Plug boards can end up with 1/2 a volt of noise on the power and ground rails. It's good to see that you are doubling up the wires. The scope grounding can influence the signal. one quick way to check this kind of noise is to put the scope probe tip on the ground pin. Also on logic gates you can get outputs that are not fully up or down if the inputs are not all the way up or down. logic gates are amplifiers and have a narrow range where there output can be in the middle. And one way to check for a short is to use a thermal camera.
@LaserFur -- Thank you for the reply! Yes, the breadboards for the clock module are getting worn out. I need to convert that, if not to PCB, to a semi-permanent solderful breadboard. I have a few I can recover from the failed EEPROM replacements. I have been concerned with the ground noise as I have been going and do check that quite frequently -- just not on camera. I will look your other recommendations. Thanks again!
Not sure if your scope settings changed between the signals that were good and the 4V signal (QR). If it did not (and appeared to be 1v/div until near the end), notice that the low pulses do not actually go fully low. If the 1V/division is correct (which it appeared to be on the other signals), the lows are about 3V. I would suspect another output driving that line. possibly a short between pins on one of the other points it connects to. Also, there appeared to be some missing solder mask on a few of the traces that went between pins on the 'HC74. that could also be a source of the short, but I would look elsewhere first as the mask damage probably occurred during the rework. Is there an LED on that signal? If so, it could be in backwards. In the last shot with the control board connected directly to the breadboard, the scope was set 2.5V/div. I personally prefer 1V/division and move the trace so that the 0V reference to be the 3rd line below center on the graticle. That makes it easy to see if anything is out of voltage spec which indicates bad outputs, bad inputs or conductive dirt/etc. Sometimes you can look at the other ends of the driven trace and see if there is any noticeable voltage difference from the output pin. If there is, that is frequently where the problem lies. I know you checked that the cap connected to the power rails of the '74 IC, but do those rails connect to the rest of the board?
The traces headed to the '193s from the '74 are really close to the pin header. Check for shorts to the header pins. Your scope trace setup appeared to be pretty close to what I like for most of the video, though I would zero out to the graticle line rather than being a bit low.
@grumpyoldman5368 -- thank you for taking the time with the replies. I really need a proper oscilloscope. This cheap thing no longer references 0V properly. I'm sure there is a reference voltage somewhere on the board which is messed up. Relatively speaking, I can see the shape of the signal, but I cannot necessarily count on the magnitude of the signal. Therefore, I would not get too hung up on 1V/div or 2.5V/div. These settings all reset each time I power on the scope and I missed setting it in the last clip. Since the DMM is in DC mode, it is averaging the voltage which is coming out to be 4V. I agree that it would imply that the voltage is bouncing somewhere between 5V and 3V, which the oscilloscope seems to imply based on the shape of the signal. The PCB is 4 layer, with a GND place on the second layer and a VCC plane on the third. Signals are on the top an bottom only. So that cap is connected to the entire PCB circuit. For the missing solder mask, that may be a flaw on the PBC manufacture. I will look at that again. But there was no change in behavior, so my assumption at the moment is that the solder mask was missing before I started soldering the first component. Recall that there was no change in behavior when I changed the '74 IC.
I'd bet that 2 outputs are short together. Not sure what SW you are using, but I think most of them have possibility to check schematic/board for such think. It sould be of course as well not design-error, but solder bridge. If schematic/PCB check does not help, I'd measure voltage on all pins on PCB and find where else you have those 4V...
@hwmland -- Thank you for the reply! I use KiCad. I have the latest version for the Ubuntu distro installed. It is pretty good about identifying traces which are driven by multiple outputs -- or more to the point it has warned me about that in the past on earlier versions. But, I do agree as I think about it more that I have 2 outputs which are driving the same signal somehow. As I have been thinking about this over the last couple of days, the edges of the incorrect signal feel crisp to me, which I interpret to be a solid connection. I'm probably dealing with a good old-fashioned bridge somewhere else on the PCB. The question is how to find it. Exhaustive search....? Or is there something more efficient?
@@eryjus4594 - I'm too lazy to check your schematic, but I suppose this 'problematic' pin is not connected to too many other pins. So I'l switch multimeter to short detection, one probe on the pin (it's on header, so there is easy access) and go on pins of other IO. You don't need go pin-by-pin in this mode, just firm (to be sure you really touched this pin) swipe. On areas where it beeps you can go pin-by-pin to find exact location. You have to 'check' of course as well other components (LEDs, ...) And yet another fault-stete that came to my mind is that this output pin goes into input of faulty IO... Here I'm afraid you'd need to remove those IO one by one...
@@eryjus4594 Oh no, I genuinly felt sorry to be lazy, with checking schematic and board design I could help better. I'm full of admiration for amount of work you must put into creating your videos. I'm designing myself 74HCxx CPU, but to make videos... well - simple word - lazy :( BTW: I had as well problem in my design because of counterfit chips from aliexpress. It were chips from old electronic (no problem, it's only toy construction on breadboard), but it was 74LS re-marked as 74HC. My special-purpos 16bit register contains four 4-bit adders with one operand tied to single pin (==16 input pins). It as well lead to poor levels and very slow edges when driven from HC...
@hwmland -- While it is not impossible, I do not believe that the ICs are counterfeit as they came from DigiKey. Good luck with your own build and videos.
I'm trying to build a logic analyzer in my spare time using a Raspberry Pi. I don't have a ton of spare time these days. Interesting thought, though, to use a logic analyzer to capture analog voltage levels in relation to other signals. I'm going to have to think on that thought a bit. (I apologize if Google Translate sent me down the wrong path here.)
Plug boards can end up with 1/2 a volt of noise on the power and ground rails. It's good to see that you are doubling up the wires. The scope grounding can influence the signal. one quick way to check this kind of noise is to put the scope probe tip on the ground pin. Also on logic gates you can get outputs that are not fully up or down if the inputs are not all the way up or down. logic gates are amplifiers and have a narrow range where there output can be in the middle. And one way to check for a short is to use a thermal camera.
@LaserFur -- Thank you for the reply!
Yes, the breadboards for the clock module are getting worn out. I need to convert that, if not to PCB, to a semi-permanent solderful breadboard. I have a few I can recover from the failed EEPROM replacements.
I have been concerned with the ground noise as I have been going and do check that quite frequently -- just not on camera.
I will look your other recommendations. Thanks again!
Not sure if your scope settings changed between the signals that were good and the 4V signal (QR). If it did not (and appeared to be 1v/div until near the end), notice that the low pulses do not actually go fully low. If the 1V/division is correct (which it appeared to be on the other signals), the lows are about 3V. I would suspect another output driving that line. possibly a short between pins on one of the other points it connects to. Also, there appeared to be some missing solder mask on a few of the traces that went between pins on the 'HC74. that could also be a source of the short, but I would look elsewhere first as the mask damage probably occurred during the rework. Is there an LED on that signal? If so, it could be in backwards. In the last shot with the control board connected directly to the breadboard, the scope was set 2.5V/div.
I personally prefer 1V/division and move the trace so that the 0V reference to be the 3rd line below center on the graticle. That makes it easy to see if anything is out of voltage spec which indicates bad outputs, bad inputs or conductive dirt/etc. Sometimes you can look at the other ends of the driven trace and see if there is any noticeable voltage difference from the output pin. If there is, that is frequently where the problem lies.
I know you checked that the cap connected to the power rails of the '74 IC, but do those rails connect to the rest of the board?
The traces headed to the '193s from the '74 are really close to the pin header. Check for shorts to the header pins. Your scope trace setup appeared to be pretty close to what I like for most of the video, though I would zero out to the graticle line rather than being a bit low.
@grumpyoldman5368 -- thank you for taking the time with the replies.
I really need a proper oscilloscope. This cheap thing no longer references 0V properly. I'm sure there is a reference voltage somewhere on the board which is messed up. Relatively speaking, I can see the shape of the signal, but I cannot necessarily count on the magnitude of the signal. Therefore, I would not get too hung up on 1V/div or 2.5V/div. These settings all reset each time I power on the scope and I missed setting it in the last clip.
Since the DMM is in DC mode, it is averaging the voltage which is coming out to be 4V. I agree that it would imply that the voltage is bouncing somewhere between 5V and 3V, which the oscilloscope seems to imply based on the shape of the signal.
The PCB is 4 layer, with a GND place on the second layer and a VCC plane on the third. Signals are on the top an bottom only. So that cap is connected to the entire PCB circuit.
For the missing solder mask, that may be a flaw on the PBC manufacture. I will look at that again. But there was no change in behavior, so my assumption at the moment is that the solder mask was missing before I started soldering the first component. Recall that there was no change in behavior when I changed the '74 IC.
Saw your short. Am subscribed.
Awesome, thank you!
I'd bet that 2 outputs are short together. Not sure what SW you are using, but I think most of them have possibility to check schematic/board for such think. It sould be of course as well not design-error, but solder bridge. If schematic/PCB check does not help, I'd measure voltage on all pins on PCB and find where else you have those 4V...
@hwmland -- Thank you for the reply!
I use KiCad. I have the latest version for the Ubuntu distro installed. It is pretty good about identifying traces which are driven by multiple outputs -- or more to the point it has warned me about that in the past on earlier versions.
But, I do agree as I think about it more that I have 2 outputs which are driving the same signal somehow. As I have been thinking about this over the last couple of days, the edges of the incorrect signal feel crisp to me, which I interpret to be a solid connection. I'm probably dealing with a good old-fashioned bridge somewhere else on the PCB.
The question is how to find it. Exhaustive search....? Or is there something more efficient?
@@eryjus4594 - I'm too lazy to check your schematic, but I suppose this 'problematic' pin is not connected to too many other pins. So I'l switch multimeter to short detection, one probe on the pin (it's on header, so there is easy access) and go on pins of other IO. You don't need go pin-by-pin in this mode, just firm (to be sure you really touched this pin) swipe. On areas where it beeps you can go pin-by-pin to find exact location. You have to 'check' of course as well other components (LEDs, ...)
And yet another fault-stete that came to my mind is that this output pin goes into input of faulty IO... Here I'm afraid you'd need to remove those IO one by one...
I’m sorry. I did not mean to imply I wanted you to check my schematic.
(Hits send too quickly here)
Thank you for the suggestions!
@@eryjus4594 Oh no, I genuinly felt sorry to be lazy, with checking schematic and board design I could help better. I'm full of admiration for amount of work you must put into creating your videos. I'm designing myself 74HCxx CPU, but to make videos... well - simple word - lazy :(
BTW: I had as well problem in my design because of counterfit chips from aliexpress. It were chips from old electronic (no problem, it's only toy construction on breadboard), but it was 74LS re-marked as 74HC. My special-purpos 16bit register contains four 4-bit adders with one operand tied to single pin (==16 input pins). It as well lead to poor levels and very slow edges when driven from HC...
@hwmland -- While it is not impossible, I do not believe that the ICs are counterfeit as they came from DigiKey.
Good luck with your own build and videos.
J'ai aussi un analyseur logique et je ne l'utilise presque jamais
I'm trying to build a logic analyzer in my spare time using a Raspberry Pi. I don't have a ton of spare time these days.
Interesting thought, though, to use a logic analyzer to capture analog voltage levels in relation to other signals. I'm going to have to think on that thought a bit.
(I apologize if Google Translate sent me down the wrong path here.)
@@eryjus4594it's old analyser HP de 1983 UP to 2O MHz state with 68000 engine ;-)