Orton 3C input/output troubleshooting

[James Harland]

Hello everyone,

I have soldered together the input/output board and front panel for the Orton 3C.

To test it out, I ran the RAM/IO echo code from here.

It kind of works, but there is a bug.

Basicically flipping switches 1-7 work as expected, and the corresponding light switches on, though there is a kind of ghostly echo of the same effect on the data lights on the right hand side, not sure if that is part of the problem.

The real problem is switch 0. If I flip it on, then all the other output lights go out, though the ghostly echo continues on the data lights, and the 0th output light does not go on.

While I was soldering I had real trouble with the transistor - I managed to glob all three legs together and it took me ages to separate them out, and resolder the holes (with thinner diameter solder). Otherwise I'm wondering if there just might be a problem with the 0th LED, though if that is the case, why do the other lights go out too?

I'll attach two photos, one of it working and one of the problematic switch.

Working (with ghostly echo):

Not working (0 switch on)

Thanks everyone,

James

[Mark T]

The ghostly echo on the data bus display should be expected, you have a tight loop reading and writing that pattern from the switches to the output leds.

Check for shorts between switch 0 and the enable of the output latch or address decoding.

[James Harland]

Thanks Mark. I need someone to break down how I can do this in a concrete way. I've only been soldering for a few months, and I don't know how to read electronics diagrams.

As I understand it, I should first check to see that there is not a short circuit at the switch, ie that there is not a solder bridge there. I did check the back of the switch already, though I suppose I should also check that no solder has seeped throught the board and made a connection on the other side.

I'm not sure where the enable of the output latch or address decoding are, so I don't know how to check them. I wonder if someone could show me on the boards, or explain how I can follow this route visually?

[James Harland]

Hi again,

I have done some software testing of the bug, and can conclude the switch is working fine, but there is some problem when I try to se the 0th LED. The LED itself is working (when I first switch the machine on, all output LEDs light up). 

If I just try to output an arbitrary number, everything works unless the 0th bit is set, in which case nothing happens. The output lights stay in whatever condition they were before.

If I try to double the input, the first switch works fine, and in fact nothing breaks at all, as the 0th bit is never set.

So this does not work:

1     0000                                     org 0
2     0000
3     0000 00                 START:  nop
4     0001 3E 01                          ld a,$01      ; odd number
5     0003 D3 00                         out ($00),a
6     0005 C3 00 00                    jp START
7     0008

but If I change line 4 to ld a,$02 or any other even number, it works.

Conversely, this code works, as does the 0th switch, as an odd number is never output:

1     0000                                     org 0
2     0000
3     0000 00                START:  nop
4     0001 DB 00                         in a,($00)
5     0003 87                               add a    ; result will never be odd
6     0004 D3 00                         out ($00),a
7     0006 C3 00 00                    jp START
8     0009 

Please could someone advise how I should troubleshoot this issue?

Thanks, 

James

[Mark T]

Check for shorts and soldering around the IC labelled U3 on the input output module. Pay particular attention to pin 2 and 3. Note pin 1 is the one with a square pad and the pins are numbered clockwise when viewed from the component side. Make sure U3 is fitted in the socket correctly and none of the pins are folded under the IC. Follow the tracks from pin 2 and 3 and check for short or open circuits anywhere else on the board.

[James Harland]

Thanks Mark. I checked the U3 connections and couldn't see any problems there. I also reseated the chip - no bent pins.

I did find something which didn't look good around the solder points of the resistors R9 and R10. To start with I thought it was just loose solder, but it did not come up easily, and then I began to wonder if I had either stuck some solder wick to the board or scratched the surface of the PCB.It was hard to see with my rather weak magnifying glass.

In any case, in the end, I went from this:

To this:

I hoped I had broken the connection, but when I put the computer back together and powered up, I noticed there is still the problem with setting the 0th bit as described above (and that trying to do so stops all other LEDs lighting up), and in addition the 5th switch is now not working, in that switching it to on has no effect when using an echo routine.

So now it seems I have an input and an output problem...

[Spencer]

Hi James,

I assume that you have a continuity tester there?  If so, take a look at the schematics for both parts of the Front Panel I/O Kit here https://rc2014.co.uk/backplanes/front-panel-kit/ and check continuity on every joint along the LED0 / D0 path - in particular, check if any of those have direct continuity to either ground or 5v (One side of the LED should have continuity to ground, but nothing else should).

Also, by the sound of things, you might want to do the same thing with the D4 path for the switch too.

Spencer

--
You received this message because you are subscribed to the Google Groups "RC2014-Z80" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rc2014-z80+...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/a980ecc7-b8d2-40eb-b27a-505ab00a1a29n%40googlegroups.com.

[Mark T]

That second photo still looks like a problem. Sometimes the easiest way to remove a solder short is to add more solder, the flux from the added solder wire should help the surface tension pull the solder short back to the pad. Also clear any excess solder from the tip of the iron and some of the added excess solder should stay on the tip. Make sure you are applying solder wire to the joint with the iron on the opposite side of the joint. If you apply the solder wire to the tip of the iron it will burn off the flux before it has chance to flow on the joint.

[James Harland]

Hi Spencer,

OK so I tried to follow this diagram:

I set my multimeter to continuity testing.

As I understood pins 17 and 18 on the 40 pin (39 pin) connector are GND and 5V respectively. So I touched one probe to each of the following locations and the other to GND and 5V, and got the following results. "Nothing" means that the multimeter did not react, and stayed on "1" which is what it shows to start with.

I did this on the front side of the PCB. Now I wonder if I should have been doing it on my solder joints instead.

Point                                       GND            5V

J1 LED0                                 Nothing       Nothing

U3 LED0                                 400               Beep
R1 (on the side of U3)         400              1200

R1 (on the other side)         Nothing       Nothing

U3 D0                                     Nothing       Nothing

U2 D0                                     Beep            400

Please let me know if I did the right thing, and if the results tell you anything. I'll look at the front panel tomorrow or another day.

Thanks,

James

[James Harland]

OK, I'll try adding more solder and see if that helps me remove what's there. I don't understand what you mean when you say "Make sure you are applying solder wire to the joint with the iron on the opposite side of the joint. If you apply the solder wire to the tip of the iron it will burn off the flux before it has chance to flow on the joint."

[James Harland]

After my next attempt to fix that solder bridge, I think I only made things worse. I'm fairly sure now that I've dug up the surface of the PCB and am only exposing the metal layer beneath, and on this attempt the issue with the first LED is still there, and now switches 1 2 and 3 don't work (but 4 suddenly does again...)

[Mark T]

From about 24 minutes onwards

https://youtu.be/GAjuZoyAhgI

[James Harland]

Thanks, I think generally it wouldn't hurt me to watch some more soldering tutorials. I have been lucky so far, but probably need to get better at certain techniques. 

By the way, now when I power up, LED #3 does not light up on output port 0. I wonder if the main problem is the connection between the front panel and the backboard PCB?

[Doug Jackson]

Just as a thought, have you been using rosin cored leaded solder, or a lead free variant?  Often lead free can be problematic for people starting as it requires higher temperatures and spotless technique.

--

You received this message because you are subscribed to the Google Groups "RC2014-Z80" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rc2014-z80+...@googlegroups.com.

To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/0274228e-c20f-4bf6-ae28-a5eb2103efb6n%40googlegroups.com.

[James Harland]

I'm using leaded. I think there are several problems. The main one is that I hurried on this board, I was impatient to finish the Orton 3C. This led to me being particularly careless when some solder wick got stuck to the board, and that's when I started to damage things, I think. I also think the wedge tip of my cheap battery powered soldering iron has worn out. I have another, wired, soldering iron which I am going to switch to for which I have a bunch of spare heads, doing some other cheaper soldering projects to get my hand in and see what that soldering iron is like, and I am also considering investing in a better soldering station and a more expensive iron.

Also this board has the closest together holes I've ever worked with. I wonder if I should have used thinner solder than 0.8mm (I did end up using 0.3mm solder on the transistor), and if I shouldn't have folded the wires over to fix them in place before soldering, but soldered them straight up through the holes.

One general question, when you use a wedge shaped tip on through holes, are you supposed to put the flat side of the wedge flat against the pad, and then touch the wire with the tip? That's what I've been doing. I notice in the video Mark shared that on the contrary, the person put the flat side of the wedge against the wire, and touched the pad with the edge of the wedge. That was with a wire sticking up vertically, not folded back.

To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/CAM_9E61PFXwhBaDY5skve5RDMCUS7NOts-c3H34J5aOVptnT-w%40mail.gmail.com.

[sunnyboy010101]

I have a weller plug-in soldering station that I bought in about 1990. It was the most basic model at the time (no adjustments) but works. My solder has always been MG Chemicals Sn63/Pb37 , 2.2% flux (rosin core), 0.025in (23 ga). I find even the next size up (0.032in or 32 ga) is too big for any delicate work.

I use the solder bulb sucker for simple solder removal and I bought a really cheap ($25CDN) desoldering unit that's hollow-tip soldering iron plus push button vacuum all-in-one for the really difficult desoldering. That inexpensive unit has saved me too many times to count, and in all but one instance left things as good as original when the desoldering was done.

I like the pointed tip because you can apply the heat really precisely using the tip point or side of the tip (for more heat). I've never had a problem with tiny, tiny things nor with big areas that hold things down (like ground pads on say a DB-9 connector).

When I recently built the PiDP-1 kit (after I did my Orton 3-C) I put together a website on my build, and one of the pages shows and discusses all the tools I used: https://huntrods.com/pidp1/tools.html

Oh yea, one of the most important tools shown on that page is my illuminated magnifier. It's essential for checking soldering joints. Many is the time I see something that requires a quick touch of the iron after I've just done a big bunch of LEDs or switches.

I also made a page (after the fact) on my Orton 3-C build but it's mostly just the photos of my finished build: https://huntrods.com/orton3c.html

-R

[James Harland]

Thanks R - I saw your photos before when I was building, that was helpful. I probably will get an entry level soldering station (or maybe a slightly better one with a hot air soldering option) when I next get paid. The "desoldering unit that's a hollow tip soldering iron plus push button vaccum all-in-one" sounds good, as sometimes there are some fiddly ones.

Spencer - is there any chance you could give some feedback on my attempts to continuity test? I don't know if I was doing the right thing and/or if my results tell you anything.

[Spencer]

Hi James,

I looked back at your testing results, and struggled to understand them really.  This is the method that I would use for the D0 LED though;

Remove the I/O Module from the backplane. 

Put one probe on pin 27 of the header (D0) and check that you do get continuity to U2 Pin 2 and U3 Pin 3. Check that you do NOT get continuity to Gnd or 5v.  This checks that the Data 0 going in to the module is ok and getting where it needs to.

Put one probe on U3 pin 2 and the other on the end of R1 it connects to. This should get continuity, and there should NOT be continuity to Gnd or 5v.

Put one probe on the other end of R1 and check continuity to Pin 4 of the 2x10 header. Again, this should have continuity, but not to Gnd or 5v.

Keep the probe on R1 and check pin 4 of the 2x10 on the front panel. This will confirm that the singal gets through to the front panel. (still noting to Gnd or 5v)

Keep the probe on R1 and check there is continuity to one side of LED0. Again, still nothing to Gnd or 5v

Finally, check that the other side of LED0 does​ have continuity to ground, but not 5v.

If all of that checks out then the circuit for that LED is correct.

To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/e915f729-6815-4a4d-a6e4-4b8fcf057aa2n%40googlegroups.com.

[James Harland]

Thanks Spencer,

So just to check I understand your instructions.

I should unplug the I/O module from the backplane, but leave it connected to the I/O front panel, is that right?

Whenever you say Gnd or 5V you are talking about pins 17 and 18 on the I/O module 40 (39) pin connector?

When you say continuity, you mean my multimeter should beep. If I see a resistance value but get no beep that counts as no continuity?

Thanks again,

James

[Spencer]

Hi James,

Yes, ideally unplug the module from the backplane. You can disconnect the Front Panel module for some tests, but it is worth at least checking it connected once to make sure Pin 4 on one side connects to Pin 4 on the other.

Yes, Gnd and 5v are on those pins. They appear at many places on the module, so sometimes it's more convenient to check continuity to those instead.

And, yes, continuity means zero ohms which will normally beep on most multimeters. Testing with components fitted might give various resistances to ground or 5v, so you might not want to worry about those.

Also, I took a closer look at the photos of the solder bridge earlier in this thread. There is some PCB track running between those joints (I think that's a row of resistors which all have one side commoned together), so whilst solder bridges are generally bad, and should be avoided/fixed, if those two pads should be connected electrically, then it doesn't matter.

Spencer 

-------- Original Message --------

To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/fdc122b7-8665-424b-9dc2-ab666e89ba0cn%40googlegroups.com.

[James Harland]

Hi Spencer,

Thanks for this. Ok so I went through your procedure and everything checks out. I had a couple of false starts, as I was trying to read where the pins of the chips were off of the circuit diagram, but then I realised I should probably be using a datasheet for those chips instead, and then things made more sense. (Is there a hard-and-fast numbering scheme for all ICs I wonder?)

I didn't always know what you meant when you repeated about testing there was no continuity to ground or 5v, but I think I checked everything. I suppose you meant from each additional point you mentioned.

I also wasn't sure about where to find pin 4 on the front panel, but I guess it is this one, as this is the one with continuity:

Now I look at this picture, I don't like the look of two of the solder joints. I will probaly try to fix them up now I have it apart.

I'll let you know if my soldering has any positive results, otherwise I'll be asking for next steps :-p

Thanks,

James

[James Harland]

No change after re-soldering those pins. For now LEDs/switches 2 and 3 not working, and 0 still acting to switch everything else off:

[Doug Jackson]

They are out of focus in the image, but are you 100% sure that all LEDs are inserted the same direction?

Sometimes the flats are hard to see, but the internal structure is quite visible, just make sure that one does not look out of pl;ace.

Kindest regards,

Doug Jackson

em: do...@doughq.com

ph: 0414 986878

To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/147363c0-33f5-4294-ac31-f8dafd1b1856n%40googlegroups.com.

[James Harland]

Hi Doug,

Thanks for taking a look. Yes. they're all the right way round. They all light up when you switch the machine on, or at least they did to start with. Now #3 or #4 (I can't remember which offhand) doesn't light up when you switch the machine on, but as LEDs I think they are all installed correctly. During different tests I have been able to activate all of them except for #0 which has the odd effect of not switching on, and switching all the others off, but only when addressed to do so, it always switches on after powering up.

James