Posted By gdancer on 09/01/2008 11:14 AM
snip...
, it appears that Bachmann has now solved most if not all the serious design issues.
1) Am I correct on this? Unfortunately, I have several curves on my layout where I had to use 8' diameter track.
2) Will the K-27 easily take 8' diameter without larger diameter transition track? Finally, I had to quit using a Bachman Annie on my Aristo stainless 4 years ago as I am convinced that some sort of electrolysis occurred between the cast drivers on the Annie and the stainless track: required serious cleanup every 10 minutes of operation to the point of serious frustration. (I had no problem with the turned steel drivers of my LGB locos.) Of course, that was just on DC, not DCC which I am presently using.
3) Does the K-27 have power pickup thorough the drivers or through a 'shoe' a la LGB?
4) Are the drivers of the K-27 cast rather than turned?
5) Has anyone with stainless track experienced unusual black deposit on stainless while using the K-27?
6) Any recommendations of DCC decoders for ease of installation on the K-27? Thanks for your input, guys!
1) No
2) No, (define easily)
3) No shoe, sorta through the drivers.
4) I don't know what process was used, but, they are definitely not nickle plated or polished like the LGB wheels
5) I don't have stainless track, but I did notice the black "deposit" on both nickle silver and brass rail. I would suppose it will on stainless, or aluminum also. (See "Arcing Modification" below)
6) My final configuration ended up being a SFX064D and a K1 in the tender and a DG583S, TF4 and K1 in the loco.
First, some videos of the Kay in action on the ALLY...
This was it's first run.
Actually it took a dozen attempts to get that shot. The Kay kept locking up and would not start when commanded. I wanted the camera dolly loco to be moving when the Kay started. (See "Bind Fix" below)
Here is a couple showing the different whistles I loaded into the SFX sound card.
Here are some of the improvements I came up with:
In no particular order:
Tender Wire Improvement
The wiring diagram that Bachmann published shows wires from each truck to the plug board.
They also show wires between the two trucks. Those wires DO NOT EXIST! Had they been there, they would shunt high currents around the lands on the plug board should the front and rear trucks happen to be on reversed polarity track. That could happen if backing through a reverse loop, or if the rear truck derails when backing through a turnout.
Unfortunately, the loco wiring does not match the drawing (in this and a few other respects)
I redrew the drawing, corrected the errors and colorized the wires to make the drawing easier to follow.
(Full size version can be downloaded
HERE )
Notice that there are no wires connecting the front and rear tender trucks. If there is a derailment and the rear truck polarity is opposite the front truck, the traces that run the length of the circuit board will have to carry all the current.
If you run the Kay on a power source greater than 5 amps, and uses slow blow fuses, thermal cutouts or DCC delays longer than 500 ms, I recommend you make the following improvement.
Unsolder the red and black wires from the rear truck where they enter the board. Extend them and solder them to the same solder pads that the front truck uses.
Note: do not move the front wires to the rear to accomplish the same thing. If the entire tender is at the opposite polarity from the loco, then the current will still go through the lands of the board to the loco.
Tender Wire improvement 2
On my Kay, after a few days of testing, the loco began applying intermittent shorts to the rails on curves. I noticed sparks coming from the area of the tender trucks pivots.
It turns out the wires to the trucks go through small holes in a METAL part of the trucks. The holes are too small and have sharp edges. They wore through the track pickup wires.
Replace the wires from the tender trucks with good quality rubber insulated test lead wire. Drill the holes in the trucks out larger and file edges smooth. Install Teflon selves where the wires pass through the trucks.
Arcing Modification
On both DC and DCC the wheels on my Kay would pit and crud up really fast. In less than an hour of running, I had to clean the drivers, trailing truck, and to a greater extent the tender wheels.
I was also noticing burn marks on the tops of both my brass and Nickle silver rails. Some might mistake these burn marks for a crud build up. In reality they seemed to me to be caused by arcing. By watching the loco at night I discovered severe sparking on only one side of the loco when running on DC. On DCC the arcing was less, but equal on both sides of the loco. By running the loco in one direction for an hour on on a freshly polished test track, I discovered the black "deposits" were only showing up on one rail, the one where I observed the sparking. I therefor theorize that these black "deposits" are related to the arcing and may be carbon deposited from the pits in the wheels.
To find out why this loco arcs so bad, I traced the circuit boards and drew a schematic of the majority of the boards. This was weeks before Bachmann published a schematic. When the Bachmann did publish the schematic, I compared mine to theirs. This lead me to complete mine which corrects some errors on theirs, (not the least of which is that they had the + and - outputs of the bridge rectifier reversed.
As far as I can tell, this is a correct schematic. Thanks to Tony and George for their assistance.
( FULL SIZE VERSION
HERE )
Notice that there are two 470 micro-farad capacitors on the output of the bridge, These act as a dead short until they are charged up and will therefor absorb the maximum instantaneous current your power source can provide. ( This can be considerably higher than the rated sustained current or the value of the fuses. )
I could not eliminate these capacitors because it would result in immediate failure of several regulator chips. Worse still, the regulator chips themselves draw their max current until the output voltage reaches the desired output voltage.
To determine if the arcing was due to the motor itself, or the capacitors and regulators on the boards in the loco and tender, I temporarily unplugged the connector on PCB 8 and connected the loco track pickups directly to the motor. This bypassed all the circuitry.
Running the loco on DC with all the boards bypassed, the arcing was almost gone. After three hours, both rails appeared equally clean with no black carbon deposits or sign of pitting on one rail. If there was any carbon or pitting going on, the loco was wearing it back off as fast as it was depositing the carbon.
The best fix would be to rewire the filter circuit to form a twin "T" filter. However, I did not go that route because of other issues. That leaves only four other solutions, remove the boards with the capacitors and all the boards that depend on them and wire the loco for DC. Or, do a normal wired DCC decoder install. Or convert to battery/RC. Or clean the wheels and rails every hour or so.
Tender Shell Repair.
There have been many reports that screw towers inside the tender break. Three of the four in my tender were broken. In addition, part of the tender shell was broke out at the rear.
Apparently the factory tightened down the tender shell but did not check to see if the cutout lever was in the correct position. It was captured between the tender shell and tender floor.
This caused a big chunk of to be broken out of the shell. Fortunately, the broken part ended up inside the tender. A couple other broken parts were in the plastic bag.
The breaks were fairly clean with no deformation of the plastic. I was able to fit the part tightly back in place, then apply a drop of CA on the inside to hold it there. I then reinforced the part with Epoxy on the inside of the tender.
Because My Kay is an undecorated one, it was easy to spray primer, sand and then repaint the entire shell. The repair is invisible thanks to the hardness of the plastic Bachmann used. Had the plastic been soft, it would have deformed and the repair would have been a lot harder to hide.
The screw towers can not be glued. They are slightly shorter than needed, so when you tighten the screws they will pop again. If your screw towers are not broken, be careful not to over-tighten the screws.
Several fixes have been published. However, I opted to cut the towers in half and then shorten them by taking 1/4 inch out of the center, I then rejoined the two halves with a short piece of surgical tubing. I epoxied the lower halves. This provided a spring loaded hold down of the tender shell when the screws are tightened, which stretches the tubing.
Smoke Improvement
The smoke element runs at 12 volts. I have not had mine burn out yet. However, the 12 volt regulator did quit after an hour or so. I by-passed it with a 12 volt regulator and heat sink from Radio Shack. I drove the regulator directly from the yellow lead of the decoder which I remapped for the smoke unit. I also set a qualifier so the smoke would be on when moving to conserve fluid when I parked the train.
When the smoke is on, no smoke will come out of the stack. When I looked down into the stack , I could see that the element was smoking, but all the smoke was being sucked down inside the loco.
To get the smoke to come out the stack, remove the four screws holding the fan in place. Discard the two longer screws. Flip the fan over and re-install the two shorter screws.
With the fan blowing upward, this is what it looks like.
Improving lighting effects.
Despite assurances that the fancy current regulator would not interfere with FX effects like firebox flicker, they did. There is no way to get the two phase flicker effect included with Digitrax and SoundTraxx decoders to work through the Bachmann circuits. I bypassed them by wiring them through 1 meg 1/8 watt resistors to the decoders.
If you want the headlight to brighten and dim properly in sync with the generator sound on a Tsunami sound card. you will have to bypass the regulator circuit. I replace the Bachmann LED with a 5 volt bulb. The regulator will not provide enough current for a bulb, so I used a 270 Ohm 1/4 watt resistor and drove the bulb from a K1 function decoder mounted on the back side of the boiler front..
I did not like the orange LED in the front markers. I used three bulbs per marker, red, yellow and green, and wired the front markers to a Digitrax K1 decoder so I could operate the Kay in sections or as specials.
I did the same thing for the tender, as seen here:
Improving Sluggish Operation.
I have no issues with the high top end speed. I normally try to find out what the prototype typical speed is, and set the decoder vMax to that speed. This provides a full 126 speed steps over the entire speed range for more accurate control.
However, a low gear ratio, (one that makes the loco run too fast) also sacrifices pulling power and smooth low speed performance.
As a typical train for the Kay, I decided that I liked the looks of one 1:20 Bachmann box car, 12 LGB logging disconnect pairs and a caboose.
This would be typical of the length and weight of a train I would pull with the Kay on the ALLY. It is not so long that it can pass it's tail, and there are places where the entire train can bee seen. Yes, Ideal train length is purely subjective and largely a matter of personal preference.
Before setting up back EMF, I always run the train with BEMF off. It is always best to get the loco running smooth prior to setting up BEMF.
I noticed that this train lugged down the Kay in the curves. I pulled the same train with a LGB Forney Also with BEMF off) for comparison and there was no drastic slowing in the curves. (I have no grades)
After looking at several possibilities I finally discovered that with the decoder removed, the voltage at the motor under moderate load was 2-1/2 to 3 volts lower than the voltage on the tracks. So somewhere in the loco wiring the Kay was dropping a lot of voltage as the amps went up.
OK, Let's trace the current as it flows from the left front driver through the motor and out the front right driver.
right front driver
pickup brush
front brush board
red jumper wire
through printed circuit board (PCB) 07
red jumper wire
through PCB 05
8 pin connector
ribbon cable
through main loco circuit board
red wire to tender
7 pin connector
through PCB 09
red jumper wire
through PCB 01
the DCC socket Track R
PCB04 or decoder
the DCC socket Motor +
through PCB 01
orange jumper
through PCB O9
7 pin plug
orange tender jumper
through main loco circuit board
ribbon cable
8 pin cable
through PCB 05
red wire to motor
black wire from motor
through PCB 05
8 pin connector
ribbon cable
main loco circuit board
gray jumper to tender
6 pin connector
through PCB10
gray jumper wire
PCB 01
DCC socket Motor -
through PCB 04 or decoder
DCC socket
through PCB 01
black jumper
through PCB 10
6 pin plug
black tender cable
through main loco circuit board
ribbon cable
8 pin plug
through PCB05
black jumper wire
through PCB 07
black jumper wire
front pickup board
brush
front left driver
Try to count how many solder joints.
Try to count how many mechanical contacts ( don't forget the four on the direction switch, which I didn't list).
Just how many time does it have to go through a circuit board?
I estimated over 9 feet of thin stiff brittle wire.
OK, so I figured I'm losing three volts somewhere, but where?
The problem is, the more resistance, the less able the motor will be able to maintain speed. The problem is worse because the motor has to run too slow due to the gear ratio. I could either change the ratio, or reduce the resistance.
As an experiment, I disconnected the ribbon cable from PCB 05. Then I jumped the red and black from the loco pickup to the red and black motor leads. Now the wiring is reduced to 6 inches and all plugs are eliminated. I ran the Kay again. Pulling the same box car, 12 disconnects and caboose, the loco breezed through the curves without slowing down.
To prevent the Kay fr lugging down, eliminate as much wiring and as many connectors as you can.
Installing a decoder.
I tried 6 different motor decoders, (three different brands) using the socket. I tried 7 different sound decoders, (three different brands)
I tried all of the tricks published on the Bachmann site to get them to work. I tried little transistors, big transistors, three transistors.
I followed all the Bachmann web site instructions.
Did they work? Well some people might say so. In most cases the Kay would run, sort of. In many cases the chuff would work, if you weren't very particular.
I wasn't satisfied. The Kay did not run or sound like a model should. Certainly not nearly as well as my stock LGB Mikado.
Along the way, I hooked up a scope and really experimented with the chuff. I discovered that a 1 k resistor from rail A to + out worked just as good as any of the transistor "fixes". I also found out that a Phoenix 2K2, a SoundTraxx and even the three different Digitrax SFX cards would chuff perfectly using the optical chuff, if they had a battery, and a battery for the optical circuit. But, only if you connect the rail inputs of the sound cards to the + and - outputs of the dummy card INSTEAD OF THE RAIL PICKUPS.
Of course if all you want is perfect chuff's, that is great. but the sound card always thinks the loco is going forward, so the whistles, uncouple and some other features are messed up. That was about the closest I came to a perfect sound installation using the optical chuff.
I had to stop and evaluate the situation.
I really wanted to use a standard DCC interface. It would be easy to try new decoders. Easy to swap sound cards on a whim.
But, I had already bypassed the flicker circuit to get an acceptable flicker. I bypassed the headlight, to get a generator effect, the markers to get rid of the LED, the smoke because it burned out, the capacitors were causing arching issues, the rats net of wires were causing performance e and reliability issues, and all the handling had introduced a couple connector failures and broken wires.
I removed everything.
Here is what I installed:
I positioned a DG583S in the loco with the connector next to the motor. I re-wired the pickup boards and trailing axle to the track inputs. I connected the existing motor leads directly to the decoder. Total wire length from driver to driver through the motor, 10 inches.
I connected a two conductor cable with a plug on it to the decoder track terminals. This carried DCC up into the loco and provided a plug to make it easy to disconnect the boiler. I mounted a K1 to the back of the boiler for the six bulbs in the markers and the headlight.
I mounted another K1 on the blackhead for firebox flicker and cab light.
I connected the smoke unit to the F0F screw terminal and common on the motor decoder. The fan is powered by the F0R screw terminal and set to turn off when the loco stops to reduce noise when the loco is standing..
In the tender, I installed two 4 ohm speakers in series, a Digitrax sound bug and a K1 on the rear tender wall for the six bulbs in the lanterns. My Kay came without a backup light, so I installed an LGB headlight which I connected to the K1.
Why did I use the sound Bug? Well It cost less than $50.00. Go back and listen to the videos, was it worth it?
For the chuff, I installed two reed switches in the cylinders and glued magnets to the optical flags. I wired the switches to the draw-bar and the draw-bar pin in the tender to the Sound-Bug.
Why separate decoders? Simple, I now have a two wire track/data buss going to all the major components of the Kay with a two pin plug in each wire. Disassembly and maintenance is a snap. Fewer wires equals fewer failure points. Total cost of all the decoders was about 120 dollars street price. If I want to upgrade the sound, motor a function decoder in the future, I don't have to replace a big expensive multi-purpose decoder and do extensive rewiring. I can pick and choose features to "build" a decoder that does everything without sacrificing some feature just because the super decoder does not have it. Some will point out that a separate sound decoder can not be aware of the load on the motor. To that, I'll just say that those people do not understand just what decoder feedback is all about. Both the sound card and the motor decoder have Transponding, a Digitrax decoder feedback technology.
All wires between the loco and tender are totally eliminated with only the chuff input going through the draw-bar. I did not feel the need to feed the pickups in the tender to the loco. They both are heavy enough to work on the ALLY. The tender pickup is probably not as good as the loco pickup so eventually I may get sound dropouts. If that happens I can add a super cap to the sound, jumper the loco to the tender, or change out the Bachmann axles with LGB BB axles with electrical pickups.
Bind Fix
There was a lot said about the Kay locking up due to the loose eccentrics. The drive rods get out of wack and the Kay won't move.
Mine did it from the very start.
I also noticed that the front left wheel had a slight wobble.
TOC published a fix for the loose eccentrics, which I immediately applied.(shims and shortening the shafts a little.)
The lockups obviously caused by the eccentrics went away, but, the loco still galloped. It would seem to bind one per revolution.
I waited for my replacement eccentrics.
They came in, I put them on. they were tighter than the originals, but still a little loose and so the sidrods did some funny jumpy things..
I still had the galloping.
I could not get shims in the smaller space of the flat spots, so I took off the new eccentrics and put the old ones on and the shims back in.
I still had the galloping. and once in a while, the loco locked up when changing direction.
This time I noticed that the third right hand driver had a wobble.
I knew I had not moved the front wheel to the third position, so I removed the CW. and found... A BENT SCREW.
Well DUH! I should have wondered why one screw was always harder to tighten.
The loco still locks once in a great while.
If I am not pulling cars, it is OK
If I stop suddenly while backing, and the cars coast and create slack, it is OK.
If I slow very gradually, backing, and keep a small amount of resistance against the rear coupler. It locks UP.
I took out and disassembled the motor and gear box.
Found nothing.
I ran the loco on rollers for 24 hours. 10 seconds forward, 10 seconds backward (DCC and a program on my desktop)
The problem seemed to go away. still, every once in a while, It won't budge going forward after a smooth gentle stop when backing a string of cars. When that happens, the only way to get it to move, is to throw it in reverse, back up slightly and then go forward. When it is locked, all the junk on the sides are free of binds, only the geared driver is tight.
I was told the loco would take five foot curves.
It will.
It will navigate my R3 turnouts at each end of the passenger sidings.
The only problem is, the loco AND TENDER WON'T.
It throws the tender off the tracks at both ends of both sidings. I looked at possible modifications, but decided they weren't worth the effort. and would make the loco look bad.
Sure is a nice looking loco.
I canceled my order for a second Kay. Bought another Forney with factory DCC and sound (for about the same price)
I found a shelf in my den where I can admire how great the Kay would have looked on the ALLY.
