Permanent Killer Poke Mod and CRT Spot fix on the Commodore PET 2001

[@gorber81]

Adrians vids... Pretty much the only thing that can make me super excited and relaxed at the same time. This is clearly therapy for me

[@jackywhite880]

Back in the early 1980s, I signed on for a computing course with the UK Open University. We operated mainframes that input via teletype keyboards and output on paper tape - no screens at all. Although expensive, I quit the course within a few months and wrote off my money. Partly because during the course I'd encountered no equipment less than a decade old, and partly because of an open policy that home computing was a fantasy, and the new, increasingly available devices like the Pet, TRS-80 and others would NOT be included in the course as they were "toys that could never have any place in computer science." I subsequently learned more from a year's subscription to a computing magazine than I might ever have learned from that course.

[@Sheevlord]

Looks like the concept of disabling VSYNC to improve performance is older than most people realize. However, nowadays it doesn't require a hardware mod to achieve it. And yet some people complain that editing a game's INI file by hand is annoying.

[@jengelenm]

@@YoureUsingWordsIncorrectly Well that's sarcastic comedy obviously, but funny!

[@jengelenm]

@@YoureUsingWordsIncorrectly Which comedian was it?

[@macdaniel6029]

@@YoureUsingWordsIncorrectly So true. Just let Darwin do his job!

[@squirlmy]

IDK. In COVID-19, this is less funny, because it's too close to the truth. People are refusing to wear masks, they're infecting their family members and each other (motorcycle gathering in S.Dakota) and people are dying. Ironically, this group of people are most likely to believe in Eugenics and letting this kind of thing happen. They don't understand that they're Mother Nature's targets; that they are, in fact, the "unfit", not liberals in NYC, as their fantasy imagined.

[@camillecirrus3977]

Imagine being able to edit an .ini file to change a game's settings. This comment was sponsored by encrypted config file gang.

[@stylesoftware]

The persistent dot in the middle of the screen, is because the emitter in your CRT is experiencing two things. Firstly the magnetic coils that focus the beam has collapsed, so the emitters focus is only in the middle, so energizing the phosphor for longer, so seeming brighter for less current (effort). Secondly there is a tendency to for emitter to continue to emit when they are hot. Basically they are amplifiers, the heat means the electrons are excited, and even when the power is removed, they continue to emit due to heat. If your valve emitter system is in good condition, they will persist for longer. Answer: You have a good condition CRT.

[@TzOk]

I think you could safely connect that pin of VIA chip, by the 510-2k resistor to where it originally was connected. Resistor would be enough for separating the VIA from the rest of the circuit in case of level conflict. The POKE would then work as it was supposed to.

[@eDoc2020]

That's what I was thinking as well. Adrian could use the stacked socket trick to do this using the original VIA. Also, Commodore surely knew about this trick, why didn't they add a resistor when they designed the newer models?

[@jaycee1980]

I was thinking the same - a 1K resistor between the VIA PB5 and the rest of the circuit would do the job perfectly well. The rest of the circuit would be relatively unaffected by the 1K to ground, and the 1K series resistance when the VIA is configured as normal would still result in a correct logic level being seen by the VIA

[@iz8dwf]

@@eDoc2020 yes this is correct, a resistor in series can also "do the job" but you have to "poke" everytime you want the "fast" speed. With Adrian's mod there's no need for any poke anymore. It depends on what you want to accomplish. Most strange thing is that Commodore corrected the "snow" effect on 2001N motherboards, so the VRAM can be safely accessed outside the vertical retrace periods without causing any snow, but they didn't change the firmware or changed the VIA input circuitry. So 2001N remain unnecessarily slow.

[@timsmith2525]

Flashback to when I was a kid, and the power-off dot on our B&W TV mesmerized me. Fascinating video. Thanks for sharing!

[@tiporari]

Normal behavior for a CRT. You would need to discharge the HV before you lost deflection to avoid it. It doesn't hurt anything. If your CRT has a burnt spot, it is because it was run for an extended time without horizontal and vertical deflection. Eventually I suppose the power down spot could burn in, but it would take 40 or 50 years... Oh wait...

[@iz8dwf]

I have to disagree. Commodore itself tried to avoid the big spot you see on Adrian's PET with changes made in the later monitor revisions. And no, you don't want to discharge HV in a single spot, when it's done on purpose, the energy is spread with raster oscillator still running. This can't obviously be done on a PET monitor since it doesn't have raster oscillators.

[@Walczyk]

@@iz8dwf 100% this

[@shadowflash705]

@@Walczyk I never had a PET but last time I've seen that spot on a properly working CRT was early 60s TV. All TVs and video monitors from 80s i had never had that issue. And it's better to be fixed.

[@kenkobra]

Every time I see the PET it reminds me of Twiki and Dr. Theopolis from Buck Rogers.

[@mikesilva3868]

Agreed 📼

[@DankNoodles420]

It reminds me of the fallout games when I see a pet.. it makes me think of the terminal hacking in the games.

[@mikesilva3868]

@@DankNoodles420 😎

[@JapanPop]

Holy cow-I was about to say THE SAME EXACT thing!!!

[@squirlmy]

@@HutchCA I think Silent Running was simultaneous to PET, B.Rogers was later. The fact that there really are aesthetic "fashions" in computer designs, that are picked up by Sci-fi, is funny. It's an area that prides itself on form following function. We had beige boxes for a long time. But there are always little design features that poke through.

[@TechBench]

The early IBM CGA adapter had a similar issue with snow. In 80-column mode, the CPU and the video controller chip would not be able to read/write to the memory at the same time without causing snow. The BIOS routines would wait for the screen retrace before updating the video memory, but it had to poll for the retrace status, which was terribly slow and caused the same flicker you see on the PET. So you could have slower/flickering text or faster snowy text. Some DOS programs even had options for these two "modes". Great video, thanks!

[@TheDiveO]

next episode after fix: Adrian testing new retro device that emulates CRT switch-off behavior on HDMI screens.

[@nilswegner2881]

I want that!

[@corygraves2402]

Me too

[@dennis8196]

I so want this.

[@StrangelyIronic]

Nice RPN Calculator app. Also really nice collab and deep dive into the CRT issue, my Uncle was an electronic engineer building TV's way back in the B&W and early color era and I remember him explaining this problem to me in words I could understand at the time (he was 20-30 years older than my father, can't remember exactly, and was already retired by this point). I was learning to program Assembly and C on an Apple IIGS and got the //e and display as a gift/challenge to fix...I fixed the computer but stayed clear of the CRT and watched him repair it closely as he explained how it all worked in words a slightly educated ~7 year old me would understand in the late 90's.

[@simonbaxter8001]

Love retro PET video's. My first exposure to computing was as a 14 year old (back in 79'') when my dad bought a 2001. Later upgraded to a 4016 with external dual floppy drive (4040) which I built a mod to make the top half of the character ROM switchable to RAM so you could create your own character sets and draw basic graphics. Happy days.

[@Obie327]

I messed with some programming on this very machine back in the day. The dot was normal with these PET's. Thanks Adrian for the video!

[@adriansdigitalbasement]

Yeah definitely a design flaw from the factory :-)

[@mikesilva3868]

@@adriansdigitalbasement 😎interesting

[@Fifury161]

@@adriansdigitalbasement Not a design flaw - a design choice, there's no need to provide a different discharge path as the damage to the screen would not occur during the warranty period/expected service life of the machine. Most mono CRTs of the day have a similar power down dot...

[@iz8dwf]

@@Fifury161 There's no need to provide a discharge path in the first place. Commodore made a few mistakes and they even tried to correct them as soon as the second monitor revision came out (I have a 1978 and a 1979 monitors).

[@stevesstuff1450]

Yes, and it was standard with every black & white TV set too at that time.... everytime you switched it off, it had that collapsing dot that finally faded to nothing! It was nothing unusual. Every monochrome CRT and TV did the same thing. Perfectly normal! 🙂 And yes, I was around during the 60s and 70s... pretty much every CRT image will collapse to a small dot when turned off, but some go much faster than others - 80s and 90s TV/monitor types especially....

[@k_kon131]

13:00 since i have a PET 2001 I can agree as well, that the dot on the screen appears when you power the computer off, in my case sadly the CRT even has a burned spot in the midle

[@retropuffer2986]

Your pet has 1459 ticks. Time for a collar. :D

[@williamsquires3070]

Or Hartz Flea/tick powder... or a trip to the vet! 🤣

[@mrnmrn1]

Well played. xD

[@georgechambers3197]

I never used that killer poke the entire time I owned that PET and the "dot" was always there as far as I can remember. Thanks for the video, seeing the old gal alive again is great!

[@greendryerlint]

Thanks for a very informative video. I have a 3032 PET with the full-size keyboard. I have used the 'killer poke' for many, many years on my machine. It does speed things up with the penalty of 'snow' on the screen at certain times, same as you experienced. It has never harmed any of the chips. Many early Cursor games incorporated the killer poke as a speed boost. You really notice the speed up when listing a program, for example. However, trying it on a 4032 (fat 40) with the CRT controller chip, it makes very unhappy noises from either the flyback or deflection circuitry. I quickly turned off the machine before any damage could occur. Also, as you probably know, they fixed the dot issue on the later PETs. I've never had it happen on mine.

[@HeyBirt]

"How could I possibly figure out where R2 is?" R2 is located on the exterior of the X-Wing fighter behind the pilot. Jeesh, everybody knows that :)

[@Dukefazon]

*screams in astromech*

[@erik365365365]

I don’t have a PET and I honestly never will but I love your willingness to play around with these issue and love watching along.

[@raymondheath7668]

I'm an old tube guy, so that also includes CRT's. Back in the early 80's there was always the standing argument of turn off dot "burn in", or whether or not a shadow of the turn off dot would be permanately burned in the phosfor coating on the viewable face of the screen. I was always with the group that the turn off unfocused beam did not have enough current to create the shadow spot. Can't wait to see the next installment

[@Fifury161]

There is good logic to your argument. When you consider the duration there is likely more damage to be done whilst the tube is on! Damage will occur over time for any CRT - most are rated for 20,000 ~ 30,000 hours...

[@EmergencyChannel]

I’ve seen thousands of CRT tv’s over the years and I’ve never seen one with burn in or a super dim picture. Usually the most common problems I saw was one of the guns not functioning so you’d either get a all purple or all blue/green picture. I’ve only seen a few monitors with burn in, mostly business terminals.

[@Fifury161]

@@EmergencyChannel You realise you have contradicted yourself?

[@EmergencyChannel]

@@Fifury161 TV's and monitors are two different things. I never claimed to know anything about CRT's, but I did run the front desk of a repair shop and done paper work while the technicians fixed the tv's.

[@Fifury161]

@@EmergencyChannel Erm, in this analogy they are not as they both have CRTs. The best example of screen burn in could be found back in the day in airports. Both the flight timetables shown on CRTs and the local news channels on CRTs all suffered from screen burn in. The timetable where on monitors, the news channels on TVs, all CRTs back in the day...

[@bborkzilla]

Even after all this time, I still use peek and poke commands during initial board bring up. In this case they're Linux applications I write to allow direct access to kernel memory and PCI space from the command line. It's amazing how much of the hardware you can wring out with old BASIC commands!

[@steveb9501]

My screen was starting to burn the screen with a spot. Thanks to your hard work guys, the mods worked great on my 3032 :)

[@NicoDsSBCs]

I love nerding out on little details like that. Can't wait for the next video. I also always had the spot on my old crt in my bedroom when young. When to bed I could see it for quite long. It's those little things that bring back big memories. Thank you!

[@jcbecker2800]

I don't know how to fix any of this stuff, but I enjoy watching your videos.

[@EsotericArctos]

The dot does not represent a dying CRT. In fact, it usually happens with strong CRTs. Zenith TVs in particular use to get it a lot, but you are right with saying a lot of early TVs did. The explanation provided is very good.

[@eDoc2020]

I've found that on TVs the dot happens more the darker the picture is. I presume this is because the current draw of a bright picture quickly drains the HV before deflection fully stops. The association with the dot and weak CRTs may be just because their picture is dimmer.

[@EsotericArctos]

@@eDoc2020 I think Zenith use to recommend turning the brightness to maximum just before turning the tv off to minimise the dot, but the dot doesn't do any damage. It's just the last burst of energy from the guns as the cathode cools. It creates a bright dot as the horizontal and vertical is already shutdown, so the beam just hits the middle of the screen

[@terosaarela4555]

PET stuff is always fun to watch! Thanks for sharing!

[@crunchyfrog555]

Funnily enough I was talking to an old school friend of mine recently about this. We used to spend our lunchtimes going to his dad's accountancy firm to play his PET. He had a couple more at home, and it was there we first tried this. It did the same as you experienced. Blank screen, only cured by powering off. mind you, I'm British so I don't know if that's relevant (50Hz and all that) - and yes, the dot appeared all the time too. It seemed to be on the early models - the chiclet keyboard ones, but I can't remember if it happened on the later ones. The only other thing I remember for sure, is that the chiclet keyboarded variety one (like yours - I'm sorry I can never remember the model numbers) was a really early model one. They get it at the business very soon after they came to these shores. Iirc, it was through Germany as well.

[@sierraboney1394]

I remember seeing the spot starting about 3 inches in diameter gradually decreasing down to nothing over about 30 seconds or so after switching off our old wooden cased 20" ITT colour TV. Was quite a strange effect watching it in the dark after you'd switched it off at night! I had a feeling it was just due to the design of the chassis, as i'd seen other TV's doing it back then as well. That was a good TV, had valves (tubes) inside, shame though that the vertical hold went on it (I think the vertical hold pot crumbled). My dad chucked it away in the end (it had already been replaced with a newer TV, I had the ITT in my bedroom when it still worked). Wish I still had that TV, if I knew then what I know now about repairing arcade and monitor/TV chassis I could've repaired it in a few minutes!

[@terrythomas1343]

As several people have said, many CRTs used to show a spot when turned off in the "olden days". When manufacturers started to sell CRTs/TVs with the spot eliminated they were referred to as having a "spot swallower".

[@androgyne_sander]

I've owned several of these PET 2001 with the blue (white) on black CRT, that all have the power off spot. My current one, by far the best one I've ever had, also has it. It's also common to see on vintage television sets. I can't remember though having owned one with the green on black CRT (later model PET 2001) that had it too.

[@williamsquires3070]

The “dot” is inherent to the design of CRT monitors/TVs; it’s because the horizontal and/or vertical oscillators shut down, but the electrons are still coming off the hot cathode in the electron gun. With no force in the horizontal or vertical direction, the beam just shoots straight forward and hits the center of the picture bulb, causing the dot, at least until there’s no more emission from the cathode, or the HV drains off (or both.) another common symptom is having the vertical shut down while the horizontal (oscillator) is still running; then you just get a horizontal line vertically centered (roughly) on the screen. This then collapses (usually) to a faint dot, before fading away. The only real way to stop it would be to: 1) kill the HV at the anode cap by discharging it through a high-power transistor, or 2) on set shutdown, a capacitor would dump a negative bias voltage onto G1 or G2 in the picture bulb; that would pinch-off the electron stream until the horizontal and vertical circuits shut down, or 3) on set shutdown, a capacitor would apply a positive bias voltage to the cathode, to suck up all the electrons hanging around the still-hot cathode, until the horiz./vert. circuits shut down, or 4) design the set so that when the power is switched off, the filament in the picture bulb is turned off first, then - after a short delay - the rest of the circuitry shuts down, or 5) wire in a switch in series with the CRT filament and bring it outside the set. Now flip the switch first to kill current to the filament, then turn off the CRT. Just remember to flip the switch back on when you turn on the CRT or you’ll never get a raster! This is probably the simplest solution.

[@mal2ksc]

Couldn't you also cut the power to the cathode heater a few seconds before turning off the raster circuitry? Even though there will still be a high voltage present at the end, it won't have the thermal excitation to go anywhere.

[@Stoney3K]

That's also why older CRT's have an ion trap to deflect stray electrons from hitting dead center every time but instead, scatter them around the center area.

[@Stoney3K]

@@mal2ksc The thermal mass of the cathode will be too large, it will still cause emissions for dozens of seconds, up to half a minute after the power was cut.

[@mrnmrn1]

@@Stoney3K Not quite. The ion trap magnet is for trapping ions (who would have thought... :) ). Along with the electon beam, also some ions are being emitted from the cathode. The deflection system (neither magnetic nor electrostatic) has no effect on them because they have much higher mass compared to electrons, so if the cathode points straight towards the screen, the ion beam would hit the screen always at one spot in the middle. They not excite the phosphor to glow, but, over time worns off the phosphor on the screen on that area, causing a permanent dark dot at the centre. So they mounted the cathode at an angle, than used a permanent magnet to redirect the electron beam towards the screen, while the direction of the ion beam remains straight relative to the cathode, so it hits the neck of the tube instead of the screen. Voila, the screen is saved. Later they found out if they vacuum deposit a very thin layer of aluminum onto the phosphor layer, it protects it agains ion burn, while increases the brigtness in the same time (both because now the entire screen is on anode potential due to the conductive Al layer on it, and because the Al layer reflects the light from the phosphor towards the screen which otherwise would shine backwards into the CRT). It was a neat invention because it made the construction and setup of the tube simplier by eliminating the need of an ion trap magnet, increased brightness and probably the contrast ratio as well (because the light that was previously shining inside the CRT introduced some back illumination to the screen, elevating the black level).

[@mal2ksc]

@@Stoney3K You're probably right, and I remember how long it used to take for CRTs to come up to full brightness. But this isn't a very big one, so I thought the problem might be easier to control.

[@aziztcf]

V-Sync, ruining performance since '77!

[@ksaspectre]

This

[@scottlarson1548]

On the Atari 2600 vertical sync was the only time you could have performance!

[@georgemaragos2378]

Hi All Another great video Adrian i thought you would be similar age to me and remember black and white TV's up to 70's done that when switched off - i never knew why. But you have invented the turbo button for the PET This would be usefull for speed and if there is a operator watching screen you can turbo of to fix the snow This reminds me of changing the CGA cards on the IBM PC to fix the CGA snow at work ( and then also at home at the time ) A lot of people complained about headaches with the machines with CGA snow at the time Regards George

[@bf0189]

Thanks for having a fire extinguisher just in case something were to go down so the priceless antiques can be saved and appreciate the safety even if it is unlikely!

[@tonanornottonull7132]

i remember my black and white TV i had back in the day exhibited the same behavior after you turned it off. took about 5 seconds for it to go away -- i figured it had to be some sort of discharge. and the thing worked for 30+ years with zero issues!

[@davidjames666]

I am trying to fix my Dialysis machine for my kidneys. I tried to remove a dot, and now my blood gets too warm, and burns me when it enters my body again. I will look for tips on fixing my dialysis machine in other videos of yours. THANKS BUDDY!!!!

[@amcluesent]

Tried that poke on my Surface Book 2...nothing. Guess they fixed it.

[@Dr_Reason]

Does it overheat less now that you poked it? If so I will try it on mine, when it cools off and reboots.

[@marcgforc3]

you made a mistake !, try it on your printer, it will improve seriously the ppm ratio....🤣

[@Pedro8k]

I remember the bright dot on switch off way back when I fixed TVs a lot of customer though it was a fault or it could damage the tube it was just a product of the design of the early black and white TVs although there was a similar fault call frame collapse I think it was a long time ago still don't know how you could fix it keep the videos coming I went on to be a computer engineer as well later I remember the pet 2001 it was the only computer we had at school in the uk for a long time until it was replaced by the BBC computers

[@nonchip]

given that VIA has a nice amount of space around it, i think it might be feasible to make a "killer poke adapter" by putting 2 pin headers in the right distance (to fit into the mainboard's VIA socket) and a socket (to fit a normal VIA) on a pcb, and then having a jumper to select between pin 15 being passed through or being grounded through a resistor. feels neater than your cable and pin bending solution. plus you could attach a 3 pin cable to that jumper header that leads to a switch on the frontplate if you want a fancy switch to select between "fast" and "nice" mode.

[@СергейВладимирович-б9з]

As far as i remember what my father says about that dots on CRT, he says that this dot apears for 2 reasons: first there is difference of electric charge left between anode and cathode of the tube and second cause is that cathde still emitting electrons that are fly to the screen and make this glowing dot. There is a trick to make this dot scatter over the entire screen surface: you need to support horizontal and vertical deflectors in yoke to continue to generate correct deflecting signals for sone time, about 10-15 (on a bigger screens it can be up to 20-25) seconds after shut down. On our old tv-set there is custom homebrew board for this magic. And, yes, on some early black/white CRT tubes this spot this point burned out the luminophore. Later models of tubes is less vulnerable to this issue.

[@iz8dwf]

This is right and infacts, some monitors (and even TVs probably) do sustain some raster scan after the power has been removed. However, on a PET 9" monitor, there're no vertical and horizontal oscillators, the signals to sustain the raster scanning are generated by the PET logic motherboard, and that dies instantly at power off.

[@phr3dmcc0y]

I had always assumed it was just the leftover energy being drained into the cathode ray tube, since there was no signal coming in but there was still energy inside the apparatus, so I thought that was just the way to drain the capacitors. I grew up in the '70s and '80s I had an Atari 2600 with a black and white television set. I fondly remember turning off my television set and noticing the Little Dot in the center sometimes up to 20 seconds. As I got older and began to understand some electronic principles I understood things like batteries and capacitors and draining power and etc. So I began to realize, at least I thought so, that those old CRT screens were just draining leftover power, and we could see that residual power being discharged against the phosphorus.

[@PipThorne]

Be very careful when powering up a PET with the monitor unmounted from the main case. ALWAYS make sure the earth strap is connected between the monitor and the main case. If you don’t EHT transient spikes can travel down the video cables and take out components on the motherboard. I found this out at cost a while back.

[@eDoc2020]

When Adrian tested this monitor off of the main case I spotted some clip leads clipped to the chassis. It probably would've been good for Adrian to mention this. I presume the problem is the chassis is the only thing grounding the 'dag' of the CRT. Commodore should've put a ground wire between the CRT board and the CRT itself to avoid this problem.

[@rivards1]

The PET 2001 is my grail computer. Despite being a mostly Apple guy, the look of the PET has an unsurpassed for retrofuturism factor.

[@chrisnewman2008]

i want one just so I can display it as art!!

[@DukeFawks]

A quick and dirty fix for the spot is to put a cap across the G1 on the CRT. This will keep the negative bias on G1 alive while the cathode is cooling down. The only downside this has is that your brightness control will be slow to react as the cap needs to charge/discharge. A cap 100uF 50V should work fine and remember positive to GND as there is a negative voltage on G1. You can tack the cap on the brightness pot legs between the middle and GND legs.

[@GadgetUK164]

Excellent video!!! Looking forward to seeing Franks follow up vid!

[@nevellgreenough404]

It is often said that the spot is an indication of a weakening CRT. Weak CRTs get a soft cutoff characteristic and thus can't kill the beam as the supplies fade away. One possible fix: Ensure that the G1 grid has negative bias on it for as long as possible. A schematic in words.... Connect a diode in series between the -30V end of the brightness pot and the -30V supply, anode on the pot. Replace/shunt SG1 with a 47uF/50V or so electrolytic, positive on ground. This will hold the grid negative for 5 seconds or so upon turn-off. Hope this works!

[@adriansdigitalbasement]

I've actually finished part 2 and it's all solved. You're into something with some of the actual fixes :-)

[@nevellgreenough404]

@@adriansdigitalbasement Can't wait to see what you did!

[@unconformity5727]

I can confirm via some older CRTs I have even in color that the dot occasionally happens on some of those when you turn off the power, I've never seen it as an issue because I just thought it was normal behavior, but interesting to see it happening on a PET as well.

[@garthhowe297]

I remember as a kid, always staring at the dot on the tv until it went away. There wasn't much else to do back then, so you took your entertainment where you could get it ... lol

[@youtube-ventura]

That's the cleanest PET I've ever seen, bravo!

[@HAGSLAB]

Just what I needed right now Adrian! Thank you for keeping the content flowing. Have a great weekend!

[@McTroyd]

I also remember a black and white set we used to have would do exactly this. It ran fine for decades.... I think we only got rid of it when TV broadcasts went digital-only. I have heard spots like that can eventually burn the phosphor, but as old as this PET is, I'd have expected that to happen by now.

[@pegtooth2006]

I was l'Enfant Terrible in my computer science class in maybe 1984. It came to my knowledge that poking a certain byte would collapse the screen size enormously. I also found out that a three line programme that flipped the screen large to small repeatedly killed the crt on the School's PET. My friends never ratted me out. 😊

[@geofftottenperthcoys9944]

I remember lots of old TV's having the "white spot" when turning off.

[@grumpyoldwizard]

That machine is a beauty. What a great nostalgia blast! Thanks.

[@williamsquires3070]

Another idea is to write a 6502 machine-language program for the PET that’ll tell that 6522 to not generate interrupts when that pin goes low; test it, then burn it into an EPROM somewhere there’s space. Then, when the Pet BASIC comes up, you just call the decimal address where you hid the program in EPROM, and no hardware modification necessary; just keep the old ROM in static-proof black foam and stick it inside the case somewhere, should you want to return the machine to the stock ROM. 🙂

[@surebrec5113]

13:05 We used to get little white dots on CRT TV's back in the day after switching the TV off. Never put too much thought into why though. It was one of those things that just happened.

[@admirerofclassicalelectron2858]

You are a typical HP calculator user installing a RPN calculator app on your phone. Same as me. By the way a very interesting video.

[@ntsecrets]

I remember the dot appearing and staying on back in the day with my dads PET in 1981.

[@paulgray1318]

Lots of system you could cause physical damage via software back then: RM 380Z just single stepping thru the high resolution graphics driver would cause physical damage BBC Micro - loop with *motor on;*motor off off would turn the relay on the cassette port on and off so fast - bzzzz smoke. But anything that controls hardware that has limits, potentially cause issues.

[@FixedUntilBroken]

I'd love a video all about CRT's. Most channels won't touch them due to the danger of them.

[@rfmerrill]

A modern computer that's designed well would have a controlled shutoff that turns things off in the right sequence to avoid that kind of brownout behavior. To go further, some products I've worked on (datacenter network switch) have a control signal from the main power supply telling the power control circuitry when the power input has been removed, and it then has just enough time to shut things off in the correct sequence before all the energy is gone.

[@g0bzy]

Very late comment here.. Yes, the spot is completely normal. Many TVs and other CRT equipment like oscilloscopes have it. And its due to a cathode path to ground while the heater is still warm. A fix? any method that breaks the circuit from cathode to ground on turn off. (like a mini relay going open?). I'm sure there is a more elegant way to do it but thats the idea. Lots of fun!

[@whiskerlesswalrus]

In older TV's B&W and Color of higher price-they used to employ a "spot killer" circuit to dissipate the spot because it would bother people-the reason the spot happened is the cathode stays hot after the TV was shut off and it was still streaming electrons and with the deflection circuits turned off the electron stream is not being deflected so the electrons just stream towards the face of the tube until the cathode cools enough to stop emitting electrons and the HV dissipates-one technique the used to do is when the set was turned off the picture tube would be biased on full so that the fully biased picture tube would conduct heavily for a split second and dissipate the HV rapidly and with the HV dissipated there was no acceleration voltage to the face of the tube and no spot--just research CRT spot killer circuitry--the spot was a concern on older TV's because when it happened it could burn the phosphor on the face of the tube and it was a permanent dark area in the center of the face of the tube when B&W TV's used Ion trap magnets and was lessened when the aluminized picture tube came out

[@Fifury161]

Screen burn in occurs on CRTs when the phosphor coating burns off, this process will always happen given enough time. Screen brightness & constant displays will accelerate this. Seeing that dot at power off is not really a fault and a recognised design choice (safer to discharge in the tube than to chassis - saves component count to). Given that most CRTs are rated with between 20,000 & 30,000 hours it wouldn't normally be a problem. You would need to perform many power cycles to cause a permanent dot.

[@Stoney3K]

In Commodore's defense, they didn't expect their CRT's to last almost half a century.

[@principals16842]

I'm not sure if it has been mentioned before, but Kirk has a PET in his apartment at the beginning of The Wrath of Khan when McCoy shows up with a birthday gift. I don't think it's possible to tell the model, but I love the idea of a PET still going after 300 years.

[@dodgydruid]

My first ever "proper" computer I managed to salvage from a fire damaged business and it was toasted good and proper that Pet was. Luckily the unique chips had survived and the unit was little damaged except it didn't work and over the following months using a hoary old "How to: Electronics" book and salvaging components from broken stuff I slowly put it back together and got it working. I had a nasty glazier's soldering iron that stank horribly and back then solder wasn't cheap and the fluxed solder costly but it was a lovely moment after many long long hours it sprang into life but did pong a bit of burning smell from the fire and I sold it quite quickly using the money to buy my first IBM PC and something Sinclair prob the Spectrum 48k and because the IBM was also broken I started again with my little box of harvested components set about getting that up and running and was far easier than the Commodore imho. Has to be said in those days it wasn't the easy architecture we know and love today, none of this swapping out of cards etc and you could kill your motherboard quite easily through heat death, lightning spikes, telephone bell too close etc and because it was new technology the quality control wasn't what it is today with CPU's taking an early shower due to poor manufacture of components.

[@bigalejoshileno]

The "solution" for this is to enable some sort of horizontal blanking interrupt. However this would make the software consume a little bit more of cpu time, with some extra latency costs because of the interruption, but characters for the next line could be updated in the previous line by the cpu. This is why old video game consoles had registers to enable it only at required moments, but this allowed lots and lots of precise animations or pallette changes. One thing that made the pet a better performer, but less flexible computer at lower cost, was the absence of a character rom remapping mechanism. Probably something like that would make this computer almost perfect, and the inclusion of some dedicated ram for video, additional to both interrupts, would made this almost unbeatable. Sadly, memory and buttons were expensive, but is considered an error from commodore to no design this computer to support more "high end" models.

[@lelandclayton5462]

The yoke is like a Inductor. When powered on it has a magnetic field and when you power it off the magnetic field collapses and creates a high voltage. It's visual since the heaters to the CRT are still have emission. Atari with their Vector Arcades by putting protection Diodes on one side of both yoke coils for X and Y. However it's more then a few simple Diodes and requires adding Transistors and crap. In all honesty this will not hurt the Monitor, unless you constantly turn the System ON and OFF forcing it to burn a hole into the CRT phosphor.

[@marcppparis]

I have the same 2001 Chiclet PET (blue frame monitor) and mine does the same with the dot when turning off the computer. I did install a Tynemouth PET ROM/RAM board so I can run BASIC 1,2,4 and 32k on it. I purchased a dedicated 6502A as well so i have a spare if I want to go back to stock config and not worry about mucking up some pins. I do get a shimmering "jumpy" screen though - will need to investigate that

[@nickbnash]

Thanks for the video! It was great. Looking forward to the next one.

[@rastislavzima]

I'm still watching, I'm on 10th minute but already can tell this is very good interesting video.

[@cppsage]

I am half-tempted to send a nicely-made bag of rye for Adrian so at the beginning of every video he can put it next to the thing he tests as he says “Let’s get rye to it”. :D And I still think the plushy ram should be named Eccy.

[@spacewolfjr]

a bag? why not a bottle of rye?

[@cppsage]

@@spacewolfjr USPS is able to break titanium in shipping, so a malleable container seemed more appropriate.

[@spacewolfjr]

@@cppsage Ahh, as a fellow Canadian to Adrian I was thinking Rye whisky, thus the bottle :)

[@cppsage]

@@spacewolfjr Hmmm. I am type 2, I don’t know how whiskey and type 1 diabetes works. I just thought of the grains and the horrible pun. :)

[@spacewolfjr]

@@cppsage Either way don't inject it into your body (except by mouth)

[@BrianRRenfro]

Me: "OOOOHHHHHH this is starting with the 70's trash towel! This is gon' be good!

[@mikesilva3868]

kzbin.info/www/bejne/pGTQaIulrJWMrqc

[@ThePillenwerfer]

I seem to recall with CRT TVs from the late-'80s that if you switched one off properly the picture would quickly collapse to a dot that would more-or-less instantly vanish. If you just pulled the plug out though they'd do the same as your PET.

[@Samuel-ge7im]

Great video, love both channels

[@AdamBouzaneOfficial]

RealCalc in RPN mode! a True Nerd! love it

[@lindoran]

Awesome! Love commodore stuff!!

[@reeeeeeeeemmmmmmmmmm]

Seems to me that if you permanently insert a resistor between the VIA pin and the sync signal, you get a PET that behaves as normal but which supports the killer poke without damaging components. In default behavior the resistor should not have any effect on the signal going into the VIA, and after applying the killer poke the resistor will take the voltage drop. 1k should do it, then the two outputs would only have to sink/source 5mA.

[@dragonhed123]

Man I enjoyed this video the pet is such a beautiful computer!!

[@DrHouse-zs9eb]

Love your short but interesting intros!

[@stuartirwin3779]

Although you've apparently solved this now, it is perhaps noteworthy that many B&W TVs had what they called a "spot swallower" circuit, which turns the CRT full on after switch off, so the HV is discharged quickly while the deflection circuits are still operating. They did this by making sure the negative bias on the control grid is the first voltage to collapse, turning the gun full on. Hope that helps. 😊

[@stuartirwin3779]

@@HutchCA Please re-read my comments. The answer to your objection is there.

[@stuartirwin3779]

@@HutchCA Perhaps you should re-read your own comments too! 🤣

[@eDoc2020]

I've never seen the poweroff flash on BW TVs but only on color TVs.

[@stuartirwin3779]

@@eDoc2020 I do actually remember seeing the switch-off flash on B&W TVs. Usually reaches highest brightness when the scan has collapsed to about half the screen area, from memory.

[@iz8dwf]

The problem on a 9" PET monitor is you can't be (easily?) faster than both deflection waveform to collapse, since there're no deflection oscillators, only pulse shapers and deflection drivers. The oscillation pulses come from the PET logic board. Deflections collapse as "ringing" waveform very fast, so you wouldn't discharge all the stored charge into the anode. Also, on earlier 9" revisions they made some real mistakes. However, you're free to experiment with the G1 "fast bias decay" and see if you can outrun the collapsing raster waveforms :)

[@thepcman]

If the performance almost doubles, may be you can connect a pll to double 60 hz signal to 120 hz and feed it to vla, that way you will still get high performance and minimized tearing

[@kirbyyasha]

Leave the spot. It was made to do it, and just very iconic of turning them off.

[@Turnbull50]

that was my first computer i received it on 14th of March 197. Mine also showed the dot as well and it never caused any problems.Some kids used to use that killer poke in computer shops to destroy the odd Pet computer.

[@fitfogey]

New subscriber here. Love your channel and content!

[@DiazFelix]

So add a swich to enable an disable the "Turbo mode"

[@dennis8196]

Just remember to disable it when saving to cassette

[@FordGT40MkIV]

Eliminating the 'snow' by only writing to video RAM during VSYNC always struck me as a cheap 'fix'. I designed a video card for my Altair 8800 using DRAM for the video memory running at twice the processor frequency, with alternating phases dedicated to the CPU and the screen refresh. It worked great - unlimited CPU access and no snow.

[@Wenlocktvdx]

CoCo (Tandy/Radio Shack Color Computer) had a double speed poke which doubled the clock frequency. It only worked for a few seconds as the RAM refresh was too fast. It has another option that only doubled clock speed during ROM access. The ROM only setting was less than double speed but still faster.

[@leonardochiruzzi7642]

Another computer that I would like to have in my collection, but difficult to have !!

[@GigsTaggart]

you wont believe what I did in 1995. A guy had a garage full of IBM PC and PETs, from a school auction i assume. I bought the 20 PCs to resell and passed on the Pets as worthlessly obsolete.

[@leonardochiruzzi7642]

@@GigsTaggart Nooooo !!! Why did you tell me this? now I want to hit my head on a sharp edge. How sad. Why don't such fortunes happen to me? Why?

[@patprop74]

13:30 my older Analog oscilloscope does that as well, I think it may be normal for older CRT's to do that.

[@stevef6392]

This reminds me of the old 12" b&w TV that I used as a kid (and still have stashed away in a closet). It does some weird "CRT stuff" as well. For example, if I plug it in to the same outlet as an 1800w toaster oven, the image on it shrinks. Newer CRT displays, like a 21" NEC monitor and a 34" Sony TV do not seem at all affected by the toaster oven. Better power supplies, I'm guessing.

[@eDoc2020]

Older TVs usually have unregulated power supplies and newer displays have regulated supplies. The picture shrinking is because of decreased voltage to the deflection circuits. Monitors tend to use tightly regulated switching supplies and TVs apparently tend to use relatively inefficient and slightly less stable series regulators. TLDR. Yes, better power supplies.

[@needgamesnow3466]

My black and white from the 70’s always did that.

[@The_Studioworkshop]

I’ve recently re found your channel again, and I have been absolutely loving all of your videos - but this particular one shows just how misinformation can spread, about the spot! For years people on the Facebook groups the “experts” have been saying spots are caused by bad this, and low emission that... in fact, this particular way of doing things goes right back to the 30s at the start of television!

[@moshly64]

My 9"PET has no dot on power off, PCB is marked CDU-B79-CMD, the tube is an AMPREX M24-392GH. More modern CRT monitors have a spot killer circuit. (google: CRT spot killer) Long persistence of a dot may be an indication of circuit failure.

[@akkudakkupl]

Grid 1 bias source for the picture tube needs more decoupling, it seems to collapse too fast from -30V, and as the bias disapears electrons from the gun have free reign to accelerate towards the screen.

[@dennis8196]

Some versions of the Dragon 32 and the Dragon 64's were subject to chip damaging pokes. On some versions of the Dragons the same pokes doubles the processor speeds, there was a 2 pokes, and when used together you cooked the CPU, on others machines they has either no effect or the effect was to freeze it

[@FireballXL55]

The spot on your pet is too large to worry about spot burn, spot burn occurs when you have a stationary focused spot of around 1 to 2mm, and is worse on monochrome monitors as there is no shadow mask to take some of the energy. Monitors where the spot suppression circuit has failed will strip the phosphor in 1 turn off. There are two main ways of dealing with this; 1 you can dissipate the CRT energy while the scan circuits are still scanning but this is reducing with time from off. 2 you can cut off the CRT usually by taking G1 negative. The first option is to turn on the CRT hard at power off so the energy is discharged before the scans shrink. The 2nd was to take the G1 hard negative on power off, this was often achieved by having a transistor that either shorted part of the G1 divider chain or disconnected the G1 control GND end on power off.

[@eDoc2020]

Do you think a cap between the 12v line and the grid would provide enough of a negative voltage to cutoff the tube?

[@FireballXL55]

@@eDoc2020 No I don't think it would. I would probably increase the cap on the -30V rail and try to disconnect the 0V end of the brightness pot on power off. This could be done with a transistor and the base controlled by a zener to the 12V rail in some way. You could also increase the video HT capacitor value.

[@yaroze]

RealCalc and it looks like its in RPN mode! Cool!

[@SledgeFox]

Amazing video! Thank you very much!

[@jimstand]

The original IBM PC 5150 with a monochrome monitor could blow out the monitor's horizontal output transistor. When powered up, the video controls could come up in a state that would blow out the HOT if it was not reset. If there was a BIOS failure BEFORE the 6845 was reset, the horizontal output transistor would fail. Eventually the BIOS was modified to put the 6845 into a state that would not blow out the HOT before testing the memory, etc. I know this because a friend blew out 3 monochrome monitors before he figured this out. The BIOS of the original 5150s could not be rewritten so this time bomb was in the 5150 for ever.