19 November 2009, 22:20 | #1 |
Posts: n/a
|
Apollo 1230-16 Outperforming Blizzard 1230-50??
Yes I know the title of this thread is intriguing
I have an Apollo 1230 card. First of all... how the hell can anyone overclock a 16 mhz cpu to 40 mhz and call them self a decent man (150%), even more sell it as a proper product?? Anyway, this seems to be the rule of thumb for Apollo hardware as far as I can read from various forums. Although I must give them creds for making a crappy looking, ferociously overclocked turbo card that actually manages to be extremely fast AND stable (although I'm really feeling for that poor 68030 every time I flip the switch)... Where comes the Blizzard 1230 into play here? Well, my Apollo kitten is performing 7885 dhrystones and 8.22 Mips in SysInfo 3.24. Yes! I know SysInfo sucks but it gives a hint. So I ran AIBB 6.5 and at 40mhz it's being on par with my GVP 1230+ II at 50mhz, actually slightly faster in some tests. Which... is actually comparable or perhaps slightly faster than some Blizzard numbers I've seen out there. Mind you this is on WB 3.1 without any fancy utils or libraries loaded afaik. I don't even have the install disk for this card (if there are any?). So now I wonder what would happen if I pop a real 68030-50mhz and a 50mhz crystal on this card? It would certainly not be slower! There is a problem. As you can see on the picture, the CPU is soldered to the card. How ever the good news is that there are soldering pads for a PGA socket. My question is this: Can I simply de-solder the cpu, solder in a PGA socket and put in a new cpu in it?? And of course change the xtal to 50 mhz. Oh, by the way, I'm not even sure if this is really an Apollo board. From comparing pictures on the net I think it could as well be an Elbox 1230-II. Thanks for reading. Please comment... Last edited by DonFuego; 19 November 2009 at 22:28. |
19 November 2009, 23:03 | #2 |
Global Moderator
Join Date: Aug 2008
Location: Sidcup, England
Posts: 10,300
|
Hi DonFuego,
Are you sure that the PGA / PLCC sockets are for a CPU and not FPU? If they are indeed for a CPU, then I am sure that you could drop a socket and 50MHz CPU in there and upgrade to a 50MHz crystal. After all, 50MHz from 40MHz is nothing too drastic. If you are not worried about damaging the 16MHz CPU, you could always just cut through the legs. However, it would be a shame to ruin a CPU of such fine "pedigree". Regarding the second SIMM socket, you may find that it will no longer be possible to install or remove the card through the A1200 trapdoor without dismantling the case. If, of course, you have your A1200 towered, then this will not be a problem. However, this is not guaranteed to be successful. I have heard of some expansion cards which will not accept a second SIMM even though there seems to be provision for it on the pcb. |
19 November 2009, 23:11 | #3 |
Cat lover
Join Date: May 2004
Location: Italy
Age: 54
Posts: 804
|
SysInfo 3.24 in my A1200 reports (with 16MB RAM - 60ns - installed):
Blizzard 1230/IV: 9486 Dhrystones, 9.90 Mips, 7.79 Vs. 1,00 basic A1200 (with Kickstart in RAM) Blizzard 1230/IV: 9228 Dhrystones, 9.63 Mips, 7.58 Vs. 1,00 basic A1200 |
19 November 2009, 23:31 | #4 |
Ya' like it Retr0?
Join Date: Jul 2005
Location: United Kingdom
Age: 49
Posts: 9,768
|
@DonFuego
Your card is an Apollo 1230 Mk3, [the pcb is brown so is likely a licensed re-run] its running @40MHz which is a nice overclock for the 030 rated 16MHz =) (image from big book of amiga hardware) Leaking Battery ? yeah... replace it now, before we read another horror story.. Can I just Solder a new SIMM socket ? Yes... each port will take upto 32MB of ram. if they have programmed the CPLD properly you will have 64MB of Fast RAM Ring of Fire(c) Apollo ? Well a lot of these TQFP 030's can run at higher frequences, the 030 pictured here is a latent model and even at a rated 16mhz should run at 33/40mhz without issue (with heat sink of course) PGA Socket goes Here ? Indeed, and in here you can install a PGA 030 50mhz component PLCC Socket goes here ? Yes, this is for the FPU, a nice 68882@ 25mhz should clock to 50 =) @fc.studio its worth mentioning that the Blizzard Mk4 tested is infact running at 50MHz I was truly shocked to read some of the results from AIBB showing the Blizzard 030Mk4 actually slower than the Apollo 1230@50Mhz in the following crucial tests - reading chip to fast writing fast to chip reading fast to fast writing fast to fast I cannot remmeber whom pointed this out to me as I had always thought the Blizzy Mk4 had faster fast ram access... for a good 40 minutes i read in dis-beliefe that infact the Blizzy 030Mk4 (Mhz for Mhz) is nearly 10% slower than the Apollo 030 in the above tests.... |
19 November 2009, 23:35 | #5 |
Global Moderator
Join Date: Aug 2008
Location: Sidcup, England
Posts: 10,300
|
@Zetr0
|
20 November 2009, 00:16 | #6 |
Cat lover
Join Date: May 2004
Location: Italy
Age: 54
Posts: 804
|
About SysInfo 3.24, is possible to know how many drystones reports for the Apollo 1230/50?
|
20 November 2009, 00:31 | #7 | |||||||
Posts: n/a
|
Thanks to everybody for replying so quickly, especially Zetr0 for such detailed and insightful post!
Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
Quote:
|
|||||||
20 November 2009, 04:35 | #9 |
I hate potatos and shirts
|
Or scavenge a heatsink from an old PC motherboard. Catch the low profile ones used over the northbridge, then use it with a thermal pad.
|
20 November 2009, 09:10 | #10 |
Ya' like it Retr0?
Join Date: Jul 2005
Location: United Kingdom
Age: 49
Posts: 9,768
|
@DonFuego -
MMU gone lala...? its likely you need to use maprom command to initiate the MMU, otherwise it just wont be visible as registry flag for its use is not set =) Insofar as adding an extra PS2 72pin SIMM, IF they stuck to the original sources for the CPLD chips (most more than likely) then you should be fine with adding the extra SIMM socket, obviously no garantees, however it should work =) btw, amigakit actually sell the angled 72pin SIM socket. The 122pin PGA and TQFP are directly connected pin for pin, you would only need to disable the original TQFP cpu, you can remove it of course, however you can just lift (DONT CUT) the +5v pins and it should not be a problem. I will say that you CANNOT use the RST or HALT signals on this CPU to disable it in favour of the PGA socket CPU. as the signal lines are tied it will just HALT/RST the PGA CPU. A 122pin PGA socket cn be sourced for around £5 including postage, you may get a cheaper deal on fleabay. As far as the jumpers / resistors are concerned as both CPU socket signals are tied its transparent in this instance. Cooling is always imporant, In the case of this I would look at a handfull of possibilites - as my good friend Rkauer states, a low-profile heatsink with thermal grip pad would provide suffcient cooling, Also have a look at a laptop cpu cooling solution ( although you may need a little extra hackery). But by far the most awesome I have seen in a while was a fellow forum members method of Heat-Piping and using the back of the keyboard as a giant heatink, using very inexpensive components. have a look out for a thread by my fello forum chummy Amenos, he made a very cool (literaly) heat ducting system. |
20 November 2009, 11:40 | #11 |
Registered User
|
If you guys say that an 68030fe16 easily overclocks to 40 mhz, what about the 33 or 40 mhz fe-cpu's... will they go even higher?
|
20 November 2009, 12:50 | #12 |
Ya' like it Retr0?
Join Date: Jul 2005
Location: United Kingdom
Age: 49
Posts: 9,768
|
@Nikolaj sofus
If I am not mistaken, when Motorola produced these latent TQFP version of the 030, they were all made from the same die and mask, as such the later revision are all inherrently 40mhz compliant, although I would suggest you check the mask revision of the chip to be 100% sure. Motorola, in the interests of cutting costs, would of only produced one chip and labled them according to order and or limmited testing. i.e. produce 50,000 68030fe TQFP... of wich 15,000 tested and stamped at 16mhz 10,000 tested and stamped at 25mhz 20,000 tested and stamped at 33mhz 5,000 tested and stamped at 40mhz The above are all the same chip from the same production run, only tested at that particular frequency. thats not to say some might not function correctly at hight frequences however. |
20 November 2009, 12:50 | #13 |
Registered User
Join Date: Jan 2007
Location: EU
Posts: 146
|
How about OCing standard A1200's 68EC020?
|
20 November 2009, 12:54 | #14 |
Ya' like it Retr0?
Join Date: Jul 2005
Location: United Kingdom
Age: 49
Posts: 9,768
|
@kixs
the 68ec020 is infact highly over-clockable. it should be able to clock upto 28mhz without to much fuss, the only problem is, at that speed, without extra electronics it wont function with the motherboard. Argualby you could build a small XOR'ing board to communicate properly at 14MHz, (wit hthe CPU running at 28MHz) however this is absolutely pointless without fastram (and not the PCMCIA kind) a 28mhz 020 is only 15% slower than an 030 of the same speed (although if the 030-CPU *and hardware* support 32bit burst mode access it will spank the 020 a fair bit!) |
20 November 2009, 13:26 | #15 |
Registered User
Join Date: Jan 2007
Location: EU
Posts: 146
|
Thanks for clarifying.
|
20 November 2009, 13:42 | #16 |
Posts: n/a
|
I have another turbo card with PGA type 68EC030 and a 68882, both rated 40 MHz running at 50 MHz without any problems. Although this is only a 25% OC, it runs hotter than the TQFP type 68030 (non-EC) 16 MHz running at 40 MHz (125% OC)
|
20 November 2009, 14:20 | #17 | |||
Posts: n/a
|
Quote:
Quote:
Quote:
|
|||
20 November 2009, 20:45 | #18 |
Doogster
Join Date: Dec 2007
Location: Dorset
Posts: 1,069
|
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Just found an Apollo 1230/40 | Deaths_Head | support.Hardware | 1 | 21 October 2012 21:51 |
WHDLOAD Freezes on Apollo 1230 LC | desantii | project.WHDLoad | 7 | 20 July 2010 00:11 |
Apollo 1230 mk3 woes | emuola | support.Hardware | 9 | 29 October 2009 20:10 |
Apollo 1230 problem | macce2 | support.Hardware | 20 | 03 July 2009 02:03 |
Apollo 1230 jumper settings | philau | support.Hardware | 1 | 05 December 2002 12:45 |
|
|