GRJ, that SEEED pricing is quite good! I'd have to go back and look but I think you posted that one of the later PCB designs at OSHPark was $13 and something for 3. That is a big reduction, 30 PCBs for $13.   I doubt that I will ever be needing that many, but it's good to know anyway.

The Asian ones I got were about the size of your lighting boards in the pic above. I think it was the small DC-DC regulators. I can see where combining the PCBs like that would make things much easier to create and populate. 

Soldering iron is sitting in it's stand idling and ready to go! Now, I just have to figure out how to set them up after making them...I think Adrian or maybe it was you posted something earlier about that. Maybe a thread re-read is in order soon. 

The setup is pretty simple, you have one adjustment.  I'll be doing mine with a signal generator and a 'scope.  However, failing that equipment, connect to a good TIU channel all set up to get,and then "tune" the pot to turn on both lights.  You'll probably want a "tiny" bit more toward the lights on to compensate for small variations in the outputs, but not enough to let a bad TIU pass as good.

I was hoping to hear from Adrian and Stan on the "final" version to see if we've missed anything before I go to press on a bunch of PCB's.

Thanks, I think I can handle that. I have one TIU channel (Var 2) that has never been used, so that should be a good one, I hope.

I imagine Adrian and Stan will be along before too long. It's always good to have them (as Stan described the process above) look at things before production. I have noticed it's really difficult to get something past Stan! (He must get up really early! ) Also, I really like all your cooperative design projects, very interesting and informative to follow along with. Especially for a rookie greenhorn like me!

rtr12 posted:

GRJ, that SEEED pricing is quite good! I'd have to go back and look but I think you posted that one of the later PCB designs at OSHPark was $13 and something for 3. That is a big reduction, 30 PCBs for $13.   I doubt that I will ever be needing that many, but it's good to know anyway.

Well, it wasn't quite that good a deal.   Turns out, they take exception to multiple designs on one panel, so I couldn't do my trick of having three on one "board".  Also, since ten of these boards exceed the maximum first class shipping weight, the freight is more as well.

Still a lot better deal than the $13 for three of them.

Attachments

Photos (1)

The latest incarnation hasn't wandered too far from Adrian's original design.  As I look at the latest schematic I believe you can even implement his exact original circuit by selectively populating the board with the original parts.  The modified version does add a half-dozen or so line-items but it still looks to me like $10 or so in parts.

Some of the components are needed in quantity 1 whereas even at DigiKey you can save quite a bit per unit by going to quantity 10.

MFG_PRT-12978_sml

Adrian shows a red-black pair of alligator-clips to attach to the TIU output jacks.  Obviously they don't have to be red and black, but try finding "just" 2 alligator-clips leads.  Or maybe everyone already has these clips lying around.  At DigiKey it looks like you buy a set of 10 double-ended cables for about $3... of which you only need half of a red and half of a black.

https://www.digikey.com/produc...1568-1582-ND/7056460

My guess is most guys would prefer a built-and-tested version...rather than firing-up the soldering iron, ordering DigiKey components, ordering bare circuit board(s) from OSH Park / SEEED, etc..   I might have missed it in the conversation but has anyone stepped up to providing completed units...or if not then at least a kit of parts?  

 

Attachments

Photos (1)

Are you suggesting alligator clips just to keep the cost lower?

Since the TIU has banana jacks built in, and this project is intended to just test at the TIU itself, I'd be inclined to put banana plugs on the end intended to mate to the TIU posts.  Just the plugs seem to be about a buck a piece.  Using a pre made test lead would raise the cost of the project significantly though, I guess (well, maybe $5 per lead - not expensive, but double the $10 we started with )

-Dave

stan2004 posted:
My guess is most guys would prefer a built-and-tested version...rather than firing-up the soldering iron, ordering DigiKey components, ordering bare circuit board(s) from OSH Park / SEEED, etc..   I might have missed it in the conversation but has anyone stepped up to providing completed units...or if not then at least a kit of parts? 

You're right, most of the folks probably do want a completed unit.  That was my experience with several other projects, I started them out to be kits, and almost everyone wanted assembled and tested.  I may consider doing a kit of parts, but I'm not willing this time to have to put it together.  Perhaps we can recruit someone else to build them.   They're fairly easy as they're all thru-hole.

I'll probably buy ten sets of parts for the ten boards I'm getting.  Actually, I really only need the HCT123's, and the 2K pot, I think I have all the resistors, caps, and diodes in quantity anyway.  I know I have a ton of 22uh chokes left over from a quantity buy for another project.  LED's?  Tons of them.

FWIW, I agree with the dual banana plug, or at least individual banana plugs.  No sense in using alligator clips, they're not nearly as easy to use.  Now that it's just one connection to the TIU channel, it'll be a quick PnP test setup. 

I think the alligator clip leads are cheaper than a couple of good banana plugs by themselves. I think Stan's linked set of alligator clips are less than a pair of the banana plugs I have been using, which are heavy duty and fit tightly in the jacks. They don't loosen up like many of them do these days.

I have alligator jumpers, and was originally thinking of just cutting a red/black pair in half as Stan says above. The banana plugs sound like a good idea too (hadn't thought of that) and I also have plenty of those on hand.  

The thing about alligator clips is the small one Stan illustrates don't clip will the the posts on the TIU.  I have a bunch of similar ones, and they don't work well with that style of post, they keep popping off.

GRJ, let me see how I do on mine? If all goes well, I might consider building some if there are not too many orders for them. I like the kit idea myself, but my direction writing would probably be a major disaster. They would all be throwing shoes at me!

Maybe a small quantity of assembled units and a kit option would work? Hopefully more kits than assemblies.  

gunrunnerjohn posted:

Track power MUST be provided to the channel under test or it won't generate the DCS signal.  If you're seeing the DCS signal on a TIU output, that channel is powered.

Actually if you have voltage on the aux power in, and no channel power the packet still comes out... I figured this out like a day ago.

gunrunnerjohn posted:
stan2004 posted:
My guess is most guys would prefer a built-and-tested version...rather than firing-up the soldering iron, ordering DigiKey components, ordering bare circuit board(s) from OSH Park / SEEED, etc..   I might have missed it in the conversation but has anyone stepped up to providing completed units...or if not then at least a kit of parts? 

You're right, most of the folks probably do want a completed unit.  That was my experience with several other projects, I started them out to be kits, and almost everyone wanted assembled and tested.  I may consider doing a kit of parts, but I'm not willing this time to have to put it together.  Perhaps we can recruit someone else to build them.   They're fairly easy as they're all thru-hole.

I'll probably buy ten sets of parts for the ten boards I'm getting.  Actually, I really only need the HCT123's, and the 2K pot, I think I have all the resistors, caps, and diodes in quantity anyway.  I know I have a ton of 22uh chokes left over from a quantity buy for another project.  LED's?  Tons of them.

FWIW, I agree with the dual banana plug, or at least individual banana plugs.  No sense in using alligator clips, they're not nearly as easy to use.  Now that it's just one connection to the TIU channel, it'll be a quick PnP test setup. 

Ive updated the title of the forum thread accordingly based on all the design progress...

gunrunnerjohn posted:

The thing about alligator clips is the small one Stan illustrates don't clip will the the posts on the TIU.  I have a bunch of similar ones, and they don't work well with that style of post, they keep popping off.

Well, for less than $1 shipped (from Asia) you can choose your poison!    This pair has bananas on one side and alligators on the other.

alligator conundrum

I have no problems making a firm, solid contact using the alligator style to the TIU jacks.  Additionally, though not the intended application, you can easily attach alligators to the track which would be quite the hat-trick with banana plugs.  

The key is the strain-relief hole to loop the red/black wires thru so that the soldered-wire connections don't fatigue/fail over time.

And if you don't mind waiting for shipping from Asia, what appears to be the same set of 10 double-ended alligator cables is available for about 1/2 the price on eBay.

 

Attachments

Photos (1)
Adrian! posted:
gunrunnerjohn posted:

Track power MUST be provided to the channel under test or it won't generate the DCS signal.  If you're seeing the DCS signal on a TIU output, that channel is powered.

Actually if you have voltage on the aux power in, and no channel power the packet still comes out... I figured this out like a day ago.

Isn't that the definition of passive mode?

Chris Lord posted:
Adrian! posted:
gunrunnerjohn posted:

Track power MUST be provided to the channel under test or it won't generate the DCS signal.  If you're seeing the DCS signal on a TIU output, that channel is powered.

Actually if you have voltage on the aux power in, and no channel power the packet still comes out... I figured this out like a day ago.

Isn't that the definition of passive mode?

Sort of.  The channel will output a WD signal only when power is first applied to a channel, and not again until a power cycle on that channel.  Truthfully, I never checked to see if the signal was coming out with just aux power, but it's an interesting tidbit.

FWIW, the tester should work exactly the same, connect to the port and you'll see a signal.  The filter allows the DCS signal to pass, so it's pretty much transparent whether there is 60hz to filter or not.

Obviously banana jacks can't connect to the track.   Considering the discussion that resulted when that topic came up a few times earlier in the thread, I considered that a benefit rather than a hindrance. 

If one can't easily attach the probes to the track, one is less likely to try to use it that way and then wonder why the results aren't what's expected.

I used regular Digikey prices when suggesting what the jacks would add.  I didn't try to go to eBay for direct from Asia pricing or anything like that.

-Dave

You'll note that all my BOM stuff is Digikey, but I don't get all the stuff from them.  I typically buy some stuff in bulk from other on-line sources, eBay, AliExpress, etc.  The nice thing about specifying Digikey components is you can get a full spec sheet and know what to substitute from other sources if you desire.

John:

Like many others, I have been watching this thread and enjoying the discussion.  Please count me in for a kit, bare board, completed unit or whatever you decide to make available.  Setup for testing of DCS signal output on customer TIU's has been time consuming and inconvenient.

Thanks to all the electronics gurus for all the time you put in here.

Pat Marinari

 

Northern Central High Railers

The long wait is over, I finally got the 10 boards from SEEED Studio.

Here's the tester after I put it together, one small glitch in the layout, I got the pot backwards.  Didn't have one to look at and thought the legs went the other way.  Still works, you just have to adjust it from across the board.

I connected it to a TIU and fired it up.  It calibrated just fine, so I think it's probably doing it's intended purpose.  I have to do some more testing with various signal levels to see how it reacts, but it seems to function as expected.  I didn't have the 330 ohm resistors for the LED's, but it's perfectly happy with 470 ohm resistors.  The large leads go to the TIU, the board is powered through the TIU, so that's the only connection required.

Attachments

Photos (2)

Slow down guys.  

My contribution to this effort was to lay out the board and build a prototype for testing.  I don't have the time right now with all my other projects to actually produce these.  However, I will make the board layout files, schematic, and BOM available when I get them all gathered up in a coherent form.  I'll also make the DIPTRACE files available in case anyone wants to "improve" the design.   You can work on this board with the free version of DIPTRACE, so no cost to tinker.  All the parts are commonly available, so there should be no issue in putting these together.

Perhaps someone else would like to pick up the baton and build some of these?

I like the red boards, yours looks neat and works. Much more colorful too!

I got my order from OSHPark, the v1.0 boards, assembled one of them and so far no worky? I haven't tried to see what's wrong yet, but upon examination nothing appears to be out of place? Or maybe my TIU is not performing properly, but I have only tried one channel so far.

I did manage to goof up my capacitor order, I got some funny looking tantalums instead of the ceramic I thought I had selected? Oh well, I still have 2 more boards to play with and I do have some ceramic caps around here somewhere. 

I still want one of your current versions though, and I'll be watching for the updated Diptrace files to order from.

Thanks again for doing all this GRJ, it's really a big help for us (me anyway) still trying to learn! 

 

Edit: If I figure out how to make one work and then am able to test them somehow, I would be willing to try and make a few for some of the other forum members.

I'll turn the pot around in the design files and then post the whole package.  They are easy to assemble, they're all thru-hole parts.  If you pick up the parts in quantity, they're also pretty cheap, even from Digikey.

Here are the design files for the TIU Signal Tester.  Included are the Gerber files for PCB fabrication, the Diptrace files to tinker with the design, and the BOM for ordering parts.  There are several other ZIP files in the overall ZIP file, the Diptrace files and the Gerber files are all separated out.  I did flip the pot around in this design as it was backwards in my original boards.

TIU Signal Tester Design Files.zip

You don't have to use the exact parts I used, they're all sourced at Digikey.  Most of the parts are very generic, and for any kind of quantity, you can probably find cheap eBay parts.  However, the BOM does give you industry standard parts and specs so you know what to look for.  Even sourcing them from Digikey, they'd be pretty cheap.  Stuff like the 2K pot will probably want to be the exact part as it fits the PCB footprint.  Have fun!

Attachments

Yes, thanks GRJ!  Think I'll try out the SEEED place this time and see what happens. I take it that is where you got the red ones? I like those! 

Digikey is fine by me for parts, I like them too. Good place and fast shipping.

Add Reply

Post
The DCS Forum is sponsored by
OGR Publishing, Inc., 1310 Eastside Centre Ct, Suite 6, Mountain Home, AR 72653
330-757-3020

www.ogaugerr.com
×
×
×
×
×