Cassette Interface now works!!

What have you built or do you want to design to go with your replica 1

Postby vbriel » Oct Sat 20, 2007 9:12 am

Radio Shack sells adapters. I'm even using stereo 3.5mm cables! I have a lot more tests to run before I can determine the boards are ready.

Vince
User avatar
vbriel
Site Admin
 
Posts: 1184
Joined: Jul Tue 19, 2005 1:10 pm
Location: Ohio

Postby Jeroen » Oct Sat 20, 2007 9:47 am

Radioshack is that some kind of computer store. (I know it's an AMERICAN electronics store don't worry)
I'm sure I can find adapters somewhere else. (and if all else fails, Make them ;-) )

edit: also i'm willing to test them for you ;-) . I'm even thinking of building my own! (looking at the schematics now!)
Jeroen
 
Posts: 38
Joined: Sep Fri 28, 2007 3:47 am

Postby vbriel » Oct Sun 21, 2007 10:53 am

Well, I've been having problems getting anything over 1K to load, so I tried the b side of my cassette tape. What do you know, it doesn't work. I flip back to side a and I can almost read programs. I tried a metal tape, does not work. I'm going to try a new everyday recording tape and see what I get with that. If it was media all this time, I'm gonna be sick. I don't know why anybody would want this as a storage device, it is frustrating beyond belief.

Vince
User avatar
vbriel
Site Admin
 
Posts: 1184
Joined: Jul Tue 19, 2005 1:10 pm
Location: Ohio

Postby hanso » Oct Sun 21, 2007 11:37 am

vbriel wrote: If it was media all this time, I'm gonna be sick. I don't know why anybody would want this as a storage device, it is frustrating beyond belief.

Vince


Welcome to the 80ties! Before slow floppy drives became affordable we had to live with these misery all the time, and he, it made us computer nerds!
hanso
 
Posts: 71
Joined: Sep Sat 22, 2007 1:35 pm
Location: Weesp The Netherlands

Postby daustin777 » Oct Sun 21, 2007 3:16 pm

vbriel wrote:I tried a metal tape, does not work.


I don't think metal tape was available at the consumer level back then, if I remember correctly. Audio-wise it's much brighter.
--David Austin-- replica 1 se, CFFA1, MicroKIM, [kits]
daustin777
 
Posts: 54
Joined: Nov Sun 12, 2006 4:03 pm
Location: Southern California

Postby vbriel » Oct Sun 21, 2007 5:54 pm

Well, I'm getting a strange data readback. If I backup BASIC ($E000-EFFF) and read it back into $1000-1FFF it appears to work. But, only the first 1K, then around 1500-up the code is incorrect. If a byte should be $02 the value will be $01. If a byte should be $40 it will be $20, exactly half of what it should be. Not sure why it is not reading correctly but I'm working on it. Curse of the cassette board continues...

Vince
User avatar
vbriel
Site Admin
 
Posts: 1184
Joined: Jul Tue 19, 2005 1:10 pm
Location: Ohio

Postby iam7805 » Oct Sun 21, 2007 6:07 pm

It seems most logical to think it's a firmware problem.

Also, daustin seems to be on to something with metal tape not working. It does a better job of capturing all the high frequencies, which Woz said need to be rolled off. I haven't looked at your schematics, but have you put a low-pass filter on the inputs? That could help a lot, particularly with newer tape formats.

-iam

franz wrote:i asked Steve Wozniak last month about this problem.

Here his answer:

----------------------
I would suggest a low-pass filter, cutoff around 5 kHz., on the signal from the tape recorder. Such a filter would be beneficial for the data to the tape as well. I may have myself included a filter or I may have assumed that the cheap cassette decks of the day did enough filtering along those lines. Oddly enough, it implies that a lower quality voice quality recorder might work better with the Apple I than a better cassette might.
----------------------
IRC Channel: homebrew-computer-club on SorceryNet. Discuss all retro computers and replica/tribute computers!

ROSES ARE #FF0000
VIOLETS ARE #0000FF
ALL MY BASE
ARE BELONG TO YOU
User avatar
iam7805
 
Posts: 8
Joined: Oct Sun 21, 2007 11:30 am

Postby Jeroen » Oct Mon 22, 2007 2:21 am

Iam7805 is a friend of mine. (he's the brandon in Jab ma computer, lol)
Now you mention it it DOES sound like a firmware problem! Maybe the rom got corrupted over the years (bit rot) and when read to a computer nobody really tested it well (under 1k progs)?
Jeroen
 
Posts: 38
Joined: Sep Fri 28, 2007 3:47 am

Postby iam7805 » Oct Mon 22, 2007 10:04 am

I thought that Vince used the source code written by Woz on paper, and assembled. It could very well just be small translation error.

-iam
IRC Channel: homebrew-computer-club on SorceryNet. Discuss all retro computers and replica/tribute computers!

ROSES ARE #FF0000
VIOLETS ARE #0000FF
ALL MY BASE
ARE BELONG TO YOU
User avatar
iam7805
 
Posts: 8
Joined: Oct Sun 21, 2007 11:30 am

Postby Jeroen » Oct Mon 22, 2007 12:36 pm

He did? Then it definitly could have a mistake.....
Jeroen
 
Posts: 38
Joined: Sep Fri 28, 2007 3:47 am

Postby vbriel » Oct Mon 22, 2007 5:54 pm

No, I had to piece the code together from 2 sets of nibbled files. Took an hour by hand to do it but I did it like 3 times to make sure it had no errors. 256 bytes isn't much to patch.

With tape recording media without error correction there is a LOT of room for error. It could be media, recording device, cables, outside noise, etc. On this prototype board I added buffers to the address bus to see if it helped things out. What is really strange is that the first 256-1K bytes play back just fine, after that, the data changes.

I'm not using metal tapes, just standard everyday tapes appear to get the best results. I've tried 2 different tape players.

I have several EPROM's and I've introduced different ones in to check for different results.

One thing I've noticed is that I get an exact incorrect result. For example, if a byte is suppose to be 20 and it comes back as 10, it does it every time. What could be the cause in my opinion, a lot of things.

A least I'm not getting just straight garbage. I'm going to work on the low pass filter and other adjustments to see what results I get.

On the last test I just ran it made it over 2K without error so it is probably something simple I'm missing. I hate analog.

Vince
User avatar
vbriel
Site Admin
 
Posts: 1184
Joined: Jul Tue 19, 2005 1:10 pm
Location: Ohio

Postby iam7805 » Oct Mon 22, 2007 9:43 pm

http://www.sbprojects.com/projects/apple1/aci.htm

^ Perhaps you'll want to try assembling the source code on the page linked too above and see if the code works. As Jeroen said it's possible that the files you used were dumped from chips that had been through some form of bit rot. That actually seems likely, as only 50 Apple 1 computers survived, and the board didn't have any case on it. UV rays could have started to erase the EPROMs slowly overtime.

Of course, it could even be a problem with the firmware on the Replica 1 board itself.

-iam
IRC Channel: homebrew-computer-club on SorceryNet. Discuss all retro computers and replica/tribute computers!

ROSES ARE #FF0000
VIOLETS ARE #0000FF
ALL MY BASE
ARE BELONG TO YOU
User avatar
iam7805
 
Posts: 8
Joined: Oct Sun 21, 2007 11:30 am

Postby Jeroen » Oct Tue 23, 2007 6:53 pm

I might know the problem: 65c02's were used in replica right? Guess what apple I used 6502's these had "undocumentated opcodes" most were replaced in the 65c02 by more valid ones. Problem is knowing woz he wanted to make the code as compact as possible. Often these "invalid opcodes" allowed more fast instructions cause they kinda got "shorted with each other" (best way to explain it).
Jeroen
 
Posts: 38
Joined: Sep Fri 28, 2007 3:47 am

Postby hanso » Oct Wed 24, 2007 3:19 am

vbriel wrote:
One thing I've noticed is that I get an exact incorrect result. For example, if a byte is suppose to be 20 and it comes back as 10, it does it every time. What could be the cause in my opinion, a lot of things.

Vince

I have looked at the source (and no illegal opcdes there ;) ) and the way it constantly checks for changes on the input makes it very noise sensitive.
More robust algorhythms will wait until around the middle of a bit to check.

So my assumption is that some noise triggers these false readings. Noise on a cassette recorder can come from the tape itself (the cassette noise, dropouts) or that spikes from the digital circuits reach the input of the opamp via the 10K resistors on the inputs.

I would suggest:
- decouple the +5V before it reaches the 10Ks at the inputs of the opamp (resistor - capacitor)
- narrow the band of frequencies reaching the opamp
- put a lowpass filter in front of the opamp,
- the capacitor present hopefully filters out low frequencies

Hans
hanso
 
Posts: 71
Joined: Sep Sat 22, 2007 1:35 pm
Location: Weesp The Netherlands

Postby Mojoseph_12 » Nov Sat 17, 2007 1:59 am

Will the Interface be ready to sell before X-mas? I know I've been asking when it will be ready alot but im just getting anxious i think it would be nice to have some kind of mass storage that is not a computer even if it is not the best.
If not thats fine :) i waited for a year a little longer won't hurt. 8)
Mojoseph_12
 
Posts: 46
Joined: Jan Fri 05, 2007 4:07 pm
Location: New Mexico,USA

PreviousNext

Return to Creation Room

Who is online

Users browsing this forum: No registered users and 2 guests

cron