SIDE 3 preorder

[TGB1718]

There really is a thing called "Sod's Law", started testing everything, loaded a few Cartridge files, but couldn't get

any to play, pressing Start didn't work, Select was OK, Option didn't work, all was fine before SIDE3.

 

Went into basic wrote a few lines to check the buttons and would you believe it, Start and Option didn't work ?

Opened up the keyboard, cleaned the contacts and hey presto all working, what a relief, what were the chances

of that fault happening just then.

 

Have to say just love this SIDE3 (now my 130XE is playing ball) again @flashjazzcat thanks for this update

[flashjazzcat]

3 minutes ago, TGB1718 said:

There really is a thing called "Sod's Law"

Absolutely. And it applies 100 per cent of the time with old equipment like this. At least dirty keyboard contacts can be empirically considered a malfunction on the base machine.

6 minutes ago, TGB1718 said:

Have to say just love this SIDE3

Glad you like it! Candle has designed a fantastic piece of hardware, and I can guarantee the already nice user experience will just get better and better.

[flashjazzcat]

Regarding the aforementioned RWCRC, here's a link to the download page:

 

http://drac030.krap.pl/en-inne-pliki.php

 

I should point out that - unlike the IDE drivers for SIDE2 and compatibles - the SD card drivers do not currently include any kind of 'error correction' or retry logic, so if RWCRC completes an error free run on your SIDE3 APT partition, this means it works and the loader should not experience errors either. I may add some kind of retry logic to the PBI driver in the future (for example, if the data token never arrives or a write is never acknowledged), but such measures have the adverse affect of obfuscating minor transmission errors. In the past, this could mean that system could JUST manage a successful RWCRC run, but the user would still see transmission errors in the loader.

 

Of course, errors are detected and reported, but currently if IO fails on the first attempt, you'll receive an SIO error.

Edited October 24, 2020 by flashjazzcat

[candle]

On 10/22/2020 at 11:39 PM, dostrow said:

I got my hands on a known working Dig Dug cartridge and it works perfectly, lunching both with SDX disabled and by typing CAR in SDX. This tells me pretty conclusively that it is the SIDE3 cart itself. I take it this is something I am going to have to take up with @lotharek once he is back from vacation. Unfortunate as sending stuff back and forth from Poland to Texas is less than awesome. Is there anything else I should be trying before then? My Fujinet is supposed to be arriving this weekend in case that helps get data to the system at all...

I was postponing my answer, since lotharek is still on his vacations, and there is a lot going on in Poland unrelated to all this, but still keeping me distracted

Anyways. From my experience based on units Lotharek sent me to investigate their abnormal behaviour it comes down to bad U11 chip on SIDE3 cartridge

this is 74LCX04 chip and can be replaced with exact the same chip or 74LVS04 chip, but doing this by yourself would void the warranty - so i'm not advising anyone to try this 

best would be to return it to Lotharek, unless he would agree on something else

depending on which gate would fail results might be as follow:

loader doesn't start, athough you can use side3 as mass storage under SDX and using PBI BIOS - RD5 control fails

atari doesn't start, but SIDE3 led flashesh during power up - CCTL control fails

some cartridges will fail to load, that are supported - RD4 controls fails

nothing works - PHI2 control fails

 

Some people, that are known to copy other's people work and sell it prevent me from releasing schematics for SIDE3 hardware - this was the case for SIDE2 that was cloned, Ultimate1MB, and even VBXE

I don't like this anymore than any of you, and I'm all for the right to repair, but some people don't know how to behave in modern world, thus my decision

 

[mdivancic]

14 hours ago, DrVenkman said:

Apologies if I've missed it, but does this machine have an Ultimate 1MB installed? If so, you might consider installing a small 100pF capacitor between the RESET and GROUND pins on the U1MB board itself. I started having similar configuration resets of my U1MB on a 1200XL, and it was suggested to me by @tf_hh to mod my U1MB with a capacitor like that my my problems went away. I'm pretty sure @flashjazzcat has suggested a similar mod to others in the past as well.

Another thing to try is replace the 74LS08 with a 74F08 to fix stability issues with flash devices. My 800xl went from crazy unstable with a Side2 to completely stable with simple fix. 

[TGB1718]

2 hours ago, candle said:

I don't like this anymore than any of you, and I'm all for the right to repair, but some people don't know how to behave in modern world, thus my decision

Totally agree, I will never buy from any other source, buying anywhere else will surely stop people like you developing more

super gear.

[+Stephen]

2 hours ago, candle said:

Some people, that are known to copy other's people work and sell it prevent me from releasing schematics for SIDE3 hardware - this was the case for SIDE2 that was cloned, Ultimate1MB, and even VBXE

That is a real shame.  I try to buy at least one of each new upgrade even if I don't need it, to support you guys that are still making hardware.  I've never been able to figure out why some people are all about creating, and others just want to tear stuff down.

[Marius]

2 hours ago, candle said:

Some people, that are known to copy other's people work and sell it prevent me from releasing schematics for SIDE3 hardware - this was the case for SIDE2 that was cloned, Ultimate1MB, and even VBXE

 

I noticed a while back someone posting a picture of a 'fake' device. Can't remember whether it was SIDE2 or U1MB, but it was an clone indeed. I responded to that post that it felt wrong that someone did this. I got a pretty annoyed and abusive answer. 

 

Pity that people do not simply buy the hardware. It is not that you or Lotharek make a fortune out of it. 

 

Good that you write about these things on the forum. I wished it wasn't necessary.

 

 

 

[dostrow]

On 10/23/2020 at 5:21 AM, flashjazzcat said:

In the absence of a any other commentary, I can only refer again to my experience with a 65XE which worked with every cartridge except SIDE3 (which was either not seen at all, or displayed a black screen). This turned out to be a timing issue on the host machine, fixed by replacing 74LS08. I am not saying this is the issue with your machine, but if it happened with one machine, it can happen with two machines. I agree that actually returning the item to Poland should be a last resort. Perhaps Lotharek will advise on his return if you explain the issue to him by email.

Taking a look at the 74LS08 on my board (it is a Rev. B board with the bodge wires connecting pin 8 to a via between GTIA and ANTIC and pin 10 to pin 15 on the 74LS138 next to it, these seem pretty standard for all the pics of Rev. B boards out there...) the only other notable things are that pin 9 had a GIANT solder blob on it on the top of the board (which I removed with my desoldering gun, to no effect) and pin 9 and 10 are jumped on the bottom of the board with solder, given the flux residue this was clearly intentional, unsure if like the bodge wires this was standard practice but jumping the two inputs to the third gate of this chip sure seems odd. I haven't been able to find any pics of the underside of this board but at least in Rev 4 (which also lacks the bodge wires) those pins are not connected. I have a spare 74LS08 sitting in a dead 800XL that I could scavenge but I want to make sure I understand exactly what of these changes I am supposed to re-create after replacing the chip.

[flashjazzcat]

1 minute ago, dostrow said:

unsure if like the bodge wires this was standard practice but jumping the two inputs to the third gate of this chip sure seems odd.

Yes: this is a standard factory bodge and nothing to worry about. On the machine that ended up with a socketed 74LS08, I simply replicated the bridged pins and jumper wires on the socket. The machine worked with SIDE3 after the original chip was swapped with another randomly selected 74LS08.

[dostrow]

5 minutes ago, flashjazzcat said:

Yes: this is a standard factory bodge and nothing to worry about. On the machine that ended up with a socketed 74LS08, I simply replicated the bridged pins and jumper wires on the socket. The machine worked with SIDE3 after the original chip was swapped with another randomly selected 74LS08.

Perfect! Thanks for the info. I don't currently have any 14 pin sockets on hand so ordering those right now (gotta love Amazon Prime). Also going to just get a few brand new 74LS08s from Mouser while I'm at it just in case the one in the 800XL carcass that I have is questionable. Will report back based on what I find.

[+remowilliams]

8 hours ago, flashjazzcat said:

Again: clearly the cards initialise, accept commands and attempt to deliver data, so perhaps we have a secondary issue with some card/host machine combos.

Understood, I'm no stranger to the potential timing weirdness in A8 machines (I don't recall this machine ever exhibiting timing weirdness)  Though in this case a 'newer' SD card being more stable seems to make sense. 

 

8 hours ago, flashjazzcat said:

You can try running Drac030's RWCRC tool (available from his website) on APT partitions. This tool writes a 1MB or so file full of CRC checked data, then reads it back and scrupulously tests for errors. If the file reads back perfectly, it's likely that your system is stable. This tool will work just fine with the SIDE3.SYS driver if you don't have U1MB.

Thanks, I've run that a bunch of times with no errors shown.  Now that this setup seems to be quite stable, it's time to give it to the 1200XL and see what it thinks... 

 

Random question - does the SDX FAT driver still only work with FAT16?

[flashjazzcat]

15 minutes ago, remowilliams said:

Random question - does the SDX FAT driver still only work with FAT16?

It does, to the best of my knowledge.

[flashjazzcat]

Thanks to @TJ76 for pointing out an issue with '180' failing after the first round of play. This issue was caused by the game's odd reliance on the lower part of the stack area being set to $00, and the fix is common to the SIDE3 loader, SIDE loader and XEL Loader. Here's an update for SIDE3:

 

s3loader.xex

 

The updates for the other hardware targets will be released as part of the forthcoming U1MB/Incognito/SIDE firmware update, which largely addresses bug fixes.

[dostrow]

 

On 10/24/2020 at 2:18 PM, flashjazzcat said:

Yes: this is a standard factory bodge and nothing to worry about. On the machine that ended up with a socketed 74LS08, I simply replicated the bridged pins and jumper wires on the socket. The machine worked with SIDE3 after the original chip was swapped with another randomly selected 74LS08.

I got the sockets in today and swapped the 74LS08 out with the one that was in the 800XL carcass I have (the original was a Mitsubishi 74LS08, the one I pulled from the 800XL was a TI) but it exhibits the same behavior. I have nothing left but to assume that the hardware issues that @Candle referred to have affected me. This is the last day of vacation for @lotharek so once he is back I'll get the RMA process started, I've already sent him an email for when he returns. Thanks for all the help to both of you in troubleshooting.

Edited October 25, 2020 by dostrow

[+mytek]

Replace the 74LS08 with a SN74F08N. This has fixed many a timing problem related to adding upgrades to your A8.

 

[Roydea6]

7 hours ago, flashjazzcat said:

The updates for the other hardware targets will be released as part of the forthcoming U1MB/Incognito/SIDE firmware update, which largely addresses bug fixes.

I know that it is not a bug, but will the Side3 Clock be fixed soon.

Edited October 26, 2020 by Roydea6
miss spelling

[Jimbaloid]

8 hours ago, mytek said:

Replace the 74LS08 with a SN74F08N. This has fixed many a timing problem related to adding upgrades to your A8.

 

I've ordered one for my marginal machine too - I will report back with results later this week. 

 

In the meantime, the other new SD card that I ordered has turned up and it works as expected in the two known good systems. (SanDisk Ultra 16GB.) 

 

?

 

[flashjazzcat]

5 hours ago, Roydea6 said:

I know that it is not a bug, but will the Side3 Clock be fixed soon.

You mean the 'Z:' driver? As I explained: there is simply no room for it in the current implementation of the U1MB SIDE3 PBI BIOS.

 

[Roydea6]

 @flashjazzcat  NO no not the z.sys driver I mean the stand alone side3clock.sys like ultime.sys

 

side

[flashjazzcat]

6 minutes ago, Roydea6 said:

NO no not the z.sys driver I mean the stand alone side3clock.sys like ultime.sys

Ah - I see. Well, if the SDX clock driver doesn't work, it's a bug, but unfortunately I have no idea when it will be fixed since it's DLT's responsibility.

 

[+ZuluGula]

On 10/24/2020 at 11:02 AM, Marius said:

I noticed a while back someone posting a picture of a 'fake' device. Can't remember whether it was SIDE2 or U1MB, but it was an clone indeed. I responded to that post that it felt wrong that someone did this. I got a pretty annoyed and abusive answer.

Could you share a link to this post? I was searching in your recent posts, but I have trouble finding anything. In the last six years I started following Atari again, I haven't seen cloned or copied U1MB, SIDE2 or SIDE3.

 

On the other hand, in my opinion, we could all benefit more if the creators of such devices were willing to licence them to more people and allow modifications. Think about 1088XEL with integrated U1MB and UAV without need for separate boards. The other benefit would be better prices and service for people in US if those devices were distributed locally. It has been shown multiple times that Atari community will respect and reward creators even if you could get something for free.

[flashjazzcat]

Next U1MB SIDE3 plugin update runs the SIDE3 loader regardless of whether the PBI HDD is enabled or not. This seems a little less confusing and accounts for any situation where the PBI HDD needs to be disabled during normal operation. Only when SIDE3 isn't actually plugged into the machine will 'L' run the SIDE2 loader built into the U1MB.

 

Edited October 27, 2020 by flashjazzcat

[dostrow]

On 10/24/2020 at 2:25 PM, dostrow said:

Perfect! Thanks for the info. I don't currently have any 14 pin sockets on hand so ordering those right now (gotta love Amazon Prime). Also going to just get a few brand new 74LS08s from Mouser while I'm at it just in case the one in the 800XL carcass that I have is questionable. Will report back based on what I find.

Welp...I got it working....and how is somewhat annoying. I had purchased a new power supply from lotharek when I bought the U1MB and the SIDE3. The moment it came in the mail I switched to using it exclusively. When I switch back to using the original power supply from the mid 80s...everything works. Swap out to the power supply I bought from lotharek and the cart no longer is detected.

[+DrVenkman]

12 minutes ago, dostrow said:

Welp...I got it working....and how is somewhat annoying. I had purchased a new power supply from lotharek when I bought the U1MB and the SIDE3. The moment it came in the mail I switched to using it exclusively. When I switch back to using the original power supply from the mid 80s...everything works. Swap out to the power supply I bought from lotharek and the cart no longer is detected.

I wonder if that PSU is putting out a ripple current that’s affecting bus signal levels or timing. Do you have a scope or digital multimeter you can use to check for ripple or other noise from the PSU?