Skip to main content

RJR posted:

Question, would calibrating the unit on a Rev G or Rev H, which apparently didn't suffer from signal degradation, be satisfactory for use on a Rev L?

It makes no difference, the DCS signal is the DCS signal.  However, as George says, if the older TIU doesn't have full signal strength, that could be a problem, you don't know what results the calibration gets.

Thanks RJR, you are very kind. But without Adrian, GRJ, Stan and anyone else that helps with the designs of these things I would not be able to put any kits together on my own. These guys need the big thanks, they are the idea and circuit design folks doing all the heavy brain work. I am quite thankful for them as well.

I will admit the kits are kind of fun to put together and send out, and good experience too. Learn something every day! Plus they are educating me on much better shipping methods as well. Maybe an old dog can still learn a new trick or two? 

I know GRJ, having met him at a few Yorks.  Never met Stan and Adrian, but they, like GRJ, have always shown themselves to be true gentlemen on the forum, always ready to lend their deep knowledge to someone floundering.

I recall a few years ago when I was designing a system to use light rather than track shoes to activate operating cars, Stan provided invaluable help.

Dave, I said that half in jest.  I have no idea what those little surges would do to the tester (if anything) and I don't intend to try to find out.

I do see this device as having the potential to ascertain the source of doubletoot.  I noticed that no signal went out from my TIU when the second toot occurred.  Haven't had a chance to see what happens on the other TIU.  This does require a loco to be powered up.

With help, suggestions and even an example I thought maybe something should be included with these kits that might help with assembly. I'm including this with each kit now. For the ones that already have their kits (and for all), I thought I would post it here. (Probably too late for some.) It is really just the info already here in this thread, but now some of it is all in one place. 

Attachments

For those that don't have an oscilloscope, here's a video showing the calibration method I suggested earlier using the on-board 5V DC voltage as a reference level:

As it turns out, you can simply short the wire-leads from two adjacent components (D1 and R6).  Or as engineer-joe suggested, you could attach a momentary SPST push-button switch to these points if the circuit is mounted in a box whereby you don't have access. 

DC%2520trigger%2520calibration

Note I added a drop of white paint to the trim-pot to better show the adjustment angle.

This sets the threshold to about 10V.  You can then further refine/trim the threshold to a known-good TIU output.  Note that turning the trim-pot CW increases the threshold.  The idea is without a starting point (10V in this case), you don't know if your "known-good" TIU output is at the desired 12V (or whatever) signal level.  As Adrian previously reported, you can get all 10's on the DCS signal test with a signal level of less than 5V; but a TIU port that only puts out 5V is probably damaged.

Again, thanks to rtr12 for putting together these kits!  

gold star small

Attachments

Images (2)
  • DC%2520trigger%2520calibration
  • gold star small
Last edited by stan2004

Aha.  I see how it can be confusing.   The "red wire" is meant to be a momentary connection.  Each time you make the connection, it generates one +5V trigger pulse.

My observation after looking at the actual board is that you can make this same momentary connection by shorting (using a metal screwdriver tip in the video) two adjacent components.  The "orange line" in the schematic shows this momentary connection.  Again, each time you make the connection, it generates one +5V trigger pulse.  It is just happen-stance (dumb luck?) that the board was configured in this way such that you don't need to add a temporary wire for calibration.

Not sure if I'm being clear 

Stan, Thanks so much for the calibration update and video. That really helps things for calibration and makes it easier as well. And you are most welcome. I've actually been enjoying putting the kits together. But it wouldn't be possible without Adrian, GRJ and your help with calibration. (I'll still take the gold star though, but I'd be happy to share. )

Last edited by rtr12

Dave, I think Stan is saying that we don't need to pay attention to the arrows anymore, just follow his calibration video and all should be well. I believe the only wires in his video are going to the TIU output to power the tester at it's power inputs.

Edit: The arrows in Stan's mark up schematic would still tell you that 5vdc was present, but I believe the blue light does the same thing. 

Maybe Stan will clarify this, or add to it, he words things much better than I do and he knows for certain what's happening too!

Edit again: Sorry Stan, I didn't see your post above? I must have been posting at the same time you were?

Last edited by rtr12

I tried Stan's calibration method, but I wasn't totally happy with the outcome, so I dropped back to my 'scope method.

I start out with a good signal and make sure both LED's are blinking.  This signal is measured with a separate hi-pass filter right off the output of the TIU.

TIU Full Voltage Output

Next, I put 3.3uf of capacitance across the TIU output pins and fire off another command.  In this case, I "tune" for just a couple winks of the green LED, most of the time it's off.  I consider this signal marginal.

TIU Attenuated Output 10V

Finally, I put 5.0uf of capacitance across the TIU outputs and make sure I never see the green LED.  At this voltage, I figure the TIU has lost at least one of the ACT244 gates.

TIU Attenuated Output 9.28V

I then marked the pot and support with a marker line so it can be returned to this setting without a total calibration again.

Attachments

Images (3)
  • TIU Full Voltage Output
  • TIU Attenuated Output 10V
  • TIU Attenuated Output 9.28V
stan2004 posted:

For those that don't have an oscilloscope, here's a video showing the calibration method I suggested earlier using the on-board 5V DC voltage as a reference level:

As it turns out, you can simply short the wire-leads from two adjacent components (D1 and R6).  Or as engineer-joe suggested, you could attach a momentary SPST push-button switch to these points if the circuit is mounted in a box whereby you don't have access. 

DC%2520trigger%2520calibration

Note I added a drop of white paint to the trim-pot to better show the adjustment angle.

This sets the threshold to about 10V.  You can then further refine/trim the threshold to a known-good TIU output.  Note that turning the trim-pot CW increases the threshold.  The idea is without a starting point (10V in this case), you don't know if your "known-good" TIU output is at the desired 12V (or whatever) signal level.  As Adrian previously reported, you can get all 10's on the DCS signal test with a signal level of less than 5V; but a TIU port that only puts out 5V is probably damaged.

Again, thanks to rtr12 for putting together these kits!  

gold star small

Dahhhhhhhhhhhh…… I did good? 

I still haven't actually used or even calibrated my device fully again. I'm doing other things and I'll get back to this when I have the extra patience. I got frustrated that my outdoor RR keeps going south. I'm guessing that it's a wiring issue that the ends are degrading. Maybe the TIU is weak? I haven't started full trouble shooting. I feel that this device will help me in the coming years to save some time.

I thought having a switch built inside my device's project box would help me calibrate it. Knowing that Stan posted calibration should last, I could have done it better before mounting it permanently. I had my O scale layout connected to the TIU's output so I don't know if the unit was calibrated correctly.

stan2004 posted:
rtr12 posted:

...I am going to ask Stan if he minds me including his instructions and his picture showing component orientation in with the parts list.

Anything you think would help is fine by me.

swise posted:

… "AC Track voltage" means hook 18VAC up to the test board?

Yup.

 

 

Stan, when you did your Calibration Video what power did you connect the tester to?

While calibrating with your video my tester was connected to an 18V AC TIU output.  Was that correct?  I'm getting all reds or nothing on my outputs.  My blue Power LED works fine.

Correct.  Apply 18V AC TIU output.  The 18V AC should be from a train transformer - not the chopped-sinewave variable output from a conventional controller.

Not sure what you mean by "all reds".  There should be one red and one green LED.  If you are hooking up to a TIU output that has just been powered you may get some LED flashing as the TIU channel sprays the watchdog DCS packet for tens of seconds.  But after this initial activity, the red and green LED should be OFF.  Or, for calibration purposes, it might be simpler to just hook the tester to 18V AC (without the TIU present).

With board powered (blue LED on), referring to 9/21 video:

- rotate trimpot full CCW.  Tap (bridge/short together) the two indicated points on the board using a metal screwdriver.  Both LEDs should flash on each tap.

- rotate trimpot full CW.  Tap the two indicated points.  Neither LED should flash.

- if you don't get above two behaviors, then there's a troubleshooting issue - something is not right.

- otherwise back off the trimpot from full CW to the point where just the red LED starts flashing on each tap.

 

Thank you, Stan.  I re-calibrated with 18 v from postwar ZW.  (with about the same results).

While testing a new Rev L TIU, all 4 OUPUTS (1 at a time) Blue On and at the same time the Blu and Red fast blip once and go off.  Does that mean very good signal or NO signal?

I posted test results toward the bottom of this topic, Post for Use and Interpretation of TIU Tester Results, for 4 TIUs.  I didn't get any green red combination results, just red.  I'm puzzled why I don't see any red green combos.

 

Last edited by Susan Deats

Susan,

To properly use the tool:

   Turn on speed Mode in the remote.

   Do not have an engine on the track.

   Connect the test tool to a TIU channel.

   Put power to a track on the TIU’s channel.

   Start up an engine in the Active Engine List (that isn’t on the tracks).

Send commands.

Monitor the tool’s responses.

stan2004 posted:

Correct.  Apply 18V AC TIU output.  The 18V AC should be from a train transformer - not the chopped-sinewave variable output from a conventional controller.

Not sure what you mean by "all reds".  There should be one red and one green LED.  If you are hooking up to a TIU output that has just been powered you may get some LED flashing as the TIU channel sprays the watchdog DCS packet for tens of seconds.  But after this initial activity, the red and green LED should be OFF.  Or, for calibration purposes, it might be simpler to just hook the tester to 18V AC (without the TIU present).

With board powered (blue LED on), referring to 9/21 video:

- rotate trimpot full CCW.  Tap (bridge/short together) the two indicated points on the board using a metal screwdriver.  Both LEDs should flash on each tap.

- rotate trimpot full CW.  Tap the two indicated points.  Neither LED should flash.

- if you don't get above two behaviors, then there's a troubleshooting issue - something is not right.

- otherwise back off the trimpot from full CW to the point where just the red LED starts flashing on each tap.

 

My tester required me to back off the trimpot a bit more until the red and green LEDs did 2 long flashes when testing my best TIU output (sometimes I cheat).

Now I have red with green for very good and red for okay.  I will retest all five TIUs to verify.  My newest TIU has absolutely no signal outputs and I have a Return Authorization from MTH.  It also has the tiny TVSs.

gunrunnerjohn posted:

Perhaps, I tested a known good TIU (checked with my 'scope), and the variable channels (in fixed mode) tested just the same as the fixed channels.

Here's a test of the board using the Lionel 1033 to power the TIU, tester is connected to Variable #1 (in fixed voltage mode) to illustrate the variable channels work the same as the fixed channels.

During your tests, with your tester connected to the TIU variable output, what was your Remote "talking" to when you activated the whistle and bell?

Susan, the remote is only talking to the TIU.  The tester should be used with ONLY the TIU connected, no track, no engines, etc.  Also, you need a pure sine wave transformer to power the TIU as the sawtooth from most electronic transformers will screw things up.

I put the remote into SPEED mode so that it doesn't stall waiting on the engine when I press keys.

gunrunnerjohn posted:

Susan, the remote is only talking to the TIU.  The tester should be used with ONLY the TIU connected, no track, no engines, etc.  Also, you need a pure sine wave transformer to power the TIU as the sawtooth from most electronic transformers will screw things up.

I put the remote into SPEED mode so that it doesn't stall waiting on the engine when I press keys.

What engine?  When I bench tested I connected the tester to the Output and just tested for the watchdog signal.

I have an engine loaded, it can be any engine.  If you're in SPEED mode, you can do stuff like STARTUP (does a nice long string), WHISTLE, BELL, etc.  You are addressing the engine, but since there's no track or engines connected, obviously it doesn't answer.  We're just looking at the outgoing signal strength here.

Didn't you see my video of my testing?  This shows you what I'm doing.  The extra board on the output of the TIU is my filter for the 'scope, it doesn't affect what the test board is seeing.  This test was on variable #1 in fixed mode with a 1033 transformer powering the TIU.  Note that the transformer voltage isn't significant as long as it's enough for the TIU to run, it can be anything from around 12 volts and up.

Last edited by gunrunnerjohn

Add Reply

Post
The DCS Forum is sponsored by

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.
×
×