fujinet #FujiNet - a WIP SIO Network Adapter for the Atari 8-bit

[tschak909]

@TheMontezuma that brings timing back into spec on the esp32, but the loading timings are noticeably slower on the ESP32 because of the flush, compared to the ESP8266:

 

-Thom

[TheMontezuma]

Yes, I realized that flush() is a lengthy call. It adds ~10ms delay.

I'm afraid that the only way to be faster and yet still compliant is to re-write the code with ESP32-IDF.

It this phase of the project (prototyping use cases with Arduino) we may have to live with that.

Edited January 10, 2020 by TheMontezuma

[tschak909]

I was afraid of that.

 

I noticed the same thing when using flush() in the exact same places.

-Thom

 

p.s. and really, that's okay, because IDF is where the production software should go anyway, it'll give us much tighter control across the board, and we'll need it once we have all these virtual devices chirping.

 

Edited January 10, 2020 by tschak909

[tschak909]

ok, so @TheMontezuma's fixes have been folded into multilator-rev2 and multilator2modem tests.

 

The multilator2modem test fuses modem850 and multilator-rev2 together to provide "D:" and "R:" devices. with the current bugs squished out, it's usable...

 

 

[Bee]

Is there someone I can PM on the 8266 pinout - there are some things that are unclear to me.  IE.  IO label "USB" on the Atari side in the spreadsheet?

 

Thank you

 

[mozzwald]

13 hours ago, Bee said:

Is there someone I can PM on the 8266 pinout - there are some things that are unclear to me.  IE.  IO label "USB" on the Atari side in the spreadsheet?

Just post your questions here in the thread. Someone else may be wondering the same thing. I removed the USB label from the Atari column since Atari obviously has no USB port .

 

Perhaps the schematic is more clear? Notes: SJ1 (solder jumper 1) and it's corresponding resistor is not required and only allows optional connection of GPIO2 to Atari CKO. Debug serial output from GPIO2 is optional and requires a USB/UART bridge (ie, https://www.sparkfun.com/products/9873). The diode below is SMD. You can use any schottky diode with as low voltage drop as possible. The pinout image in the spreadsheet will reference the GPIO pin numbers to the PCB Silk Dx markings.

 

Hopefully this helps.

[Bee]

38 minutes ago, mozzwald said:

Just post your questions here in the thread. Someone else may be wondering the same thing. I removed the USB label from the Atari column since Atari obviously has no USB port .

 

Perhaps the schematic is more clear? Notes: SJ1 (solder jumper 1) and it's corresponding resistor is not required and only allows optional connection of GPIO2 to Atari CKO. Debug serial output from GPIO2 is optional and requires a USB/UART bridge (ie, https://www.sparkfun.com/products/9873). The diode below is SMD. You can use any schottky diode with as low voltage drop as possible. The pinout image in the spreadsheet will reference the GPIO pin numbers to the PCB Silk Dx markings.

 

Hopefully this helps.

 

[E474]

Hi,

 

   Would a 1N5817 diode be appropriate (I used one in the SIO2Pi connector I made)? I appreciate that for a lot of people it will be obvious which component(s) to get, but if you could add in "such as a <component IDENTIFIER> " it would be really helpful to the less technically qualified (such as myself), and you might get feedback about more appropriate components if you are a bit more specific about which parts to use. Just my 2 cents worth.

 

   Thanks!

[Bee]

Thanks Moswald.  I think that clears it up.  I had pulled that out but S1 & J2 did confuse me.  I'm trying to pull together a BOM and make a through hole board for folks like me who are good at soldering but not so much at SMD.  i spent a good amount of time pinning out a SIO cable to use for this.  I have a small Desktop CNC.  While I suck at making the Gerbler files I can machine PCB's on the fly.  So if my design buddy can be convinced I can get some through hole boards made and test during the week.  I have the resistors, might have suitable diodes as well if not I will order them.  My goal is a clear BOM, and hookup guide.  I wanted to work that out and was trying to avoid confusing anyone else.  I'm also making a SIO BOB. I use Evernote to keep track of things.  I will share my note here:

 

https://www.evernote.com/l/AkW4vNFIqlVPJbTPHfzhDV14_N-uNDGWyXs

 

Q - what is the function of S1? IS this what you are referring to as SJ1?  It is switched on your boards? Once I have the 8266 worked out I will do the ESSP32 as well.

 

Thank you

Edited January 12, 2020 by Bee

[Bee]

@E474 in the earlier schematic a 1N5819 is called out.  I don't know the difference between the 1N5819 and 1N5817 but in looking for the 1N5819 the 17 came up so they my be similar enough to use.  See my Evernote link.

 

Thank you

[mozzwald]

44 minutes ago, E474 said:

Would a 1N5817 diode be appropriate (I used one in the SIO2Pi connector I made)? I appreciate that for a lot of people it will be obvious which component(s) to get, but if you could add in "such as a <component IDENTIFIER> " it would be really helpful to the less technically qualified (such as myself), and you might get feedback about more appropriate components if you are a bit more specific about which parts to use. Just my 2 cents worth.

The 1N5817 should work in this situation. The important factor for D1 is that it have a low forward voltage and schottky diodes are best for this. The forward voltage is how much voltage loss there is after the diode. So, if you have 5V coming out of the Atari on SIO-5V and it goes thru D1 with a Forward Voltage of 450mV then you will get 4.55V entering the FujiNet. So you see that a lower forward voltage is better (you want to be as close to 5V as possible). Hopefully that explains it.

42 minutes ago, Bee said:

Q - what is the function of S1? IS this what you are referring to as SJ1?  It is switched on your boards? Once I have the 8266 worked out I will do the ESSP32 as well.

S1 is a switch to enable or disable powering the FujiNet from SIO 5V, this could also be optional. SJ1 (solder jumper 1) is described in my previous post and is unrelated to S1 (switch 1).

[Bee]

@mozzwald How are your powering your boards?  SIO or USB currently?  I thought I saw earlier it was better to power over USB.

 

Thank you

[tschak909]

I'm testing both ways, with no ill results, on both my Mars Bar and Ice Cream sandwich prototypes.

 

My early rev2 lacks some circuitry to allow power over USB to work correctly.

 

-Thom

[mozzwald]

1 minute ago, Bee said:

@mozzwald How are your powering your boards?  SIO or USB currently?  I thought I saw earlier it was better to power over USB.

 

Thank you

I have tested both and they work. But most of the time I have the USB plugged into my computer for debugging.

[Bee]

Fww - I noticed you have a newer schematic then I was referencing - my bad.  Could you include a rev #  on the  schematic for clarity?  I also see the part count went up by a few resistors and 2 transistors.  The reason I didn't ask for a board and I'm making my own is my time is a variable this time of year.  The company I work for has a holiday sale and that creates a wave of support needs that may not end until May.  Some 18 hour days leave me wrecked.  So I don't know how much I can contribute beyond documenting the build and trying to keep up with the changes.  Once I clicked on the new schematic and zoomed SJ1 became obvious.  

 

I appreciate all the guidance.

 

Thank you 

[a8isa1]

I tried uncommenting these lines:

#define DEBUG_N
#define DEBUG_HOST "192.168.1.188"

I get this error: 

 

multilator:1664:26: error: 'DEBUG_PASSWORD' was not declared in this scope
   WiFi.begin(DEBUG_SSID, DEBUG_PASSWORD);
                          ^
exit status 1
'DEBUG_SSID' was not declared in this scope

What am I doing wrong?

 

-SteveS

 

P.S. This for an ESP8266, nodeMCU dev board v0.9

P.P.S. This with mozzwald's multilator-rev2.

Edited January 12, 2020 by a8isa1

[tschak909]

1 minute ago, a8isa1 said:

I tried uncommenting these lines:

#define DEBUG_N
#define DEBUG_HOST "192.168.1.188"

I get this error: 

 

multilator:1664:26: error: 'DEBUG_PASSWORD' was not declared in this scope
   WiFi.begin(DEBUG_SSID, DEBUG_PASSWORD);
                          ^
exit status 1
'DEBUG_SSID' was not declared in this scope

What am I doing wrong?

 

-SteveS

 

P.S. This for an ESP8266, nodeMCU dev board v0.9

add 

#define DEBUG_SSID "your ssid"

#define DEBUG_PASSWORD "your password"

 

These need to be in there because the network bring-up has to happen before everything else in debug mode.

-Thom

[tschak909]

1 minute ago, tschak909 said:

add 

#define DEBUG_SSID "your ssid"

#define DEBUG_PASSWORD "your password"

 

These need to be in there because the network bring-up has to happen before everything else in debug mode.

-Thom

You should also be using tests/esp32/multilator2modem

 

(yes, things are moving very quickly)

 

-Thom

[E474]

Hi @mozzwald,

 

    thanks very much for the explanation, looks like I will have all the parts ready to go when the ESP32 and breadboard arrives!

[tschak909]

#FujiNet #Atari @mozzwald has been hard at work fusing together my work on the Multilator and the work he's done implementing an 850 compatible RS232 interface and integrated WIFI modem. Here you see it effortlessly loading modem programs over the network and connecting to BBSes, but wait until the end, for a surprise!

[tschak909]

@Albert ^^

[+Allan]

I'll be waiting for the Official AtariAge Atari 8-bit BBS with a forum and store.

[E474]

Hi,

 

    Would it be possible to map the Boot button to "mount next ATR image in folder locally or on server" in a round robin type setup?

[tschak909]

That would result in very ambiguous behavior. This is why adding the web server front-end is on my list.

 

-Thom

[E474]

Use case is only several related ATRs in same directory, and original 8-bit coder checked correct disk has been inserted.