Any Success Stories with Rapidus?

[lotharek]

ps. very interesting effects gives pull down of o2 signal at ls08 pin 1 -- 3k -- gnd

 

in simple word : solder resistor of 3kom betwen first and last bottom pin of ls08

test please and report

 

ps. by the way, it is quite good to replace it witg f08

 

why ? in most recent FJCs SIDE firmware problems may appear with cf recognition

 

PICTURE

Edited June 8, 2017 by lotharek

[flashjazzcat]

Yep: I discovered CF card issues on my 600XL and 800XL (both without Rapidus), and fixed both by replacing LS08 (although I recommend a simple like-for-like swap, i.e. another LS08). The fix is equally valid when Rapidus is present, and can improve stability regardless of the presence of SIDE2.

 

I'll try the pull-down on the current "problem" Rapidus machine tonight and see if helps with the last of the reset issues. Note: pullups on A0-A7 are absolutely essential on this machine, otherwise it suffers immediate reboots on reset in all modes. Note the CPU in this FREDDIE 800XL is NCR, and it has the "reboots on reset" issue, so use of the NCR CPU still isn't a guaranteed cure-all, I'm afraid.

 

BTW: Thanks for corroborating overheating issues. That's welcome after the degree of ridicule my findings received from some quarters last year.

Edited June 8, 2017 by flashjazzcat

[lotharek]

i think, when temp goes high; o2 signal gets some disturbance. anyway, i stat to believe it is not cpu-related problem, but general trash on atari data lines or production quality differcence between factories or.. whats the worst: TIME

 

anyway; on my test 65xe machine : i have tested all cpus i have for last two weeks and except higher temp than usual; no porblems with rapidus. Even once has not progress bar hung up.

 

PS. no pullup on data; no other mods. pure pcb with with vbxe, ultimate, stereo..

Edited June 8, 2017 by lotharek

[lotharek]

Yep: I discovered CF card issues on my 600XL and 800XL (both without Rapidus), and fixed both by replacing LS08 (although I recommend a simple like-for-like swap, i.e. another LS08). The fix is equally valid when Rapidus is present, and can improve stability regardless of the presence of SIDE2.

 

f08 is quite faster and give more juice to cart slot, when buffering o2 signal from cpu. speed - doesnt matter of course here, but overall is better

[drac030]

Thanks for corroborating overheating issues. That's welcome after the degree of ridicule my findings received from some quarters last year.

You have possibly received a "degree of ridicule" from "some quarters last year" not because of stating that 6502 is overheating but because of failing (or neglecting) to prove that statement. Long story short: the observation that 6502 is hot during operation (or even when it is halted) does not prove that it is overheating, simply because the 6502 is usually hot whatever it is doing (or even when it is doing nothing), even in an unmodified XE.

 

Statements (not yours, if I remember correctly, but not explicitly denied by anyone) that 6502 is known to get overheated when held in the reset state for a longer time did not help either, because this has nothing to do with the Rapidus (where the 6502 is kept in the HALTED state, when not used, not in the RESET state).

Edited June 8, 2017 by drac030

[lotharek]

Getlemen, please... we have problem to solve and i think is hardware related not personal

 

ps. i think Pasiu mentioned about pull down of o2 signal once here..sorry, didn`t take care faster and hope it solves all issues (even on buggy 800xl F)

 

proof:

Edited June 8, 2017 by lotharek

[+Stephen]

Hopefully everything can be worked out, this accelerator is a really cool design which I'd like to try at some point.

[flashjazzcat]

i think, when temp goes high; o2 signal gets some disturbance. anyway, i stat to believe it is not cpu-related problem, but general trash on atari data lines or production quality differcence between factories or.. whats the worst: TIME

 

anyway; on my test 65xe machine : i have tested all cpus i have for last two weeks and except higher temp than usual; no porblems with rapidus. Even once has not progress bar hung up.

 

PS. no pullup on data; no other mods. pure pcb with with vbxe, ultimate, stereo..

Interesting - thanks. And I must offer a correction to my previous post: it's a NON-FREDDIE 800XL.

 

f08 is quite faster and give more juice to cart slot, when buffering o2 signal from cpu. speed - doesnt matter of course here, but overall is better

OK - worth a try too. Maybe I have some lying around here.

 

You have possibly received a "degree of ridicule" from "some quarters last year" not because of stating that 6502 is overheating but because of failing (or neglecting) to prove that statement. Long story short: the observation that 6502 is hot during operation (or even when it is halted) does not prove that it is overheating, simply because the 6502 is usually hot whatever it is doing (or even when it is doing nothing), even in an unmodified XE.

It may be convenient to accuse my fingertip of lying or incompetence, but the important finding was that higher running temperature (whether higher than normal or not, although we now find that it is) resulted in reduced stability. It seems this just falls into the category of inconvenient truth.

 

Statements (not yours, if I remember correctly, but not explicitly denied by anyone) that 6502 is known to get overheated when held in the reset state for a longer time did not help either, because this has nothing to do with the Rapidus (where the 6502 is kept in the HALTED state, when not used, not in the RESET state).

Well, anyone with the expertise to do so had over a year to come here and definitively and publicly clarify the state in which the 6502 is held. Without feedback from the developers and with the impression actual findings are falling on deaf ears, those (like me) tasked with actually building working systems for people who have bought the hardware have no option but to try anything and hope it works. Of course, I had the option to refuse installation, and it's certainly one I wish I had exercised at the time (as Lotharek himself reserves the right to do "if there are other upgrades already in the machine").

[drac030]

It may be convenient to accuse my fingertip of lying or incompetence, but the important finding was that higher running temperature (whether higher than normal or not, although we now find that it is) resulted in reduced stability. It seems this just falls into the category of inconvenient truth.

You are still to prove this "incovenient truth" with a thermometer, not with just a finger of yours. This is, if you still did not notice, the very problem with your findings: they may well be honest and true, and I by default trust they are just such, but you fail to prove them by anthying besides your statement that they are true. Sorry, people are error-prone, so no proofless statements are to be accepted without reservations.

 

A pratical example: "overheating 6502" - ok, what is "over"? How many Celsius degrees over the norm, which is how many Celsius degrees and the circumstances to collect those data are defined as how?

 

Well, anyone with the expertise to do so had over a year to come here and definitively and publicly clarify the state in which the 6502 is held.

Anyone also had over a year to ask in what state the 6502 was held. Noone asked that, just some rubbish was assumed.

 

And obviously, you were not obliged to install anything.

Edited June 8, 2017 by drac030

[flashjazzcat]

You are still to prove this "incovenient truth" with a thermometer, not with just a finger of yours...

Although I now have the equipment needed to accurately measure the temperature, I suppose it's no longer necessary since Lotharek can presumably provide accurate readings for you. All I'll say is: what do you think prompted me to take the time and trouble to place a heatsink on top of the 6502C? Was it a vision that appeared in a dream, or some voices coming from under the floor? No: it was because I happened to touch the 6502C and noticed it was much hotter than normal. Take it or leave it.

 

And obviously, you were not obliged to install anything.

Yeah: I could have knocked fifty per cent off the agreed installation fee and returned the machine to the owner with the Rapidus in a separate bag, but I wanted to make it work and make the owner happy. My bad.

 

Anyway: end of off-topic and I apologise for using trigger words like "corroborating" and "ridicule" in the same paragraph.

 

ps. i think Pasiu mentioned about pull down of o2 signal once here..sorry, didn`t take care faster and hope it solves all issues (even on buggy 800xl F)

Excellent. That's three things to try (pullups on A8-A15, pulldown on LS08, and pulldown on 02).

Edited June 8, 2017 by flashjazzcat

[drac030]

Although I now have the equipment needed to accurately measure the temperature, I suppose it's no longer necessary since Lotharek can presumably provide accurate readings for you.

I take it as your confession that you did not really measure the temperature before stating on the forum that the 6502 was overheating. No problem. Statements are quite different from proofs.

 

Still, it would be better to provide a verifiable proof. I think noone could then have a problem with accepting it.

 

what do you think prompted me to take the time and trouble to place a heatsink on top of the 6502C?

Your fingertip, that is out of the question.

 

it was because I happened to touch the 6502C and noticed it was much hotter than normal

This is still to be proved with a thermometer. You know, I happened to touch the 6502C many times in the stock Atari (at least I have been an Atarian since late 80-ies), and every time I was noticing it was very hot. In Rapidus - no difference attracting attention. "You do not believe me?" Then why I have to believe you? I am afraid you have to provide VERIFIABLE and accurate figures in Celsius degrees together with the method you measured them. Because a finger, you know, is a subjective method of measuring anything (the history has recorded people who did not find the Bunsen burner hot enough to prevent them from inserting fingers into it - and not getting hurt - so you know).

Edited June 8, 2017 by drac030

[Keatah]

So, assuming I'm competent enough to have pointed the hair dryer at the correct component and my observations are not clouded by the symptoms of mental illness: looks like the 6502 overheats!

 

Could be the 6502, or a nearby component sinking or sourcing X amount of heat. And that could be causing subtle timing/spec changes someplace.

 

Not having real Atari 8-bit hardware at my disposal, let alone a Rapidus board, that's all I could say. But it's sound reasoning.

 

Usually circuitry that's sensitive to temperature changes is designed from the get-go with insufficient tolerances someplace. Could be as simple as an RC network variance someplace.

 

---

 

Anyways, the only reason I'm reading this thread is a friend's rig isn't working right and I'd like to get the chance to work on it.

[lotharek]

Anyway: end of off-topic and I apologise for using trigger words like "corroborating" and "ridicule" in the same paragraph.

 

 

Excellent. That's three things to try (pullups on A8-A15, pulldown on LS08, and pulldown on 02).

 

puldown ls38 pin 1 = pulldown o2, but in different place

would start from that without others mods.

Edited June 8, 2017 by lotharek

[flashjazzcat]

I take it as your confession that you did not really measure the temperature before stating on the forum that the 6502 was overheating. No problem. Statements are quite different from proofs.

 

Still, it would be better to provide a verifiable proof. I think noone could then have a problem with accepting it.

I never claimed to have measured the temperature with a thermometer in the first place, but I infer from what you're saying that you don't believe Lotharek either when he writes that he discovered overheating issues. In which case, please argue it out between yourselves. I can no more go backwards in time and use a thermometer to measure something which apparently turned out to be correct about than anyone else can go back in time and ensure the issue didn't exist in the first place.

 

This is still to be proved with a thermometer. You know, I happened to touch the 6502C many times in the stock Atari (at least I have been an Atarian since late 80-ies), and every time I was noticing it was very hot. In Rapidus - no difference attracting attention. "You do not believe me?" Then why I have to believe you? I am afraid you have to provide VERIFIABLE and accurate figures in Celsius degrees together with the method you measured them. Because a finger, you know, is a subjective method of measuring anything (the history has recorded people who did not find the Bunsen burner hot enough to prevent them from inserting fingers into it - and not getting hurt - so you know).

Ask Lotharek what he used to measure it. Perhaps he used his face. It seems to me that I suggested the world is spherical and now that it turns out that it's round, you're wondering where I got my evidence from.

Edited June 8, 2017 by flashjazzcat

[lotharek]

 

Ask Lotharek what he used to measure it. Perhaps he used his face. It seems to me that I suggested the world is spherical and now that it turns out that it's round, you're wondering where I got my evidence from.

 

i use cheap trorec bp21.temp measured of 65c02 with Rapidus are average 8 degrees higher and rounds 45 degrees, surr temp was 23. Supply voltage 5.26V to atari PCB. At Rapidus 0,2 lower. will prepare tommorow measuring results with photos

Edited June 8, 2017 by lotharek

[drac030]

I am just repeating and emphasizing what I found obvious about your "findings" in the past, that these are unfounded, unsubstantiated and not proven in any way. You were just saying things, as "the 6502 is overheating" but "I never claimed to have measured the temperature with a thermometer". And of course, from Lotharek I expect the temperature figures as well as from anyone else. Because, in the first place, I have never seen a correlation between my Rapidus getting hot and its stability getting impaired. And I tend to believe my personal experiences than anyone else's statements AS LONG AS THESE ARE NOT SUBSTANTIATED ENOUGH TO BE VERIFIABLE.

 

is it unreasonable, or what?

Edited June 8, 2017 by drac030

[+mytek]

I am just repeating and emphasizing what I found obvious about your "findings" in the past, that these are unfounded, unsubstantiated and not proven in any way. You were just saying things, as "the 6502 is overheating" but "I never claimed to have measured the temperature with a thermometer". And of course, from Lotharek I expect the temperature figures as well as from anyone else. Because, in the first place, I have never seen a correlation between my Rapidus getting hot and its stability getting impaired. And I tend to believe my personal experiences than anyone else's statements AS LONG AS THESE ARE NOT SUBSTANTIATED ENOUGH TO BE VERIFIABLE.

 

is it unreasonable, or what?

 

Not unreasonable, but something I learned with the BETA testing of the 1088XEL, is that just because my board worked, it didn't discount that others on my team were having problems with their's not working. If I had taken the stance that what they reported was unfounded and/or unsubstantiated just because I wasn't seeing it on my board, the final version that I am working towards would be flawed, since given time and good detective work by at least one of the beta testers a problem was discovered and solved. In the case of the 6502 running hot and possibly causing an issue with Rapidus, could be related to only certain batches of 6502's being on the edge in some regard, or just attributable to age. My suggestion would be have someone who is consistently reporting this issue to send you their 6502 so that you can verify this on your board. Perhaps this would lead to a solution .

 

- Michael

[flashjazzcat]

 

i use cheap trorec bp21.temp measured of 65c02 with Rapidus are average 8 degrees higher and rounds 45 degrees, surr temp was 23. Supply voltage 5.26V to atari PCB. At Rapidus 0,2 lower. will prepare tommorow measuring results with photos

 

Just measured surface of NCR package in 800XL (non-FREDDIE) at 56 degrees C using cheap Proster meter. I already fitted the O2 pull-down but unfortunately this has not eliminated the consistent reboot behaviour in Classic mode. CPU VCC reads 5.06V (3A switching PSU). At the end of the day, the simplest workaround I can see here is to simply not use Classic mode at all, since everything else appears to work on this machine.

 

As for personal experiences: it's likely that everyone believes their own personal experiences, and the fact that someone else's Rapidus works just great does not necessarily mean that mine does too. Moreover, I have observed similar behaviour in some five machines now (2 x 130XE, 2 x 1200XL, and 1 x 800XL) over the past year (with the XE's issues fixed by the firmware update), and I think those observations made by someone who's been upgrading Ataris for six years or so are at least as valid as the observation of one machine by someone who does not do their own soldering.

 

Clearly the issue is repeatable, but at the same time the fix that works on one or more machines does not necessarily work on all of them. So unless every problem machine is to be sent to Poland for diagnosis going forward, it seems there is still a little more troubleshooting to do.

[lotharek]

can You send me Your atari ?

 

or just please, explain : how shell i proceed to duplicate Your atari act ?

[flashjazzcat]

can You send me Your atari ?

 

or just please, explain : how shell i proceed to duplicate Your atari act ?

 

I doubt the owner will feel like sending the machine to Poland after paying me to install VBXE, U1MB, Stereo, U-Switch, Rapidus and Adaptus, but if he really wants Classic mode to work without the reboot issue, I'll let him know the offer is there.

 

There is really nothing unusual about the machine. It's a decent 800XL board (later revision, I guess, since RAM has no sockets) with all cabling kept as short as possible, precision sockets used for your upgrades, NCR CPU. If anything else dawns on me, I will certainly let you know of anything which might help. For now, I'm gonna forget about it for the rest of the evening.

[lotharek]

There is really nothing unusual about the machine. It's a decent 800XL board (later revision, I guess, since RAM has no sockets) with all cabling kept as short as possible, precision sockets used for your upgrades, NCR CPU. If anything else dawns on me, I will certainly let you know of anything which might help. For now, I'm gonna forget about it for the rest of the evening.

FJC;

i really insist on having ATARI here.. i will pay postage of course. please, contact owner fro approval.

[flashjazzcat]

FJC;

i really insist on having ATARI here.. i will pay postage of course. please, contact owner fro approval.

 

Of course I will let him know of your generous offer. Once things are settled up here, I certainly have no objections.