Junior Computer

[frankdh]

BigDumbDinosaur wrote:

For a bit of nostalgia, you could read his page about slide rules. 

I've never owned a proper sliderule, I was intrigued by them though, and might pick one up in the (near) future.
I did own one of these addiators at the bottom of the page though (don't know where that got to, being a pack-rat I can't imagine throwing it away. Maybe it will show up some day...).
And later a good old TI-59 with thermal printer and extra modules, which I didn't use very much I'm ashamed to say (big plans, scarcely any accomplished )

[frankdh]

BigDumbDinosaur wrote:

frankdh wrote:

Now that I'm retired and starting to feel nostalgic I dug up the cardboard box and am now in the process of restoring it...by the clock circuit of Ben Eater...

I should caution you that Ben Eater’s videos will, at times, ignore design and/or construction details that are important in producing stable operation.  Quite often, we have new members sign up here who tried to build something that was depicted in an Eater video, only to not have it work due to omission of an important detail that was not brought up in the video. 

I already made a change (improvement?) building the clock module. Before the clock pulse leaves the board to whatever it is driving, it goes through a Schmitt trigger (74LS132, 'cause that's what I had laying around at the time) to produce a crisp square wave.

[BigDumbDinosaur]

frankdh wrote:

I already made a change (improvement?) building the clock module. Before the clock pulse leaves the board to whatever it is driving, it goes through a Schmitt trigger (74LS132, 'cause that's what I had laying around at the time) to produce a crisp square wave.

I hope it isn’t a 74LS132.  Its TTL-level output doesn’t meet the specs for any modern 65C02.

If you are clocking the WDC 65C02, I recommend you use a 74AC device so as to meet the VCC × 0.7 minimum, as well as the =< 5 nS rise/fall time spec.  You can fudge with the latter at lower Ø2 rates by using 74HC or 74AHC—both have slower edges than 74AC.  However, not meeting the VCC × 0.7 minimum may cause baffling stability problems.

Ringing can be a problem with 74AC logic due to the extremely fast edges.  You can mitigate this problem by keeping the clock circuit connections as short and direct as possible, and by inserting some series resistance between the clock generator and anything that is being driven by it.  Somewhere in the range of 100-150 ohms usually does the trick.  A metal film resistor may work better than a carbon film—YMMV.

In general, 74LS logic should not be used to drive the input(s) of a CMOS device.  The voltage levels at which a typical CMOS device (other than 74xCT) distinguishes between a logic 0 and logic 1 are markedly different than what a typical TTL device produces as output with any kind of loading.  On the other hand, having a CMOS device drive the input(s) of a TTL device doesn’t pose a problem.

[frankdh]

@BigDumbDinosaur
Thanks for the advice.
At the moment I am using an LS version, but that can easily be remedied. Just need to source the AC version.

[frankdh]

OK. Next step in the - starting to be a never ending - story.
I built the clock circuit as per Ben's video instructions (https://youtu.be/kRlSFm519Bo?si=mSM3vmfU6GLigqLs & following episodes) and it all worked just fine (I swapped the second and third NE555 by an NE556, but it all worked as expected). Put it aside for a while (more like a couple of years). Dug it up a couple of days ago, and hooked it up to a PWR supply, and ... nothing. The PWR supply even went into protection mode and switched off. Couldn't find what went wrong during the time of inactivity, so decided to rebuild the thing re-using the componest on the failing circuit.
The first circuit works fine (blinking LED, speed adjustable with the trimmer).
Added the second circuit, and that's where it gets flaky. The LED lights up and remains lit, when I approach the board to press the button, the LED turns off, with my finger still about an inch away (the LED also doesn't behave the way it should when I do manage to press the button). Seems to be picking up some interference (I used to work in a mainframe environment that got an upgrade from really old mainframes to state-of-the-art stuff like VAX cluster, 100 terminals, etc. One of my colleagues could crash such a terminal by just sitting down in front of it... extremely strange IMO) ? It didn't do that in the first version, that for some misterious reason now has a short somewhere.
Back to the old drawing board I guess. The quest continues.

[gilhad]

It sounds like there is some connection missing, or some component destroyed.
If the first circuit works and second do do not, I would try rebuild the first one with components from the second, ideally swaping one component at time and chceck, if it still works.
Then rewire it to test also the other half of NE556.

[Dr Jefyll]

frankdh wrote:

The first circuit works fine (blinking LED, speed adjustable with the trimmer).
Added the second circuit, and that's where it gets flaky.

First circuit? Second circuit?

Remember please that not everyone has seen those videos, or will make a special visit to YouTube on your behalf. Can you post the schematics of these two circuits? Also, it'd be good to see some photos of the clock modules (or whatever it is you've built). Making lots of info easily available will increase the number of individuals who can assist you.

That said (and based on limited info), I agree with gilhad that it sounds as if a connection is missing (or perhaps the connection does exist but is flaky, as can easily happen if you're using a solderless breadboard).

-- Jeff

[gilhad]

I had seen that video few years ago and did the same breadboard clock as first step.
The first 555 is just astable oscilator with adjustable speed to create some kind of clock.
Then goes another 555 as monostable debouncer for button to create single pulse per press for single step.
Then goes third 555 as bistable debouncer for switch and its output is used for selecting between those two previous -
the actual selection is then realised by AND/OR/NOT gates and result is digital clock source for the rest of computer without bouncing and with selecting single-step or continuous run.

It is used to show different uses of 555 as well, and beginer have fast something, what visibly works, before the rest of computer is build.

Anyway my point was to test all three 555 in the first schema, which worked for frankdh - so it proved, that used components and schema are OK.
Testing rest of the components in this should reveal, if there are any defects. And is as easy as the first part, which is already completed. Which is easy and fast way, how to test it without need for more complicated tools (osciloscope etc.).

[frankdh]

Dr Jefyll wrote:

frankdh wrote:

The first circuit works fine (blinking LED, speed adjustable with the trimmer).
Added the second circuit, and that's where it gets flaky.

First circuit? Second circuit?

Remember please that not everyone has seen those videos, or will make a special visit to YouTube on your behalf. Can you post the schematics of these two circuits? Also, it'd be good to see some photos of the clock modules (or whatever it is you've built). Making lots of info easily available will increase the number of individuals who can assist you.

That said (and based on limited info), I agree with gilhad that it sounds as if a connection is missing (or perhaps the connection does exist but is flaky, as can easily happen if you're using a solderless breadboard).

-- Jeff

I apologize for, apparently, providing insufficient info. I have no drawings of the circuits, since the drawings provided by Ben Eater are clear enough to interpret. Building the circuits is just a matter of following the videos (3 of them: part 1, part 2 and part 3) and pauzing whenever necessary to insert a component. Since I follow the instructions in the videos my circuits look exactly the same as his. So photos of my circuits are, IMO, not necessary.
Of course, I can not, and will not, force anybody to do anything they are not prepared to do. So if they are not prepared to use the info I provided that is, of course, their choice. That is what it is. No hard feelings.
I appreciate any input. Gilhad did an excellent and clear explanation of the circuits I'm talking about. Thanks for that gilhad.

Anyhoo, I'll try a different breadboard, and if that still doesn't solve the problem I'll solder the circuits on a perfboard...

[BigDumbDinosaur]

frankdh wrote:

Dr Jefyll wrote:

frankdh wrote:

The first circuit works fine (blinking LED, speed adjustable with the trimmer).
Added the second circuit, and that's where it gets flaky.

First circuit?    Second circuit? 

Remember please that not everyone has seen those videos, or will make a special visit to YouTube on your behalf.  Can you post the schematics of these two circuits?...Making lots of info easily available will increase the number of individuals who can assist you.

I apologize for, apparently, providing insufficient info...I can not, and will not, force anybody to do anything they are not prepared to do. So if they are not prepared to use the info I provided that is, of course, their choice.

Further to Jeff’s comment, it’s not necessarily a case of someone not wanting to view the videos.  During the day, some 6502.org members will be at work and their employers may have their Internet service configured to block sites such as Youtube (I have two clients who have such an arrangement).  When said members get home, they likely will have other things that need to get done and may not want to expend time chasing down a video.

You are more likely to get assistance if you do everything you can to make it easy for someone to assist you.  After all, you are asking that someone to expend some of their valuable time on your behalf.  Expecting someone to go off-site to view a video posted by an individual who occasionally leaves out important details is not fair to the bloke who is trying to help you.

You can make it a lot easier for everyone who is trying to help by sketching up a schematic showing exactly how you wired your unit and attaching same to one your posts here.  Photos of the construction of your project may also be helpful, although, as Jeff implied, anything built on a breadboard is inherently prone to trouble.

Speaking from my perspective, I try to stay away from websites run by Google.  Aside from the annoyances of Youtube and their endless commercials, I don’t relish having to sit through an otherwise-uninteresting video just to glean a few bits of information about a project on which I am trying to offer assistance.

Incidentally, what Jeff and I have said about going off-site to watch a video is not unique to us.  That post has some valuable advice.

[barrym95838]

I watch youtube literally every night, and Pie Adblock has done a seamless and thorough job for months on my behalf. I haven't turned my television on in months, because the relentless and repetitive commercials spoil the experience for me.

[BigDumbDinosaur]

barrym95838 wrote:

I haven't turned my television on in months, because the relentless and repetitive commercials spoil the experience for me.

Other than an occasional news broadcast, I don’t watch TV at all.  My wife is the TV addict around here, and watches endless game shows.  If it weren’t for her, there would be no TV service—on-air broadcasts are good enough for the minuscule bit of watching I do.

As for Youtube, the commercials are less a problem than Google’s intrusiveness.  There was a time when Youtube was actually pleasant.

[AndrewP]

barrym95838 wrote:

...Pie Adblock has done a seamless and thorough job for months on my behalf.

BigDumbDinosaur wrote:

...Youtube, the commercials are less a problem than Google’s intrusiveness

Also let me introduce you to Sponsor Block, advert and sponsorship blocking for YouTube. Search up "Sponsor Block Firefox" if you're interested. And if you don't have a Pie network blocker (or pfSense or similar) then "Ublock Origin Firefox". It's a fork of AdBlock after AdBlock went dodgy. I've no idea if these work under Chrome anymore.

[barnacle]

For Linux (I don't know about other OSes), Freetube works well, though it does have bursts of not working for a few days while it catches up with Google's latest changes to the youtube API.

Ublock Origin works well with YouTube, but sometimes YouTube gets sniffy about it.

Neil

[BigDumbDinosaur]

Gentlemen, I’m wondering if maybe what I (and Jeff) said is being buried in the discussion about Youtube ad blockers (I do use one, BTW, although it falters at times).

To reiterate, we shouldn’t have to go off-site to look at videos, pictures, diagrams, etc., on a quest to offer assistance with people’s misbehaving projects.  My attitude is if you can’t expend a modicum of effort to post it here, I can’t expend a modicum of effort to help you.  On occasion, a video of one’s unit starting up or doing something bizarre may be helpful—that, of course, will involve an off-site link.  However, a reasonably-clear schematic, memory map, and perhaps a photo of the unit, will usually be more useful.

Garth gives some good advice on this on his site, especially in comments about what constitutes a good schematic.