Hmm... I've tried the hot air, I obviously would need more practice! I actually have a hot air rework station. Of course, the shaped hot air tips are where they kill you in price, depending on the different chip patterns you're using. I fooled around with the solder paste, but again I didn't have the syringe and tips.
Maybe I'm better off contracting the builds out. 
Chuck, that's exactly what I do with the small resistors and caps. For the multi-pin parts, I tack one lead down and then one on the other side, finally I finish them off. As long as I stick with 1.25mm lead spacing, I can solder them. The really small fine-pitch parts I ended up doing something like you do. I blob the solder on one side then the other. Then I go back, heat up one side, and just use the solder sucker to slurp up all the excess. It's never failed to just leave a neatly soldered part, pretty slick.
I try to stick with 0603 or larger discrete parts, and 1.25mm lead spacing or larger on other SMT parts.
gunrunnerjohn posted:I tend to doubt the back-EMF could accurately enough generate the chuff. Every different motor or locomotive would have a different chuff rate, that would be difficult to deal with.
How does it accurately determine speed then? It is touted as just as good as ODY/Legacy,. Seems to me it is programming. Motors are similar. Mabuchi or Pittman, typically 12-15VDC at 6000 rpm. Soft set tech is all ready there. K-Line you could change the Cruise board gear ratio and wheel dia setting. But we are really after a basic effect and since most engines have ratios between 20 and 40, being able to (nudge) the chuff rate and set 2, 4, 6 per revolution should be possible (not by me though
). I really wonder if this has to do with MTH patent on integrated chuff and puff. Hopefully you won't get a cease and desist letter. My understanding is the whole integrated issue is what MTH has patented. G
G, Jon Z has weighed on this before. He uses a combination of commutator pulses at low speed with back emf at higher rpms. Commutator pulses could be used like any other of pulse counting schemes used by DCS, K-Line Cruise, or TAS EOB but how do you accurately synchronize a DC level with wheel rotation? It would be like a VCO. Chuff rate would increase but never be synchronized.
Pete
What he said. 
John another option would be to use an op-amp setup as a buffer to tap into the existing locomotive speed sensors. You may not even need the op-amp. This would reduce your parts cost and eliminate alignment problems with your sensor.
Very cool! 
Matt, trying to tap into the existing sensor seems a recipe for disaster. The signals are pretty low level, and adding any load is sure to screw things up. Also, there are very few locomotives that I'd be putting this into that would have a speed sensor that I could tap into. Since this is powered from the Super-Chuffer, it's DC ground is also locomotive frame ground. Many of the other systems would have a DC ground that isn't also frame ground. Another issue is that something like the TAS EOB cruise has one scheme, but Lionel Odyssey has a different scheme. I'd have to engineer for both of them at least, and since many locomotives have no tach, I'd still have to have my own sensor for a vast majority of the installations. Way too much work to save a $1 tach sensor. Too many variables, much better to do my own thing, no chance of screwing things up that way.
Norton posted:G, Jon Z has weighed on this before. He uses a combination of commutator pulses at low speed with back emf at higher rpms. Commutator pulses could be used like any other of pulse counting schemes used by DCS, K-Line Cruise, or TAS EOB but how do you accurately synchronize a DC level with wheel rotation? It would be like a VCO. Chuff rate would increase but never be synchronized.
Pete
Gear ratio. You know how many poles, you know the ratio. It is a matter of programming to generate the "approximate chuffs per drive revolution. I am sure some of this logic is used for the speed steps. Again, softest tech would allow user adjustments. Certainly a reader works, but it has similar issues. How many stripes used, size of Flywheel, driver size and gear ratio. All the stuff MTH factors in on the software for the speed and chuff control. Your just changing less pulse for stripes. Certainly not precision, but I bet a 35SMPH no one would notice. Except a small few. . G
Well, it seemed that the flywheel tape was the easy solution, and I have an easy way to calibrate for any flywheel size and gear ration with one simple calibration procedure, so I think I'm set.
Of course you are. I was just pondering why not built into ERR. And folks that say it can't be done I question. IF you can measure and adjust for 100 speed steps, I am sure you can measure, calculate and adjust for chuff rate.
As I have stated before, clearly direct measurement by counting (tach reader) would seem more accurate and consistent. Otherwise why even use ODY at Lionel. G
I could borrow 5 volts off something, no problem. This is a very elegant solution to a common problem. I hate setting up reed switches too and always end up using tender wheels, so I'm never quite getting the right rate.
5V is pretty easy, you can get it with a few components. Just assemble these on a little perf-board and you're all set for 5V.
Sorry if this has been answered but will this work with previous installs of the Super Chuffer?
No problem Marty, this just replaces the existing chuff switch. It takes 5VDC, GND, and has a chuff output. The "standard" instructions for it will be to just wire it to the Super-Chuffer. The chuff output drives up to 25ma, so you can parallel it with the Super-Chuffer chuff input and the chuff input to the R2LC as well. Like I posted above, if you wanted to use this without a Super-Chuffer, a 5V power supply is simple to do.
You need someone to build them up for I'm game. I'm sure we can work out arrangement to our mutual liking
Do you have the artwork done for your PC board?
The artwork is all done and I'm waiting on the first set of "production" boards. I made all the corrections from the prototype, so hopefully when I get these, they'll run with no mods to the board.
superwarp1 posted:romiller49 posted:I use err cruise without fly wheels. What do I do.
Romiller
magnets and a reed switch or John's optical sensor he came up with last year.
If I'm reading this correctly, for my TMCC engines without a tach reader and all my ERR upgraded engines this Super-Chuffer Chuff Generation Enhancement board is of no benefit. I really liked the idea of an optical sensor to sense reflective spots on the locomotive driver for perfect sync with sound and easier installation, but as mentioned had some issues and is now dead. Oh well.
CAPPilot posted:If I'm reading this correctly, for my TMCC engines without a tach reader and all my ERR upgraded engines this Super-Chuffer Chuff Generation Enhancement board is of no benefit. I really liked the idea of an optical sensor to sense reflective spots on the locomotive driver for perfect sync with sound and easier installation, but as mentioned had some issues and is now dead. Oh well.
You can add this reader to any TMCC engine, all you have to do is print out the tach tape and wrap it around the flywheel. If you don't have a flywheel, you're out of luck, but that's only a handful of steamers, at least any that I've worked on. The Super-Chuffer and this chuff generator were specifically developed to enhance ERR installations, so at least in my estimation it's certainly a benefit!
I liked the optical sensor with the spots on the wheel idea as well, but in practice the limitations of sensing them and the different mechanical issues made it problematic. This gets the job done with no issues with mechanical tolerances.
John,
Thanks for the clarification on the fly wheel; I missed the part on printing out a tach tape. I guess I was not reading this correctly
No problem, just didn't want anyone getting the wrong idea.
One more step along the road. I have the "final" circuit, no cuts and jumpers on the board, and it appears to function just fine. Next step is to install it into a locomotive and give it the real "action" test.
Inventive could well be your middle name John.
But... what to do next?
R3 looks a little lop sided. Need me to straighten those out for you.
For prototypes, my criteria is they work. I let production get all the parts straight.
gunrunnerjohn posted:But... what to do next?
Fastrack switch controller color flipper bank...might have a limited following.
As long as you commit to buying the entire production Chuck, I'm all ears. Unit price is looking like around $60/ea. 
gunrunnerjohn posted:As long as you commit to buying the entire production Chuck, I'm all ears. Unit price is looking like around $60/ea.
A bargain...but I already did my two I needed.
gunrunnerjohn posted:But... what to do next?
I'm 'Imagining' Stan coming along shortly to help you with this.
I can get into enough trouble alone I can assure you.
I understand, I can get into trouble just reading about stuff here. 
Hi John,
Quick question - in your picture, the flywheel is roughly the same size as the motor. For some steam engines, this won't be the case. Will this come with a bracket or some other way to allow this to work with a motor that is bigger than the flywheel?
Funny you should ask. The sensor has fairly long pins, and the game plan is to space it correctly and solder it to the board at the correct distance. It just happened that my test subjects were the right size that I didn't have to do this.
Unfortunately, this is one reason that the installation could be complicated for folks that can't solder. I don't know of an easy way around this issue. When you install the MTH PS/2 upgrade kit, you have a similar issue with their tach sensor. They make them with a couple of different sized spacers, but I've had to ad-lib in a couple of cases for oddball locomotives.
I could see two possibilities for a way for those who can't solder to do this (then again, most people doing these upgrades can solder I would think).
1. Instead of using a solder connection for the sensor, use a socket similar to the LEDs on DC1000 switch motors and have it as a pass-through type so the sensor leads can be pushed into the socket as far as needed.
2. Include a bracket with double-sided tape that would be long by default (for small flywheels) that could be clipped to size for larger ones (or smaller motors.
Those were just ones I thought of off the top of my head.
BTW, I have a Cab Forward that needs upgrading - I'm happy to help test it.
The socket is too easy to dislodge the sensor from, especially one with the contacts that allow it to go all the way through. The machine pin contacts might be an option, but even then I see more issues than I solve.
Not sure what the bracket solves, but it certainly adds a level of complexity and reduces reliability I suspect.
I thought about various socket options, but in the end I really think I'd regret making it more complicated.
Do what TAS did back in the day with EOB. Supply different diameter shims to be glued to the tac reader for the correct spacing. Your never really going to find a perfect answer to this. The least is leaving the sensor un-soldered for the installer to solder when installation is completed.
You don't really need shims, you just have to stick something about 1mm thick on the flywheel and move the sensor down to touch it. Solder the leads and it's done. Trying to stack spacers just makes the installation more complicated.
As stated above in my opinion that's a bad idea. Fine for you or me but as you stated, for someone who doesn't have much experience soldering could end up damaging the board.
Somebody with little experience is always at risk in a highly technical hobby and that's how they learn.
Access to this requires an OGR Forum Supporting Membership
