Skip to main content

With some TMCC upgrades, or even adding the Cruise Commander M to some locomotives, the serial data stream from the R2LC (or R4LC) gets overloaded and doesn't properly trigger either the motor control or the sound.  I previously created a transistor buffer to correct this, but that doesn't always work.  I set about finding out why it would sometimes fail, turns out it's the amplitude of the serial data, not the actual drive impedance.  Since the transistor buffer was an emitter follower, it didn't increase the amplitude, which turns out to be the bigger issue.  So, back to the drawing board.

I came up with this.  It's a 2x amplifier using an op-amp that also provides at least 20ma of drive if needed.

R2LC Buffer-Amp Prototype

Simple circuit, any caveman could do it.

R2LC Buffer-Amp Schematic

It accomplishes the goal, with a 2.5v P-P input signal, it outputs a rail-to-rail 5v P-P replica.  Even adding a 470 ohm resistor load didn't phase it, the amplitude remains almost exactly the same.  The 5VDC required can be stolen from the R2LC 5V lines on that board's pin-19 & pin-20.

R2LC Buffer-Amp Waveforms

The actual board is pretty tiny, .3" x .4", but I'm thinking I may actually slightly increase the size to make it easier to assemble.

R2LC Buffer-Amp 3D N1R2LC Buffer-Amp 3D N2

I believe this will for once and for all put to bed the issues occasionally encountered with loading on the R2LC serial data signal.

Attachments

Images (5)
  • R2LC Buffer-Amp Prototype
  • R2LC Buffer-Amp Schematic
  • R2LC Buffer-Amp Waveforms
  • R2LC Buffer-Amp 3D N1
  • R2LC Buffer-Amp 3D N2
Last edited by gunrunnerjohn
Original Post

Replies sorted oldest to newest

Hey John,

I'm thinking that this may be the solution to a problem I've been having with a K-Line GS4 that I upgraded the motor driver board to Cruise Commander M unit. When I power up the transformer, if the voltage is set to 0 and I bring it up to operating track voltage which is what I've done for years, the locomotive has only one speed and its pretty fast and no matter what I do, the sound stays at one chuff rate and the speed is either off or on. But, if I preset the voltage on the transformer to what I run at and then put power to the transformer, it works exactly as expected with all speed steps and chuff rates. From what You posted up top, this sounds like it may be the solution.

Are  you going to be selling these? 

I use it whenever I am doing an installation and upgrade and I run into serial data issues.   A common place is when trying to use an IR tether and the ERR Cruise Commander M.  A serial port buffer circuit has been around for years, but none of them amplified the serial port data.  I found that attenuation of the serial data stream was a primary cause of issues with seeing the serial data.  This is my attempt to put that issue to bed.

Are  you going to be selling these? 

Norm, I doubt I'll be building them for sale, I can't imagine getting enough sales to justify having them built in quantity.  I may do a re-spin of the board a bit larger and make it easier to assemble, then I could offer the boards and maybe the parts.  As you can see from the actual board, I may have gone overboard making it small, the little resistors and cap are 0.06" x 0.03" in size, I suspect most folks would have issues soldering them.  The op-amp is also pretty tiny, the spacing of the leads is less than a millimeter.

This is probably what the one for the masses to build would look like.  In order for it not to get really large, I did keep SMT for the op-amp, but it's a SOIC with leads spaced a 1.27mm, they're pretty easy to solder.  Functionally, it would be identical to the SMT version I posted previously.  It also is the same schematic as the SMT version, just the device footprints are different.

R2LC Buffer-Amp Thru-Hole 3D N1R2LC Buffer-Amp Thru-Hole 3D N2

Attachments

Images (2)
  • R2LC Buffer-Amp Thru-Hole 3D N1
  • R2LC Buffer-Amp Thru-Hole 3D N2
Last edited by gunrunnerjohn
@GregM posted:

Not that I think I will ever build one, but this caveman would need to know the value of C1 first.

Yep, forgot to turn on the value display.  As you can see, it's just a power bypass, so a .1uf ceramic would do fine there, that's what I'm using.

@DaveGG posted:

I have more than a couple of the MTH PS1 large steam and large diesel engines that I am considering converting to ERR.  Would you think one of these would be needed in those "upgrades"?

Only if you have issues with loading the serial data line.  It's somewhat hit or miss.

I will be interested in several of these built ups or kits when you finish up your prototyping. I've seen some odd problems with Cruise M installations and this might be the solution. In any case it would be good to have several on hand in case of unexpected problems with future installations.

Also, wouldn't this be a good addition to Scott's ERR lineup? If there are some Cruise M installations that will not work because of serial data issues, it seems like he would need to be able to sell a solution.

Ken

I have the thru-hole boards on order now, and the parts are also coming.  I'll offer some kits that include the board, parts, and instructions of how to wire and use the buffer.  Obviously, they won't be that expensive considering the few parts needed.

I build up a handful of the small SMT versions and realized that was probably going to challenge many folks putting tiny .06" x .03" parts on the a .3" x .4" PCB.   The little SOT-23 op-amp is actually easier than the tiny chip parts.

Well, I got my prototype boards back from OSHPark, and built a couple of the thru-hole versions.  Note that it's not totally thru-hole, but I went with a larger footprint surface-mount op-amp to make them easier to solder.  The resistors and caps are thru-hole.  This one is quite easy to build, so I'll order a larger quantity of PCB's now that I've actually seen it function.  The actual board is only .4" x .7", and about 1/4" thick at the thickest part.

It took a little longer as the first set I somehow got the wrong footprint for the op-amp and it was too wide!  Had to deep-six those and try again.

Here's a couple of the assembled boards.

To finish the build, I encase them in heatshrink so they can just tucked in where they fit.

This illustrates the performance with a 220 ohm load, still has almost a 2:1 gain.

Attachments

Images (3)
  • mceclip0
  • mceclip1
  • mceclip2

The boards are on order, to keep the costs reasonable, they're shipping by slow boat from China, so it'll be a few weeks.

@Jim Sandman posted:

I just went digging for a previous posted design I'd made a few of.  Still have about 8 of them, took me almost a hour to find what I'd done with them.  Thanks for killing my time today guys!

We love to waste time, especially if it's not ours.

I have some previous buffers that didn't amplify, but the silver bullet for these issues turned out to be boosting the amplitude of the serial data signal so that everyone receiving the data gets the word.  I had failures with the previous designs that didn't boost the data, lesson learned.

Jim, the old one was a repeater, but it didn't address one of the significant issues, that being the amplitude of the serial data stream.  I finally realized after doing a little detective work, that was the one issue that kept the previous design from solving some of the problems.  So I figured it was time for a "final" cut to address that issue, and then the request for a thru-hole version came along and I went half-way.   The one SMT part has pretty wide lead spacing, and shouldn't be a major issue to solder.

I made up a few of the buffer kits, contact me on my profile address if you'd like one or more. 

The kits come with everything you need to build one buffer board.

R2LC Buffer Kit N3

All the parts and PCB are included.

R2LC Buffer Kit N2

And an instruction sheet with a schematic on the back.

R2LC Buffer Kit N4

You should end up with something that looks like this.   If you stick to the color codes shown, it'll match the diagram and the connection table.

Attachments

Images (4)
  • R2LC Buffer Kit N2
  • mceclip0
  • R2LC Buffer Kit N3
  • R2LC Buffer Kit N4

Add Reply

Post

OGR Publishing, Inc., 1310 Eastside Centre Ct, Suite 6, Mountain Home, AR 72653
800-980-OGRR (6477)
www.ogaugerr.com

×
×
×
×
Link copied to your clipboard.
×
×