Page 2 of 66

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 13:55
by Badwolf
exxos wrote: 10 Nov 2022 13:53
Badwolf wrote: 10 Nov 2022 13:50 Pin 1 is on the silkscreen & there's a little bit of text to to say "oscillator goes this way -->" :smackbum:
*squints* ah yes, just need a magnifier and good to go.
My best one is the one that says 'do not fit, for expansion use only' which is actually too small for the silkscreen printer to make anything of it bar Morse code :?

BW

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 14:41
by exxos
First up to establish a baseline speed test in GB6.

IMG_0175.JPG

Then next test with DFB1 fitted.

*hammers in* :D

IMG_0176.JPG

First test run.

IMG_0177.JPG

*Grabs some random version of BLTFIX and MAPROM*

IMG_0178.JPG

Did noticed on the first test that some of the icons are missing :shrug: but really need to know what software I am actually supposed to be running @Badwolf :lol:

IMG_0179.JPG

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 15:00
by Badwolf
exxos wrote: 10 Nov 2022 14:41 Did noticed on the first test that some of the icons are missing :shrug: but really need to know what software I am actually supposed to be running @Badwolf :lol:
Absolute bloody hooligan! :P

https://github.com/dh219/DFB/tree/main/DFB1#software

You *need* NVDI (or another VDI replacement). Blitfix doesn't work on TOS4.

MAPROM2.2 is probably the best compatible -- the latest version requires cold boot each time. I'm a bit of a fan of plain FASTRAM.PRG, to be honest, but if you want that 350% ROM figure, then MAPROM is your friend. :)

BW

EDIT: you don't have Flash fitted, but if you decide to add it later, this is useful:- viewtopic.php?p=84589#p84589

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 15:14
by exxos
Badwolf wrote: 10 Nov 2022 15:00 Absolute bloody hooligan! :P
:excited: :yay: :excited: NEW TOYS!! :crazy: :crazy: :excited: :yay: NEW TOYS!! :excited: :crazy: :excited: :yay: :yay:

Badwolf wrote: 10 Nov 2022 15:00 https://github.com/dh219/DFB/tree/main/DFB1#software

You *need* NVDI (or a VDI replacement). Blitfix doesn't work on TOS4.

MAPROM2.2 is probably the best compatible -- the latest version requires cold boot each time. I'm a bit of a fan of plain FASTRAM.PRG, to be honest, but if you want that 400% ROM figure, then MAPROM is your friend. :)

Well I tested BLTFIX anyway :P

Without it this happens.

IMG_0180.JPG

With it, it looks normal other than the GB6 first test problem.

IMG_0181.JPG

Guess I will go find NVDI now then...

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 15:45
by exxos
@Badwolf This is MAPROM from Anders repo and NVDI. Nothing else. All looks fine now :D

IMG_0182.JPG

Going by the time for alt-ram test, looks about the same as the TF536, so things seem to be running as expected :) Not sure why ROM speed ends up slower than the TF536 though ?

In fact slightly puzzled by results overall :shrug: Integer division should be a lot higher running from alt-ram.. I will check GB6 flags make sure they all set properly.. but should be more like 1700 kinda mark..

ST536_NVDI.jpg

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 15:57
by exxos
Oh nevermind I was being dumb there.. The previous image is the speed boost ONTOP of a stock falcon ...

This is the speed of the DFB1 compared to a stock STE.

IMG_0183.JPG

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 16:00
by Badwolf
exxos wrote: 10 Nov 2022 15:45 @Badwolf This is MAPROM from Anders repo and NVDI. Nothing else. All looks fine now :D
Cool, although please note it may not work after a warm reset.
Going by the time for alt-ram test, looks about the same as the TF536, so things seem to be running as expected :) Not sure why ROM speed ends up slower than the TF536 though ?
Actually my AltRAM speed is slower than the 536 as I have a 50MHz clock rather than Stephen's 100MHz. My speeds are 8.575/9.345 ~= 92% of his and the baseline for ROM access is higher in the Falcon so the absolute score will be around that 90% figure but the relative percentage will be lower under half. (90% / 2x initial ROM speed).
In fact slightly puzzled by results overall :shrug: Integer division should be a lot higher running from alt-ram.. I will check GB6 flags make sure they all set properly.. but should be more like 1700 kinda mark..
1.82s is bang on for a 50MHz 030 with instruction cache on. The figures are identical.

Remember you're comparing with a base level Falcon on one hand with a base level STF on the other.

BW

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 16:05
by exxos
Badwolf wrote: 10 Nov 2022 16:00 Actually my AltRAM speed is slower than the 536 as I have a 50MHz clock rather than Stephen's 100MHz. My speeds are 8.575/9.345 ~= 92% of his and the baseline for ROM access is higher in the Falcon so the absolute score will be around that 90% figure but the relative percentage will be lower under half. (90% / 2x initial ROM speed).
Makes sense.
Remember you're comparing with a base level Falcon on one hand with a base level STF on the other.
Yep. I just mentioned that in previous post. So things all look as expected now :)

So all other than the FPU mystery, is there anything you want me to specifically test while it is connected up ?

Not sure if there is a different firmware for with and without FPU ? I did quickly plug in one earlier and it crashed on the first FPU test. maybe it is not getting clock for DTACK is broken :shrug:

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 16:21
by Icky
exxos wrote: 10 Nov 2022 15:57 This is the speed of the DFB1 compared to a stock STE.
Interesting the suggestion is to use NVDI as scores are always going to be higher than without it.

Can we get a test of a stock falcon with NVDI installed to compare the DFB1 falcon with NVDI? Not to give you more work @exxos

Re: exxos's DFB1 trials

Posted: 10 Nov 2022 16:29
by exxos
Icky wrote: 10 Nov 2022 16:21 Can we get a test of a stock falcon with NVDI installed to compare the DFB1 falcon with NVDI? Not to give you more work @exxos
Yeah I can do that..

IMG_0185.JPG

Oddly it seems to have "died" somehow. Now it gets stuck exact same place...

IMG_0184.JPG

:shrug: