English Amiga Board


Go Back   English Amiga Board > News

 
 
Thread Tools
Old 30 April 2011, 03:13   #2061
FOL
PSPUAE DEV

FOL's Avatar
 
Join Date: Nov 2006
Location: Barry / UK
Age: 39
Posts: 5,591
Send a message via MSN to FOL Send a message via Skype™ to FOL
Quote:
Originally Posted by Mr B View Post
I'm not sure i would consider noise issues a fault. Jens stated it's a rare combination of factors that makes it an issue, eve with his cards, and even then, a firmware update resolves it. If anything it's an issue with the cards, not being properly designed to suit the hardware it's meant for. And, as a proof of Jens dedication, he came up with a solution. (At least i think he said it was sorted, and that now shipping cards had the fix?)

What i'm sort of wondering. This fix, as i recall, is a firmware update. Those of us that doesn't suffer from the issue, but might want to sell our accelerators, or perhaps put it in a different A1200 later. Can we obtain the fix, and apply it on our own? Does it affect performance?
B!
Really, well Jens has not communicated that to me at all regarding fix.

Quote:
Originally Posted by Claw22000 View Post
Well so far so Good ran a bunch of demos and played Alienbreed 3d for a wile and now I'm here on my Amiga! This is sweet never though I would be using this to serf the web. haha. Well now all I need is wireless and a portable DVD player with video in so I can sit in a coffie shop and blow peoples minds.
Glad its all working for you.
FOL is offline  
AdSense AdSense  
Old 30 April 2011, 10:16   #2062
Claw22000
Registered User
Claw22000's Avatar
 
Join Date: Mar 2009
Location: St. Paul, MN/USA
Age: 39
Posts: 222
Quote:
Originally Posted by FOL View Post
Really, well Jens has not communicated that to me at all regarding fix.



Glad its all working for you.
Yes I'm very glad about it just wish I knew a demo normally runs slow so I can see the difference. I did mainly get it for the game and maprom features but I still would like to see that 030 kick ass. Well more so that sysinfo.
Claw22000 is offline  
Old 30 April 2011, 11:20   #2063
Mr B
Registered User
 
Join Date: Nov 2005
Location: Sweden
Posts: 740
Quote:
Originally Posted by FOL View Post
Really, well Jens has not communicated that to me at all regarding fix.
Really? I figured we were talking about this:
Quote:
Originally Posted by Schoenfeld View Post
The trouble was - once again - power supply related. Some PSUs cause more noise than others, and if it's "on the edge", then you get a funny effect: The memory is stable with no keyboard connected, and it's getting unstable (showing 0MB fastmem most of the time) if the keyboard is connected.

This noise is a matter of a few nanoseconds only, and I managed to get around it in pure logic. All new 42MHz cards have this fix already applied.
Thats why i asked what end users currently without issues can / should do, and if the fix affects performance. At some point someone might start being affected, perhaps due to having to swap PSU, motherboard, or simply by selling of the ACA to someone who gets the issue. As i figure Jens isn't going to provide free firmware updates until the end of time if it has to be done "in-house" then that might be a reason to make sure there is a solution that lets owners apply the fix themselves.

Anyways, cheers.
B!
Mr B is offline  
Old 02 May 2011, 11:37   #2064
Schoenfeld
electricky.
 
Join Date: Jun 2010
Location: out in the wild
Posts: 1,243
That fix is indeed a firmware upgrade. It's clock-related, and it only affects the 42MHz version. With the multiplication factor of 3 and lots of noise on the supply lines, the memory controller locks up in an unknown state at some point. This is indeed a digital thing, and it can be fixed in the firmware.

However, while I still think it has a connection to the "graphics errors" effect, it's not the solution yet. Remember that the 0MB-fastmem-effect is only affecting the 42MHz cards. On both the 28MHz and 56MHz cards, the PLL runs at 4x speed where it's obviously more immune against noise. However, the "gfx errors" problem affects all three models.

I'm on it, and it seems that I have a combination that actually shows the effect now. It's a board that AmigaKit has sent to me, combined with one of the power supplies that Vesalia sent to me. I'm currently working out a procedure to *force* the effect, and once I have that, I can hopefully program the logic analyzer to record what's happening, so I can see what's going on.

Another issue that came up this weekend is the MMU functions of WHDload (MMU tooltype). I may be able to solve this by violating the Motorola datasheet and give a proper termination even to "unknown/reserved" function codes. It's not directly MMU-related, but the MMU-tooltype of WHDload uses those unknown/reserved function codes for "something" (don't know exactly what), and since they are not documented, my hardware does not terminate such accesses, trying to be as close to "the book" as possible.

Jens
Schoenfeld is offline  
Old 02 May 2011, 15:48   #2065
Schoenfeld
electricky.
 
Join Date: Jun 2010
Location: out in the wild
Posts: 1,243
Preliminary solution to GFX corruption problem

Hi guys,

I found a method to reliably reproduce the graphics corruption effect. It's an effect where the blitter gets the wrong data, thus copying the wrong data, or at least the wrong amount of data.

I had to place some insulating material on Alice to heat it up and keep it from radiating too much heat. With a cool Alice, or at least in free-floating air, it was hard to get the gfx corruption reliable, but with Alice in a cosy coat of styrofoam, I had reliable GFX corruption when moving windows around on the Workbench and when using the TGtest in AIBB.

At first I tried solving the problem on the digital side (within the CPLD of the accelerator), but failed due to it's high access speed to the chipset. As some of you like to call it, the ACA1230 is "maxing out" the AGA chipset, and to those of you who are a little more technical: my state machine relies on the chipset of catching the data on the falling edge of CPU_CLK right after DSACK has been detected active (low) on the previous rising edge. It's a matter of a few nanoseconds that Alice (and the blitter that's contained therein) is missing the data when hot.

I call the fix preliminary, because I have to wait until the whole system has cooled down and re-try. Note that Alice isn't the only one that's handling clocks - Budgie makes the largest part of that work, but Alice depends on those clocks, so I still need to verify that it works reliably on a cold start.

The board in question is a 1D.4. I can't tell if it's a Commodore or an Escom board. However, I have measured severe clock jitter especially during blitter operation. This 1D.4 board had no filters on the 14MHz CPU clock lines installed at all, but all my other boards have a filter combination assembled that is NOT seen in the schematics. The two 14MHz signals are ran through filter combos E121 and E122. Schematics say that it's 27R/22p combinations, but the faulty board has no caps installed - only the 27R resistors. During heavy blitter operation, the 14MHz clock (which is the basis for the whole accelerator timing) is jittering as much as 8ns, but it looks just fine with an idle workbench.

Remove resistors E121R and E122R and replace them with 0805 ferrite beads - something in the range of 60R@100MHz or more. The two parts are located on the top of the board, near the 28MHz oscillator (see first attached picture).

Next, assemble two 22pF capacitors (size 0805) in places E121C and E122C. These places were empty on the board in question, but need to be assembled with the parts mentioned in the A1200 schematics, and that's a 22pF capacitor. These two parts are located on the bottom of the mainboard, also near the 28MHz oscillator - see second attachment.

This will clean up the 14MHz clock "big time" and make it completely immune against blitter-induced noise. However, it's not fixing the whole problem. I found that when assembling E125C with a 22pF capacitor, but leaving E123C open, the machine is rock-solid, even with a heated-up Alice. With E125C removed, it's still showing some gfx corruption, but with the one part in place, the data-takeover into the AGA chipset is truly happening at the right moment, and the fast access of the ACA1230 doesn't cause trouble to the mainboard any more.

To give that final fix a face, I have taken a third picture of that filter group under Alice.

I found that some boards have a ferrite bead installed in E127R, and other boards have a 27R resistor installed there. This also makes a difference in chipset timing and may be another part that I have to look at if the board starts acting up again when cold. However, since I have boards working great with one or the other installed, it doesn't seem to make that much of a difference.

I'll let the board, accelerator and power supply cool down for a few hours now, and report back how it behaves when cold. If anyone of you with access to the required parts, tools and "bad boards" wants to give this a try, I'd be happy to hear your report(s).

Jens
Attached Thumbnails
Click image for larger version

Name:	E121_122R.jpg
Views:	799
Size:	171.5 KB
ID:	28575   Click image for larger version

Name:	E121_122C.jpg
Views:	774
Size:	81.0 KB
ID:	28576   Click image for larger version

Name:	Filter_group.jpg
Views:	753
Size:	67.2 KB
ID:	28577  
Schoenfeld is offline  
Old 02 May 2011, 16:46   #2066
FOL
PSPUAE DEV

FOL's Avatar
 
Join Date: Nov 2006
Location: Barry / UK
Age: 39
Posts: 5,591
Send a message via MSN to FOL Send a message via Skype™ to FOL
Awesome.

So alot of boards are missing these components? to make clock more stable?
FOL is offline  
Old 02 May 2011, 18:41   #2067
Amigafanboi
Registered User
 
Join Date: Oct 2010
Location: Fareham Hampshire
Age: 47
Posts: 142
Fantastic show of techie there Jens!

You should just get on and design us a new and improved motherboard for our 1200's
Amigafanboi is offline  
Old 02 May 2011, 22:23   #2068
Schoenfeld
electricky.
 
Join Date: Jun 2010
Location: out in the wild
Posts: 1,243
Quote:
Originally Posted by FOL View Post
So alot of boards are missing these components? to make clock more stable?
I would not speak of "a lot of", as most of the customers do not experience graphics errors. However, since I had not been able to reproduce the bug here until today, there may be more "unfiltered" clocks out there than I'd expect from the range of boards that I have here, which all have the 14MHz clocks filteres with the ferrite/22pF combination.

I have just switched on the "cold" computer (that is: Board, power supply and accelerator are cold), and it works great starting off with an Alice temperature of 21 degrees Celsius, quickly working it's way up to over 46 degrees Celsius, where the computer starts to become unstable without the fix (measured with an Infrared thermometer). I consider this a successful test and rate the fix as "stable", so everyone out there who has graphics problems, please check your mainboards for E121R and E122R on the top side of the mainboard, as they're easily visible without too much hassle.

As always: Do not try to modify your board unless you do have the problems. If you apply the fix to a board with a different chipset combination, you might end up having a totally unstable computer - with or without accelerator.

Life would have been too good if this was the final problem (as I mentioned a few days ago). This weekend, someone on A1K.org pointed me to a problem that the FcCheck tool from WHDload forces: The computer freezes when you run this tool on an ACA1230. What it does is to make bus accesses with undefined/reserved function codes. Since they're undefined/reserved, my accelerator does not terminate these accesses. I find it highly questionable programming practise to use such function codes and expect the manufacturer of an accelerator to support them.

The closest thing to documentation I could think of is to give a "bus error" if such an undocumented/reserved access occurs. I have attached an image of the result. If someone who owns a "compatible" accelerator could post his result, that would be great (as it would really mark all reported problems of the ACA1230 as "solved").

Jens
Attached Thumbnails
Click image for larger version

Name:	FcCheck_fixed.jpg
Views:	332
Size:	30.3 KB
ID:	28580  
Schoenfeld is offline  
Old 03 May 2011, 23:48   #2069
Schoenfeld
electricky.
 
Join Date: Jun 2010
Location: out in the wild
Posts: 1,243
MMU tooltype fix for WHDload

Hi guys,

to take care of that final problem, which is the MMU tooltype of WHDload, I have eMailed a little back and forth with Bert Jahn, the author of WHDload. He explained to me how his hack works (he himself calls it a hack), and I explained to him how bad it can be to access the bus with an undefined/reserved function code.

We both agree that with a hack on my side, the MMU tooltype will work. The hack is a CPLD update that also allows function codes 0 and 4 to access all parts of the memory. Function codes 3 and 7 will result in a bus error (after a timeout), so the machine won't freeze.

This update is not free, as it's strictly against Motorola documentation, and it's not standard "Commodore or Amiga" operations. If you have an MMU, and you truly need that WHDload MMU tooltype, send the board to me and you'll get the update for 10,- EUR plus shipping.

After the update, the FcCheck tool gives the attached output with an ACA1230.

Jens
Attached Thumbnails
Click image for larger version

Name:	FcCheck_fixed2.jpg
Views:	236
Size:	25.6 KB
ID:	28585  
Schoenfeld is offline  
Old 04 May 2011, 03:49   #2070
Mr B
Registered User
 
Join Date: Nov 2005
Location: Sweden
Posts: 740
I'm perhaps a bit of on the target here, but wouldn't it be better to fix it on the other end, the WHDload software? (However, it's not something you can decide, i know that.)
This pretty much means everyone who has a 1230/56 should send the card back to you, as well as everyone who bought a refurbished 1230/28 with a MMU, unless they intend never to use the MMU for WHDloads. OK, so i don't know how much WHDLoads gain from a MMU if at all, or, if there are any WHDLoads thet require the MMU to work.
But somehow i would rather see this "fixed" in software. And it's not only due to me not wanting to spend the 10 euro, and shiping both ways. Again it's a matter of "what if the next owner needs this to work". It's not like Jens is going to want to provide the service of updating the cards for years and years to come.

B!
Mr B is offline  
Old 04 May 2011, 04:15   #2071
NovaCoder
Registered User
NovaCoder's Avatar
 
Join Date: Sep 2007
Location: Melbourne/Australia
Posts: 3,444
I was thinking the same thing, if it's a WHDLOAD issue, then it should be a fixed there if possible.
NovaCoder is offline  
Old 04 May 2011, 06:12   #2072
Claw22000
Registered User
Claw22000's Avatar
 
Join Date: Mar 2009
Location: St. Paul, MN/USA
Age: 39
Posts: 222
What about all us registered users don't we count. I actully paid for WHDLoad. Hoping future needs would be devloped for. Maybe Jens needs us to start a petition. Since hes already been so good to us. Lets take this one on ourselves!
Claw22000 is offline  
Old 04 May 2011, 08:58   #2073
Foxman
Registered User
 
Join Date: Dec 2009
Location: Germany
Posts: 122
Quote:
Originally Posted by Claw22000 View Post
What about all us registered users don't we count. I actully paid for WHDLoad. Hoping future needs would be devloped for. Maybe Jens needs us to start a petition. Since hes already been so good to us. Lets take this one on ourselves!
This is also my opinion.
I (and many others) have paid for WHDLoad. So why it is not possible to solve the problem the software side?
By the way, 10 + shipping is not to much i think, because its also work for Jens too.

One Question: What is the disadvantage I have when I do not use mmu tooltype?
Foxman is offline  
Old 04 May 2011, 09:00   #2074
NovaCoder
Registered User
NovaCoder's Avatar
 
Join Date: Sep 2007
Location: Melbourne/Australia
Posts: 3,444
Bert Jahn -> wepl#whdload.de

To avoid spam: please send your mail to wepl#whdload.de. Replace the # sign by an @ sign. Thank you.


Quote:
MMU/S
This must be used on 68030 machines to get the MMU related features working (memory protection, improved cache managment, Snooping, resload_Protect#? functions). On 68040/060 this option has no effect because the MMU will be used by default. It is recommended to set this option in the global configuration file on all systems containing a 68030 with working MMU (i.e. not a 68ec030) because it increases stability and security a lot. If the option NoMMU/S is also set this option has no effect.
source

Last edited by NovaCoder; 04 May 2011 at 09:09.
NovaCoder is offline  
Old 04 May 2011, 11:40   #2075
FOL
PSPUAE DEV

FOL's Avatar
 
Join Date: Nov 2006
Location: Barry / UK
Age: 39
Posts: 5,591
Send a message via MSN to FOL Send a message via Skype™ to FOL
Hmmmm, this seems to have gone way off topic.

I cant see why you would need an MMU to be honest, I have never used it in WHDLoad. I have never had any problems.
Besides it clearly says its disabled by default on 68030, so 90% of users would'nt even notice.

I dont think this is berts problem. Or it would have popped up before.
FOL is offline  
Old 04 May 2011, 12:48   #2076
Mr B
Registered User
 
Join Date: Nov 2005
Location: Sweden
Posts: 740
In the event that it is an issue, it should be solved on the "other end" and is none of Jens business. As a service Jens is able, and willing, to update the firmware of the card to handle the software errors that isn't really his problem. At that rate, is 10 bucks and shipping, a lot? No. But should it have to be done? Still no. Just saying.
But yes, i can see this is a bit of-topic.
B!
Mr B is offline  
Old 04 May 2011, 13:03   #2077
Schoenfeld
electricky.
 
Join Date: Jun 2010
Location: out in the wild
Posts: 1,243
While it's technically off-topic, it was "made" on-topic with different accelerators behaving differently. Note that this has no direct connection to an MMU being there/used or not. It just happens that the MMU code also uses an undocumented function code, which happens to work on one kind of accelerator, but not on the other.

I have made a suggestion to Bert Jahn how to solve it on his side, but since I am not a programmer, I don't exactly know if it's possible that easily. Bear with us a few days - the good news is that we currently have a solution for each and every issue that has been reported. The current work is just to improve the choice of solutions.

Essentially, we have a luxury problem here.

Jens
Schoenfeld is offline  
Old 04 May 2011, 13:57   #2078
Claw22000
Registered User
Claw22000's Avatar
 
Join Date: Mar 2009
Location: St. Paul, MN/USA
Age: 39
Posts: 222
What is the mmu tool type?

What problems might occure from removing it?

Thanks
Claw22000 is offline  
Old 04 May 2011, 14:44   #2079
Schoenfeld
electricky.
 
Join Date: Jun 2010
Location: out in the wild
Posts: 1,243
Quote:
Originally Posted by Claw22000 View Post
What is the mmu tool type?
You're asking the right question, because it's the one that you haven't missed yet. In other words: Nothing to worry about.

Quote:
Originally Posted by Claw22000 View Post
What problems might occure from removing it?
This question would have to be re-phrased to "do I need it if I didn't know about it until now?" The default is "off", as it's known to cause trouble with other accelerators as well.

Just give Bert a few days to think about my proposal for a solution on the software side.

Jens
Schoenfeld is offline  
Old 04 May 2011, 16:15   #2080
Claw22000
Registered User
Claw22000's Avatar
 
Join Date: Mar 2009
Location: St. Paul, MN/USA
Age: 39
Posts: 222
I would still like to know what it is as I'm always trying to expand my amiga knowledge. So if any one knows for education sake. What's an mmu?
Claw22000 is offline  
AdSense AdSense  
 


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
Santa's come early... A new scandoubler from Individual Computers NovaCoder News 708 18 October 2016 23:43
Details for the next generation of Flickerfixers from Individual Computers MarkONE News 2438 01 February 2016 13:04
A600 board versions and accelerator cards LePiaf support.Hardware 9 03 September 2012 13:49
Individual Computers Silver Sponsor of the Revision gibs Amiga scene 1 22 April 2011 17:43
New products by individual Computers Paul News 0 30 November 2004 16:58

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +2. The time now is 05:39.


Powered by vBulletin® Version 3.8.8 Beta 1
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
Page generated in 0.45588 seconds with 12 queries