Overview Features Instructions Performance Forum Downloads Products Reseller Contact

Welcome to the Apollo Forum

This forum is for people interested in the APOLLO CPU.
Please read the forum usage manual.
VISIT APOLLO IRC CHANNEL



All TopicsNewsPerformanceGamesDemosApolloVampireCoffinReleases
Information about the Apollo CPU and FPU.

SAGA Flickering With Black Screenpage  1 2 3 4 5 

Carlos Milán

Posts 44
25 Mar 2017 03:32


I finally enabled SAGA with P96 in my Vampire enhanced Amiga 600. It looks really nice, but I'm experience the following: while I'm using Workbench, the screen randomly* get black for 1-2 seconds and then it is restored, without anything crashing.
 
  I have tried the following:
    * Different screen modes.
    * Bootup with ECS and see the HDMI output with the Vampire logo. It doesn't happen there.
 
  Has anyone experienced this?
 
  * It is not actually random, it is related to have activity in my Amiga. If I load Workbench and do not touch the computer, the black screen flickering doesn't happen. I suspect it has something to do with IDE disk (actually CF card) activity.

I'm using GOLD2 with P96 and SAGA driver 0.11.


Thomas Blatt

Posts 91
25 Mar 2017 07:10


Same here, i have figured out that it seems to be an ide problem.
  First time i had a 30cm round cable installed. I got many freezes, black screen and had to load a few times my Amiga to be in Workbench. Then i tried a 10cm ide cable. My Amiga runs much more stable but not perfect.Then i tried a 15cm ide cable, now my Amiga is less stable than with 10cm ide cable.
  I ordered me two solutions to get my Amiga 100% stable.
  EXTERNAL LINK  and this
  EXTERNAL LINK  I know i need 44 cables, i have already some here.
  Maybe this will solve it, watch this space


Carlos Milán

Posts 44
25 Mar 2017 10:07


Yeah, it looks like it is related to IDE activity. No IDE activity, no flickering. I'm using Kipper2k adapter (that is the same as the one sold in ebay that you pointed), so there is not actually an IDE cable.

As not everyone is experiencing this, I wonder what is the difference between our configurarion and theirs?

My complete hardware config:
* Motherboard rev1.5
* A604n + RTC
* Indivision ECS
* RapidRoad USB
* Sum USB keyboard adapter
* Kipper2k IDE CF adapter

Everything working fine and no flickering with ECS.


Carlos Milán

Posts 44
25 Mar 2017 13:39


After some testing, I don't have it clear that is the IDE, but for sure is related to system activity. I have uploaded a video for showing what is happening:

EXTERNAL LINK 
Ideas or clues are welcome :)


Myzar Alcor

Posts 20
25 Mar 2017 19:20


It's the ide but in a weird way , it's actually power related, it happens when the ide draws alot of power from the +5v. You can try with a shorter hdmi cable, another psu, or start disconnecting stuff , you have alot in the a600


Carlos Milán

Posts 44
26 Mar 2017 03:17


Thanks Myzar, that makes sense, but I currently have the 600 with a 60W PSU (http://www.amibay.com/showthread.php?79035-Amiga-Power-Supply-upgraded-60W). It is power related, it shouldn't be the PSU... could it be how the power is handled in the 600 motherboard? I'm curious about what has that to do with the Vampire RTG.

So far:
* HDMI cable: 0,5m long (way short)
* PSU: 60W, 8.0A in the +5v

I have also a 600W ATX PSU ready to use with the Amiga. I'll try that one in order to see if there are any improvement. I know I have quite a lot of stuff, it is truly a 600 at maximum upgrade capacity, I forgot to mention in the previous post that I also have a PCMCIA Ethernet connected, I'll also try removing that one.


Thomas Blatt

Posts 91
26 Mar 2017 06:49


This is my System:
  Amiga 600 Rev.1.3
  Vampire2 with Gold 2
  A604n+Indivision ECS+ RTC Modul
  Prisma MegaMix
  2x RYS MK2
  Gotek
  PCMCIA WLan
  5V=3A PSU from Commodore
  @Carlos i got only sometimes a black screen, not so often like you
 


Roy Gillotti

Posts 169
26 Mar 2017 18:18


I was using an 80Watt PicoPSU and was getting blanking just as described, also getting some other instabilities. I switched to Mean Well T-60B Power supply it's only 61 Watt, but my guess is an industrial PSU has a cleaner power source and can handle variations in current draw better than an ATX PSU.


Dezső Bodor

Posts 8
26 Mar 2017 21:26



Select a 60Hz screen frequencies.
I think actually the only monitor in 60Hz can work, however, tolerate 50Hz was a few seconds. However, over time, it will be a big time difference and then resets itself.


Gunnar von Boehn
(Apollo Team Member)
Posts 3350
26 Mar 2017 21:32


Carlos Milán wrote:

After some testing, I don't have it clear that is the IDE, but for sure is related to system activity. I have uploaded a video for showing what is happening:
 
  EXTERNAL LINK 
  Ideas or clues are welcome :)

Can you please give some more information:
What screenmode do you use?
What resolution?
What depth?
What is the pixel clock?
What is the refresh rate?


Carlos Milán

Posts 44
26 Mar 2017 22:16


Gunnar von Boehn wrote:

  Can you please give some more information:
  What screenmode do you use?
  What resolution?
  What depth?
  What is the pixel clock?
  What is the refresh rate?

Screenmode: SAGA
Resolution: 1280x800 16:10 50Hz 16 bit
Depth: 16 bit
Pixel clock: 56.28 MHz
Refresh Rate: 47 Hz

I have not modified any settings there, so I'm using the defaults.

I have uploaded a screenshot with Picasso96Mode and ScreenMode open with my current resolution: EXTERNAL LINK 
I also tried other resolutions (all of them were 1280x) and color depths without any improvement. I'll give a try to the 60Hz ones.

I have been testing today without PCMCIA card, but I see no improvement. Despite the black flickering, the system is really stable with the Vampire and I have been using it with the Indivision ECS for a month without problems.

Roy: did the black flickering improved after changing the PSU?


Roy Gillotti

Posts 169
26 Mar 2017 22:35


@Carlos, oh looks like I was a little vague in my post, yes it seems that Power Supply fixed my issues.


Carlos Milán

Posts 44
26 Mar 2017 23:34


Dezsõ Bodor wrote:

  Select a 60Hz screen frequencies.
  I think actually the only monitor in 60Hz can work, however, tolerate 50Hz was a few seconds. However, over time, it will be a big time difference and then resets itself.
 

  You point in the good path!
 
  I have been testing with a lower mode for an hour and I have NOT experienced a single black flicker!
 
  Here is a screen capture of the mode it's working flawlessly: EXTERNAL LINK 
  Screenmode: SAGA
  Resolution: 960x540 60Hz 16 bit
  Depth: 16 bit
  Pixel clock: 30.74 MHz
  Refresh Rate: 49 Hz (I'm confused here, ScreenMode prefs says 60 Hz)
 
  Tomorrow I'll test other modes, but it looks like just the 1280x are the ones who are causing the black flickering; at least with my monitor.


Gunnar von Boehn
(Apollo Team Member)
Posts 3350
27 Mar 2017 08:11


Carlos Milán wrote:

  I have been testing with a lower mode for an hour and I have NOT experienced a single black flicker!
 
  Here is a screen capture of the mode it's working flawlessly: EXTERNAL LINK   
  Screenmode: SAGA
  Resolution: 960x540 60Hz 16 bit
  Depth: 16 bit
  Pixel clock: 30.74 MHz
  Refresh Rate: 49 Hz (I'm confused here, ScreenMode prefs says 60 Hz)

OK, Excellent.
I would recommend you to keep this monitor setting.
In general we recommend you to use Screenmodes with Pixelclock up to 50 MHz maximum.

Screenmodes above 50 MHz Pixelcock can work but will
stress the video out of the Vampire to the maximum.
And there is also the factor of power supply.
You have to mind that you are using an 20year old AMIGA which was not designed to support such clocks rates.



Carlos Milán

Posts 44
27 Mar 2017 21:44


I see, it makes sense. Anyway, if someone told me 4 years ago that my little Amiga 600 would have more than 060 performance, RTG graphics and probably AGA I would laugh really hard. Quite a lot for a 20 years old machine :)

I'm really happy with that resolution and color depth. Thank you so much guys for the incredible work you are doing with the 500 and 600 machines.


Ma Sch

Posts 2
04 Apr 2017 16:38


Gunnar von Boehn wrote:

  Screenmodes above 50 MHz Pixelcock can work but will
  stress the video out of the Vampire to the maximum.
  And there is also the factor of power supply.
  You have to mind that you are using an 20year old AMIGA which was not designed to support such clocks rates.
 

 
  This is not true.
  The problem is not really the power supply or the 20 year old amiga.
  The design of the power supply at vampire board itself and the design of the HDMI output limits the pixelclock and maximal display resolution.
 
  1. The simple HDMI output in the FPGA and the bad termination of the TMDS signal and the tolerance at sink results in timing problems at higher resoltions.
  Shure, a real TMDS transmitter could be solve most of the problems. ;-)
  But, i use a HDMI reclocking chip for driving external HDMI cable.
 
  2. The series resistors for wave termination at HDMI output is to high. 270ohms has a Vswing (TMDS 3,3V - signal level) at 0,3-0,4V and is too low. The errata of HDMI signal set the Vswing to minimum of 0,7V and maximum of 1,4V!
  Reduced values to 27 or 30 Ohms is very better....
 
  3. the used voltage regulators has too small output capacitors of 10uF. The datasheet set the capacitors to a nominal value of 2 pcs. 22uF (because lower ESR)!
  This is the problem, when the CPU do more at the core or at bus. The current will be increased at shortest time, the regulator get power from main PCB. Spikes results in trouble of HDMI signals.
 
  The sum of all points above will limit the maximal resolution of HDMI output.
  At my very sensitive Eizo the Vampire had only stable signal at 800x600 in the past.
  Now, some modifications later, i can use resolutions like 1600x900/50Hz (72,83 MHz) without problems. (Tested at 2 Vampire boards with same results)
 
  Regards MaSch
 


Igor Majstorovic
(Apollo Team Member)
Posts 236
04 Apr 2017 18:41


Voltage regulators tool for calculating values of passive components was proposing different values than it was stated inside manual. In the same time when we were using AOZ devices others did too and on e-bay and on many other places you can find upgrade kits for equipment who used those voltage regulators. All of them had the same situation as we, various bitcoin mining equipment used those AOZ devices so you can find posts about that on bitcointalk also...
  Anyway those finds are true but something we figured out long time ago and fixed it.
  For fixing Pk-Pk values let's say for the system who sinks more and more current with this design we are limited with Cyclone III and/or SDRAM. Each component have its own limits and it seems to me Apollo-core don't. In reality transferred to MIPS card will be stable at around 120MIPS and in some situations it won't be stable at 130MIPS. If we want more we have not just Pk-Pk problems, Cyclone III can't handle more, then SDRAM then you get to the point when you need to add active cooling and so on. Why, to get 10 more MIPS. Does that really matter. Is our goal again to have fastest Amiga in the street where we live and only thing that matters is what SysInfo shows. In reality slowest Vampire, most defective is still far better than anything done for so long time. After all having retro computers and not having soldering station near your desk is strange thing for me :)
 
  Regarding HDMI, me personally never had any kind of problems because I use some LG TV who can take all and turn it to perfect picture. Maybe just maybe you should try to use 0603 100nF and report back. 


Niclas A
(Apollo Team Member)
Posts 165
04 Apr 2017 18:46


Ma Sch wrote:

Gunnar von Boehn wrote:

    Screenmodes above 50 MHz Pixelcock can work but will
    stress the video out of the Vampire to the maximum.
    And there is also the factor of power supply.
    You have to mind that you are using an 20year old AMIGA which was not designed to support such clocks rates.
 

 
  This is not true.
  The problem is not really the power supply or the 20 year old amiga.
  The design of the power supply at vampire board itself and the design of the HDMI output limits the pixelclock and maximal display resolution.
 
  1. The simple HDMI output in the FPGA and the bad termination of the TMDS signal and the tolerance at sink results in timing problems at higher resoltions.
  Shure, a real TMDS transmitter could be solve most of the problems. ;-)
  But, i use a HDMI reclocking chip for driving external HDMI cable.
 
  2. The series resistors for wave termination at HDMI output is to high. 270ohms has a Vswing (TMDS 3,3V - signal level) at 0,3-0,4V and is too low. The errata of HDMI signal set the Vswing to minimum of 0,7V and maximum of 1,4V!
  Reduced values to 27 or 30 Ohms is very better....
 
  3. the used voltage regulators has too small output capacitors of 10uF. The datasheet set the capacitors to a nominal value of 2 pcs. 22uF (because lower ESR)!
  This is the problem, when the CPU do more at the core or at bus. The current will be increased at shortest time, the regulator get power from main PCB. Spikes results in trouble of HDMI signals.
 
  The sum of all points above will limit the maximal resolution of HDMI output.
  At my very sensitive Eizo the Vampire had only stable signal at 800x600 in the past.
  Now, some modifications later, i can use resolutions like 1600x900/50Hz (72,83 MHz) without problems. (Tested at 2 Vampire boards with same results)
 
  Regards MaSch
 

All good points im sure. Just wanted to say i run my V500 in 1280x720 24Bit 50Hz and have not had any problems on my Philip 32" tv.


Ma Sch

Posts 2
04 Apr 2017 20:44


Niclas A wrote:

  All good points im sure. Just wanted to say i run my V500 in 1280x720 24Bit 50Hz and have not had any problems on my Philip 32" tv.

sorry, i am talking about problems of my Vampire 600 boards only.



Jari Eskelinen

Posts 19
08 Apr 2017 18:39


Ma Sch wrote:

    3. the used voltage regulators has too small output capacitors of 10uF. The datasheet set the capacitors to a nominal value of 2 pcs. 22uF (because lower ESR)!
    This is the problem, when the CPU do more at the core or at bus. The current will be increased at shortest time, the regulator get power from main PCB. Spikes results in trouble of HDMI signals.
 

 
  This was good info. I had some problems exactly like described in this thread - IDE activity causing screen flickering. Also PCMCIA activity did this. Even worse, some occasional system crashes, seemed to relate to flickering. This happened only after system had heated up, I initially suspected cold joints but now it sounds more plausible that heat causes voltage regulator capacitors to drop their capacitance even lower from 10uF which makes it things worse, no?
 
  Today I soldered 10uF 0805 caps (did not have any 22uF) on top of the existing ones. Ever since I've had no screen flickering nor any crashes at 1280x800@50Hz. So really, this nails it so I thank you A LOT!
 
  Ps. So I currently have 2x10uF caps installed. What would be the recommendation? 2x22uF like datasheet says is typical configuration or even 2x47uF?

posts 93page  1 2 3 4 5