Kato double crossover question

[inobu]

kvp,

 

Have you actually set up a BCD circuit? 

 

I'm finding the same results as emkay.

 

Test Results

 

Switch Unit :  Kato #4 Turnout 20-221

 

Power Source: Kato Power Pack 22-014

 

Tested 4 in parallel

 

The voltage threshold on 4 turnouts are

 

10.80 v .................100%

10.70 v...................75%

10.60v....................50%

10.44v....................3 of 4

10.20v....................fail

 

The turnouts starts to fail @ 10.70v.

 

 

Switch Unit :  Kato  Crossover Turnout 20-210

 

Power Source: Kato Power Pack 22-014

 

Tested single turnout

 

The voltage threshold on #4 turnout

 

10.90 v .................100%

10.70 v...................100% chattering

10.68v....................75%

10.60v....................3 of 4

10.50v....................fail

 

The crossover operation starts to fall off @ 10.70v with complete failure @ 10.50v

 

 

Note: Because the failure threshold can occur at +- .05 volts the failure threshold can be effected by the gauge and length of the wire. 

 

 

 

 

Switch Unit :  Kato #4 Turnout 20-221

 

Power Source: BCD 1000uf 16V with max 15.0v test limit

 

Tested 1

 

The voltage threshold on single #4 turnout.

 

11.40v ...................100%

11.16 v...................100% chattering

10.97v....................50%

10.60v....................fail

 

The #4 operations starts to fall off @ 11.0v with complete failure @ 10.60v

 

 

 

Switch Unit :  Kato #4 Turnout 20-221

 

Power Source: BCD 1000uf 16V with max 15.0v test limit

 

Tested 2 #4 turnouts in parallel

 

The voltage threshold on 2 #4 turnouts

 

14.6v .....................100%

14.00 v...................33%

13.90v....................fail

 

 

 

Switch Unit :  Kato Crossover  20-210

 

Power Source: BCD 1000uf 16V with max 15.0v test limit

 

Tested 1 crossover

 

The voltage threshold on crossover

 

15.00v .....................fail

 

The 1000uf BCD could not drive the crossover the test cases limits were reached due to components and test configuration.

Power supply limit of 15v and capacitor limit of 16v.

 

 

 

Findings

 

A single 1000uf BCD can drive 2 #4 turnouts when the voltage level is increased to 15v. There are limitations when it comes to the crossover. 

The crossover consists of 4 switched that requires a higher current draw which is contributing to the failure. Increasing the applied voltage can

remedy the issue but may not be practical for all.

 

The limitation is in the engineering of the circuit but the design of the circuit is functional. This hold true based on the fact that the test was conducted with

the Kato 22-014 power pack supplying power to tactile switch and the BCD circuit.

 

The tactile switch and power supply could throw all switches utilizing 10.5v where as the BCD required 14.6v to throw 2. Based on emKay testing the additional

2200uf did not make a difference.

 

It is my opinion that the crossover tests the BCD's in the realm of current draw.

 

Inobu

[kvp]

 

Have you actually set up a BCD circuit? 

I've been using #4 turnouts with 11.3V (12V-0.7 polarity protection diode) and throwing two with the same switch without problems. (this was the setup on our show layout) Cable type: standard ribbon computer cable (pin doubled). The trick was that i put the cap to the turnout and not the control panel, minimizing capacity problems.

 

The Kato turnouts work well with 12V if you provide the 12V and can ensure that it stays 12V during the whole throw. This means a peak load of 0.8 per coil, so the 4 coils means a peak load of 3.2A is required. Anything less and the voltage starts to drop below 12V. I've been using a 5A 12V DC switching power supply with on board buffer capacitors. (a bigger off the shelf wall wart actually) I tried switching back and forth continously very fast and it worked flawlessly.

 

The resistance of the turnout coils determine the current draw, while the pulse time determines the capacitance. The 4 coils in the crossover need 4 times the current and this means also 4 times the basic capacitance (1000 uF x 4) for the right pulse time. Less current and the voltage drops, less capacitance and it won't have time to throw.

 

ps: any hacks that use higher voltage will work but damages the coils, making them fail over time

 

Table:

signle turnout: 1000 uF @ 12V, 0.8A

double turnout: 2200 uF @ 12V, 1.6A

quad turnout: 4700 uF @ 12V, 3.2A

(capacitors should be rated at least 50% above the operating voltage, so for 12V at least 18V, but 25V is better)

[inobu]

kvp,

 

Why didn't you post this info before? Can you post a picture of your setup so we can see what you are doing right verses what we are doing wrong.

 

I guess I'm guilty of being a hack.

 

Inobu

 

waiting

Edited April 4, 2016 by inobu

[kvp]

I think i did:

Posted 02 April 2016 - 05:56 PM
Around 1000 uF is needed for a single coil. A single crossover has 2 and needs a 2200 uF cap. A double crossover has 4 coils and i would suggest either 4 x 1000 uF or 2 x 2200 uF. Parallel connection (just like with batteries) increases the capacity, while a series connection the maximal voltage.

I think you need around 0.8A per coil. The capacitor method needs a bit more power than the plain DC switch that Kato uses but throwing that too quickly also gets the same result.

 

Picture? I don't think i made any (it didn't seem important), except the exhibition photos, but those only show the control panel with a single ribbon cable and 4 dpdt push buttons. The BCD circuit is under the turnouts on a raster board along with the auxiliary circuits and the ribbon cable station control bus connectors. I'll make a photo when i go to the club next time. (i left everything there after the last exhibiton) The control panel just selects between power and ground on all channels and then they are routed to either the turnouts or the station/platform lights, by breakout screw terminals on each module.

[inobu]

So you don't have anything to show us how to do it?

 

Thanks!

 

Inobu

[cteno4]

Ray always had his BCD instrucriins their caps at the control panel. When I fiddled with the BCD to try it out I had it hooked to one of our old point wires that was 20g speaker wire probably 12' long and snapped #4 and 6 fine over that distance. I'm pretty sure I used 25v caps and I also did it on the breadboard and with some alligator clips, so not the best over all high amp connections! With the above setup it snapped the single points much more reliably than the kato controller. Nice pop and not bounce back.

 

The amp values that kvp referenced I pretty sure are what what ray references in the BCD instructions, I'll check. He had some explaination on source power as well.

 

the reason I was going to move the club to the BCD was that we were getting som bounce back and longer rattle action on the kato controllers with time. Granted our kato controllers got a lot of abuse with the transport (they were rarely packed well I began to notice), but then we tired of wiring all the points at shows (we had a plug system but required someone under the layout for 30 min and not fun) and with the iffy point firing we found it best to have a spotter anyway making sure the point was thrown properly so they could just do it manually.

 

I never did the double crossover though. I'll rig it back up this week and see my results. a single 1000uf of course won't budge it, but I can't remeber the higher value that ray had for the double crossovers. He was always measuring the cap needed to the resistance of the coil, so if higher gauge wire is used between the coil and cap then that resistance would need to be added in there and a larger uf cap used. This should be safer than boosting the voltage.

 

Might just try adding on another 1000uf cap in parallel with the existing ones to see if it can get it to snap. Might just need a bit more of a pulse either for the coil, mech and/or wire loss. Also try a few different caps from your bag, I've been burned in the past with a bad component causing the trouble thinking is was dumb me screwing up the circuit.

 

Jeff

Edited April 4, 2016 by cteno4

[inobu]

As it stated before the circuit is functional but it is not how Ray configured it.

 

It is the applied voltage that drives the circuit. As em_kay stated he needs more umphhh! Which is voltage or current.

The only way to achieve that is to change or increase the input power.

 

The power supply has 1A the question is does the cap maintain the 1A in the cap.

 

Inobu

Edited April 4, 2016 by inobu

[kvp]

Short answer: no. Long answer: caps have some resistance too so they burn some power too. This means a bcd circuit needs a bit more power but prevents the burning of the coils by limiting the pulse time. I would say 3A @ 12V with a 4700uF cap (or equivalent) should do it. Adding the double of this (parallel) as a buffer before the switch could help solve the weak source problem. (1A charge for the buffers then 3A inrush for throwing)

[cteno4]

I went thru all the stuff i had from ray over the years and he mentions that the double crossover ends up with about 5 ohm resistance, a bit lower than what 4 #4 would be in parallel so a 4700uf cap is recommended with a 12v source.

 

he also recommends a 470uf cap across the power supply to cancel any ac ripple from the supply.

 

Ray says he uses a 0.25A 12v transformer on his layout (dont know if he has any double crossovers), i believe that upping the amps just speeds the cap charge. someone once claimed a larger amp power supply would be needed, especially with double crossovers, and Ray claimed it would not. nor does he mention in his circuits with multiple simultaneous throws happening on restart reset circuits or yard ladder controls a larger power supply is needed. Im trying to get in touch with him to clear this up.

 

 

jeff

[inobu]

em_kay

 

You can use Kato's power pack to drive the BCD. The output @ 14v can drive the crossover. You can run down to 13.5 but it does not hit every time. The quick snap occurs above 14v or so. The other point is the size of the 4700 uF it's kinda big. 

 

Here is the comparison of the capacitors as you can see the analogy of the cup of water is reflected here. With the cup it is the volume of water it holds verses the size of the plates inside of the capacitor. The issue with the capacitor is you can make up for the current draw here by increasing the input voltage.

 

You are safe @ 14v as the output voltage on Kato's power pack (the side jack) is 15 volts. So the little blue switch boxes are running or throwing the switches at 15v.  The BCD will be doing the same.

 

 

Tested and verified

 

 

Inobu

Edited April 5, 2016 by inobu

[emkay_777]

Thank you, gentlemen, for all your time and effort.  Guess I'll order a 4700uF cap and integrate it.  Right now I'll use the Kato switch machine til it comes in.

[katoftw]

Any chance you can rewire it with 2 switches to control it like the Tomix crossover?  Only activating 2 points per switch?

[cteno4]

emkay,

 

you could try wiring 4 or 5 1000uf in parallel as well as a test. im going to set a set of tests up this week to see what i can come up with.

 

sad thing is no response from ray, i tried a list i know he frequented and no luck their either. 

 

jeff

[kvp]

Jeff is right, 4 or 5 1000uF caps in parallel is similar to a single 4700uF. Raising the voltage to the Kato controller's output voltage should be also safe.

 

ps: Rewiring for double or even single coil operation is also possible but requires disassembly. If anyone goes this way it's even possible to add a single dedicated 1000uF cap to every coil and have a 5 wire cable come out of the turnout, that goes down to 2 wires after the 4 cap underfloor bcd circuit.

[emkay_777]

Oh, that's good!  The 5 X 1000uF caps I got.  That just cut 10 weeks off my project :)  Just wondering if Kato makes an 'adapter' that attaches where the switch lever sits, but only has the + and - leads.  I don't believe that they do, so I might have to sacrifice a switch lever, but that's okay. 

[cteno4]

Did it work?

 

Kato does sell an adapter box. Not sure the plug that's on it though

 

http://www.1999.co.jp/eng/10003117

 

Could see if 9v battery clip works and hack two female sides together.

 

Jeff

[emkay_777]

Jeff,

 

That adapter certainly looks like the answer, and it'll only take 3 weeks from Japan . . . I tried the 9V battery clip, but they are a little too small - good idea though.  I'll modify a Kato switch and try the 5 caps tonight and let you guys know.  Thanks,

 

Evan

[cteno4]

Evan,

 

Now that you mention it I kind of remeber trying that a long time back!

 

The part might be somewhere in Canada or the us as well, modeltrainstuff.commhas it, but shipping up north may be way more than Sal from Japan!

 

Jeff

[emkay_777]

Shouldn't be a problem modifying the Kato switch lever.  I've had it apart before when I pushed too hard on the plug and something broke inside.  I can replace it pretty cheap at a flea market.

[emkay_777]

Well, it works!!!  5 X 1000uF in parallel, using the Kato accessory power supply will robustly throw all 4 points on the Kato crossover - done; finished; Yay!  Thanks again guys!

 

Evan

[kvp]

I'm glad we could help!

 

In the meantime i made a photo of the bcd under my t-trak station:

(blue-white: traction power, red-black: turnout, ribbon sockets: control bus, screw terminal sockets: traction power wires, jumpers: dc/dcc mode selectors)

Edited April 7, 2016 by kvp