a large proportion of the top of every page in the thread.

[Rallymatt]

TT, I tried uploading a post and pictures of how I set the gauge but the post was rejected by one of the moderators.

[Too Tall]

RM Maybe you can post a link to your magic pliers ? smile

[Rallymatt]

[Rallymatt]

So, photos are ok, 😁 The outer dimensions of the jaws are 10.45mm perfect for setting the gauge on wheel sets. Adjust so the jaws just engage inside wheel backs and gently squeeze, they will force the wheels out evenly and precisely each time. If the wheels are too close, roll them down the top part of jaws and the will open out. Just careful not to close fully or you may damage the plastic insulation that secures the wheel and as that has a flange, it could force the wheels too far apart. These are Draper Redline so quite wildly available.