Design of a $10-20 DCS-TIU Port Tester Tool?

$400 - Real Oscilloscope 

$90 - USB Oscilloscope like the T-train

$40 - GRJ with 4 LEDs with fixed Rs and comparators

$10 - Adrian with 1 LED and trimpot tuning

I think the market is pretty much flooded at this point?

Other participants in a typical corporate design-review would be someone from Marketing and someone from Manuals.  It's maybe just my experience but these guys were usually silent (or maybe asleep?) but in this case might suggest that you need a way to turn the widget on and off - and if using a 2-pin non-polarized connector, what happens if you plug in the battery backwards?

Here are two 9V battery-holder connectors suitable for PCB use.  The connectors themselves "guarantee" correct polarity on a 9V battery.  I suppose the on-off "switch" could be the act of inserting the battery.  Otherwise, the right example right shows a cheap (10 cent) on-off slide switch.

Adrian! posted:

$400 - Real Oscilloscope 

$90 - USB Oscilloscope like the T-train

$40 - GRJ with 4 LEDs with fixed Rs and comparators

$10 - Adrian with 1 LED and trimpot tuning

I think the market is pretty much flooded at this point?

Agreed.  Just suggesting that if the journey is to learn DipTrace and the process of fabricating PCB's via OshPark, it's always useful to bring something to the party given GRJ has already done the heavy-lifting on the $10 version.  So this would be, say, 

$11 - Adrian with 2 LED and trimpot tuning

I figured out the issue with the trigger of the 123 after I posted it, which made sense why Adrian has the pot.  So, to use more 123's, you'd need more pots, that would be ugly!  I guess the comparator is probably the way to go if you wanted to have multiple LED's.

I thought of a power switch and also maybe an input cap and diode so you could run this on AC or DC and from a plug-in power supply.

I think I’ll keep mine simple.  Go - No Go.  Although I didn’t order my components or board yet!  LOL

Adrian's design, the one I posted the build files for, will be perfectly serviceable for the simple go-no go test.

Yes John, that is what I thought.  I’ll review and order tomorrow.  It has been a very interesting and informative discussion!!  Many thanks to Adrian, John, Stan, George, and RTR

Does anyone have a list of the part numbers that are needed from Digi-Key?

I don't have an "official" list of parts as most of them are all very generic parts.  All the resistors are 1/4w thru-hole resistors, the caps are all 50V ceramic.  The 74HC123N is the standard thru-hole 16 pin part, any maker is fine.  The LED is just a standard 5mm or 3mm LED in the color of your choice.  The regulator is the LM340LAZ-5.0 or similar three terminal 100ma regulator.  The pot is the only "special" item, it's the Bourns 3306W-1-202.

GRJ, what is the lead-spacing of the caps?  It looks like 0.25" or 0.3"?  I realize it's just a place-holder but the 3-D rendering shows a radial package which tend to be expensive and harder to find with such large lead spacing.

Point being if you go to DigiKey and search for 0.1uF 50V ceramic capacitor,  you are overwhelmed with choices.  While terminal wires can be manipulated to fit in whatever hole spacing, I think some guys would rather just be given a specific part number to use.

So another participant in a corporate Design-Review process was a representative from Purchasing who, rather than being asleep, was always chomping at the bit to nail down exact part numbers for good reason.

To that end, note that Adrian in his early post suggests he used the HCT version of the 123 chip:

http://www.newark.com/texas-in...ip16-5-5v/dp/25M8579

The HCT version has a lower threshold trigger voltage than the HC version.  It probably doesn't make a difference in this application if using a trimmer to tune the voltage, but if you're after apples-to-apples, the DigiKey version would be:

https://www.digikey.com/produc.../296-2088-5-ND/38252

If anyone is putting together an "official" DigiKey parts list, 

the 1N4148 diode could be:

https://www.digikey.com/produc...8/1N4148FS-ND/458603

And the least expensive flying-lead 9V battery-holder appears to be:

https://www.digikey.com/produc...232/36-232-ND/303804

1/4W 10K resistor could be:

https://www.digikey.com/produc...4JT10K0CT-ND/1830374

Other resistor values also available in the CF14.... "family" of 1/4W resistors at same price (4 cents each at qty 10).

330 ohms

https://www.digikey.com/produc...4JT330RCT-ND/1830338

100K ohms

https://www.digikey.com/produc...4JT100KCT-ND/1830399

Good points Stan.  I actually did miss that Adrian suggested the HCT part, good idea.

In looking at the HCT part, it's current outputs are pretty low, maybe a larger resistor on that LED...

Lead spacing for the resistors is .5"

Lead spacing for the caps is .2"

Lead spacing for the diode is .4"

Here's the board with what should be all the dimensions you might need, including the connectors, the pot, etc.

Mark Boyce posted:

Yes John, that is what I thought.  I’ll review and order tomorrow.  It has been a very interesting and informative discussion!!  Many thanks to Adrian, John, Stan, George, and RTR

I certainly agree Mark! A big thanks to all the designers (RTR not included )!

I'm making the go-no-go one as well, but may do some more fiddling with some of the other suggestions above. The parts I ordered were just about the same as what Stan posted so I think I got that part ok. We will see when the boards get here.

Got it.  So with a 0.2" lead spacing, the 0.1uF cap could be:

https://www.digikey.com/produc...R/478-3188-ND/936830

rtr12 posted:

Mark Boyce posted:

Yes John, that is what I thought.  I’ll review and order tomorrow.  It has been a very interesting and informative discussion!!  Many thanks to Adrian, John, Stan, George, and RTR

I certainly agree Mark! A big thanks to all the designers (RTR not included )!

I'm making the go-no-go one as well, but may do some more fiddling with some of the other suggestions above. The parts I ordered were just about the same as what Stan posted so I think I got that part ok. We will see when the boards get here.

You are the user who asks the important questions to get the designers to give more detail on the points they took for granted, so to speak.  This is coming from the documentation guy who writes the first step; "Plug power cord into 120v AC outlet"  

When I'm doing an "official" design, I have to come up with a detailed BOM with actual manufacturers and part numbers, so normally all this information is available at the press of a couple of keys.  For something like this, I tend to just pick generic component outlines so most any part will suffice.

gunrunnerjohn posted:

...

In looking at the HCT part, it's current outputs are pretty low, maybe a larger resistor on that LED...

Well now we're entering the eyes-glazing-over zone, but the 330 resistor is no problem in this application.  The 4 mA output current characteristic for the 74HC and 74HCT family is simply a point on a curve.

For example, with a 5V chip supply voltage, the output pin driving the LED can supply 10 mA if you don't mind the output voltage dropping to 4.4V.  Or you can get 20 mA if only 3.7V is suitable.  There are considerations and equations that get messy when you push the envelope but for this application 330 ohm resistor is not asking for anything untoward.

rtr12 posted:

...

I'm making the go-no-go one as well, but may do some more fiddling with some of the other suggestions above. The parts I ordered were just about the same as what Stan posted so I think I got that part ok. We will see when the boards get here.

So as part of the learning moment are you considering modifying the schematic and the actual PCB layout?

If so, I will describe my 2nd LED idea which would "cost" you 4 resistors, 1 capacitor, and the additional LED...but using existing values so no "new" values to procure.  

Stan, if you have a "second LED" idea that's functional, I'd be glad to add it to the existing design.  The good thing is that I haven't ordered any of the boards yet.

stan2004 posted:

 Well now we're entering the eyes-glazing-over zone, but the 330 resistor is no problem in this application.  The 4 mA output current characteristic for the 74HC and 74HCT family is simply a point on a curve,

 I guess I should have actually downloaded the data sheet, I just saw that on the Digikey site and it caught my eye.

Well, I realize this is a situation where size matters since you pay OSH Park by the sq. inch.  The additional cost for components is 25 cents or so.

Add a 10k-10k divider that generates a 2nd trigger signal that is 1/2 the amplitude of the original trigger.  Add a 100K-0.1uF combo to set the blink pulse duration of the 2nd timer.  Add a resistor-LED for the 2nd LED.

The idea is as follows.  You have two trigger voltages, one being 1/2 of the other.  The 123 chip will trigger when the voltage at the "B" input is about 2.5V.  We'll call this the full-scale trigger for a good TIU output.  The other trigger would then be about double that or 5V which of course will trigger the other half of the 123.  

But if the TIU level drops, the lower trigger will fall 2.4V, 2.3V, 2.2V, etc. while the higher trigger will fall to 4.8V, 4.6V, 4.4V, etc.  So in this case only one LED will blink.

So rather than a go, no-go the creeping feature syndrome makes this a go , maybe, no-go. 

And from the peanut-gallery, if the Gerbers are still pending, how about a hole to make a strain-relief loop for the battery holder wires.  I'm not clear if you're advocating the 2-pin male-female connector (looks non-polarized which is iffy IMO) as the on-off mechanism.  But if running the red-black battery connector wires soldered right to the pads (no connectors), looping the wires thru a hole is a cheap "free" strain-relief to prevent the wires from breaking at the solder joints.

Looks doable Stan, I'll work something up.  I actually order stuff like this from SEEED Studio if I want a few, ten boards of any reasonable size are around $12.70 shipped, and it takes about a month to get them.  I just got my TIU Buffer boards for the surface mount version back, and I ordered them just about a month ago.  The restrictions are they are 1/16" DS boards, no options, and quantity 10.  However, they're two sided boards with silk screen and solder mask both sides, so they're nice boards.

OSH Park is great for little modules, but once the boards get more than about 1 sq/in, I look for other options.  I like the little boards from OSH Park as I can get them in 1/32" and with heavier copper so I can run narrower traces and still have decent current capability.

I can get boards like this from China about $40 for 100 pieces if I want a lot of them.

I updated my Digikey cart for the "go , maybe, no-go" variety!  You twisted my arm!!  

Stan,

Yes, I was planning to do some more fiddling and would be happy to add your ideas, but GRJ will probably have it done and posted by the time I finish writing this...  GRJ's design will probably be more compact and 'correct' as well. 

I plan to still fiddle with the Diptrace program some more as well, so willing to try anything at least once or twice! Unlike GRJ, I will probably need the full board review (with revisions) prior to boards and parts or Gerbers.  

Sorry, I have been busy with Dr's appts and a dental implant the last half of this week and haven't gotten much of anything else done. Good news is cardiologist scheduled me back again in one year. I'm taking that as a good health report! May even have time to finish a Diptrace drawing... 

Mark, thanks for the kind words above. Asking questions is my strong point, the rest not so much. 

Looks like I may now have one of each type. Good thing I ordered extra parts! 

RTR, Congratulations on your good report from the cardiologist!!  Yes, I would say one year is a very good sign!!

Maybe I'll add a connection for a power switch as well.

rtr12 posted:

Stan,

Yes, I was planning to do some more fiddling and would be happy to add your ideas, but GRJ will probably have it done and posted by the time I finish writing this...  GRJ's design will probably be more compact and 'correct' as well.

I'm fast, but not that fast.   However, I think this is fast enough.  

Here's the "Stan0004" upgrade version.

• Two level detection and indication.
• Strain relief for battery connection.
• Provision for external power switch.
• Cap and diode to allow power from AC or DC sources.

Needless to say, the PCB got larger since this is an all thru-hole design, currently 1.15" x 2.35".  I thought about making it for 1/8W resistors except for the LED's, but decided to stick with what I have using the 1/4W parts.

TIU Signal Tester Rev. 1.1 Gerber.zip

GRJ, are you going to make anything available as a kit, like some of your earlier excellent kits?

Thanks, GRJ, that was plenty fast!  I was just fiddling with Stan's adds to the schematic, but not finished just yet.  Now I can see how I did for correctness. I hadn't thought of the diode and cap on the supply, probably would have thought of that unless you or Stan said something. Some added learning for me for next time. 

Glad you stayed with the 1/4W resistors, I have lots of those, 1/8W not too many (if any?). 

OSH Park says the PCB order has gone to the board maker, shouldn't be long now for Rev 1.0. Maybe sometime next week? 

Also, thanks Mark! 

RJR, if my parts are all correct and fit the PCBs I ordered, I will 'Share' the PCBs on OSH park and post a link here. I can also make a detailed parts list of what I ordered. I was hesitant to do that until I see if it all fits together and works correctly with my ordered parts. And also there were still discussions of added features going on with the first PCBs.

Then I will try the Rev. 1.1 project and do the same for it. 

Only thing I didn't order was the jumper for the power switch, I have a ton of those already, from years of fiddling with computers.

I guess I can go ahead and share the 1.0 boards now, I will just mark them Rev 1.0. I'll do that later today and report back.

I've been watching sense day one of the thread. This would indeed be a handy little devise if one has any doubt as to weather any of there channels is having an issue in regards to signal strength. I was all set to build Adrian's original version with GRJ's first OSH zip file but thought I'd wait in hopes the other electrical gizmo gurus would jump in a crank this thing up a couple of notches. Low and behold, BAM!

A kit would be nice. However, a zip file and a parts list would work for me thank you very much.

RJR posted:

GRJ, are you going to make anything available as a kit, like some of your earlier excellent kits?

I hadn't considered it, I was considering maybe just ordering a bunch of the blank boards and putting together a BOM for Digikey for all the parts so it would be easy for someone to build them. 

OSH Park wants $13.70 for an order of three of the 1.1 boards, so they're more than $4/ea.  I can get them for less than a buck if I order 100.  Of course, I have to be able to recoup my money for boards, so offering kits is always somewhat risky.

milwrd posted:

A kit would be nice. However, a zip file and a parts list would work for me thank you very much.

Ask and you shall receive.  I'd wait a day or so until Stan and Adrian "review" the schematic to see if I screwed something up.  it all "looks" good to me, I don't think I scrambled anything.

GRJ, I assume that if you made boards available and a BOM, that you'd be including a schematic so us neophytes could see where to place the components, and that it would be thru-hole so we needn't have surface mount skills???

RTR, I understand from my son just south of you that cold weather has struck your area---that it's below 90.  :-}

Obviously, there would be a schematic.   I could also actually put the component values right on the silkscreen of any production board, that's sometimes useful for hand built stuff.  This is all thru-hole construction, there are no significant size restraints, and all the components are "full sized".

What would be slick is a nice little box for it so it was a "finished" unit.

As for the weather, we had great weather for quite a stretch, in the 80's, low humidity, no rain, etc.  Now it's turning wet.

GRJ, the weather comment was for rtr.  He lives a few miles from my son out in Kansas.  I was talking to my son and he said it was the first sub-90 day since early June.

Let us know when you have a package ready for sale.  I would expect to order one.   

Yes, it was only 89 so far today! Been way too hot this year. Maybe even some low 80s later in the week too, finally!

As for the boards and parts, I will defer to GRJ on all that unless he wants me to do something.

I'll wait on GRJ for the OK to share them on OSH Park too, he did do all the heavy lifting as Stan said above. And GRJ, I agree that project box would be a nice touch for these.

Here's a little enhancement.  Before I make any "final" boards, I'll spin the Gerber files from this.  Should make it easy to figure out where the parts go.

If you're game for some minor mods:

Without increasing board size, I think you could add 1/8" mounting holes in diagonal corners as shown.  This would take #4 screws.  I like to mount tiny boards like this on a piece of acrylic so the nothing shorts out on the bottom of the board.  Example:

I'm not sure what everyone's plan is for connecting the TIU input wires, but perhaps a strain-relief hole is needed for those two wires too.

I think the components in the upper left can be moved down to make room for the mounting hole.  I note that your 3-D cap body looks much larger than the actual 0.2" lead-spacing component...where the leads are actually bent out to make them wider.  Point being there is more room available than the component outline suggests.

It looks like the 3-D rendering of the LEDs shows the flat-side on the left while the silkscreen shows the flat side on the right.  Trivial but for whatever reason caught my eye right away!

I'm not sure on what the call the 2 LEDs (weak, good, whatever).  But note that for a "good" signal, both L1 and L2 should blink.  For a "weak" signal, only L1 will blink.  The trigger voltage to the #1 half of the 123 chip gets the larger trigger voltage.  The schematic calls L1 "good" and L2 "weak".  

Good input, I stuck the "weak" and "good" in at the last minute, I'll have to go back and fix that.

Here's the "latest".   Let's see how close this is.

I added holes all over the place, changed the caps to .1 spacing as those are more common, not to mention smaller footprint.  Of course, I did change the LED colors, you were right.  I figure the power switch might just be a .1" jumper pair with a jumper on them, that's the simplest case.

TIU Signal Tester Rev. 1.1 Gerber.zip

Edited 7/22/18 at 22:00 to insert latest files and graphics.

Very nice John!  I'm sold on the "no go - weak - good" model!!  I'll add to my Digikey cart tomorrow.