Frog Juicers - Light Loco Problem

[gppsoftware]

Geoff,

One of the giveaways earlier in this thread is that you mention a clicking relay being heard. The idea of introducing a physical make/break interruption to a digital circuit with all the interference that such an action would cause horrifies me! No wonder you've got a problem! I am really surprised that Gaugemaster are even marketing such kit (actually, I'm not really, because they still live in the DC era and try to apply DC solutions to DCC problems) because other 'juicers' I have come across have been electronic devices, not a relay.

And herein lies a further problem which is coming out in this thread: once a device is inserted into a layout which has the potential to destroy digital signals, certainly interfere with them and cause momentary phase issues which confuses decoders, people start observing side-effect problems and then start calling for 'terminators'/'snubbers'/'quenchers' to mask the problems caused by the 'juicer'. 'We don't know what we're doing but we do it any way!'. Please throw out the juicer - it is a waste of money and is causing you more problems than it will solve. 'Quenchers' are only necessary when we do things that shouldn't be done.

Going back to the original issue of hand operated turnouts not having a switch. One solution I observed used by the Burgess Hill Model Railway club many years ago was to use a wire-in-tube approach where the 'lever' was actually a switch. This link (https://www.modelrailforum.com/forums/i ... topic=6545) isn't a BHMRC layout, but it is someone else who has done the same thing - scroll down to the third post. The switch is wired to the crossing (frog) and the turnout is controlled mechanically via tube. BHMRC dotted these things around their layouts where they needed them, close to the point being controlled, rather than centralising them on a panel, but either method is fine.
This approach also has the advantage that you don't touch the turnout - frequent hand-changing by moving blades eventually bends them and causes other mechanical problems.

So my suggestion would be to throw out the 'juicer' and use some kind of variation of the mechanical switch. It is guaranteed to work, it guarantees that you digital signals will be kept clean, it doesn't rely on a short (bad practice) to work and it doesn't create a situation where 'bandaids' (quenchers) are needed to mask further problems and probably introduce others. Seems like a no-brainer to me!

I believe that Bufferstop's (very good) explanation says it all: the 'juicer' is really stuffing things up for you - please don't use it!!!
Flashbang's 'micros switch' approach is a variation on what I have just written - we are in agreement.

Graham P

[Dad-1]

I return to this as the layout is coming towards completion.

As I've been working on it I was not overly concerned with occasional stalls. Obviously the working created
some dust, being placed just inside my up and over garage door I also had dust particles and even fluffy
seeds blowing in.

https://www.newrailwaymodellers.co.uk/F ... 60#p683587

However.

Now nearing completion I was unhappy with the level of stalling, 5 different locomotives being used, one in
particular is well known for it's excellent slow running reliability on a fixed garage layout. Having made room to
access my other shunting layout I used the various locos without much problem. Time for a careful double-check
of everything, even replacing the Gaugemaster DCC system with a Hornby Select. The Select proved to be the best
due to a specific Gaugemaster problem and although more reliable it helped pin-point exactly what was happening.
Sadly we again return to the Frog Juicers not being suitable. The small light locomotives running at minimal speed
were stalling at the continuity break in the rails and not always switching the frog. It would also seem that the
solenoids were not creating a reliable and locked change causing all sorts of spikes within the DCC signal.

I will reluctantly have to remove the frog juicers and arrange a pair of fully mechanical switches to achieve the
level of reliable slow running required and achieved on my original HS-2 shunting layout.

https://www.newrailwaymodellers.co.uk/F ... 45#p631958


Geoff T.

[gppsoftware]

Geoff,

Interesting that Select solved a problem that the Gaugemaster (MRC) system could not.
Would you be able to elaborate on what the problem was ?
It seems strange that a change of DCC system should resolve the stalling issues. For the benefit of readers here, how did Select help pin-point exactly what was happening ?

Graham Plowman

[Dad-1]

Hi Graham,

Been doing other things for a day or two !
I didn't say that the Select solved the problem, but it helped me work out what was happening.
The Green Frog Brewery layout was built around the clubs rather old Gaugemester system, even to
the degree that it could Velcro onto two supports and sit on the layout.
However it's got a problem. If you've ever used one you'll know that the throttle control knob works
through a 'No Stop' system, by that I mean that you can continue turning the control knob even when
already at zero power. This slipping clutch effect is worn and results in very poor accuracy in either
adding, or removing power. It is so unpredictable that the best way to get accurate shunting where
you have multiple changes of direction is to set a given speed and then drive on the change direction
button. This is not ideal for identifying unexpected problems.

A Select has a controlled and fixed level of throttle, any movement up, or down is consistent. Because
of this I could vary my running speed, faster over the problem areas, but slow down to the required
coupling & un-coupling crawl rapidly and predictably. A thing to note is that I have no inertia +, or -
set in my decoders as I want instant throttle response. By increasing speed over some power zones of
track I could evaluate the Frog Juicers ability to change my two point's frogs power supply.
Then when running at very low speed the locos stalled at the interface joint. By applying a file to the
other interface join I could hear the solenoid click and the stalled loco began to move. Evidence that
the slow light loco wasn't able to create sufficient level of short to work the Frog Juicer. It also seemed
to be able to 'drop out' after having switched, leading to additional stalls away from the zone interfaces.
Back to speeding up and everything ran well again. Due to some length trains actually having to change
direction on an interface joint it was NEVER going be reliable.

I do have my own identical but Right Hand version of the layout with mechanical frog switching with
the reliability I demand, so I know it can (and will) be achieved.

That proved beyond any doubt that in this application Frog Juicers simply don't work reliably enough.

I have used many DCC systems, Lenz, Dynamis, Gaugemaster, Digitrax Radio Controlled, Digitrax Zephyr
and Selects. The Select is a very basic system, all too often criticized because of it's simplicity, but it
doesn't pretend to rival quality systems Yet I built one into my first shunting layout and still think it
works well at a price that I could afford to actually permanently build in. The only problems are Kids
pressing random buttons, but then what is Child Proof ??.

Geoff T.

[Dad-1]

Hi Graham,

One little thing I didn't include in that last message was that the 5 locomotives used
during testing had 3 different manufacturers decoders, obviously another variable.

Hornby 8 pin (Bachmann 57XX)
Hornby 4 pin (Hornby Sentinel)
Lenz direct 6 pin (Hornby Ruston 48HP)
Zimo direct 6 pin (Heljan 1361 class)
Zimo direct 6 pin (Hornby 0-6-0 Peckett)

When testing and checking it's good to have options, the variations can help pin-point
some specific locomotive, or maker focused irregularities.

For instance the Heljan 1361 has been a brute regarding any less than perfect track.
The specific reason there is too tight a manufacturing tolerance in the axle bearings.
When new it was difficult to move the wheels through any lateral or vertical movement.
This caused stalling as well as spikes in the decoders power feed. In addition to stalling
it upset the Lenz decoder originally fitted and it did some funny things. A Zimo decoder
was much better, also had much more running now. The Lenz decoder was put in the
Ruston and in that it behaves perfectly !!
IF you have the stuff substitution helps understand what is going on.

Geoff T.

[gppsoftware]

Thanks Geoff, yes, having parts to substitute is helpful.

I run my layout with a radio-control NCC system. Previously had a Lenz SET 100 which is now on the test bench layout: http://www.mrol.com.au/Pages/Vu/WorkBenchLayout I also have a Hornby Elite system which is the next step up from Select.

I am of the belief that locos all need some kind of up and down movement on some axles in order to maintain continuity. I don't believe that sideways 'slop' in axles is necessary. If fact it has been the bain of many a model over the years.
It sounds to me like you 1366 lacks up and down movement and that will cause problems on anything but perfect track.
I'm not usually a fan of 'stay alive' as I hail from the camp who believe that fixing problems rather than masking them is a better way to go, however, a chassis which is fixed really leaves you no option - I only have one myself fortunately (http://www.mrol.com.au/Pages/Vu/Airfix61xxModifications) - but I found a solution to stalling with this loco was a Zimo with stay alive. Runs a treat now.

All the best,

Graham

[Dad-1]

Hi Graham,

I have contemplated stay alive, but the locos I have that would in theory benefit
are all very small. Where some love top link traction, and don't get me wrong, I
love a good 2-8-0 freight, my main focus tends to be 0-6-0, or less. Space is then
at a premium, so is overall weight. Grind away for capacitors to improve stalling,
but loose the traction benefit of weight.

I can usually get what I'm happy with - You can't have everything in life though.

Geoff T.

[BromsMods]

Have you considered using decoders with built-in stay alive functionality such as the TCS KAM4?

[Ironduke]

Stay alive would not have helped a frog juicer to trip. Also you have momentum disabled so you won't be getting the issue of engines having to build up speed again when they hit a dead spot, which is where stay alive really comes into its own.

[Dad-1]

Hi Ironduke,

Yes I knew that, but it's easy for readers to quickly scan a thread, but
miss an important element. I just have a situation that doesn't suit
Frog Juicers.
In fact I have very few basic stalls, probably because I frequently wipe
the rail head with a clean linen hanky to remove dust which is the main
enemy. My wife relegates hankies to me and cuts one corner off. It wouldn't
be the first time she does a wash and scolds me 'Have you been using this'
meaning for my nose !! Yes it happens, but beware of leaving a dark metallic
stain on a regular hanky.

Geoff T.

[Ironduke]

My wife relegates hankies to me and cuts one corner off.

I have recently stumbled upon using a small square of denim cut from an old pair of jeans. It has a very thick, tight weave and resists getting caught on point blades and such. Also, it doesn't wear through quite as fast as anything else I've tried. In fact it hasn't worn through at all yet. Fingers still smell like alcohol but they aren't black

[Dad-1]

Rob,

Alcohol smell ? Isn't that from the drinking habit ?

Geoff T.

[Ironduke]

Rail cleaning gives me plausible deniability.

[Dad-1]

Green Frog Brewery now working as it should.
Constant and unpredictable stalls, both on the frog to track interface
and on short sections of straight track gone - As are the frog juicers !!

Polarity switching now done by microswitches operated by the point
change rodding.

I know they can work, a colleague swears by them, but I ask why take
the risk when microswitches are £2 for 5 ?????

See comments from https://www.newrailwaymodellers.co.uk/F ... 17#p685380

Geoff T.

[Dad-1]

Why, Oh Why did I even try again !!
Ahh, sorry to bring this one back to life, but my Gaugemaster frog juicers are driving me mad (Again)

Latest story - Because I run small locomotives, slowly to very slowly & I was struggling with the modellers
space problem, I fancied making my own Set-track points. In fact on another layout I have 2 set-track crossovers
that limit it's use to a couple of 0-6-0's. Making up copper-clad code 75 track crossovers is quite easy, but initially
you want to do some proof testing.

So what's the easiest way of switching frog polarities on a loose piece of track ? Ahh I have quite a lot of redundant,
& even unused frog juicers. Simply wire into the suitable copper-clad sleepers, when fresh they did work, enough to
prove the actual geometry of the points worked fine, Remember we are bending switch rails and on set-track points
these are somewhat shorter.
Now many months later the darned juicers are simply not switching power in a way that my small 0-4-0's like.
I should have remembered ! Now to adequately test display both my 2 crossovers, plus a couple of scissors I will have to
mount on some board with tie-bar activated mechanical micro switches. The trouble is I now need to make a board, which
could easily become a layout which was not what I intended !!! Short cuts frequently don't work !!

Geoff T