Frog Juicers compatible with RailPro?

[Alan]

Ah, very much so.

[G8B4Life]

ANE's documentation has always been a bit on the "to be desired" side. They do try though. Unfortunately I can't comment much on the SmartFrog v5 as I don't have one; I only have an old (v1 ??) SmartSwitch set somewhere... I could possibly get v5 fairly easily though as a friend of mine carries ANE.

Pondering Alan's comments, if it's automatic polarity wanted (not switch machine driven) I wonder if it wouldn't be better to leave the frog unpowered until something runs over it. I imagine you could then detect which polarity it needs to be easily without the "allow the mistake to happen then resolve it" as the frog would have some potential once the locomotive bridges the frog and the powered rails. This could operate one of two transistors that apply the correct power to the frog (yes I know this would probably cause a power loop which would cause the frog to remain powered indefinitely but I think you get the general idea). Only the shortest wheelbase loco's  might still an issue but a timer could possibly be thrown in somehow to delay the turning off of the power once there's no more detection.

I don't think this approach would work for DCC though so it wouldn't work for those that want dual capability. Then again it might not be a workable idea at all even for DC in which case I'm just spouting hot air.

- Tim

[Alan]

...leave the frog unpowered until something runs over it...

That's an awesome idea!

One glitch that would need to be worked out - differentiating between a short circuit and loco motor draw. Otherwise, you are effectively lower bound limited to sensing the current drawn by the heaviest locomotive in your fleet. Even a good runner drawing only 100mA is in the sparking range. Little tiny spark but still a spark. The sensing current needs to be down in the low double digit mA range to make sparking inconsequential.

Perhaps the circuit could sense current rise time instead of total current. As an aside, my block circuit breakers do this. They allow temporary current overshoots if the ramp is slow enough. If the ramp is fast then they trip right at the set point. It is not my handywork. This functionality is part of the IC I used. Hmmm, wonder if my breaker circuit could be adapted? Lest I digress. On the good idea, keeping the sensing current low in the frog would require precision measurement. A zero ohm path should have a faster current ramp than a 150 ohm path (100mA loco) but it will be close. I would imagine only occasionally will all the power for the loco be transferred through the one wheel on the frog. If only a percentage of power flow exists then it would be easier to differentiate the ramps. Still may require scary precision to work reliably. Hey! I know of a company with Precision right in their name - Precision Design Co. Maybe they can pull it off. 

Short circuits of any sort cut square against good electrical design. I am always drawn to any idea that eliminates them from the application.   

[Josephbw]

This can cause problems unless you have gaps between the main and the switch, and a gap between the diverging route and the switch.

And that's my problem. I have Peco insulfrogs and they are not gaped after the frog. I'm still mulling over the best approach to adding a gap to them without pulling up track.

Bill our club has been using Peco Insulfrogs for over 30 years. I put a small thin layer of clear nail polish right behind the frogs on the diverging rails of all the switches. That has worked just fine for all these years. We've had 2 different switches that get the most traffic wear through the nail polish in the last 3 years, so I just gave them a quick touch up and after drying we're back in business. Nail polish is pretty tough and will last through many years of wheels rolling over it.

Joe

[JJ Crooke]

Thanks everyone. Lots of good information/comments. But going back to my original question: can I power the SmartFrog (@ 12V DC, 4 amps) using my layout bus (powered by the PWR-56 @ 14.2V DC, 4 amps) without damaging the unit? Or is that something that must be tested to know the answer?

[Alan]

Sort of answered that in first reply. As mentioned earlier, their documentation leaves a lot to be desired. English is obviously not their native tongue.

From the manual:
1. Power input DC12V, AC 16V, DCC 18V
2. Acceptable for DC, AC and DCC power
3. Max current 4A for any scale layout

http://www.anemodel.com/Upload/Download/201509140818159669.pdf

This leads one to believe 14.2v should be fine. It accepts DCC 18v and since it works on DC it must be rectifying the DCC signal yielding 17.3v.

You will find out soon enough, eh?

[JJ Crooke]

Thanks for the answer Alan. I needed a little bit more explanation in order to understand the DC12v vs DC14.2v vs DCC18v thingy. And yes, I chuckle a bit every time I read their manual but at least they're trying.

[JJ Crooke]

... I could possibly get v5 fairly easily though as a friend of mine carries ANE.

Would you mind sharing the info regarding that friend? If I ever decide to go with the SmartFrog, it may be easier for me to deal with a North American-based dealer than the Taiwan-based Ane.

[G8B4Life]

Joel,

More than happy to share, but I'm afraid it's not quite what you think, myself and my friend are way over the duck pond, in Australia.

http://www.roadandrail.net

Looking at ANE's website, there are a few dealers in the US. Here is the link to them. Maybe one of them might  have the SmartFrog, or at least be able to get it.

http://anemodel.com/dealers_list.aspx

Alan,

I'm glad you liked my idea. I was thinking a bit more basic than your description though. I was thinking of simply detecting/using the voltage somehow; when a locomotive first bridges the insulating gap from the powered rail to the frog (same side of the locomotive) the frog will either have ~14v or 0v applied to it from rest of the locomotive pickups on that side. That much we know. I had thought that perhaps it'd be possible to use that voltage to simply turn on one of two transistors (one each for the correct polarity to apply to the frog). There is of course a bit more to it that would need to be worked out, like stopping the frog from being permanently on after first activation that my basic description provides.

Perhaps my way of thinking how to do it isn't workable. It sounded good in my head at least.

- Tim

[Alan]

the frog will either have ~14v or 0v applied to it

Relative to what potential? A completed circuit is required to detect current flow.

Where Power = Vcc or Gnd depending on train direction: Power --> Rail --> Wheel --> Frog --> Sensor --> ?

If "?" connected to Vcc then will be unable to sense when Gnd frog is required. If connected to Gnd then unable to sense when Vcc frog is required.

[Dean]

Hi everyone,

I had the chance to go down in my basement to work on my layout this past weekend for the first time in several months so I'd like to give a quick update on Tam Valley's Frog Juicers compatibility with RailPro. I was able to install an LM-2S decoder in one of my locomotive and the verdict is...

Tam Valley's Frog Juicers are NOT compatible with RailPro's PWR-56. The locomotive can run fine in one of the turnout's alignment but will create a short on the alternate alignment. The Frog Juicers will not change the polarity of the frog as it can't detect the short.

Oh well... I'll have to sell my Frog Juicers and look at the other options listed in this thread. 

You might try an auto-reversing device.

[Alan]

I'm glad you liked my idea. I was thinking a bit more basic than your description though. I was thinking of simply detecting/using the voltage somehow; when a locomotive first bridges the insulating gap from the powered rail to the frog (same side of the locomotive) the frog will either have ~14v or 0v applied to it

I've got it!!! The solution hit me out of the blue last night - a voltage divider network. I'll whip up a schematic later but here is the general idea: A series network consisting of 2 sense resistors sized to develop a 1.7 volt drop at 20mA and 2 limit resistors sized (combined) to limit current to 10mA at 14.2V.

Arranged thusly: V_CC --> R_SENSE --> R_LIMIT --> Frog --> R_LIMIT --> R_SENSE --> Gnd

Connect an optoisolator across each sense resistor. When no wheel is on frog then only 5mA flows across the network. Insufficient R_SENSE V_DROP to meet optoisolator LED turn-on voltage. When a wheel connects frog to either polarity rail then effectively the network is cut in half raising the current flow to 20mA creating a 1.7V drop across the corresponding R_SENSE and firing the respective optoisolator.

Now with a clean optoisolator output signal we can trigger anything we want. I'm thinking a pair of NAND gates controlling a quad switch like a CMOS4066 that disconnects the R_NETWORK from the frog to solve the permanently-on problem. The logic output would initiate the correct charge pump MOSFET combination and start a short duration timer to temporarily hold 1/2 network at 20mA. The logic gates serve to prevent both polarity transistors from being on at the same time.

Wheel bridges rail to frog --> current jumps on one side of resistor network --> optoisolator triggers --> timer starts --> frog to network disconnect --> charge pump fires transistor --> timer expires --> transistor shuts down --> frog to network connect.

Tim, you are a genius!

 

[G8B4Life]

Alan you've put a smile on my face this cold winters night  ; I've been called many things before but never a genius. I do think the title of genius should go to you though for figuring out the way to implement my idea, something I would not be able to do.

I really look forward to seeing the design you come up with.

- Tim

[Alan]

The credit is all yours. I was satisfied to explain why off-the-shelf juicers work but are not ideal. Mostly because I practice another religion where there is no need for them - switch motor control. Your statement caused me to think about the problem, a problem I don't have <snide wink>, from a different angle. It is all about occupancy detection not short detection. Kudos to you.

The dots you helped me connect were "Hey, I've already done this!" My circuit breakers sense current flow and current ramp. Accurately. My occupancy detectors use measured voltage drop to trigger optoisolators. Sensitively. Together they form the foundation of the solution. Instead of shutting down high current circuits or sending logic pulses to a signalling system, we use the same basic sensing but do something different with the output - switch circuit polarity. At least we know the sensing side is probably doable!

Circuit breaker on the right - Sensing current rise via voltage across resistor (big brown resistor) to trigger a charge pump (IC) to turn on a MOSFET (transistor with the heat sink).
Occupancy detector on the left - Using voltage drop via current across a device (3 big diodes) to trigger an optoisolator (far IC) that starts a timer (near IC) to put out a signal (LED).
Many of both devices are functioning on my layout today. Proof the sensing scheme works. Exciting.

[JJ Crooke]

I have finally received the SmartFrog V.5 (AP003) that I had ordered directly from Ane Model. It took 3 weeks to receive (Taiwan to Canada) and I was anxious to see if it would work with the RailPro system. As I explained elsewhere in this thread, I am looking for a replacement for the Tam Valley Frog Juicer as it is not compatible with the PWR-56. The wiring is the same as a Mono Frog Juicer (left rail, right rail and frog) but it must also be powered by a separate set of wires. Since my turnout was already wired for the Mono Frog Juicer, I've added two additional feeder wires (16 gauge) from the bus (14 gauge) to power the SmartFrog.

I am happy to report that the AP003 SmartFrog V.5 works like a charm with the PWR-56! The polarity of the frog changes automatically when a locomotive enters the isolated frog area and there is no jerking or hesitation from the locomotive. Everything is very smooth. You can hear a "click" when the polarity is changed or when you turn the PWR-56 on or off. Each SmarFrog is $14.95 USD (ordered directly from Ane Model's website) and shipping is free on orders over $100 USD.