How a 555 timer broke early C64 boards
HTML-код
- Опубликовано: 7 авг 2024
- The early Commodore 64 326298 boards have a flaw that prevents cartridges from using the reset line, which causes problems with the Final Cartridge, EasyFlash, Fung Fu Flash, and BackBit among others. I’ll show you how the 555 timer reset circuit works, why it causes the problem, and a simple mod to fix it.
Links
www.lemon64.com/forum/viewtop...
www.lemon64.com/forum/viewtop...
hobbytronics.home.blog/2023/0...
• Commodore 64 Rev.A 326...
Chapters
00:00 - Intro
00:25 - The schematic is wrong!
02:58 - How the 555 timer works
05:27 - Looking at signals before
06:55 - How to do the mod
08:03 - Looking at signals after
10:34 - Outro 
Наука
The 555 name is a myth. I'd love a video just about that and other weirdly named components. I heard many different reasons it's called the 555. In fact one of them is from someone who worked at TI when it was developed and alleges the name was chosen, almost at random, by some employee in the marketing department. The 5k resistors were a coincidence.
I'm not sure what to believe but I'm interested in the truth. It's like the 23 on Rolling Rock bottles. So many conspiracies about that. However it was the one brothers racing horse number. The other brother had a horse too, and they tossed a coin to decide which one got their horses number on the bottle.
Lots of cool stories out there like that.
By the way, great video, good explanations. Weird how the original development team missed this obvious (after the fact, hindsight 20/20 right) error on the board.
Rats, I was fooled! It was such a good mnemonic for introducing the 555 timer that I had to use it.
Isn't it a 556? Dual 555.
In Thailand, 555 means laughing because in Thai language, 5 is 'ha'.
@@robscovell5951 "ha ha ha 😂 " I like that. Thanks for sharing!
@@robscovell5951 What is '556' in Thailand? ;)
When you realise that 1984 is to now as 1944 was to 1984!
I know! If Marty McFly went back in time today he'd go back to 1994. That's shocking.
literally 1984 😔
Capacitors act as a short when current first flows - at t0 basically (the differential equations for RC circuits shows this). In that new configuration the threshold and trigger pin were basically shorted to +5v when power is turned on, and as the cap charges it drops all of the supply voltage. It basically inverts the RC curve when you do it like that. Somebody screwed up those version A boards.
The 555 is a pain to use for resets, I used it on a board of my own and spent way too much time debugging and tweaking it until it actually did what I expected. By the time it does the right thing, you've already added more parts than planned and wasted too much time.
I've never bothered with an IC, I just draw up an R/C tank that feeds a PNP, then feed that into an inverter (or even an NPN). Makes a negative pulse about as long as the RC constant at power up. If you need external signals to reset the computer, just have them short the capacitor via switch or transistor which will drop the reset line.
@@8bits59 -- Plus now, you can get 3-pin power supervisors. They look like transistors, but they keep the output shorted until Vcc is where it needs to be. So it pulls the reset line low.
@@8bits59 A resistor, a capacitor, and a Schmitt trigger buffer/inverter should do it.
excellent video!
The noise is probe-loop pickup
I’ve got a squirrelly Rev A in the basement, this may be useful after I finish unboxing from my move.
I have to check. Mine has the 5-pin video connector, but I'm not sure if that means it's Rev A for sure.
@@cathrynm I think there are a few early boards with the 5 pin connector (326298 Rev A through C). The only way to know for sure is to open it up and look at the motherboard. The part number is on the edge of the board closest to the user. I’m not sure what the differences are between the 326298 Rev A and B or C, though.
Sinclair never went that complex for their reset circuits, they just connected an RC network directly to the Z80's reset pin!
Yep. Genius, thats all it needed.
The early Commodore PET had a similar circuit and the capacitor had a tendency to fail stopping the machine from booting. Briefly shorting the capacitor would trigger a reset and start the machine.
Cool! That's exactly what @DavidvanDeijk was asking about
I'm pretty sure it's not 1/3 threshold in this case because the middle resistor at the input of the 555 is shorted (as you mentioned). In this case, the threshold is at the 50% point.
nice fix. 🙂
Lovely
3:05 this is a fun myth
i do wish it were true and not just because thats just the name it was assigned
its more romantic
Yes, I was fooled! If I ever have to explain a 555 timer again, I'd still say "there's a myth that the 555 timer gets its name from..." It's just too good of a mnemonic for explaining how the chip works to not use it.
@@TrevorMakes- Wait, didn't the 555 come from the 3 x 5K voltage divider/reference voltage? I might have this same Rev. A board stashed away since 1990 in a box, in will dig it up one day before I croak. lol
Nice fix. They should have done that in the first place.
Nice Video! What is the pin # you snipped? -Mark.
Pin #9, 2nd from the bottom on the right side. It's important to leave the bodge wire connected to the PCB where pin 9 goes, so I snipped it at the top close to the chip, leaving the pin in the board.
Also, check the links in the video description! There are some forum posts and another video that describe the process.
Ehh, just pick any random pin, doesn't matter. haha Just kidding, he cut pin# 9 (OUT) which was connected to pin# 13 (DIS). DIScharge is enough to reset the machine.
Is my hunch right that that reset capacitor can probably be damaged pretty far and the board will still boot?
Depends! If it was just a derating of the capacitance, it could simply make the reset pulse shorter-say it dropped to 5uF, you'd get a 0.25 sec pulse. In a catastrophic short circuit failure, it would hold the trigger pin high and stay in reset indefinitely. An open circuit failure? Who knows!
Yes was thinking of the first thing, i am not so solder iron inclined so only ever had to replace derated capacitance ones once
ElectroBoom recently managed to break a capacity so it both got a significant amount of self discharge / parallel resistance, and reduced capacity. In this specific circuit those two effects can actually cancel out each other.
So I guess anything is possible...
I just looked at my A board, it doesn't have a the bodge wire.
Mine had the bodge wire but the board had a trace parallel to it. Makes me think the bodge was added just to protect that trace in case the reset line was brute forced low by an external peripheral.
Now that's weird... It looked to me like the circuit should work without the bodge wire, since the output pin alone should be able to drive the line low, even if not as strongly as the discharge pin is able. But I doubt they'd have added that step to the assembly process unless it was a reliability issue in some cases.
I don't see a trace on the board connecting out and discharge (9 to 13), but there are traces connecting 8 to 12 (trig/thr) and 10 to 14 (rst and vcc), which could easily be mistaken as doubling the bodge wire.
@@TrevorMakes it's been a couple years since I looked at it and it's possible that I'm mistaken, but I don't think so. At the time that I did the mod, I was certain that the bodge parallels a trace. I might have to pull it out and take another look
It would certainly be interesting to find out... I've looked through the board with a flashlight, but it's possible there's another trace hiding between the chip and the top side of the board. I just don't feel like desoldering the chip to find out.
I just dug out my board. It seems there's more than one revision of the 326298 REV A. Mine definitely has a trace on the top side of the board between 9 and 13. I can fairly clearly see in the video that yours does not have that trace. I'm sure that my board had the bodge but it seems that I removed it when I did the reset modification because it was redundant and won't be seeing any significant current since the modification was performed. I also removed the cut off pin 9 leg from the board.
TheRetroChannel has a video on 326298 mods and if you look close, his board has the trace on the board as well as having the bodge.
I also did a few other mods. I replaced C10 and C11 audio capacitors with 470pf Poly caps to restore the noise channel of the SID. I paralleled R10 with a 220 to restore brightness and lifted one leg of C73 to improve the composite video. I installed a Bwack ROM switcher board to run JiffyDOS. I also put a Bwack board in a 1541 drive to switch both ROMs and Drive ID and mounted a small push button to the face of the drive. One of these days I'll get back to playing with the C64 and set up the drive for parallel data/SpeedDOS. These computers are so much fun to tinker with.
4:12 - ... the capacitor will be charged - not discharged - right❓
It is charging, but the voltage seen from the trigger line point of view will be lowering since at boot the capacitor will be empty 5V will be seen across the resistor since they are in series between 5V and ground
As the capacitor charges up, the resistor will see less and less voltage since they need 5V between the two
What he was saying at 4:12 is that when the system starts up, the capacitor will _start_ (be) in a _discharged state_ (not yet charged). I'll admit, he didn't say it in the most clear way, though...
So it _starts out_ discharged, and then it _becomes_ charged over time.
Right, I didn't explain it as well as I'd hoped. Capacitor starts out empty when the power is switched on and slowly charges up. As it charges up, the voltage across the capacitor increases from 0V to 5V, and the resistor in series eats up any voltage left over. Having the capacitor above the resistor makes it a little more confusing to explain.
By the way I have a weird issue with an Atari stf 1040. it’s super slow, the mouse barely moves. Programs barely load and are unusable. My other Atari works fine. Could this be a known issue? Or has no one heard of something like that?
2:40 Wrong reading of datasheet. Can't compare absolute max rating of DISCH pin with a OUT pin characteristic, that just specifies voltage drop at the specified current of 5 mA.
Fair enough, I oversimplified a bit to not have to bring up datasheets for the 6510, CIA, and SID chips on the reset line. The rated max for low input voltage on those chips is 0.8 volts, so going by the fig 7 shown in the video, the output current would need to be less than ~15 mA to pull the reset line low enough. I didn't see a similar graph for voltage at the discharge pin vs discharge current, but I figure it's much better suited for this purpose than the output pin.
Will you pls pls repair my 2x PC-XT motherboards? pls
Hi Ayan, are you still having a problem with the 4-bit DRAMs, or have you found other issues?
Something like an incompatible joystick broke the 64 and 16 and vic20
I've heard that with the Sega controllers. Commodore used mostly the same pinout as Atari. Sega was close enough that the D-pad would work work but certain button presses could break the interface chip. Others like TI-99 used the same connector, but completely different pinout.
Sega controllers use a parallel-in/serial-out shift register to transmit button states via clock and data line. The C64's joystick port doesn't work like that. Amiga CD32 controllers also use a shift register but are backwards compatible.
@@Kobold666 it's funny when you was a kid and you had a commodore 16 and you would find a game for a vic 20 and you would still try to load it anyway you would think its a commodore a least the make of machines right oh please let the game work
@@Mr.1.iAs a kid I tried to run the MSDOS version of Leisure Suit Larry on the C128 because I thought CP/M and DOS were basically the same PC thing.
I never thought that the cheap skates at Commodore were put in their place some forty years after they committed their crimes against computing by misusing and mistreating a 555 timer. Only to save a few cents by not using an inverter they willingly reversed this 555 timer to become a trans-times and let it work the other way around, against his will, and against his nature!
How totally hopeless. Did they perhaps keep giving this reset circuit problem to the newest unpaid intern to solve?
Is that a NE555 or a 556?
The 555 only has a DIL4 housing, the 556 a DIL8
It's a 556 (in my case an MC3456). The left side is an NMI pulse driven by the Restore key on the keyboard, the right side is the reset pulse described in this video.
555 is in a DIP-8 package, 556 is a DIP-14 pin package, with the two timer circuits sharing Vcc and Gnd. en.wikipedia.org/wiki/555_timer_IC
R34 is connected to ground? It should be.
Yes, R34 goes to ground and C24 goes to +5V. In the schematic and in later boards this is reversed.
I was trying to make a joke about the naming of R34
how could voltage increase on the cap and decrease on the res while they both are tied together. this makes no sense
What he meant was that the voltage _across_ the capacitor increases and the voltage _across_ the resistor decreases. That is, the voltage measured between the two pins of the capacitor (+5V and the common point) increases, and the voltage measured between the two pins of the resistor (the common point and GND) decreases.
Keep in mind that voltage is always relative, not absolute. Voltage is actually a difference of potential _between_ two different points. By convention, when measuring in a circuit, one of those points is usually GND, but it doesn't have to be (and in the case of the capacitor in his example, neither one of the points being measured are actually connected to GND).
Exactly! I need to come up with a better way of explaining (and animating) the potential difference on series components.
Each part in series gets a share of the total voltage difference from top (+5V) to bottom (GND). Capacitors "resist" changes in voltage, as in they take time to adjust to a change in voltage while they charge/discharge, like a water tower getting filled up by a narrow pressurized pipe. When the power is switched on, the capacitor starts discharged (0 volts) and slowly acclimates to the +5V. Resistors don't care about changes in voltage and happily pick up the slack while the capacitor is charging.
not to many people would still have a comadore C64 these days
Not many, but they're still surprisingly popular. I think 8-bit computers (including more recent things like Ben Eater's 6502 on a breadboard series) are very powerful at teaching low-level fundamental details of computers in a way that's lost with more modern systems where everything is a black box built on multiple layers of virtual machines. I'm doing my part to try and keep it relevant!
@@TrevorMakes excellent
Search RUclips for c64 or Commodore and you’ll be surprised.
my gawd. just rip that crap out and insert a dallas econoreset. instead of 'patching up' tramiels el-cheapo screwups. lol.
Kinda funny that the original C64 breaks modern cartridges, which are designed to basically jailbreak the C64..
Jailbreak out of what exactly? What are you being prevented from running on a stock C64??
@@8bits59 cartridges and software you don't own, alternative ROM code without changing the EPROM, .... Did you think this was low-hanging fruit?
@@rockosgaminglogic The C64 did not have any such protections in it to begin with. Replacing the built-in ROMs with cartridge ROMs was actually _an explicitly built-in feature_ of the system (there were signal lines specifically intended to do that). No need to "jailbreak" anything to do it. And the system ROMs didn't have any kind of copy protection mechanisms in them to circumvent anyway. That kind of hardware-level copy protection stuff didn't come along until much later.
The cartridges using the reset line weren't trying to "jailbreak" anything (there was no jail to break). They were just using it as a convenient way of signalling from hardware to software that an event occurred (similar to a hardware interrupt), that's all.
Right, there is some interesting/annoying copy protection on some disk drive software, like checking for intentionally corrupted disk sectors that might be skipped by disk copying software. Some software even used dongles you'd plug into the joystick port.
The reset line trick is just a convenience to switch between ROM images while browsing through the fancy menus built-in to these kind of cartridges.
@@foogod4237 A modern hardware jailbreak is a process that involves physically modifying the device's hardware to bypass or remove its security and restrictions, allowing the user to have full control and access to its features and capabilities. This could include removing or bypassing manufacturer-imposed limitations, such as installing unauthorized apps or customizing the operating system. In terms of a C64, which *obviously* does not need jailbreaking, "basically jailbreak" would mean being able to use the hardware to the full extent it was designed, which you can do with something like a 1541U2+.
THEY WERE PROS 😊😊
Imagine having 2 comments but 0 likes on either.
JUST REPLACE TIMER AND FIX DRY JOINTS THEN CLEAN ALL IC SOCKETS 😊😊
Nah it’s more entertaining to understand this factory bodge