Anonymous
User login
Please support the defense of Ukraine.
Direct or via Unclutter App
Active forum topics
Recent content
Navigation
No Ads.
No Trackers.
No Social Media.
All Content Locally Hosted.
Built on Free Software.
We have complied with zero government requests for information.
Spaciba Macintosh_nik !
I tried to type the hexa codes of the test found here https://www.willegal.net/appleii/6502mem.htmbut I must have made an entry error that I cannot locate.
IF that the woz mon include a checksum during memory dumps should be better.
(I know...there is only 256 bytes in proms.)
Finally I was able to load Mike Willegal's Memory TEST on the russian web site :
https://disk.yandex.ru/d/PkeDgYyDib1gqw
I had some difficulties with the AIFF. under android but VLC player solve it.
For load Mike Code from ACI :
C100R
0.3R 280.3A1R
The 4 first byte of the ram indicates the range of ram to be tested
You will need to configure them like this
To test W0-W7 RAM bank :
0: 00 E0 00 F0
To test X0-X7 RAM bank0: A2 03 00 10
Lauch the test :
280R
enjoys...
we cannot test the lower part of the ram X0-X7 because the code is housed there.The best for an exhaustive test is to swap the X and W chips and only test the W
You're welcome, glad to help. I use this free VLC player too.
By the way, I had already seen some of the Apple 1 programs but in the form of a book that I found at the municipal library around 1985...This book is available in pdf here:http://www.bitsavers.org/pdf/dec/_Books/101_BASIC_Computer_Games_Mar75.pdf
Correct link is .....................
http://www.bitsavers.org/pdf/dec/_Books/101_BASIC_Computer_Games_Mar75.pdf
Indeed the link was invalid.
thank you Tulack
Here is what Woz wrote about this book:
After a procrastination worthy of the North Pole* I started to edit the rushes
here is the first episode in French but I added English and Russian** subtitles.
On the menu: unboxing “Uncle Bernie kit” and additional components
with cost (with 2023 prices)
https://www.youtube.com/watch?v=y-iw9zcwsjo
*the next day is in 6 months.
** I don't read Cyrillic, I did the AI translation with Reverse Translation to check that the shifts in meaning are minor
Good video, it was nice to watch. I even wrote you a little bit there, which I do very rarely. Thanks for the Russian subtitles, they are much better than the ones created automatically by the YouTube algorithm. The last ones are closer to the Elvish language, or the language Master Yoda speaks, but it's definitely not Russian. )))
A new episode:
Building the 240v/50hz power supply for my Apple 1 clone.
https://www.youtube.com/watch?v=jclH6x4ZCH8
I changed the workflow to produce the subtitles
1st episode French -->English via copilot-->russian-->google translation
2nd episode French -->English via copilot-->Russian-->via copilot
@macintosh_nik can you tell me if the Russian subtitles seem improved to you?
Very detailed video, thank you! The Russian subtitles are clear, but are missing in the little bit after you cleaned the soldering iron complaint.
I would suggest to put some music on in the places where you don't say anything, because it's quite sad in silence....
@Macintosh_nik
Very detailed video, thank you! The Russian subtitles are clear, but are missing in the little bit after you cleaned the soldering iron complaint.
actually 3 minutes of subtitles had been forgotten from the Russian version. so sorry ! This has just been corrected.
I would suggest to put some music on in the places where you don't say anything, because it's quite sad in silence....
On the other hand, for the music I am very surprised because the musical track is on 100% of the video.
But I attenuated it by -30db so that the voice part is more audible on laptops.
I checked on my 2 laptops it is audible without speakers on both but the sound is better on acer than on the thinkpad
Do you have amplified speakers?
A new épisode of the construction of my Apple 1, about the Carbon vintage Resistor :
Miniature.png
https://www.youtube.com/watch?v=HfF43WmhIos
As usual with subvtitle English en Russian
Thank You!
French: Super vidéo, j'ai bien aimé les astuces de soudage et pour l'espacement des résistances. Je suis également en train de travailler sur une réplique. J'ai presque toutes les résistances au carbone également mais je n'ai pas encore commencé le soudage, donc cette vidéo arrive au bon moment.
English: Great video, I liked the soldering tips and tricks and for the spacing of resistors. I am also working on a replica. I have almost all carbon resistors as well but I haven't yet started soldering, so this comes in at a good time.
Hello there !
I am preparing my next video on assembling the Apple 1: The Diodes. I want to make sure I don't say too many incorrect things. I am well aware of the Dunning-Kruger effect and had slowly climbing the mountain of stupidity ;-).
DIODES CR1-3.png
The electron gun has to move back up to the top of the cube. For its internal organization, the counter must start at a specific value.This value is indicated by the inputs P0, P1, P2, and P3, some of which are connected to ground. These starting values are activated when pin 9, which is the load pin (indicated here as /PE), briefly goes to zero. I suppose it has an internal pull-up, so its normal state is 1.
We have three vertical signals: V5, V3, and V4, which are the respective bits of the vertical counter, plus a VBL signal, which indicates the resetting of character values contained in the 2904s, for example, during text scrolling upward or a carriage return. The values V3 and V4 are combined by an OR gate, so the value of output 13 can go to zero when the required conditions are met.
Diodes CR1, CR2, and CR3 form an OR gate (indicated as "or" in English) that isolates these signals from each other. If one of these signals goes to zero, /VSYNC goes low, causing the LOAD function of pin /PE to go to zero via the diode, thus loading the counter to its initial state. This transition to zero will not be hindered by the state of other pins, which, if at level 1, are blocked by the diodes.
In summary, these three diodes indeed form an OR gate, which would have otherwise required purchasing a three-input OR circuit, increasing the board's cost. It's a "poor man's OR gate," allowing Steve Wozniak to eliminate a circuit.
Thank for your help !
Lost me here:
"If one of these signals goes to zero, /VSYNC goes low, causing the LOAD function of pin /PE to go to zero via the diode, thus loading the counter to its initial state. "
I would think "If one of these signals goes to zero", the /PE (or LOAD pin) will go low, causing the preset on pins P0-P3 to be transferred to the outputs on Q0-Q3. If P1 and P3 are not connected as shown in the schematic, maybe they float high? In which case P3 -> Q3 which would cause /VSYNC to go high.
Yes, Retro-Tinker, I think you are right. The floating inputs of the 74166 should be 1.
However, for the video signal to reach a true zero volts, /HSYNC and/or /VSYNC need to go to zero so that the output 8 of ICC15 goes to 1 and is inverted to 0 by the output /Q1 of the 74175.
I simulated this part of the circuit in Proteus.
video.png
When this output is at zero, with the QH output of the 74166 also at zero, we get a true zero volts on the video output for synchronization.
I must humbly admit that something eludes me...
simulation timing.png
I added some circuits in the simulation.
Unfortunately, DS0025, 2504, 2519 and 2513 are missing in the Proteus nomenclature to make it a complete Apple 1.
This helped me to better understand the role of diodes CR1 to CR3. This is a logical AND.
The conditions are realized between scan lines 224 to 237.
For the other lines, pin 9 (load) is 0! And the value "$A" cf D0-D3 which is loaded on outputs Q0 to Q3 of D15(74161).So the /VINH (Q1) and /VSYNC (Q3) signals are permanently at 1.
From line 224, pin 9 goes to 1 and the counter can start counting A, B, C, D, E, F, 0, 1, 2, 3, 4, 5, 6, 7.
At this point, pin 9 goes back to 0 and the preset value $A is presented on Q0-Q3
summary in table below
Count.png
When the counter takes the values C,D,0,1,4,5 the output Q1 is 0 and therefore /VINH is active and blocks the vertical counter which will repeat the values E2,E4 and E6 twice. Thanks to this repetition, we artificially obtain 262 lines numbered from 0 to FF with only 8 bytes.
and when the counter is 0,1,2,3,4,5,6,7 Q3=0 therefore /VSYNC is active on lines 230 to 237.
In post #66, 'wirehead' wrote:
" For its internal organization, the counter must start at a specific value.This value is indicated by the inputs P0, P1, P2, and P3, some of which are connected to ground. These starting values are activated when pin 9, which is the load pin (indicated here as /PE), briefly goes to zero. I suppose it has an internal pull-up, so its normal state is 1."
" Diodes CR1, CR2, and CR3 form an OR gate (indicated as "or" in English) "
Uncle Bernie comments:
"floating" inputs of TTL logic count as "H" = high = logic 1, and this is a trick that has been used since TTL appeared in Y1963. But contrary to popular belief, there is no internal pull-up. What really is there at the inputs of classic TTL gates is a multiemitter transistor with the emitters being the inputs. The transistor is turned on when enough current is being sucked out of the emitter. This then is the "low" level on that input, and the reason why it needs so much current being drawn. On the other hand, the "high" level on the input, once stabilized, only needs to deal with the reverse leakage currents of the base-emitter junction of this input transistor. Crude 1960 bipolar process technologies had high leakage currents, especially when hot, and this is where the double digit microamperes "H" level input current specifications come from. Early crude Schottky TTL was even worse in terms of leakage currents, due to the input circuit topology being different und using leaky Schottky diodes. Most of these issues with leakage went away with improved wafer base materials and improved fabrication processes - the early TTLs were not really made in what we call "clean rooms" today. Look on youtube for videos showing early IC fabrication from the 1960s. You could do this in your garage if you had the "tools" and the chemicals (toxic !). This nonchalant attitude towards clean room cleanliness of the early days of the semiconductor industry probably was the reason why they were unable for a long time to produce useful MOS transistors, despite these had been patented in the 1920s, and their theory was well understood since the early 1960s when ICs took off.
It was soon found out that the floating input trick does not lead to reliable "H" levels and thus compromizes noise immunity, especially when it's not just a floating pin directly at a TTL IC, with no PCB trace attached to it, but a floating input line having a PCB trace and/or a wire wrap wire connected to it. This may pick up crosstalk from other signals and then the "H" of the floating input is not a "H" anymore, but a "maybe", often denoted as 'X'. This is somewhat OK as long as the input having logic level 'X' still produces the correct function - see older TTL data books which still had truth tables containing "0", "1", "X", "H", "L", "Z". Later TTL data books (beginning with early 1970s, Texas Instruments databooks) discouraged the practice to cobble together gates from diodes, and they also warned against floating inputs, of which the Apple-1 has many. They also discouraged direct connection of these floating inputs to VCC (+5V) and recommended a series resistor of about 1kOhm for each 10 such inputs. The concern obviously was what could happen during power up and power down cycles in a system having multiple +5V regulators, which may not come up or turn off at exactly the same time. The resistor would limit the current through the inputs to a safe value where the ICs would not get damaged in case of an input having power while the VCC pin of that IC still was at a lower voltage level. This is also related to the junction isolation of the TTLs and LSTTLs - and this junction isolation is the reason why they get so hot when being plugged in turned 180 degrees, with destructive consequences.
The three diodes in the Apple-1 sync circuit are not an "OR", they are an "AND" - one of their cathodes being low makes the output of this diode gate low and this is logical AND.
Although this particular diode gate (you could name it a "DTL" gate for "Diode Transistor Logic", because the anodes feed the emitter of a transistor within the TTL IC) never has caused issues with any of my Apple-1 builds, it is recommended to replace the PN junction diodes (1N914 or 1N4148) with Schottky diodes (1N5711) ... this will improve the noise immunity for the "low" level significantly, by about a factor of two. All my kits came with Schottky diodes, but just as a precaution.
As a final note, TTL and its sub-families like S, AS, LS, ALS, F, etc. are quite robust and forgiving and hence, the better choice for hobbyists working with vintage computers. All CMOS logic IC families meant to replace TTLs (like 74C, 74HC, 74HCT, 74AC, 74ACT ...) are much more sensitive and finicky and rude to the weakly designed power and ground lines of vintage TTL based computers. Only the 74HCT and 74ACT families have the correct logic level thresholds to work together with TTL. But all cause a lot of bounce and ringing on the power and ground lines and none of these CMOS families tolerates floating inputs well --- they can get quite hot when an input floats, by drawing excessive supply current, and worse, this floating input can assume any logic state depending on which vagabonding charges the input pin and / or its PCB trace did collect: literally, feeble leakage currents in the nA (nano-Ampere) range can play a role here. This is why the Apple-1 is unfit to take CMOS logic unless ALL floating inputs are connected to a solid logic level. And then, the problem with the VCC/GND bounce and ringing still persists.
Hope this helps you with what you are trying to investigate.
- Uncle Bernie
Thanks for the clear and detailed explanation! It’s interesting to see how you break down the specifics of TTL logic, floating inputs, the historical context, and tips like using Schottky diodes. It really helps with understanding the quirks of the Apple-1.
Pages