I use the ATTEN ADS1102CAL, but any 'scope with 20mhz or more bandwidth would be sufficient to do the testing.  Not much I do on the bench taxes the 100mhz bandwidth of mine.  You can probably pick up a decent 'scope for not much money if all you are interested in is using it around model train electronics.  Here's a decent one that's similar to mine, $245 shipped.

Hantek DSO5102P USB Digital Storage Oscilloscope 2 Channels 100MHz 1GSa/s

eBay: 223444992742

I ordered 2 kits from Tom a while ago and finally got around to assembling one. Other than the pads on U1being really close, soldering was straight forward.

I powered it up and all seemed to operate fine. I started Stan's calibration and verified 5 volts. I set the pot in the middle and momentarily touched I had a red light flash. I rotated the pot to both extremes and got the expected results of neither flash and both flash.

The problem - while making adjustments, I apparently had my probe slip from one side of R11 to D2 while I connected to D1. Now when I power up with no DCS signal the red light comes on and stays on. The power LED comes on and there is still 5 vdc.  Figured I blew one or more components.

Does anyone have a thought on which one or other simple measurements to isolate?


GRJ, the diodes are 200v 1A (1N4003, same as what you used), does that make the HCT chip more suspect? And there is a path directly to the HCT's Vcc when connecting D2 to R11 and D2 is track voltage. 

Very Scary, I know, but this has gotten me to print out the schematic and look around a little. Trying to gain better understanding here myself.

Ken, Didn't see your 2nd post above when I was posting earlier, I would wait for GRJ's reply to be safe here. Good luck and I hope you get things going again. Sorry, but I don't think I have any extra parts other than for the remaining kits I have, but if you need Digikey numbers let me know.

Connecting R11 to D2 would be very bad for the health of the IC, so replace that first.  If D1 is bad, it won't cook anything, the unit just won't work.  Since you have 5V, D2 and U1 are still OK.  I'd say U2 and D1 are the top suspects.

Glad you got it all going again and the CDHCT123E is not too expensive either. I looked around but didn't have any extras, they were all accounted for in the remaining kits I have left. 

John or RTR- I finished assembling my board recently.  I'm having a bit of trouble getting +5V out of U1.  Appears I may have a short from U1 pin 3 to 2 somewhere. I did verify no solder betweeen the leads of U1 as pointed out they are close. I'll double check that and some other stuff.  

My question is when powering up for calibration using the 'Stan procedure' should the power supply to the signal tester board be current limited if using only an external 18V AC power supply?  or (and) is TIU necessary for calibration?

Thanks Mike  

My guess would be the solder connections on U1 are touching somewhere. I had the same problem myself with one PCB I assembled and a couple of others have had that problem as well. I removed all the solder from U1 with some solder wick and started over with much better results the second time, as did the others I know of that were having the same problem. 

As for the power and calibration, I also used Stan's method detailed in his video posted in the thread. I use alligator clips to connect the tester to track power (with nothing on the track) and did the calibration the same way, as memory serves. I use PH-180s for track power so no current limiting here and they go through my TIU to get to the track. As I recall from following along in the design process here, GRJ's final design drawings were meant to use track power to drive the TIU Tester so no external power source was needed.

For the definitive answers on calibration I would wait for GRJ or Stan's comments just to be sure (and safe). I am by no means even anywhere close to being as knowledgeable as either those guys, I'm still just a greenhorn.

I'd be looking at U1, it's a problem area.  I've actually started to use a new footprint I created for TO-92 packages to spread the leads out a little, the library pattern is not very easy to deal with.

I'm glad you did that. I was going to try adjusting that myself, if I ever needed to order more PCBs for the testers. I have quite a few on hand so it will be a while if they ever sell out. Orders have dropped off the last few months.

I was just trying to make a different PCB the last couple of days (Diptrace) and the component libraries and patterns are a struggle and quite a bit over my head so far. 

Add Reply

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