Quick Question about VBXE

[drac030]

I, as everyone else, usually miss "important" meetings nobody told me about. But if candle finally has done what I asked him to do several times and diagnosed the problem (which according to you was not yet diagnosed 25 posts ago, quotation: "Nobody knows what causes the glitch yet" (c) 2020 Fjc), then I will be actually more than delighted, because this finally get us closer to the truth and the solution.

 

Edited December 5, 2020 by drac030

[flashjazzcat]

Oh, so there is a canonical issue at least?

 

'Nobody knows what causes the glitch' was an attempt at protecting the feelings of the easily triggered. I am told the problem is Rapidus.

[drac030]

7 minutes ago, flashjazzcat said:

Oh, so there is a canonical issue at least?

You are saying so. I cannot observe this.

 

The point is that candle in (IIRC) August 2020 had my computer in his home, for two weeks. Along with Rapidus. So, there was chance to trace problems down if there are any, and officially pronounce the offending party.

 

I guess nothing like that was done.

 

As for "protecting feelings", if I was malicious, I could say that you do not need protect your feelings so carefully. But I am not malicious, so I will not say that.

Edited December 5, 2020 by drac030

[candle]

7 minutes ago, drac030 said:

The point is that candle in (IIRC) August 2020 had my computer in his home, for two weeks. Along with Rapidus. So, there was chance to trace problems down if there are any, and officially pronounce the offending party.

 

problem with this machine were found without even installing u1mb in there, thus, futher investigation was pointless

from where i stand, rapidus works only if every part in given machine was touched by the god himself, and have his blessing

appearantly, he haven't touched all parts there, since problem would manifest if you would hit reset button repetivly while on rapidus boot menu

 

all communication with the creator of rapidus, moderated by lotharek, was declined, or laughted at, so i didn't care to persue this futher - this is pointless... 

 

perhaps, in 20 years this community will have a trully working 816 solution, but since then, all we have is semi-working one that might just work for someone, and certainly not to another

 

i won't comment on this ever again

 

[drac030]

If anyone wants to see the Atari motherboard with all parts touched by God himself, I can provide some photos.

 

Also, there is nothing like Rapidus Boot Menu. You probably mean Rapidus Configuration Menu, but then again, hitting Reset while in it results just in a reboot.

 

By strange coincidence, I currently have two Rapidus boards. If there is some impartial engineer in Warsaw, who has the required skill and equipment, and also the williingness and time to make research on the possible issues, which reportedly occur, when one installs both U1MB and Rapidus in one machine, I am willing to give the board to him for a period of time at least needed to complete the research.

 

Edited December 5, 2020 by drac030

[candle]

i do belive that 816 less cores will have far less compatibility issues than the ones runnig with overclocked 14mhz rated part

i also know for the fact, that pasiu doesn't acknowledge results of simple test with 20% variance in power supply voltage, and rapidus usually doesn't work with anything lower than 5.2V on power supply

from my point of view, if it doesn't work from 4.2 to 5.5V it doesn't work at all

 

[drac030]

I am currently using the production Rapidus with the production core 054e (i.e. it is running on hard-core 65C816). Besides, there are no noticeable differences in operation between the hardcore and 20 MHz softcore. Neither one is less stable than the other.

 

Anyway, the Rapidus in my offer will have the standard core flashed in.

Edited December 5, 2020 by drac030

[candle]

make it run on 20 diffrent machines with 20 diffrent power supplies, then we talk

single point data set is not important

you know well, that if i connect your computer to my laboratory power supply and set 5.000V @ 3A it won't work at all

 

[drac030]

If you know for sure that the power supply is the issue, then could you please make one which supplies the required voltage and send it to some place in the Britain? I will pay for it, if it helps saving everyone all that unnecessary overhead.

 

I myself do not have 20 computers nor power supplies, but perhaps the impartial engineer has one power supply which can be adjusted. Who knows.

 

Also, if someone needs a photo of my power supply, I can post one.

Edited December 5, 2020 by drac030

[candle]

you can't blame power supply there - they are from difrent manufacturers, eras and generally speaking - not important

important is, that power supply variances influence rapidus being functional, or not so much

to end this debate, i've sent you final words of the creator to your email box, please read it, and let us forget whole thing

it is not worth it - really

 

[drac030]

Well, no. No forgetting. If there is an issue, it must be fixed, and if there is no issue, then the reports on the alleged issue must be explained.

[+tf_hh]

Dear Rapidus-Developer, U1MB Developer and all other involved guys!

 

First: I don´t want to run into the battlefield here. I don´t want to confirm or falsify anything. These are my personal experiences and insights. I wrote them NOW, because somebody points me to this thread with issues I try to solve for more then 2 years.

 

This week I installed the 3rd Rapidus XL so far. All problems I discover I also have with the both installations before. All three systems were stock 800XL PAL machines. All of them were tested before for a minimum of 24h with the CPS SuperSALT test cartridge plus external test unit. At all machines I´ve installed new machine head precision sockets, not dependent if the chips were in sockets before or not. During all tests...

 

- more then 6 different CPUs

- more than 8 sets of DRAMs

- more then 3 different ANTIC and GTIAs

- different PIAs

- different POKEYs (or POKEYMAX)

- different power supplies

 

are used for testing. Because "power supply" was a keyword here: I tested:

 

- Genuine Atari power supplies with 1,5A (enough, even fully expanded I never go above 1,32 amps), of course no Ingot ?

- My labour power supply (here I tested from 4.85 to 5.40 volts)

- Several switched power supplies with 2 or more amps (mostly Raspberry PI PSU)

- A self-built classic power supply with 2A versions of 78xx standard regulators, highend electrolytic caps, filters and massive metal cooling (800 grams heat sink each 78xx) and a fan.

 

It´s absolute INDEPENDENT which power supply, which chips etc. are used. Some combinations make more trouble, some less. But NO combination works really flawless. And yes, I´ve used standard ribbon cables, I´ve used shielded cabels, I´ve made my own shielded U1MB cables, I´ve done two dozens things, but there´s no effect. All signals tested with a scope are normal - except the RESET pin at 6502C CPU mounted on Rapidus, it has max. 3.5 volts for high-level, while the 65C816 CPU has 4.8 volts at RESET. But 3.5 volts is enough, so I wonder about it, but I won´t see a problem here. And: As long you didn´t press or perform a reset, it works - classic mode or Rapidus...

 

I´ve tried all the patches suggested by Lotharek, plus 100pF between RESET and GND, in combination or single, not dependend...

 

Current setup which all videos and pictures made:

 

- Atari 800XL PAL

- Ultimate 1 MB (first tests with Candle BIOS as shipped out from Lotharek, later with FJC BIOS 1.30 - there IS NO DIFFERENCE!)

- VBXE 2 XL version

- Rapidus XL

- PokeyMAX Dual Stereo with Covox Auto (all tests also with one standard POKEY... no difference!)

 

The machine works relative good when Rapidus config ist set to "Rapidus". But talking about the 6502C "standard" mode... it´s annoying. The machine starts of course, you can run anything you want, but pressing RESET mostly ends up in a hang-up of the machine. With some reset-proof programs you can watch, that every keypress of RESET makes it more worse. Mostly after 2-3 times pressing RESET you got no picture anymore or disturbing graphics ("wild displaylists"). Coding is a hard job this way.

 

But more odd issues happen. I was not able to run any standard Atari 16 KB game cartridge when Rapidus mode is active. In standard mode they work fine, but with Rapidus... no chance. Also some 8 KB carts won´t run. AtariMax carts run fine as long the standard AtariMax software is used, but for example the "KOMA" images (60 games in one or several disk based games in one cart) for AtariMax flashcarts won´t run or make a lot of problems when trying to start one game.

 

ACP - Atari Control Picture - won´t run at Rapidus mode. SysInfo V2.25 makes wrong characters at the GTIA test. And so on... several programs like simple sector copiers etc. won´t work.

 

Some examples:

 

 

ACP.COM loaded with DOS 2.5, no fancy stuff.

 

The following example shows my setup to have no discussion about "what you´ve done":

 

 

 

 

 

 

 

 

This setting was all used with the machines I´ve build. Of course I´ve tried several other settings, too.

 

Now some videos to show issues more better than words...

 

 

This video shows one of the RESET issues. Standard XL OS, standard Atari-Basic C in U1MB selected. Rapidus settings to "Classic mode". Rapidus core via U1MB BIOS disabled. Happy DOS loaded. Go into BASIC. Press RESET... hang up. Second try without DOS menu call. Same. Later I test completely without DOS... same. You press RESET one, two or three times and machine hangs.

 

 

Same like above, this time using DOS 2.5 unpatched and Rapidus settings remain at "Classic mode", but Rapidus core using U1MB BIOS set to active - you see the loading bar of Rapidus Core.

 

 

This video starts with disabled Rapidus (set to Classic Mode). The game - Rescue on fractalus - starts as usual. After activating the Rapidus core (Rapidus settings set to "Rapidus mode" before and save this setting, of course) the loading bar appears, but nothing happens. Tried with several other standard Atari game carts and the CPS SuperSALT diagnostic card. All the same.

 

 

Next example of unexpected behavior: Running the sector copy tool from Speedy 1050 tooldisk. By default it´s reset-proof whatever you to, coldstart possible by special keystroke. When I press RESET with Rapidus core active, an endless reset loop appears. No chance to enter the U1MB BIOS nor the Rapidus setting. Must power cycle.

 

 

Copy2000 from is the next example. It´s also reset-proof. As you can see, with every RESET the display gets more and more disturbed (look at the single big characters and the menu line 2 and 3), After some RESETs only garbage comes. BUT: You can do (!) copies w/o failure! If you know how to use the program "blind", it works. Copied disks are intact and will work, byte-by-byte verify is okay. You didn´t see anything, but can copy disks.

 

BTW: All pictures and videos are made using the VBXE RGB output, but I always switched to standard S-Video output... same behavior.

 

 

This video I made in Rapidus-Mode, full speed, standard setting "Rapidus". The little listing runs perfect, until you press RESET one or two times. After this, the listing produces these "output"...

 

I could sent another 20 examples I needed, but IMHO there must be an big problem anywhere. All these issues are EXACTLY the same with all three systems. The first system (800XL, Rapidus, U1MB, classic Gumby Stereo, and a black PCB VBXE 2.0 made by Candle) build September 2018 makes exact the same issues. The 2nd machine I´ve upgraded was 800XL, Rapidus, U1MB and PokeyMAX V1 (first batch from FOFT). Changing to a standard POKEY haven´t had any impact.

 

3rd and currently at my desk system have the same issues. This system was used for the pictures and videos.

 

So now, dear developer... please stop ranting and disputing. From my view I can´t recommend Rapidus for our ABBUC members and other friends of mine. All these systems run perfect without Rapidus, and when the U1MB is really the problem, then - that´s my view of good service - a solution must be found regarding the fact, that the U1MB is some hundred times in use. Myself already equipped more then 50 machines with that expansion - until now without ANY problem.

 

Thanks to the audience.

 

 

 

[drac030]

Well, finally someone did something constructive instead of "get more info from Lotharek" and such nonsense. Thank you very much.

 

To my knowledge, though, in 6502C mode ("Classic") the Rapidus board is in idle state (I know this by accident, because in one of the prototypes there was an issue in this mode and logic state analyzer was used to check what is wrong). I also do not observe any problems with Reset in this mode (in any mode, to be specific). It is certainly a step forward, then, but we are still lacking conclusion, i.e. what is the immediate cause of the phenomena observed on some machines, or, if anyone prefers, why other computers do not show them.

 

EDIT (apologies for bad quality, made with a phone being held in one hand):

 

 

Quote

ACP - Atari Control Picture - won´t run at Rapidus mode. SysInfo V2.25 makes wrong characters at the GTIA test. And so on... several programs like simple sector copiers etc. won´t work.

 

This is absolutely nothing odd: programs which use the so called cycling (to switch modes or colors in the middle of a scanline) or delay loops or such techniques based on assumption that certain amount of code makes certain delay - you cannot reasonably expect them to run as on 6502/1.77 MHz when the CPU is 11 times faster.

 

Quote

Rapidus core via U1MB BIOS disabled.

 

What does this mean/do?

 

Edited December 6, 2020 by drac030

[Dmitry]

Thanks for the info.  I haven't read through all of it yet, but I got some important info.  The main thing I didn't realize is the Rapidus card supplies its own OS.   Somehow in my previous configuration DracOS was one of the ROM slots in U1MB.  I noticed it, it definitely had been flashed previously, but if that wasn't needed, then it was just harmlessly not being used from that card, and loading from the Rapidus.

 

And I'm probably going to update the Rapidus firmware as well later today.  Not because I believe it will solve the problem.  I'm flashing the latest because I feel it is pointless to troubleshoot older versions.  Just generally speaking I like in situations like these to know i have the latest, and that would be what to troubleshoot.

 

I said it once before, but to be clear, I own an Atari 1200XL, and I'm not willing to give up the "feature" that it doesn't have an internal basic.

And therefore I should never have to disable something which it does not have.  I will not be holding down option to start the machine.

 

 

Despite the fact that I initially didn't use U1MB for anything, even its namesake of supplying 1mb of memory isn't something i needed since my understanding is both the VBXE (with correct core) or the Rapidus, could supply 256K of extended memory.  I do have the desire at times to simulate 130XE banked memory at up to 128K, so I'm covered. I never use any game loaders, that isn't of interest to me.  The other excellent features of the U1MB - are very interesting, but at this time, not requirements.   However, I've come to use SDX, and at this time, U1MB supplies my copy of SDX.

 

I mostly dink around with coding, not playing games.  Despite the fact that I'm coding a game this go around, I may not on the next project.  Anyway...gotta get back to that.

 

I am just going to use the PBI bios, with basic disabled.  If that works, great....   If not, I will have to figure out what compatible device wil supply SDX - it could be just the SDX cartridge, or maybe I do get a SIDE or something like that.

 

 

Edited December 6, 2020 by Dmitry

[drac030]

50 minutes ago, Dmitry said:

If not, I will have to figure out what compatible device wil supply SDX - it could be just the SDX cartridge, or maybe I do get a SIDE or something like that.

SIDE, SIDE2, SDX Supercart. Also IDE+, but I am not sure on availability.

 

[xxl]

does Rapidus BIOS initialize Atari hardware registers during reset?

[drac030]

No, it does not have any reason to do that.

 

@tf_hh Your videos do not show U1MB settings. So one question: do you have (or did you have during the tests) the U1MB PBI enabled, and if so, what is the ID of the U1MB PBI ROM? The symptoms - problems during reset, core loading and overall initialization, I/O etc. - are similar to what one gets when there is a conflict on the PBI bus.

[xxl]

There are reports that Rapidus works very stably with Ultimate, provided that it is not Ultimate with FJC soft. It will work with the original U1MB software.

[+Stephen]

Is this another booby trap?  I don't recall fjc's software making repeated writes to PIA

 

It's also been WELL established that it is NOT his BIOS causing the issue.  What the absolute fuck is with the fragile ego syndrome over there?  Finding a bug doesn't imply a small penis.  If there is a bug in software, fjc fixes it and THANKS the person that found the issue.  When there is a hardware issue found, grown men scream and cry like little 5 year old babies blaming everyone but themselves.  I don't get it but I am finding it increasingly difficult to deal with and it's making me for the first time ever, not want to be a part of this scene.

[drac030]

36 minutes ago, Stephen said:

Is this another booby trap? 

I realize well that you may have great difficulties believing his words, but it is actually true: @krap the admin of krap.pl (the server which is hosting AtariArea, among several other Atari-related websites) owns a machine which, according to his words, was behaving like that, i.e. was not stable until the U1MB firmware has been reflashed to the version which U1MB contained when it first was being distributed.

 

But if I understand correcty, @tf_hh above wrote that he tested this combination too and it made no difference for him. So I my understanding that krap's case is the classic post hoc ergo propter hoc, and in fact some other factor played a role, which has been overlooked.

Edited December 7, 2020 by drac030

[xxl]

another user reports that soft FJC works in 65816 mode and not in Classic mode

[xxl]

7 hours ago, Stephen said:

When there is a hardware issue found, grown men scream and cry like little 5 year old babies blaming everyone but themselves.

I totally agree with you, all the crying on the forum and then the drama on YT ... if it wasn't pathetic it would be even funny.

 

 

[+tf_hh]

Hi!

 

first: Thank you very much for a fast response!

 

13 hours ago, drac030 said:

EDIT (apologies for bad quality, made with a phone being held in one hand):

 

This video shows Rapidus, but not together with an U1MB - right? As sometimes already stated, the issues are only when both expansions are build together. Later I will remove the U1MB and perform some tests - I´m curious about the results.

 

13 hours ago, drac030 said:

This is absolutely nothing odd: programs which use the so called cycling (to switch modes or colors in the middle of a scanline) or delay loops or such techniques based on assumption that certain amount of code makes certain delay - you cannot reasonably expect them to run as on 6502/1.77 MHz when the CPU is 11 times faster.

 

Ok, something I can understand and accept. I was wondered about the wrong graphic display in SysCheck - my assumption was, that an utility from the same coder like the Rapidus software would handle this anyway ? - This point is fine for me.

 

13 hours ago, drac030 said:

What does this mean/do?

 

I´ve connected the "S0" pin of U1MB to the "GND" (most left from normal sight) pin auf the Rapidus three-pin-connector at the lower right side. In conjuction with FJC´s BIOS you enable/disable Rapidus here. But this makes no difference, when signal is set to GND (Rapidus enable, as standard installation shows) and Rapidus mode set to Classic, the problems remain.

 

I have a 32 channel logic analyzer and a scope, of course. When I can do anything to test, let me know.

 

[+tf_hh]

13 hours ago, Dmitry said:

And I'm probably going to update the Rapidus firmware as well later today.  Not because I believe it will solve the problem.  I'm flashing the latest because I feel it is pointless to troubleshoot older versions.  Just generally speaking I like in situations like these to know i have the latest, and that would be what to troubleshoot.

 

Which new version? IMHO there´s nothing new since 3 years...?

 

[+tf_hh]

12 hours ago, drac030 said:

Your videos do not show U1MB settings. So one question: do you have (or did you have during the tests) the U1MB PBI enabled, and if so, what is the ID of the U1MB PBI ROM? The symptoms - problems during reset, core loading and overall initialization, I/O etc. - are similar to what one gets when there is a conflict on the PBI bus.

 

No, at all times the PBI BIOS is disabled, because I only use a SIO2SD for loading software.