Overview Features Coding ApolloOS Performance Forum Downloads Products Order Contact

Welcome to the Apollo Forum

This forum is for people interested in the APOLLO CPU.
Please read the forum usage manual.
Please visit our Apollo-Discord Server for support.



All TopicsNewsPerformanceGamesDemosApolloVampireAROSWorkbenchATARIReleases
Information about the Apollo CPU and FPU.

V600 Gold 2.7 Issuespage  1 2 3 

Leigh Russ

Posts 151
07 Mar 2018 17:30


Ok, so back to my original issues as I have been performing more testing today.

The SAFE core ran pretty much the same as the x11 core. Same issues with random locking up and freezing. So neither the x10 (safe) nor the x11 core run very well.

The following has already been performed:
Amiga has been recapped 6 months ago.
Upgraded PSU (Meanwell)
Sanded PLCC socket down to improve fit (this solved early stability issues with Gold2)
Caps modded on C3 and C6 (2x 0805 22uf 16v on each)

So next move was to troubleshoot the connection. This was always fine previously so have removed the V2 several times, cleaning the pins on the socket and CPU each time and re-attaching it. It did make some improvement. With the x11 core I have been able to play ADoom for 30mins with only one freeze/restart occur, and have also been able to run four cycles of all tests on SysSpeed without issue, where it froze on only one cycle. This was alot less than the freezing/crashing/restarting constantly.

So far so good, or so I thought. Loaded Eagleplayer and it crashed in seconds (rebooted), tried it again, this time it froze with a stuttering sound. Third time, same thing again, it froze. Eagleplayer was fine on Gold2.5 so thought I would try SysSpeed and ADoom again, and same again, random freezes and restarts.

Went back to re-seat the V2 again as it made an improvement last time, and this time no boot and green screen on the RGB connection. Check the pins and two of them looked a bit lower down than the rest so used a pin to push them back as best as I could. The green screen went and it again boots but is still not very stable.

The socket used on the Vampire has never been a very strong and tight fit, whereas the Furia card that I used to have felt like it snapped onto the CPU and took a lot of effort to remove. No other method of fixing the card down was used with the Furia and it was always stable.

For this reason I am starting to consider replacing the 68pin PLCC socket for the same one as the Furia uses if I can find one (tough it looks to be this one which is now obsolete EXTERNAL LINK ). So maybe have to order several different connectors to find one that the best fit.

Before doing this however, if anyone can suggest anything else to try. And the best utility or tool to perform a stability test then it would be much appreciated. I would rather not have to stick it on ebay and go back to a Furia if I can get it working stable. Annoying when Gold2.5 is/was fine.


Tony Day

Posts 62
08 Mar 2018 20:44


Im on 2.5 at the moment no issues. I guess I just have to try and see what happens. I dont like the idea of soldering the thing to be honest and never thought that would be something that I would have to do?
 
  Do you have a link to 2.7?
 
  Ive registered my card but i didn't receive any news regarding the firmware update.
 
  Thanks.
 
  P.s. I had the caps done not too long ago. I have had stability problems in the past such as it turning off and rebooting during a game or app but it was rare.


Leigh Russ

Posts 151
13 Mar 2018 16:38


UPDATE on my issues.
 
  So after getting fed up with it just not working, and repeatedly re-seating the card was hit or miss as sometimes it improved stability and other times it made it worse, I found two pins had become rather out of shape.
 
  Decided to change the socket to the same on as on the Furia card. Old socket desoldered really easily, new socket was fitted after sanding it down and cutting a notch for the resistor. Now the Vampire clips in place with a slight 'click' and I can lift the whole A600 by the Vampire alone without needed any additional screws or mounting brackets. The socket is the only thing holding it in place.
 
  On to testing... so far it has been running ADoom in demo mode for an hour and not one hang or crash. Whereas before it would hand and reset after a few seconds.
 
  Seems to be that the socket used is not ideal and the poor connection may be causing a lot of peoples issues.
 
  Old socket - EXTERNAL LINK 
  New socket - EXTERNAL LINK 
  Old socket removed - EXTERNAL LINK 
  Vampire with new socket lift test - EXTERNAL LINK


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
13 Mar 2018 16:43


Many thanks for sharing your test experience!


Leigh Russ

Posts 151
13 Mar 2018 19:06


So ADoom ran fine for about 2 hours before it froze so a great improvement.

Decided I should fix it down even though it is a snug fit so 3D printed some spacers the perfect height and some small thumb screw things to hold the tiny nuts and make it easier to get them on and off. So going to leave it testing again.

Does anyone have any recommendations for what to leave running to test it out well? As I don't think ADoom is ideal


Leigh Russ

Posts 151
13 Mar 2018 23:58


So results of testing so far are positive. It still crashes/freezes/resets but nowhere bear as frequently nor as soon as it did.

First test - Ran for 2 hours before freezing
Second test - Ran for just under 2 hours before freezing
Third test - Ran for almost 3.5 hours before restarting and getting stuck in a boot loop on the vampire screen.

Not sure how to resolve that one as it no longer seems to be the connection that's the issue



Knight Stone
(Needs Verification)
Posts 136/ 1
14 Mar 2018 03:20


TBH, the Socket on the V2, is the only thing that lets its down. I have an A600 with a Furia, and an A600 with a V2, i find on the A600 with the Furia, it's "SNAP", and it's on there, and it's not coming off, and stable, but on my A600 with the V2, i can't even screw the Amiga back together, as i am constantly having to open it up, and mess around with it, to get it to work.

I might fit a new socket myself.


Eric Gus

Posts 477
14 Mar 2018 06:59


Leigh Russ wrote:

So results of testing so far are positive. It still crashes/freezes/resets but nowhere bear as frequently nor as soon as it did.
 
  First test - Ran for 2 hours before freezing
  Second test - Ran for just under 2 hours before freezing
  Third test - Ran for almost 3.5 hours before restarting and getting stuck in a boot loop on the vampire screen.
 
  Not sure how to resolve that one as it no longer seems to be the connection that's the issue
 

Maybe its running a bit too hot? try cooling the FPGA?



Marlon Beijer

Posts 182
14 Mar 2018 11:11


Could possibly be a power draw issue? Maybe the PSU is too weak, or near the end of its lifetime or possibly needs a recap?


Gunnar von Boehn
(Apollo Team Member)
Posts 6207
14 Mar 2018 11:34


Leigh Russ wrote:

So results of testing so far are positive. It still crashes/freezes/resets but nowhere bear as frequently nor as soon as it did.

Is your Amiga open or closed?
I recall that Kippa once did temperature measurements in the A600.
His finding was that the A 600 mainboard chipset can get very hot.
Some A600 get hotter than others.
Those mainboard chips can heat of the whole AMIGA and also heat up the Vamp.

Kippa uses some very small fan which he placed behind the Vamp blowing down on the edge of the Vamp - to eliminate this heat up by the mainboard.


Leigh Russ

Posts 151
14 Mar 2018 11:51


Marlon Beijer wrote:

Could possibly be a power draw issue? Maybe the PSU is too weak, or near the end of its lifetime or possibly needs a recap?

Definately not a power issue as I have upgraded my PSU with a brand new MeanWell PSU in a self designed 3D printed case.

Gunnar von Boehn wrote:

  Is your Amiga open or closed?
  I recall that Kippa once did temperature measurements in the A600.
  His finding was that the A 600 mainboard chipset can get very hot.
  Some A600 get hotter than others.
  Those mainboard chips can heat of the whole AMIGA and also heat up the Vamp.
 
  Kippa uses some very small fan which he placed behind the Vamp blowing down on the edge of the Vamp - to eliminate this heat up by the mainboard.

It was running closed so currently got it running with the case off and a small heatsink temporarily on the FPGA and the top RAM chip as that seems to the one that gets hottest (chips under the Vamp seem to be the main cause of the heat).

If it runs better with the case off and heatsink on, then I'll get a better size heatsink and design a 3D printed fan mount to clip in the top of the case and vent it towards the edge of the Vamp so it also blows air underneath.




Leigh Russ

Posts 151
14 Mar 2018 15:10


Think you are onto something with the heat issue. With the top of the case off it has been running for over 5 hours, looping the Doom demo, and no issues at all. Still going strong.


Jari Eskelinen

Posts 23
14 Mar 2018 17:25


Leigh Russ wrote:

Marlon Beijer wrote:

  Could possibly be a power draw issue? Maybe the PSU is too weak, or near the end of its lifetime or possibly needs a recap?
 

 
  Definately not a power issue as I have upgraded my PSU with a brand new MeanWell PSU in a self designed 3D printed case.

I had problems with RT-50B, it was unstable while re-capped original 2.5A PSU is stable. I suspect MeanWell is putting out unstable power due to it being operated out-of-specs (min draw).



Leigh Russ

Posts 151
14 Mar 2018 17:33


I'm using the PT-45B and all seems fine.

Almost 7.5 hours now and running stable still with the case off.



Marlon Beijer

Posts 182
14 Mar 2018 20:32


Leigh Russ wrote:

I'm using the PT-45B and all seems fine.
 
  Almost 7.5 hours now and running stable still with the case off.
 

That's nice to hear. :)


Jari Eskelinen

Posts 23
15 Mar 2018 11:15


Leigh Russ wrote:

I'm using the PT-45B and all seems fine.
 
  Almost 7.5 hours now and running stable still with the case off.
 

PT-45B seems to have more suitable specs than RT-50B. PT-45B has 0A min draw for -12V and RT-50B has 0.1A (Amiga draws less). So I imagine PT-45B works better (I think +12V is still out of spec).

Which PSU do you have, Marlon Beijer?


Roy Gillotti

Posts 517
15 Mar 2018 14:29


Jari Eskelinen wrote:

 
Leigh Russ wrote:

  I'm using the PT-45B and all seems fine.
   
    Almost 7.5 hours now and running stable still with the case off.
   
 

 
  PT-45B seems to have more suitable specs than RT-50B. PT-45B has 0A min draw for -12V and RT-50B has 0.1A (Amiga draws less). So I imagine PT-45B works better (I think +12V is still out of spec).
 
  Which PSU do you have, Marlon Beijer?
 

 
  Interesting, I've been running RT-50Bs on my Vamped A600 and a  stock A500 for a while now, haven't noticed any issues yet. Generally the +/-12V rails are used for audio too.

Could I just add a small load to those rails to ensure the PSU is functioning properly?
 


Jari Eskelinen

Posts 23
15 Mar 2018 15:15


Roy Gillotti wrote:

  Interesting, I've been running RT-50Bs on my Vamped A600 and a  stock A500 for a while now, haven't noticed any issues yet. Generally the +/-12V rails are used for audio too.
 
  Could I just add a small load to those rails to ensure the PSU is functioning properly?
 

Could also be that my RT-50B defective, who knows? Or perhaps different units will behave differently when operating out of specs. If you have not noticed any problems then I would not do anything as you :)

Anyways adding dummy load you suggest should make it operating withing specifications though. Something like 50 ohm power resistors with heat sinks to both + and - 12V rails should do it, but that will generate 6 watts worth of extra heat.


Eric Gus

Posts 477
15 Mar 2018 21:45


Roy Gillotti wrote:

  Interesting, I've been running RT-50Bs on my Vamped A600 and a  stock A500 for a while now, haven't noticed any issues yet. Generally the +/-12V rails are used for audio too.
 

Same here for both my Vampirized A500 and my more-or-less stock A1200 and no problems either.


Leigh Russ

Posts 151
17 Mar 2018 19:30


For anyone wanting a neat solution to help cool their V600 then I have designed a fan mount/vent to draw in cool air and blow this over and under the Vampire.

https://www.thingiverse.com/thing:2829586

posts 48page  1 2 3