The Dan ][ SD card disk controller

dabone wrote:

CVT wrote:

dabone wrote:

Ok, how about this for 28cXX support?

 

It's pin 27 that the problem,  it's actually A14 on 27256s and 27512s, but the 28c series uses it as the write enable pin.

So for proper address decoding on standard eproms, this pin should be held low, but on the 28c series it needs to be held hig

 

The original design only went as high as 27C128 for the 27-series EPROMS:

Apple2Card.png

There is no need to go higher. The 27C128 chip fits 16K, while the firmware that needs to be written is only 512 bytes: https://github.com/ThorstenBr/Apple2Card/blob/main/eprom/bin/eprom.bin

I currently have around 100 25c512s, and 20 or so w27c512s, and 10 or so 27256s.

So supporting eproms that people may have or use is a good thing. 

It's true that you can just tie that pin high and offset the rom into the bigger chips, but easier if you don't need to.

dabone wrote:

I currently have around 100 25c512s, and 20 or so w27c512s, and 10 or so 27256s.

So supporting eproms that people may have or use is a good thing. 

It's true that you can just tie that pin high and offset the rom into the bigger chips, but easier if you don't need to.

 

You might happen to have lots of those, but larger chips are more expensive and harder to find. So, perhaps going lower instead of higher might be something else to consider, like AT28C16 in DIP24 package.

FYI I've just ordered a batch of boards - they should be available in the UK in the coming weeks. If anyone's interested please dm me. I can also probably prepare full kits (assembled or not).

CVT wrote:

dabone wrote:

I currently have around 100 25c512s, and 20 or so w27c512s, and 10 or so 27256s.

So supporting eproms that people may have or use is a good thing. 

It's true that you can just tie that pin high and offset the rom into the bigger chips, but easier if you don't need to.

 

You might happen to have lo

AT28C64 are super easy to find and reasonably priced.  I've used them in the original DAN ][ design.  Most of the ones I have built I used an AT28C256 ot CAT28C256

I'm doing one more respin of this, and defaulting to pin 27 being tied high for 28cxx compatiblity, but allowing a solder jumper to handle 27 tied low so it can work with 27256s, 27257s, 27512s, and w27c512s.

These will also work with 27 tied high, but then you have to figure out your offset when you burn your image.

That gives the most options and ease of use.

dabone wrote:

I'm doing one more respin of this, and defaulting to pin 27 being tied high for 28cxx compatiblity, but allowing a solder jumper to handle 27 tied low so it can work with 27256s, 27257s, 27512s, and w27c512s.

These will also work with 27 tied high, but then you have to figure out your offset when you burn your image.

That gives the most options and ease of use.

 

&n

Sounds like a plan!

tamsoftware wrote:

FYI I've just ordered a batch of boards - they should be available in the UK in the coming weeks. If anyone's interested please dm me. I can also probably prepare full kits (assembled or not).

Received the boards - all good. Very precise design @profdc9, looks wonderful.

TL;DR: I'm getting random results, each boot returns something new - sometimes I see the menu, sometimes I can navigate, sometimes I see the files, sometimes I don't,.

Sometimes I see the files of slot 1 in slot 2, although it's empty, sometimes in both. It always freezes.

The best I've had so far is the welcome screen of the total_replay disk before freezing. Once in a while I fall back to the monitor after hitting Return.

Jumpers:

• JP2: closed
• JP5: closed
• all others: open

Slot 1: 128MB class 4, FAT 16

• VOL00_TotalRecall_5.2.po
• VOL01_ip65.po

• Slot 2: empty

Ethernet: not yet

I've used @MacFly 's 1.1.1 iBom to get the latest on resistor values: https://htmlpreview.github.io/?https://github.com/ThorstenBr/Apple2Card/blob/main/Apple2Card/bom/ibom.html

(btw, D7 is missing)

IMG_1394.jpeg

• Chips:
  • ATMEGA328P U
  • CP82C55AZ
  • AT28C256
  • 02AF65M
  • 2AA6X9M

I've programmed the ATMega and EEPROM with a T48 as such, using @MacFly 's repo https://github.com/ThorstenBr/Apple2Card.

I'm putting the different atempts for the EEPROM in case it can help others.

$ minipro -p 'ATMEGA328P@DIP28' -c code -w Apple2Arduino.ino.328p.with_bootloader.hex -f ihex -e

Found T48 01.1.03 (0x103)

Warning: T48 support is experimental!

Warning: Firmware is out of date.

  Expected  01.1.31 (0x11f)

  Found     01.1.03 (0x103)

Device code: 48B20792

Serial code: 20EFIUY6TXOA3UVA1KPY5736

Chip ID: 0x1E950F  OK

Found Intel hex file.

Writing Code...  1.81Sec  OK

Reading Code...  0.66Sec  OK

Verification OK

$ minipro -p 'ATMEGA328P@DIP28' -c config -w fuses.cfg

Found T48 01.1.03 (0x103)

Warning: T48 support is experimental!

Warning: Firmware is out of date.

  Expected  01.1.31 (0x11f)

  Found     01.1.03 (0x103)

Device code: 48B20792

Serial code: 20EFIUY6TXOA3UVA1KPY5736

Chip ID: 0x1E950F  OK

Writing fuses... 0.06Sec  OK

Writing lock bits... 0.05Sec  OK

$ minipro -p '28C256' -w ../eprom/bin/eprom.bin

Found T48 01.1.03 (0x103)

Warning: T48 support is experimental!

Warning: Firmware is out of date.

  Expected  01.1.31 (0x11f)

  Found     01.1.03 (0x103)

Device code: 48B20792

Serial code: 20EFIUY6TXOA3UVA1KPY5736

Use -P if you want to write-protect this chip.

Incorrect file size: 512 (needed 32768, use -s/S to ignore)

This chip may be write-protected. Use -u and try again.

$ minipro -p '28C256' -w ../eprom/bin/eprom.bin -s

Found T48 01.1.03 (0x103)

Warning: T48 support is experimental!

Warning: Firmware is out of date.

  Expected  01.1.31 (0x11f)

  Found     01.1.03 (0x103)

Device code: 48B20792

Serial code: 20EFIUY6TXOA3UVA1KPY5736

Use -P if you want to write-protect this chip.

Warning: Incorrect file size: 512 (needed 32768)

Erasing... 0.20Sec OK

Writing Code...  0.09Sec  OK

Reading Code...  0.00Sec  OK

Verification failed at address 0x0000: File=0xE0, Device=0xFF

This chip may be write-protected. Use -u and try again.

$ minipro -p '28C256' -w ../eprom/bin/eprom.bin -s -u

Found T48 01.1.03 (0x103)

Warning: T48 support is experimental!

Warning: Firmware is out of date.

  Expected  01.1.31 (0x11f)

  Found     01.1.03 (0x103)

Device code: 48B20792

Serial code: 20EFIUY6TXOA3UVA1KPY5736

Use -P if you want to write-protect this chip.

Warning: Incorrect file size: 512 (needed 32768)

Erasing... 0.20Sec OK

Protect off...OK

Writing Code...  0.10Sec  OK

Reading Code...  0.00Sec  OK

Verification OK

Apple //e non-enhanced FR/US

• Slot 1: serial card
• Slot 2: A2DVI
• Slot 3: 80-text + 64 MB RAM
• Slot 4: Mouse
• Slot 5: Z80
• Slot 6: Disk II
• Slot 7: Dan II

My soldering looks ok to me:

IMG_1400.jpeg

IMG_1396.jpeg

IMG_1395.jpeg

Any suggestions as to what I can check? Thank you very much!

You may need to change the value of some of the resistors depending on which 82x55 chip you are using.

You may need to change the value of some of the resistors depending on which 82x55 chip you are using.

If you are going to play around with the resistors on the 82c55 chip, you only need to try changing R2 since you only have JP2 installed (but 4.7K has always worked for me.)

Intermittent problems with similar symptoms that I have had were due to Vdd at the uSD cards being too low. Instead of the two IN4148 diodes in series I changed out one of the diodes with a lower Vf schottky diode (about 0.3 instead of 0.7v) The uSD cards were much happier after that. So, check the voltages between pins 4 and 6 on the uSD socket.

I use a different T48 type programmer for the 328p (with Win 10) so can't help much there. I do know the correct fuse settings are vitally important however.

Thank you

I've identified and fixed one issue already - I had soldered the four LEDs the wrong way.

Now it works better as in I often have the menu, the 4 arrow keys are properly detected and usually work, and the boot often starts.

Now it's upon reading the boot drive that it crashes "202C - A=04 X=04 Y=00 P=34 S=EB".

I do get the occasional boot but then freeze while selecting the game. The first initial cold boot is the one that worked the best.

Hmmm I tried putting a second card in slot 2 in case the empty slot could be source of problems.

It would appear the clicking mechanism doesn't work properly. Maybe I over-heated it.

Snif snif ...

Do you see the message "DAN II: PRESS RETURN" every single time you turn on the computer without exception? Because it you don't, the problem is in the two chips to the right or the EPROM. In fact you can remove the two chips on the left and leave just the three on the right and you should still see this message every single time when you restart the machine even as it jumps into the monitor.

This is one divide-and-conquer approach that can help in isolating the problem.

tamsoftware wrote:

Thank you

 

I've identified and fixed one issue already - I had soldered the four LEDs the wrong way.

Now it works better as in I often have the menu, the 4 arrow keys are properly detected and usually work, and the boot often starts.

Now it's upon reading the boot drive that it crashes "202C - A=04 X=04 Y=00 P=34 S=EB".

I do get the occasional

Make sure the polarity of the other diodes and the electrolytic are correct too.  Your pic is a little fuzzy so I would have to blow it up to see.

Do you see the message "DAN II: PRESS RETURN" every single time you turn on the computer without exception? 

yes the "DAN II: PRESS RETURN" is here 100% of the time

Make sure the polarity of the other diodes and the electrolytic are correct too

The 5 diodes match the silkscreen, going to check with the circuit. I believe I've been careful with the electrolytic cap but will double-check as well.

TYVM

DanIIv1.1bCR.jpg

Got the new prototype board in, and I'll tell you moving 0402 components from one board to the other is not fun.

The new board works correctly with Eproms and the 28C256 eeprom.

And Programming works with the jumpers installed for the 28c256.

Ethernet is also working, so success! 

Cutting the Eprom jumper isn't needed if you have an offset of $4000 (Decimal 16384) when programming the Eprom.

Another thing worth noting is that there are PLCC versions of the 82x55 chips which may be even easier to find and possibly cheaper than the DIP-40 versions. I know because I accidentally boutgh like 10 or 15 of them by mistake because I didn't notice that the place carried different package types.

softwarejanitor wrote:

Another thing worth noting is that there are PLCC versions of the 82x55 chips which may be even easier to find and possibly cheaper than the DIP-40 versions. I know because I accidentally boutgh like 10 or 15 of them by mistake because I didn't notice that the place carried different package types.

 

I ordered some dip-40s for under $2, the other versions seem to be much more expensive.

I am trying to get one of these cards to work as well. I am having similar issues to above. 

I'm away from the computer/ card at the moment. But can I ask what is the significance of R2 ? On the circuit I have it’s just pull up for a jumper and does not connect to the sd-card at all.

 I have ended up jumpering in place 2 sd-card adapter boards (the very cheap ones for arduino) these made a massage difference, it went from a flop to fully working! Almost anyway, the issue I have now is on first power up just after the dan ][ selection it always fails to boot and drops me in the monitor. After that one keyboard reset fixes it. Works 100%.

I’m pretty sure the issues are related to the resistor divider chain for the sd-card signals. Or the sd-card voltage. I note above you mention the voltage being low? I think it maybe the opposite too high. I say this because I substituted a RED led in place of the two 1n4148’s and got far better results, not perfect but it did boot at lease 75% of the time.

Based on 5v supply – 2x0.6v gives 3.8v the SD spec seems to indicate 3.6v MAX voltage so the RED led gives around 5v – 1.6v = 3.4v but not entirely sure how stable it is!

I’m going to try a 3.3v regulator when I get chance.

Derek

R2 is the pull up for U2 28C256 eeprom's A8 address line, activated by U4 PC0 pin when jumpered with JP2 .

on page 5 of this thread Mac-fly answered the use of JP2 as follows

"The bank jumper (JP2) is required to allow the controller to switch between the first and second 256byte page of the ROM. With DAN's original firmware the second page is only ever needed when you press RETURN to activate the configuration. It's the only thing which is stored on the second page.

With my firmware (yesterday's post) the second page is required as soon as you press any key during the "DAN II PRESS RETURN" phase."

SD card instability on power on

I have had the same issues with my cards not initalising both SD cards on power on boot followed by enter and needing a second CTRL -open apple- reset followed by an enter to get both SD cards displaying content.

Like you I was begining to think this was an over voltage and possibly a current issue with the SD card supply circuit. I have today removed the 4 * IN4148 diodes and patched in a 3.3 LDO voltage regulator.

I have retained resistors R12 and R11 so the LDO has a load when an SD card is not inserted I have also retained C6 and C5 decoupling capacitors.

On a power on reset followed by a return both SD cards immediatly show thier content.

I intend to run this modifcation for a bit longer before making new boards with a dedicated LDO on the PCB. 

I hope this helps 

Crusty

Crusty wrote:

R2 is the pull up for U2 28C256 eeprom's A8 address line, activated by U4 PC0 pin when jumpered with JP2 .

on page 5 of this thread Mac-fly answered the use of JP2 as follows

"The bank jumper (JP2) is required to allow the controller to switch between the first and second 256byte page of the ROM. With DAN's original firmware the second page is only ever needed when you press RETURN

I've added the 3.3v regulater and it's the first time i've been able to use the SD-Adapters on the card! testing for a while...

What github site is considered the best to use?

Derek

Hi,

I'd like to get or build a Dan][ card to run GSOS or Total Replay from an SD card on an apple ii GS. Is there anyone selling them on ebay, tindie or other ? (or PM for direct contact)

To produce it, can anyone make the PCBWay or other project public so it can be easily reused ?

Thanks

G.

gf65 wrote:

Hi,

I'd like to get or build a Dan][ card to run GSOS or Total Replay from an SD card on an apple ii GS. Is there anyone selling them on ebay, tindie or other ? (or PM for direct contact)

To produce it, can anyone make the PCBWay or other project public so it can be easily reused ?

 

Thanks

G.

Dan, who is a member here (not to be confused with the creator of the Dan ][ Controller) sells them on eBay: https://www.ebay.com/itm/126946444491

Also there is a slightly better alternative: https://maceffects.com/products/xdrive-mass-storage-solutions-for-the-apple-ii-ii-iie-and-iigs

Thanks for the information. I already saw the XDrive which is so far the cheapest option compared with the FloppyEmu or ReactiveMicro solutions.

I was hoping the Dan][ would work to be cheaper option of the XDrive. If a PCBWay project was available, I could check if it's worth producing it.

CVT wrote:

gf65 wrote:

Hi,

I'd like to get or build a Dan][ card to run GSOS or Total Replay from an SD card on an apple ii GS. Is there anyone selling them on ebay, tindie or other ? (or PM for direct contact)

To produce it, can anyone make the PCBWay or other project public so it can be easily reused ?

 

Thanks

G.

 

 

That new XDrive looks nice, but to be fair the Dan ][ has its advantages too.  The XDrive is supposed to be faster and it offers the USB option which is sweet.  However the Dan ][ can be built for less than 1/2 what the XDrive is selling for, and because the XDrive uses SMD parts, it isn't a very DIY friendly design, plus I am not sure if it is even open and you can build it yourself anyway.  Plus the Dan ][ also has the Ethernet option which is nice and it actually performs pretty decently, even if not as fast as XDrive, CFFA 3k or MicroDrive Turbo, it's faster than wDrive, FloppyEmu or FujiNet.

But it is really great that these days we have so many awesome options.

softwarejanitor wrote:

...

However the Dan ][ can be built for less than 1/2 what the XDrive is selling for, and because the XDrive uses SMD parts, it isn't a very DIY friendly design, plus I am not sure if it is even open and you can build it yourself anyway.

... 

The XDrive is not open source, so building it yourself is not an option. It is true that you can build the Dan ][ Controller cheaper, but only if you already have the required equipment to do all the soldering and all the programming. Even if you don't already have a way of programming an EPROM/EEPROM and an Atmega, it will be just as expensive, if not more. You can of course program the EEPROM with the card itself, but that takes a while to figure out how to do and it works only if there are no hardware or software issues leading up to it.

I would summarize it this way: you can build a Dan ][ Controller for 1/2 (or even less) of what it or the XDrive sells for, but only once you have done it before.

When we are talking about an Apple IIgs however, the XDrive has a huge advantage: the ability to mount 8 ready SmartPort volumes from WhatIsThe2GS. I use a Dan ][ Controller in my Apple IIgs and it is very frustrating that I can only mount two SmartPort volumes at a time. One of them is of course GS/OS, so it's really just one software collection at a time. I have to constantly switch off the machine to change the SmartPort volume in order to load a different software collection.

And if you decide to do a new PCB Layout, may I also suggest you add values to the sik screen.

I'm getting a batch of Dan ][s ready for sale.

They are going up on ebay soon, but I'm still figuring out my pricing.

dabone wrote:

I'm getting a batch of Dan ][s ready for sale.

They are going up on ebay soon, but I'm still figuring out my pricing.

 

 

I made a bunch (16) Dan ][ cards and some others and I was going to sell some on eBay but I got frustrated with selling there after a few bad experiences so I ended up more or less giving up on that idea.

I've been selling my Link232-Wifi series, Megabit128s, and my replacement TI99 4/a Power supplies on ebay, and the shipping has made my life much easier.

I did try selling on Tindie for awhile, but the shipping hassles and customs forms was a pain in the butt.

I was having quite a lot of issues with SD-cards none worked reliable. I suspected the power supply to them because it was running them out of spec using to diodes. So I added a 3.3v LM1117 LDO regulator. Now it works finewith any of the cards I tried before. I put the regulator in place of the diodes for one SD-card and ran a jumper wire to the other set of diodes. I also add a 100uf cap in place of the 100nf as the datasheet for the lm1117 seems to say this is the way to go.

IMG_7005.jpg

That's a pretty interesting mod using the LM1117-3V instead of the diodes.

I was having the same problem with uSD card reliablity. I did discover the voltage at the uSD connectors was a bit low. My first try was to replace one of the 1N4148 diodes (D6 for example) with a lower forward voltage drop schottky diode. This did improve the uSD read/write reliability but I was worried that a some point the flash card Vdd might get too high. I didn't have any LDO regulators handy, but I did find this assembly that has the 3v3 LDO and filter capacitors on a really tiny PCB. Similar to Coopzone, I replaced both the diodes for Slot 2 with this assembly and ran a wire for 3v3 down to Slot 1 (after removing D3 & D5.) Haven't had any reliability problems since.

http://hiletgo.com/ProductDetail/1958363.html

You may notice in the picture there is a 2 pin header in place of D9. I ran wires for all three chip select LED's to the outside world so I could see when the uSD cards were selected (sort of a disk activity indicator) or when FTP was active.

IMG_3v3LDO.jpg

I recently updated an older DAN ][ card to the most current firmware (first had to reprogram with minipro which worked) and then was able to update using Dan II tools.  I know its working properly becuase I'm able to use the new "VOLXX" naming convention and have updated to v3.4.1.

For the life of me I can't figure out why I still see the old DAN ][ boot message which looks like this:

DAN ][ PRESS RTN

CARD 1 (0-9,!):

What am I missing?

schafferm wrote:

...

What am I missing?

Seeing the old message means you haven't updated the EPROM/EEPROM chip with the latest.

So I have gone through this thread, and while most of the programming stuff is over my head, I find it fascinating.  Ordered a Dan ][ from ebay today.  Just wondering what with the Ethernet option would it be possible to direct a rudimentary text file to print to a network printer?  Not that I want to give anyone more work to do... Thx.

I've been able to get simple stuff like Telnet and FTP to work over Dan ][ with the W5500 Ethernet module, but network printing is more complicated.  You'd need some kind of driver on the Apple II side to be able to capture printing and redirect it to the IP65 stack.  And the app on the Apple II side would need to speak something that a network printer could understand.  Maybe PCL or Postscript.

One project that I started a while back that I haven't gotten very far on might actually do some of what you want.  I was going to use an ESP32 microcontroller module to read the data from a parallel printer card on the Apple II.  The handshaking protocol for basic Centronics style parallel is pretty easy and the right ESP32 modules can hook up to the parallel connector through their GPIO pins if they are 5V tolerant.  It should be possible to write code on the ESP32 that could decode the data stream for something like an EPSON MX/FX/RX family printer (or pretty much any other dot matrix like an Apple Dot Matrix (aka C-Itoh 8510) or Star Micronix, Mannesman Tally, etc) and turn that into a bitmap in a common format like PNG.  The ESP32 has WiFi capabilities so it could take the bitmaps it captures and upload them to a modern computer through FTP, SFTP or HTTPS, whatever.  Then it would be possible to print the images from the modern computer where support for a wide variety of netwrok printers is easy.  Since the ESP32 also has TX/RX for serial ports, it would also be possible to hook up to a serial card and accept printer codes for printers like the Apple Imagewriters.

Anyway, I built the hardware which is very simple -- Centronics uses only 8 data bits plus a handful of handshaking signals, I just haven't had time or it hasn't been a high enough priority to work on writing the software to understand the EPSON printer codes and convert the data to PNG and upload it.  It is really more of a software project than anything else.

It would also be possible to build an Apple II card that wouldn't require a separate parallel or serial card using an ESP32 interfaced directly to the Apple II bus like the A2VGA cards.  The software on the ESP32 could then emulate the parallel or serial card (software defined so you could change the card emulation just by updating the firmware) and then do the same kind of print data stream to image conversion upload as I detailed above.

Perhaps there is a way to do network printing directly from an ESP32, but I suspect it would be hard to fit something that complex into the amount of memory available.  But I could be wrong.

Thank you !  So in the 512K besides the menu software, what else is stored there?

softwarejanitor wrote:

I've been able to get simple stuff like Telnet and FTP to work over Dan ][ with the W5500 Ethernet module, but network printing is more complicated.  You'd need some kind of driver on the Apple II side to be able to capture printing and redirect it to the IP65 stack.  And the app on the Apple II side would need to speak something that a network printer could understand.  Maybe PC

Wow it seems your 50% there!  Thanks for all the info.  The concept is simple enough -- the implementation not so simple.  Could the RAM on the ESP32 theoretically be expanded?

schafferm wrote:

Thank you !  So in the 512K besides the menu software, what else is stored there?

The EPROM/EEPROM firmware is only 512 bytes and it doesn’t even contain the menu. It just has some boot code and the boot message. Capacity wise there is no need to have 512Kbit EPROM/EEPROM chip. You can fit the firmware on a 2708 EPROM chip twice, but the larger chips are just easier to program, since they don't require 25V.

mreg376 wrote:

Wow it seems your 50% there!  Thanks for all the info.  The concept is simple enough -- the implementation not so simple.  Could the RAM on the ESP32 theoretically be expanded?

 

Sort of, but with big limitations.  There isn't a high speed data bus external to the module.  You could attach a serial RAM to i2C, but it would be pretty slow to access.  Maybe that could be something you could use to temporarily store the image data so as not to use up the RAM for that and leave more room for code.  I'm not sure how much that would buy you.  I'm not sure it's even necessary though.

CVT wrote:

schafferm wrote:

Thank you !  So in the 512K besides the menu software, what else is stored there?

 

The EPROM/EEPROM firmware is only 512 bytes and it doesn’t even contain the menu. It just has some boot code and the boot message. Capacity wise there is no need to h

There are EPROMs as small as 2716 that don't require 25V programming, but the 28Cxxx chips are much cheaper and easier to source these days plus they can even be programmed in circuit in many cases.  Plus they are electrically erasable and don't require a UV eraser and the time it takes to do that process.  It's just way more convenient to use a 28C64 or larger even if you don't need that capacity.  Most of the Dan ][ cards I built I believe I used a 28C256 because those are generally one of the cheapest and easiest to get.  The 28C64 is readily available as well.

Your point about programming voltages is a good one though.  Many of the older devices which don't require 25V like a 2708 still require 21V, and some of the inexpensive programmers like the MiniPro TL866-II can only do up to 19V.  That's why I keep my original TL866CS, because it can handle 21V, and I have a lot of 2716/2732 parts that require that.