Announcement - new QL Interface

Nagging hardware related question? Post here!
User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Announcement - new QL Interface

Post by Dave »

129 / 110 * 100 = 117.27
110 / 129 * 100 = 85.27

I changed the one result to 129 because 6447 ticks is 128.98 seconds. So, the shadowed RAM is 15% faster, or the older system is 17% slower.

I'm filled with weird emotions today. I just came across the CAD files for the UltimIDE - IDE and 1MB. It's 80% complete. It's basically this board but an extra 512MB SRAM and a thru-connector. This thread's board is mighty fine so there's little incentive to finish the UltimIDE.

Everyone go buy one of the boards listed in this thread. :D


User avatar
Outsoft
Super Gold Card
Posts: 695
Joined: Sat Apr 19, 2014 1:30 pm
Location: Italy
Contact:

Re: Announcement - new QL Interface

Post by Outsoft »

martyn_hill wrote:So, just to close the loop on the 'Scroll test' results between shadowed and non-shadowed RAM QLs, with Minerva 1.98 fitted to the Issue5 BBQL with shadow-RAM, we do get slightly better performance again, thus:

Issue 7 (Minerva + TkII 2.23) 6447 'ticks' == 128 secs - as before
Issue 5 (Minerva + TkII 2.23) 5523 'ticks' == 110 secs - an improvement of 5 secs over MG ROM - about 14% speed-up over the non-shadowed RAM unit (was 10% with MG ROM.)

I've since written a short m/c routine to further test the ram speed - I'll post here once I've run it on both units.

M.
So the ISSUE 7 is performing so much rispect the 5 one? Wow!


User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: Announcement - new QL Interface

Post by Dave »

Outsoft wrote:So the ISSUE 7 is performing so much rispect the 5 one? Wow!
No, but the info is given in case it plays a part. The CPU and 8301 are clocked at the same speed so it shouldn't make a difference. Initially, the machines had MG and Minerva OS, so there was a speed difference caused by the faster screen handling routines in Minerva. The test was then re-run with Minerva o both systems to make it a more equal comparison of the two systems.

The speed difference is from following Nasta's plan of using shadow memory for video RAM. Writes go to the QL's own RAM and the external SRAM. Reads are directed only to the faster SRAM.

For a heavy load that is half reads and half writes, the shadowed SRAM system gains 15% in speed, compared to the non-shadowed system.

I hope that is clearer.


User avatar
Outsoft
Super Gold Card
Posts: 695
Joined: Sat Apr 19, 2014 1:30 pm
Location: Italy
Contact:

Re: Announcement - new QL Interface

Post by Outsoft »

Dave wrote:
Outsoft wrote:So the ISSUE 7 is performing so much rispect the 5 one? Wow!
No, but the info is given in case it plays a part. The CPU and 8301 are clocked at the same speed so it shouldn't make a difference. Initially, the machines had MG and Minerva OS, so there was a speed difference caused by the faster screen handling routines in Minerva. The test was then re-run with Minerva o both systems to make it a more equal comparison of the two systems.

The speed difference is from following Nasta's plan of using shadow memory for video RAM. Writes go to the QL's own RAM and the external SRAM. Reads are directed only to the faster SRAM.

For a heavy load that is half reads and half writes, the shadowed SRAM system gains 15% in speed, compared to the non-shadowed system.

I hope that is clearer.
All clear sir ;)


Silvester
Gold Card
Posts: 436
Joined: Thu Dec 12, 2013 10:14 am
Location: UK

Re: Announcement - new QL Interface

Post by Silvester »

martyn_hill wrote:(I used the MG ROM on this QL simply because I found the Network receiving by Minerva + TkII problematic in my setup...
Funny you should say that, I had similar trouble when I replaced Net physical layer with parallel 8 bit link (8255 PIOs back to back). Got a P_ device working upto 65Kbytes/sec between GC QL and Minerva Trump card (mk1, slow). Once in a while it would stall on sending last block, usually opening/closing something would complete it. Never got around to finding reason, the cable was cumbersome/short compared to simplicity of two-wire link, and I wanted PAR device for GC, so I reverted to using N_.

Also noticed that using Net with Minerva (N_ or P_) did strange things to IO.EDLIN (NFS translates use to IO.FLINE), the cursor flashed erratically. Worked OK with SMSQ/E and JS rom though.
martyn_hill wrote:...explore QNet in preparation for a QNet/USB adapter that I'm currently developing for my QPC/laptop setup - I'll be writing-up that project in the next week or so...)
Look forward to that.


David
robg
ROM Dongle
Posts: 8
Joined: Wed Apr 26, 2017 1:02 pm

Re: Announcement - new QL Interface

Post by robg »

That project seems ideal, what stage is it at?
If you look me up on LinkedIn you'll see I'm a very qualified and experienced tester. I'm the one from Nottingham.
I'd happily buy one and beta test it professionally.
How could I progress this?


martyn_hill
Aurora
Posts: 909
Joined: Sat Oct 25, 2014 9:53 am

Re: Announcement - new QL Interface

Post by martyn_hill »

Hi robg!

Given the mix of topics in this thread (partly my fault!), I'm not sure which project you are referring to.

If you were referring to the 'QNet-USB Bridge Adapter' that I'm currently developing, I was only ever planning on building my own and then publishing an article here somewhere for any one else interested in copying the ideas/code etc. I've no inclination in making a commercial opportunity out of it :-)

I'll start a fresh thread for this once I've made a little more progress, and of course, would then very much welcome feedback!

M.


Derek_Stewart
Font of All Knowledge
Posts: 3928
Joined: Mon Dec 20, 2010 11:40 am
Location: Sunny Runcorn, Cheshire, UK

Re: Announcement - new QL Interface

Post by Derek_Stewart »

Hi,

I have trying Habi image editor to read a CF Card formatted under Qubide v1.56 and QubATA (Qubide v2.xx).

I inserted a CF Card in to my Windows 10 PC via USB card reader. Windows said the CF Card needed formatting. But ignored and cancelled the request.

The image editor run as Administrator and opened the disk. All Qubide partitions were recognised.

I used the insert file function and copied C68, DP Collection zip files.

Executable files will lose the File Header. But there is a Header function to possibly preserve file header.

I put the CF Card back in the Trump Card Card slot and all the files I copied from the PC, were present and correct.

Nice piece of software.

The link is :
http://hardware.speccy.org/temp/qubide/ ... ftware.zip


Regards,

Derek
User avatar
Outsoft
Super Gold Card
Posts: 695
Joined: Sat Apr 19, 2014 1:30 pm
Location: Italy
Contact:

Re: Announcement - new QL Interface

Post by Outsoft »

Derek_Stewart wrote:Hi,

I have trying Habi image editor to read a CF Card formatted under Qubide v1.56 and QubATA (Qubide v2.xx).

I inserted a CF Card in to my Windows 10 PC via USB card reader. Windows said the CF Card needed formatting. But ignored and cancelled the request.

The image editor run as Administrator and opened the disk. All Qubide partitions were recognised.

I used the insert file function and copied C68, DP Collection zip files.

Executable files will lose the File Header. But there is a Header function to possibly preserve file header.

I put the CF Card back in the Trump Card Card slot and all the files I copied from the PC, were present and correct.

Nice piece of software.

The link is :
http://hardware.speccy.org/temp/qubide/ ... ftware.zip
The original software for the Qubide (Image Editor) works also on Windows 10: just set it in XP compatiblity mode and give him administrative functions.

Tested this afteroon with an external card reader: all works well as it goes on a XP system.


User avatar
Outsoft
Super Gold Card
Posts: 695
Joined: Sat Apr 19, 2014 1:30 pm
Location: Italy
Contact:

Re: Announcement - new QL Interface

Post by Outsoft »

Found another BUG.

When I made RES_128 (Trump Command) the QL reset but no one games that works "only at 128k standard ram"can be played because there are not really 128mb.

Example: CHESS, HYPEDRIVE and many more (originals with protections on MD).

I see that the problem is that on this clone card the TK2_EXT command is always "active" and so there is a portion of RAM occopied by this.

On the original Trump Card If you make a RES128 you can't access to the TK2 anymore.


Post Reply