Revised Two or Four Port TIU Output Signal Indicator Boards Are Ready To Go

Here are pdf versions of the User Notes and the version notes.

BTW I just checked and Digikey currently show stock on the CD74HCT123M SOIC chip.

Rod

In response to a couple of questions I just want to add the following comments. You will likely find that the boards if ordered through OSHPark will run you about $3.75 a piece. I purchased a good size lot from offshore and even with express shipping the all up cost is considerably less. The boards I have built so far all work well.

I will also do a quick point to point conductivity test of installed chips to provide some assurance for a successful build.

Rod

Refresh me what does this does?  It works on an operating layout or just test a TIU output on a bench?  G

@GGG posted:

Refresh me what does this does?  It works on an operating layout or just test a TIU output on a bench?  G

Both! The basic idea was to install them permanently right inside the TIU. Once installed and on the layout they continuously monitor any data packets sent out by the TIU. On the bench they give unmistakable evidence of TIU data output once the TIU and channels are powered up. If any TIU channel does not show output when any handheld command is issued, it should be considered suspect and checked further using a scope. Hope that helps!

Rod

@Rod Stewart posted:

Once installed and on the layout they continuously monitor any data packets sent out or received by the TIU.

Small correction, they don't monitor data packets received, only transmitted.

Is this different than the other version made? That one does not work for me with a lighted lock on and I assume for a layout.  It only measure transmit strength of TIU in a stand alone mode.  Which is what I need for test repair.  G

@gunrunnerjohn posted:

Small correction, they don't monitor data packets received, only transmitted.

Distinction noted John; thanks for that!

@GGG posted:

Is this different than the other version made? That one does not work for me with a lighted lock on and I assume for a layout.  It only measure transmit strength of TIU in a stand alone mode.  Which is what I need for test repair.  G

By "the other version" George I assume you mean the portable TIU tester designed by Adrian, and sold in kit form by rtr12? The boards I have are actually adapted from that same design. The big difference is that instead of a red and a green led; one to indicate high signal strength and the other medium strength; these boards use two separate circuits to monitor two TIU channels per board. There is a trim pot for each channel that can be used to adjust signal sensitivity. With a setting of about 3/4 turn CW, I find they trigger at a DCS signal strength of about 10-12 volts, which is a reasonable value to expect for any healthy TIU channel.

Rod

Right.  But my experience with previous version which I use to test TIUs is that if I connect to short test track with Lighted Lock on, it does not work.  I thought there was discussion of it not working on a layout.  So is your circuit design different?  Can it measure properly with all the inductive and capacitive loads on a layout?  G

George, I believe his detection circuit is the same as the one you have except he only has two poles on the filter instead of four.  Note that this circuit only works for a pure sine wave, a chopped waveform indicates a signal all the time due to the harmonics of the waveform getting through the filter.

Ok, my test bench is Z-4000 with 5ft track with one lighted lockon.  Even when I use just my bench test set with a circuit board attached to the test set fed by TIU and the tester clipped onto the TIU output it does not work.  Now maybe a short section is no good and a large inductive layout and it works, but I did not think this tester worked in an operational environment, only when measuring TIU output stand alone.  G

Well, I know it doesn't work on a layout of most any size, I honestly have never tried it with my test track.  I use mine strictly to do a quick evaluation of the TIU outputs, for that it's worked fine.

Ditto, so how does this work if you permanently attach to a TIU used on layout?  Rod implies it does.  Hence my questions. G

Well, curiosity got me and I tried it.  It gets a weak signal on all four channels if my tracks are connected to the TIU outputs.  It also doesn't see the DCS-RC WD generator outputs at all, I should be seeing a signal ever second.  I get the "weak" LED on all four channels.  If I unplug the tracks and test just the output, I get the full signal indication with the red and green LED's.  I didn't try with engine(s) on the tracks, I'm sure that would further attenuate the signals.

So attached is my original circuit as I revised it (from Adrian's design) to build these boards. Adrian's original is one 4 stage AC filter circuit driving a 2 resistor splitter circuit powering a red and green led, each through one channel of the HCT123. The splitter circuit produces a high and a low sensitivity output, to drive each led. My revision uses two separate AC filter sections each followed by an adjustable 2K pot, each driving a separate channel of the HCT123, each driving only one led. The pot enables setting the sensitivity to get most DCS signals. I find about 3/4 CW seems about right; this results in about 250 ohms between pins 2-3 of the pot. I have not tried anything higher than this to be honest because it seems to work well.  And I went with only 2 stages of AC filtering because when I scoped my first (of Adrian's) dcs testers I noted there was no AC noise visible after 2 or 3 or 4 stages. And I wanted to save some costs as well as board space so as to get it to fit inside a TIU.

Below are a couple of videos of the resulting boards in action. One was taken at our local Xmas trainshow just last month on a 12 x 16 layout, the two outermost track loops. It represents real live layout operating conditions, albeit not on a very big layout. The second is a test of two built up boards connected up to a Rev I3 TIU which is what I have here in AZ. It shows the TIU startup burst followed by the 2 second watchdog sequence. This is about what you would see with these boards installed in a TIU. The later Rev L TIU should produce stronger DCS signals, and the pots may have to be turned down a little. I have not tested these out on a Rev L yet.

I guess each one of us has to decide whether such a device is something we are interested in. For me being a gadget nut it was a no-brainer. Plus I wanted something that could provide a fast indication showing that each TIU channel was sending out DCS packets, and to me this serves the purpose, kind of like the red led on a TMCC base station. I fully realize that this won't have any appeal to lots of folks, but to those interested I offer the boards and chips for what I think is a fair cost. And for those who want to do it all themselves I have posted the gerber files for the latest board rev which (uses an smd chip).

Rod

Here is a short video showing 4 port activity with some random dcs commands.

So any TIU output that does not show similar indications would indicate possible problems.

Rod

I just posted a revised pdf copy of "TIU 2 or 4 Port Signal Indicator User Notes R1a" in the OP at the top of this topic. It is designated R1a, and contains some new ideas and suggestions, as I have come to know more about these boards. Any questions or comments please don't hesitate.

Rod

It has come to my attention that there is a bit of a mess up with the resistors listed in my BOM. The PN’s listed in the BOM are actually ½ watt resistors, whereas all that is needed is smaller ¼ watt.

This would not be a big deal, but the ½ watt bodies are 9mm long which is kind of big for the board patterns, though they could be used.

Here are the Digikey PN’s for the more appropriate ¼ watt versions:

They are all in stock at Digikey and costs are $.10 each for qty 1, $.067 each for qty 10, or $.046 each for qty 25. My apologies for this mess up. (I copied the BOM over from a previous project and failed to notice the bigger resistors. My bad.)

I have replaced the original BOM in the OP with a corrected version R1. The revised PN's are highlighted blue.

Rod

A couple of folks have had trouble finding a few components, since domestic sources are stocking less through-hole stuff now, and more smd. I can offer a "mini-kit" which includes the troublesome items, as follows:

Mini-Kit (unassembled)

1 x 2 port PCB; R1.23a

1 x U2; CD74HCT123M SOIC16 package

1 x U1; LM78L05 TO-92 Vreg

2 x R3/R4; 3362P 2K 1/2W trim pots

1 x L1; 22uH 290ma, 22uH axial lead inductor

Package Price: $5.75      Add $1.00 to have the U2 chip pre-soldered in and tested.

All you need on top of this is about $2 or $3 dollars worth of inexpensive readily available caps, resistors and diodes to finish the project per the BOM. Please limit your order to a max of 4 mini-kit boards, as I don't have a limitless supply of the components on hand.

The original base board/chip pricing is still in effect per the OP, as follows:

Bare board only:                             $1.60 each

Board with SOIC chip loose          $2.50 each

Board with chip installed              $3.50 each

Rod

Just for information if your TIU is mounted in a difficult to see location such as under the layout, these boards can be remote mounted in something like the example project cases shown below. They are 80 x 46 x 17 mm inside dimensions, and will just nicely fit one or two indicator boards. These can then be mounted on your control panel, facia, or wherever is handy.

They come in lots of 10 and my cost including shipping to AZ was about $21.

Rod

So for anyone out there wondering what these would look like mounted in the project boxes mentioned above, here is a pic and a video of the first test hooked up to all 4 ports of my bench TIU; just showing the watchdog signal. Next step; putting it on the layout. These project boxes are nice and small and nicely accomodate two test boards. The setup below has the two variable ports at the top of the box, and the two fixed ports at about the mid-point. They fit really well.

Sorry the darn pic got rotated 90 deg CW. The wires come out the top. I wish there was some way in OGR to rotate a pic to the correct orientation for when this sort of thing happens!

Rod