Hey John, are you still waiting for the revised boards? I am missing the updates on this great project!

Mack

I have all the parts for the first batch, and I got a shipping notice that the boards have been shipped Priority Mail.  I should have boards in a couple of days, and probably early next week I'll get a change to assemble one to insure everything is right before I send the lot off for assembly.

I've been in a holding pattern until I get the new boards and have one to test.

I am thinking of trying to do a mod to generate 4-chuffs from a two-chuff input, though I don't know how well it'll work.  Since it's strictly software, if it doesn't work out, I just won't put it in.  I'd also have to figure if I can come up with a way to program it with only the chuff input.  I'm thinking of a sequence of chuffs and delays to enter into programming mode.  I could also simply make it a programming option when I load the software.

Of course, if it doesn't look and sound good after testing, I'll probably just drop it.

What's your latest on how you're going to multiply the chuff/puff rate?

I believe you can make an adaptive "learning" algorithm that looks at the "ON" time of the chuff switch to measure how fast the wheel is spinning.  This will change unit to unit based on spacing from magnet to switch.  But PIC can learn the duty-cycle of ON to OFF after several revolutions and can store it in non-volatile EE memory.

Then, as the engine starts moving, after it passes the switch you have a measure of speed that ought to be good enough to generate extra chuffs before the wheel makes a full revolution.  Obviously the region of interest is very slow speeds so this works in your favor as the "ON" time of the switch will be longer at slow speeds so you get better timing resolution from which to estimate speed.

Anyway, that's my 2 cents.

Stan, that's a clever way of handling the timing, I didn't think of that.  I have the EEPROM, might as well put something into it.

I just got a box with the new boards, so I'll be assembling one in the next couple of days and testing the final design to see if anything fell through the cracks.

The Rev. 2 boards are done, the bench test went great after fixing a bad part.

Stan, you'll be happy to know, for whatever reason, the PWM works great on these, and the capacitor is gone.  I double-checked my old breadboard, and I can't imagine why it needed it for slow speed operation.  I had it running at 700 RPM at low speed, and it started every time when I stalled it, no sweat at all.

The only thing I can think is the specifications of the FET's I'm using on the breadboard are much lower than the SMT parts, those are hefty little items for as small as they are.  Of course, the wiring is laid out a lot different, as there are long leads everywhere on the breadboard.

The power issues were totally solved by using a little switching power module, it doesn't even get warm when I run several incandescent cab lights from the 5V supply!

Looks like it turned out very good. Nice looking board too.  

It took two tries.   The first one had reversed FET images, so I had to mount them upside down.  This one also added much better power and the dynamic braking, a worthwhile upgrade.

Yes, removing the cap makes me happy.  The ripple currents are at the PWM rates which is more stressful wrt ESR heating than traditional 60 Hz ripple.  The last thing your need are capacitor failures.

I marvel at your progress!  I'm cooling my heels waiting to spring my next idea on you.  I'm imagining you making a wireless RF link using those low cost eBay modules.  Plenty of value added. Still in the imagination phase though.

I'm still mystified why the breadboard didn't work without the cap.  I went back and verified the wiring, and it's all correct.  I had also put the FET's on the Rev. 1 PCB, and it also needed the cap to function.  I had used Tantalum caps for those in the hopes they'd be a bit more reliable.

All's well that ends well!

BTW, I did pick up an optical tach on eBay,  C:US:3160" target="_blank">Digital Optical Tach, hard to beat for $15, and seems to work great.  I had one a long time ago, but it croaked, and I never got around to getting another.

John;

Just wondering what the latest update is on this neat project?

Rod

Harry Henning kinda' jumped the gun with an announcement, but Bill Henning should be posting within a week the official announcement of availability.

PRR Q-2 project with johns super chuff board

Few questions,

Your board will control the smoke unit right? If so will it be just as strong as the video I posted on my EOB thread?

Gary, the board just controls the smoke fan operation, so the volume of smoke won't change with the board installed, just the timing of the chuffs. The smoke resistor is still controlled by the existing power connection.

The smoke motor running time for a chuff is timed based on the locomotive speed to enhance the appearance of the chuff, and not just run for the length of time the chuff switch is closed.  As the speed of the locomotive increases, the duration of the chuffs is decreased.  Also, the smoke unit motor is dynamically braked to enhance the definition of the chuff.  Finally, the smoke fan is run continuously at a low speed when the locomotive is stopped.

In addition to the smoke, the board also optionally controls Rule 17 lighting for an LED headlight and control of LED or incandescent cab lights.  These features can be used or ignored as desired.

Me bad, what I should of ask is, the fan control.  I have a few chuff n puff boards from TAS but I'm not happy with the voltage put to the fan.  Smoke output is weak in my opinion.  Just wondering how your board pushes the smoke out.  Is your board just a switch for the fan unit power or does produce it's own power for the fan unit like the old TAS EOB motherboards?

My board has a 5V DC supply for the fan as part of the board.  Other than the very small voltage drop of the FET that regulates the voltage, the chuffs are pumped out at a full 5 volts on the motor.  I don't think you'll have any issues with the fan voltage.  If you saw the demo thread of the older version, that one actually had an output voltage for the motor of about 4.4 volts because it used a transistor drive, with the FET, you get pretty much full voltage on the motor. 

I don't know if you saw this in the other thread Gary, I'll link it here in case you missed it.  The first post has the video.

That's the old board without the FET drivers, and also it doesn't include the dynamic braking for the smoke motor.  I'll be doing a new video soon with the new board and all the enhancements.

My New Super Chuffer Board Demo

No I didn't see, so thanks for re-posting.  I'm usually away from the forum this time of year and really don't get back into trains until the weather heads south in the fall but my boy has been bugging me to run trains and I've been actually working on them this summer. Hence the issues with my Niagara.  I've only picked up on your project since I posted about my EOB issues. 

Very interesting to say the least.