28 July 2017, 14:31 | #241 |
Registered User
Join Date: Dec 2010
Location: Norway
Posts: 827
|
No sure what the intention was, but I think pcmcia scsi kits, cd-roms and harddrives (overdrive etc) were a lot more common than pcmcia ram cards
|
24 August 2017, 16:02 | #242 |
Registered User
Join Date: Apr 2014
Location: Mallorca es Funki
Posts: 44
|
I finally got my Furia (Hard to push the socket into the 68000, at least for me). I can confirm that this rom patch is working flawlessly
Best regards |
23 October 2017, 19:43 | #243 |
Develop HW
Join Date: Jul 2011
Location: Nitra / Slovakia
Posts: 104
|
Hello,
Completed the update for Furia EC020 R1/28MHz (Old version), Furia EC020 R2/33,3-37,5MHz Features v12.0 - Change Overlay, more stable MAPROM, (add support AROS rom, DiagROM rom) - More support Shapeshifter Emulator MAC. - Change DRAM controller (better support 2k OKI DRAM very stable work) - Change IDE speed, very fast access storage up to 5MB/s, (works on all boards A600, does not require Gayle R1) need only scsi.device v46.0 Hitashi microdrive IDE test R1 28MHz - 4,6 MB/s R2 33,3MHz - 4,7MB/s R2 37,5MHz - 4,9MB/s The update will be made by Lotharek at the cost of postage. Boris |
06 November 2017, 11:17 | #244 | |
Registered User
Join Date: Jan 2017
Location: Glesga
Posts: 265
|
Quote:
|
|
06 November 2017, 17:13 | #245 |
Registered User
Join Date: May 2001
Location: ?
Posts: 19,646
|
|
15 November 2017, 16:55 | #246 |
Develop HW
Join Date: Jul 2011
Location: Nitra / Slovakia
Posts: 104
|
- Lotharek is currently testing a new FW. In the new batch will be Firmware v12. (May be in December or January)
- JTAG Xilinx Download |
15 November 2017, 17:30 | #247 | |
Registered User
Join Date: May 2001
Location: ?
Posts: 19,646
|
Quote:
xc3sprog: https://github.com/matrix-io/xc3sprog We use this on Raspberry Pi to flash the TerribleFire accelerators. https://github.com/terriblefire/tf32...a-Raspberry-Pi Raspberry Pi at $5 is cheaper than any Xilinx programmer cable out there and many people already have one Would you then be willing to provide the necessary jed files to flash the card with an updated firmware to existing users that are confident enough to do it on their own? |
|
15 November 2017, 17:58 | #248 |
Develop HW
Join Date: Jul 2011
Location: Nitra / Slovakia
Posts: 104
|
If the user proves to be proficient, so yes (outside EU). Learn more by email or private message.
|
15 November 2017, 18:28 | #249 |
Registered User
Join Date: May 2001
Location: ?
Posts: 19,646
|
Great news Boboo!
I'm sure it will help out all those not in EU with a Fúria. |
15 November 2017, 19:50 | #250 |
Develop HW
Join Date: Jul 2011
Location: Nitra / Slovakia
Posts: 104
|
Yes, it is my intention, help to older owners of Furia.
|
26 September 2018, 00:19 | #251 |
Used Register
Join Date: Jul 2018
Location: Liverpool
Age: 41
Posts: 437
|
TLDR: I would like to buy a blank furia PCB to transplant my working components from a broken board. Please read though: I've gone through a lot with this board, and finally had it working perfectly, and just want the opportunity to use it again. Thanks.
For reference the original thread is here, though some of the original assumptions are inaccurate. So... The Original Problem!.. As you can read in the original thread the problem I had was that the system with the furia installed would crash at random intervals - sometimes 10 minutes, sometimes 10 seconds. Once or twice as much as 20 minutes or so. Long story short i went through every bit of troubleshooting and finally using MBR2 reasoned that the RAM was likely at fault. I got consistent failures corresponding to lines 8, 9, 10, 11, 20, 21, 22, 23, 28, 29, 30, 31. I go into more detail on this in the other thready. Anyway... The solution!.. So it turns out that those lines correspond to U7, U9, and U13. I ordered some NOS chips of the same designation and replaced the above chips one by one. As i did so, the errors associated with those lines disappeared from MBR2. Finally, after replacing U9 (I also replaced U11 because i was unsure in which order the MBR2 lines corresponded to the physical lines) - it worked perfectly. I mean it was absolutely rock solid. Night and day. It was beautiful. Here is a screenshot of it having run MBR2 for i think 8 hours? I left it over night. The new problem!.. So the next day, after double checking by running (successfully) the Frontier intro for three and a half hours, I decided to clean things up, remove any excess flux etc etc. I had already replaced the 68ec020 with a genuine MC from 2000 to rule out the CPU. I noticed the pins were not perfectly aligned (though certainly usable) and the perfectionist in me decided to heat up the chip and gently nudge them to dead centre. This was my mistake. I slipped, and the air gun i was using hit the upper right pins. The solder wasn't molten yet, so the force of the slip literally ripped the pads from the first three pins. Disaster. Utter disaster. After weeks of troubleshooting and *finally* getting this thing rock solid stable, a stupid mistake that didn't even need to happen practically ruined the board! I attempted some trace repair, but at the pin pitch of the flat pack 020 it just isn't doable. I decided to give up. The plea... I know my warranty is void. I knew it was when I made the repairs, but this Furia did come with some defective RAM chips, and replacing them *did* fix this great accelerator board. If I could purchase from you a *blank* furia PCB, so that i can transplant my working parts, I would be *eternally* grateful. I can pay however you like, and of course cover shipping. Obviously the CPLD is already programmed so I wouldn't need the jed's or anything. I am competent enough to do the transfer of components. I have already built a TF530 and am working on a second. I would be asking for no help and of course would expect no warranty. Thanks for reading my rambling post, and for designing a great accelerator for the A600! John |
26 September 2018, 13:00 | #252 |
Registered User
Join Date: Apr 2018
Location: Wakefield
Posts: 203
|
I can only get through one loop of the frontier intro with my Furia card installed before it crashes. After that it gets shorter and sorter before crashing to the point it won't even boot until I've switched off the power and left it alone for about an hour.
I thinking cooling is my issue. I've bought some thermal compound to replace the sticky pad for the CPU and I'm going to experiment with a desk fan and the lid off the Amiga to see If that improve things to probably fit a fan. If not I might be sending Project 23 it as I'd love for this thing to be usable as what a difference it makes to the speed of the A600. |
26 September 2018, 13:12 | #253 |
Registered User
Join Date: Aug 2018
Location: Italy
Posts: 52
|
I have the white version, and don't seem to generate too much heat.
Maybe is a good thing to measure the temperature directly on the heatsink. If someone has some instability I also suggest to monitor the power rail. |
26 September 2018, 14:17 | #254 |
Unregistered User
Join Date: Sep 2012
Location: Copenhagen / DK
Age: 44
Posts: 4,190
|
Remember that a low temperature on the heatsink is not necessarily a good thing as it could just mean that the thermal resistance between the CPU and the heatsink is high resulting in a cool heatsink but very hot CPU underneath. So if you replace the thermal pad in order to improve the cooling, you want to see increased heatsink temperatures afterwards.
|
26 September 2018, 15:53 | #255 |
Registered User
Join Date: Aug 2018
Location: Italy
Posts: 52
|
True, I'm just suggesting to keep on eye on the temperature, to check if raise up and how much when the system start to become unstable.
I'm not expect too much heat from a CPU like that. I'm also do not expect too much difference between a thermal compound and a thermal pad here. |
26 September 2018, 20:49 | #256 |
Used Register
Join Date: Jul 2018
Location: Liverpool
Age: 41
Posts: 437
|
Guys, you need to do a more systematic check of the board. If you have diagROM that will highlight any persistent memory issues. If the issues are not immediate (sometimes 5 minutes, sometimes 10 etc etc) you need to run MBR2-Test on the system, full test minus 'float' (just slows things down) continuous with max errors set to 8 and pause on error.
When memory errors are found, you'll get a hex representation of the data inconsistency, as well as the physical data lines that this error corresponds to. You have to understand that these are *old* chips. They are not new. Most likely they are pulled from simms from 20 years ago or more. I used this method to track down the chips that had intermittent problems. Replacing them left the system absolutely rock solid. I ran the Frontier intro for 3 and a half hours without heatink and with a simple desk fan forcing air over the bare 020 and the CPLD. I also ran MBR2-Test overnight for almost 9 hours without a single error. This is a damn good board built with some questionable parts, its as simple as that. The 'freescale' 020 that comes with it is fake. The CPU mask is fake and the date code is impossible. I have even researched whether any company was manufacturing the QFP 020 during the time of the date code and there were none. This does not mean that the CPUs are bad - just that they are counterfeits. They may work just as well as the originals. It is hard to say. Some users have found they do not overclock as well as the original motorolas. The problem with my Furia was not CPU based. I swapped out for a genuine MC from 2000 and the symptoms were the same. Anyway, unfortunately after fixing the ram issue i then made the mistake of leaving the house whilst running a CPU heavy task, and my desk fan proved insufficient at cooling the chip. A mere 24 hours after finally fixing my furia, I had fried the CPU. I was devistated. To make matters worse i unfortunately then slipped during CPU removal and ripped three of the 020 pads from the board. I actually plan on putting the fake freescale CPU back onto the board if i can either fix the board (unlikely) or purchase a blank as requested above. Counterfeit doesn't necessarily mean bad. What is much worse is the reliability of 18 year old memory chips, especially when they are pulled from simms that could have been lying around anywhere before being purchased. It could be that these chips are all new old stock - that is entirely possible and i'm in no way trying to suggest that any of these unreliability problems are a result of such corner cutting. But it doesn't help that there are plenty of Furia pics out there with either different form factor chips, or different manufacturers altogether. They're recycled chips - which is fine - but ultimately unreliable in mass production. I replaced my three bad chips with older genuinely new old stock chips from 98, and the board worked perfectly. Look, its a great board, it really is - but between unreliable ram chips and fake CPU's it seems difficult to get hold of one that'll run for more than twenty minutes. I'd love to get hold of a new blank PCB, or fix the one i have, because when i finally got it working it really was incredible. But so far i've received no response, and as a hobbyist i do what i can, but repairing those pads its kinda difficult. No hard feelings to anyone here, i really mean that and i hope its clear. It can't be easy catering to such a small niche demographic with a commercial product. John Last edited by project23; 26 September 2018 at 21:13. Reason: for the love y'all! |
26 September 2018, 23:05 | #257 | |
Registered User
Join Date: Aug 2018
Location: Italy
Posts: 52
|
Quote:
I was reading it on the other thread. A counterfeit 020, very strange indeed! I agree about the clock. 25 MHz are more than enough for a card like this. There is no need to overclock the CPU. |
|
27 September 2018, 08:47 | #258 |
Used Register
Join Date: Jul 2018
Location: Liverpool
Age: 41
Posts: 437
|
Counterfeit chips is a huge problem, getting worse by the year. Most of the time the chips in question are older or lower spec chips that have been ground down across the top and re-printed with a higher spec. Most of the high end 060 chips on ebay are like this i believe.
I'm not sure about actual fabrication and how often that goes on, i'd have to look into that, but i'm sure it goes on within china for chips that make sense to counterfeit financially. Copyright enforcement in china are lax to say the least. |
14 October 2018, 23:06 | #259 |
Registered User
Join Date: Sep 2018
Location: Doncaster/England
Age: 52
Posts: 5
|
Hi, I am/was looking at getting one of these cards. After reading this and the other thread, I am now left wondering if it is a viable option or whether to look for another option?
Have the issues been resolved by the board maker as they are currently still for sale on ebay. Last edited by Neez; 15 October 2018 at 00:14. |
15 October 2018, 09:39 | #260 | |
Registered User
Join Date: Apr 2018
Location: Wakefield
Posts: 203
|
Quote:
Though I regret it and wish Id have invested in my A1200 instead :-( |
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
2x New Low Cost Turbo Card for A600 "Fúria" | venox | Amiga scene | 80 | 14 November 2012 11:38 |
TotalCalc and EC020 FPU bug? | Leandro Jardim | support.WinUAE | 5 | 25 February 2012 20:25 |
The ST-01 Project | HideawayStudio | Amiga scene | 16 | 29 July 2008 19:54 |
Project One | rygar | request.Old Rare Games | 2 | 27 November 2007 01:20 |
|
|