slow runby
fast runby (nice and quiet now!) 
Good job on the re-gear, glad to see that's a practical solution. 
How does it do with a train behind it Joe?......looks like you’re taming the beast!...great work!
Pat
I have to confess that I haven't attached a train to her yet. At first, I was worried about any drive failure. Then came trying to achieve smoother control. Now it will be if my pulleys slip under the load of a train. I came up with solutions for securing the pulleys better. I didn't do them as other things kept failing. So far, the pulleys haven't slipped.
I grinded a slight flat on each shaft under each pulley. If they were metal, I'd tap them and add set screws. I also considered casting new ones with tiny holes and drilling them for a tighter fit. I had a mis-communication over the shaft size they were designed for.
If they slip as is, I may convert the top shaft to 4mm, and add that Pittman you sent that has a larger shaft to match up. They are a friction fit right now. I tried to drill pin holes to insert pins but the shafts wouldn't drill easily. I need new micro (wire) drills I guess?
Joe, consider Loctite Retaining Compound for the friction fit pulleys, I can about 99% certify they won't slip! I've used Loctite 680 and 638, not sure what the major difference is between them. I used the 638 to lock a slipping gear inside a K-4 locomotive drive block, so far it's holding, that was a pretty good test.
Heheh, we have one of those...noisy gearboxes. Not a challenger, I believe its a Niagara. We call it the coffee grinder at NYSME.
@Sarah posted:Wow guys, that's advanced stuff! Two motors is surely the coolest thing in an articulate.
So far I tried a Faulhaber 2342 but found it a bit small.
If you think two motors is cool, check out the Westside Triplex. I used to own one and it had 3 independent motors which ran nicely. I probably have photos deep in archive, think it was a decade ago I played with it.
Don't let size fool you, coreless motors can be pretty powerful. I've put one of those faulhaber to good use with an NWSL drive and its been running for years. It's so powerful it can run on 1 volt and has great low speed control because of all the torque. Same goes for the tool industry if you look many of the premium brands top line use coreless, eg Milwaukee Fuel. Not many brushed motors, they're bigger, less torque.
Plastic pulleys on metal shafts:
Yes, Loctite said it would work, and "comped" me some "type N" prep fluid.
Nope - did not hold. You need to "pin" or set screw. I can assure you Pittman and NWSL shafts will drill. Use a drill press, a new drill, and cutting fluid.
Usually, plastic pulleys will hold threads. You just need a big enough "flat" to avoid distortion under load.
Thank you! I already put the Loctite on the shafts and will test. I don't have faith in it because of the actual issue. The shaft is only 3mm and the hole is 3.8mm. Too big on it's own for positive locking. A 4mm shaft would have been ideal.
I did not machine a bushing to fill the difference. I used what I had laying here and is not ideal. The only thing going for me is that the wheel slips easily on the rails. When the power is killed, the engine probably slams to a stop. So the pulleys won't stay locked for long.
I did look for new tiny bits while shopping yesterday and came up empty. I need to order some and pin the pulleys. I do have spare lock screws from the universal joints that I could install in the pulleys to give them help locking on the shafts.
I installed a MTH PS2 board set back into the engine and will test it today. I am always amazed at what these toys actually go thru, trying to do their job for us.
@Engineer-Joe posted:Thank you! I already put the Loctite on the shafts and will test. I don't have faith in it because of the actual issue. The shaft is only 3mm and the hole is 3.8mm. Too big on it's own for positive locking. A 4mm shaft would have been ideal.
WOW, too much slop for retaining compound, you're right! Retaining compound is only for close fitting parts, it does have limits.
I don't have that big of a gap John! The problem is I did not have a proper way to fill it though. That's why I thought about casting new pulleys. When you cast, that stuff is so sticky, I figure it maybe enough to lock it to the shafts. Add to that, I grinded the shaft to have a flat for locking.
I also considered just pouring resin in the gap. I figure that might just be a mess and still not get in there enough to do the job. There might also need to be some type of recess in the pulley to allow the resin to lock on better?
A proper (brass or plastic?) shim machined and heat fit onto the shaft I feel might do better?
I was originally instructed that I may have to drill out the pulleys to fit. I believe that was just a misunderstanding on what sizes these are. You don't drill out a 3.8mm hole to fit on a 3mm shaft. This should not be a big deal overall. I just need to do it properly.
So the pulleys are holding and even though I have a PS3 steam kit just about here in the mail, I stuck another PS2 board in for testing. Of course, the engine is way too fast. So I put a tach tape on with way more stripes and still too fast.
I'd probably need a custom sound file for much lower gear ratio to get it right. I tested the tach and it is working. I got a good change when I added several stripes. I don't think I can go much smaller on the stripes?
Does the PS3 HO board set use the same tach? I wonder how small are those stripes? and how is their gear ratios compared to O scale? 
I take it you do not have a lathe. I can make you a bushing and knurl the outer diameter if you want. You would owe me postage.
I'm not sure how many stripes you can put on the tape, but I suspect you could probably go for about 38-40 on a 30mm flywheel and still have the sensing work.
The HO boards use a true PS/3 tach sensor, totally different than any of the O-gauge ones. The HO motor on the PS/3 test set has the sensor under the flywheel like a Legacy, the stripes are actually facing the motor, very compact design.
The tach sensor used for the PS32 board in the steam upgrades is just the PS/2 tach sensor.
Attachments
@bob2 posted:I take it you do not have a lathe. I can make you a bushing and knurl the outer diameter if you want. You would owe me postage.
Thanks for the great offer! It's actually working right now. I would need 2 bushings for 3MM shafts and 3.8+ holes. Maybe I could heat the plastic gears and cool the shafts for tighter fit on those bushings?
It's just around 4 times too fast! It was around 4 times too slow. I over shot it! At 10 MPH it seems like it's doing 40MPH. At 15 MPH, she's flying and maybe 60MPH or more?
John,
It think if I had some type of adjustable command control, or for conventional use, it would be fine. I noticed that when I added stripes, the sounds think the engine is running much slower? (chuffing slower, engine's fast) So maybe they're not all being read at higher speeds? I did just change the sound file.
I may have to work on the tach reader and flywheel?
My first tests had the diesel sound set inside so I don't have the sound comparison to go by. I only had a Challenger sound set for a G scale on my computer to try. Every time I try and get a sound file, I get a page of gibberish on my search from MTH.
This engine runs smooth and fast as I want now. I just have to get the control speed synced right.
Spacing is critical on the tach sensor. The Omron EE-SY124 sensor has a fairly tight optimum sensing distance specification. The "sweet spot" is around .8mm to 1.0mm, vary from that and the sensitivity drops pretty quickly. Half an mm in either direction really drops the sensitivity, especially on the closer side!
Attachments
Yeap!
I answered my own question when I typed it above. It hit me that the tach was missing the count at higher speeds. So first, I colored the stripes darker. No improvement. I pushed the tach closer while the engine ran and it slowed down!
So I closed the gap quite a bit. Now the engine seems like it might be twice as fast as it should be. However the chuff seems to match up much better.
Another note, I have a bad habit of saving everything. I had a tach go bad years back and I saved it. I didn't know for sure if it was bad, or it's circuit?
Now this used one I put in, seems very weak reading the stripes. I bet I have this in about twice as close as normal.
Whoever posted the thickness of a dime before? I disagree as all the ones I put in seem to like less than .3"
A dime on my bench measure like .5!
Actually, the thickness of a dime is about 1mm, the diameter of a dime is about half an inch! .3" will not work at all, spacing that wide would be a runaway. I space my sensors and the MTH ones a .9mm with a home made feeler gauge.
Ooops. The dime is .05" thick. or like 1.18 mm? I set the gap at .03 at first and that was too wide. I bet it's like .015 to .02 now?
I have other issues popping up and I'm ready to scream. I got the speed closer and the engine runs rougher? I even hear the belt skipping? I checked for binding and there was the usual that was there before. The whole gear assembly seems to be wobbling around much more. I had left it unbolted from it's spring mech and that was a mistake. I just reattached it now. and will see if that helps. The pulleys are still tight!
I hear some odd chuffs. The headlight goes dim with the smoke on and flashes on full like a beacon? I hear the belt skipping. I fix one thing and three more break?
Urrggg!
I can't believe the overall traction this thing has. I tried to hold it in place while running. It pulls pretty darn well. Remember, there's no traction tires!
You really want the tach reader spaced between 0.9mm to 1.0mm from the tape. If you go too close, refer to the previous plot I posted, the sensitivity falls off VERY FAST below about 0.8mm, so I shoot for around 0.9 to 1.0mm for my spacing, that gets me near enough to the max on the peak of the response curve. If you're still stuck in English measurements, 1.0mm is .039", and 0.9mm is .035". 
