QSI decoder Install in the Aristo Mikado Overview This page documents the things I found and changed for installing a QSI decoder (even though I am using the Aristo plug and play socket!!) As stated elsewhere, you cannot "trust" Aristo wiring to be consistent, or even make sense, especially since they have never tested the wiring of the plug and play socket, only on DC iwth the "shroting plug" installed. Note: this install is for the second generation Mikado, with the socket in the boiler, but the newer board. Improving the power pickup This is a plug and play decoder. When I first installed it, I was not happy with the performance, I kept resetting the decoder. When operating on DC, short term power interruptions are often not noticed. With DCC sound decoders, these interruptions often reset the decoder to some state like stopped. I investigated the power pickup on the tender and found that there was only pickup from 2 wheels on the tender. I also surmised with the loose fit of the axle to the brass axle bushing was not giving me consistent power pickup. (Note that while the mikado has pickup on all 8 driving wheels, very often you are only getting power from 4 or less wheels, due to the rigid placement of the drivers, and the long wheelbase) I added pickups to all 8 axle ends, thus picking up power from 4 wheels but with additional redundance to offset the loose "connection" between axle tip and brass bushing. See the section "Improving Power Pickup" under the main Mikado section. Also read the "prime mover basics" page where the poor quality control and assembly often result in not all wheels picking up power. Understanding the Aristo wiring From looking at the design, the goal appears to be that in track position, the power is picked up from the motor block in the locomotive and from the tender pickups. In the battery position, the design APPEARS to be that the battery connector on the board in the tender is connected to the locomotive. In the design, it appears that when correctly wired, all that happens is in track position, the loco and tender pick up power, and in the battery position, the battery connected to the plug in the tender is just ADDED to the track pickups. In actual practice, there are many variations, tender power wired backwards that creates an immediate short, track/power switch wired so positions are backwards, etc. Also, it appears that the board in the tender also could be used to power a backup light in the tender, but not at the same time as battery power is available. There is also no wire from the Aristo socket in the locomotive boiler to the speaker in the tender. So, for the QSI install, and since there is no backup light in the tender, I decided to rewire things to allow use of the 2 sets of wires between the loco and the tender to bring track power from the tender pickups and to route the speaker connections. Starting the installation For reference, the aristo board will be viewed this way, so directions will be in reference to this picture: Notice the board is labelled: "DCC-VERSION, MIKADO PCB02" This is for a newer version Mikado, notice one switch on top of the board. There are also 2 DPDT switches on right hand corners of the board, but soldered to the bottom, you can see the 6 pins on the top corners of the board. Also, note well the 4 solder lands to the right of the right hand socket strip, just above the "cab lamp" socket. This is where the 4 wires from the motor block connect. The outer 2 wires are the track pickups, and the inner 2 are the motor. The motor was wired backwards in my installation (and also probably the track pickups), I swapped the inner 2 wires so that the DCC decoder runs the motor in the proper direction. I don't run DC, so I did not care about the track pickup wire polarities, but they are most likely wrong also. Sigh, how many years did we ask Aristo to test the socket wiring? Here is an older board, which looks similar, but has 2 switches stacked on top of the board: (courtesy Ed Silvis) There are 2 connectors on the loco, one is a male (has 2 pins sticking out and visible, and one is a female, all plastic, just 2 holes for connecting. Remember this is the locomotive side. First, I removed the battery cable from the front of the loco, desolder the 2 wires from the upper left corner where it says "BATT" and it has white lines from the text to the 2 solder pads. If you pull one wire out at a time, and wiggle them a bit you can pull them out without further disassembly. The wires have a piece of heatshrink on them, and it catches a bit, so do them one at a time, once you get the first one, the second one slips easily. Next, it is time to figure out the wiring. The track/battery switch is on the bulkhead, and you can see that it is a double pole double throw switch, but is wired as a single pole double throw, and you can see that the circuit board has 6 holes for a dpdt switch. Looks like this was originally the NMRA "direction" switch, where you could pick NMRA or conventional polarity, i.e. what direction the locomotive goes compared to track polarity. To figure out the wiring, I started with the motor block, it has 4 leads, 2 for track pickup, and 2 for the motor. It's pretty easy to figure out, and the track pickups are black, the motor are blue and brown. The black leads are the outer of the 4 leads on the board just to the left of the track/battery switch, and they are matched to which rail picks up, i.e. looking down the length of the locomotive, the right hand wire is the right rail pickup and the left wire is the left rail pickup, and the center 2 wires are the motor. You can see this easily from underneath the board, and then you can relate it to the line of 4 soldered wires on top of the board as previously mentioned. Now comes the fun part. Looking carefully, you see the word BATT marking 2 leads from the tender, and it's obvious they connect to the track pickups, you see the traces clearly. So, the track pickup leads to the tender are labelled BATT here. Cute. Following this from the underside of the board, they connect to the male plug on the loco. OK, now we know this. "New" sound wire: The other connector, the female one, comes up to the board. One side connects to the track/battery switch, and if the switch is "on", connects to a motor lead (In my case the blue one), and the other lead connects to the brown motor lead. So here it becomes clear that the track/battery switch only connects or disconnects the battery, it leaves the track pickups in the motor block and the tender connected at all times. I see that this was easiest to do at the manufacturing site, but a little creativity could have switched things differently, i.e. disconnecting the track pickups, so running battery power on a powered layout would not be a possible disaster! OK, great, now all I have to do is remove these 2 wires and connect to the speaker jack on the QSI board. This has the added benefit of disconnecting the track/battery switch so it's position is of no consequence. To re-iterate, from the loco side, the female connector has the speaker. (on the older board, with 2 switches on the bulkhead, the wires are soldered to the underside, where it says SND/Power) Here's the speaker connector cut from the QSI harness, being soldered to the cable that goes to the plug at the back of the loco (female connector). Another note here, the "battery wires" connect to one end of the switch, and the other one, the one that goes to the brown motor lead, has "SND-PWR" silkscreened by it. There is another pad that is part to the 2 "SND-PWR" pads that goes directly to the blue motor lead. So it looks like originally the battery/sound wire was directly connected to the motor with no switch. Maybe the SND refers to the input to a sound board to sense motor speed. So with these wires removed. I can put the board back in. I dressed the 2 sets of "plug wires" over to the right side, and the 4 wires from the motor block out to the left, keeping them away from the mechanicals of the 2 slide switches on the board. Be sure to engage the switches with the 2 sliding levers as you replace the board, check them again after the board is tightened down. I put the reed switch down the right side, in the right side, as far forward in the firebox. I dropped a spot of hot glue down there, and then held the switch in it. Fast, and easy to remove later, and easy to find from the outside. Here's the final setup ready to close up: Update, this install was done with the first version of the QSI DCC decoder. Since then I have changed over to the QSI "Titan", and the install is the same, except that the sound wires connect to the screw terminals of the decoder, no special plug is required: Below is a picture with the Titan decoder: Now on to the tender: Since the female from the loco has the sound, it's the male on the tender that wants to be connected to the speaker plug that is left from cutting the QSI speaker cable in half. The 3 pin half will connect to the speaker jack, and the speaker is already wired there. You might take this opportunity to turn the speaker so the terminals are closest to the board, and to re-solder the leads into the circuit board holes, not just laying the wire on the board. The wires from my male plug (tender side) were connected to the "MOTOR" traces (silkscreened on the bottom side of the board. You might notice here that if the 2 sockets were soldered to the side with the words silkscreened on them it sure would make things clear. I did not want to try to desolder the 2 sockets, but if I had spares I sure would do this. Another manufacturing shortcut, or strange assembly of a clever design? Solder the 3 pin connector to the leads you removed. Double check visually that these leads lead to the male connector on the tender side. I would recommend removing the truck to be sure you see the wire route, or use a ohmmeter. Here's how the tender board looks when complete. I think it looks a lot nicer, and all the wires are in the circuit board holes, nice secure connections. This is all you need to get going, and you did not add any wires between the loco and tender, the tender board is cleaner and more mechanically secure, and you have used the standard plugs on the decoder and speaker for disassembly. In the future, I will probably replace the speaker with one that has more bass. Also, the QSI has a 5 watt output, and the Aristo speakers are one watt, so running at full volume often will distort the Aristo speaker. QSI Programming notes: Original software version: 7.0.45 (beta, announced version on startup) programmer reported: model: 3010 G-scale Mikado class: steam soundse: 0 version 7.0.45 build date: 7/23/07 last modified: 7/23/07 hardware: 4000 on 12/22/07 I loaded version 7.18.2, type# 3010, sound# 551, soundset 2 on 10 Mar 08, loaded 7.20.2, type 3010, sound 551, long address 3283, short 3 mute volume 40 (cv 51.1) bemf chuff on, set to 8 (cv 56.0) set status report to speed and bemf : 3 (55.178.0) set minimum bemf to 3 (56.5) select STC set to 0 (56.4) run loco at SS 8, now set CV2 to just enough to run, then back off until it just stops set chuff interval scale factor to 45 (56.12) This method synchronized the drivers and pistons to exactly 4 chuffs per revolution. Wow. I did not evenhave to use the "chuff interval trim" setting, which can further fine tune the chuff rate. One thing I do not like as much, is this mode uses the "short steam release" sound in the sound file, not the chuff sound. It is not as full bodied as the normal chuff. We'll see how it goes, but the synchronization is pretty cool. If the sound could be improved, then it would be perfect. At this point, no need for a chuff switch, which is a royal pain in a loco with no factory chuff switch. Putting magnets on the tender wheels and a reed switch to trip them is not a good solution in my mind, since the tender wheels are not the same diameter as the drivers.