QL test points...

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

QL test points...

Post by Dave »

Admit it. You're a bit of a hardware hacker, and the QL is your machine of choice. The old girl, she ain't up for much, but she purrs and is a joy to drive.

Sometimes, though, she gets to sounding a little off and needs a little work, and that's when you need... TEST POINTS!

Bearing in mind most signals are available at the DIN connector where would you, humble reader, like to see test points on a QL so that they would help you to quickly check a specific function or locate a faulty component or problem area?

Currently I have a few VIN, +5V and GND TPs spread around the board. What can I add that would be useful to you?


User avatar
1024MAK
Super Gold Card
Posts: 592
Joined: Sun Dec 11, 2011 1:16 am
Location: Looking forward to summer in Somerset, UK...

Re: QL test points...

Post by 1024MAK »

A secondary diagnostic microcontroller.


:!: Standby alert :!:
“There are four lights!”
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :!:
Looking forward to summer in Somerset later in the year :)

QL, Falcon, Atari 520STFM, Atari 1040STE, more PC's than I care to count and an assortment of 8 bit micros (Sinclair and Acorn)(nearly forgot the Psion's)
User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: QL test points...

Post by Dave »

1024MAK wrote:A secondary diagnostic microcontroller.
This is a cleaned up "Issue 7.5" based off the US.5 PCB mods to the Issue 6/7 schematic. It fixes the worst sins of the 80s designs by incorporating the small and simple changes everyone always wanted.

Changed:
4 layer board with full ground layer for clean signals/power.
The old board suffered from ground bounce, plus poor power design, mostly originating from the 16 RAM ICs.
8301/8049 moved to the CPU side of the data bus.
Reduces opportunity for CPU to be halted during 8301 video generation.
Switch mode power supply.
The 78XX are gone. A standard 2A+ 9v DC adaptor with 2.1mm barrel connector feeds into a very quiet and cool buck regulator circuit.
16x 64Kx1 DRAMs replaced by faster 4x 64Kx4 DRAMs.
The 16 1-bit DRAMs drew a lot of power and were a large switching load on the power supply. They have been replaced by 4x 4-bit DRAMs of a lower power design. The 8301 runs a lot cooler, and system current draw is cut by almost half.
MC1488/MC1489 replaced by ADM208EAR 5V part with onboard charge pump.
Part of this is that the MC1488/89 required +/-12v, and using a more modern part simplified power supply requirements.
All ICs are DIP and socketed in good quality sockets.
Highly serviceable. Designed to be easy to maintain or repair for the long haul.
All passives are now SMD.
This cleans up the board, reduces component and assembly costs, and frees a lot of board area for.... future projects.
Cleaned up microdrive circuitry on main PCB.
It had a hard time, with the power supply and video circuitry right next to it. With that gone, it has room to be more modular and organized in layout - not just crammed into the remaining space.
Changed how microdrive PCB fits on motherboard.
Microdrives now plug directly onto PCB, no cable. They have a pin array on the back which pushes down onto a socket on the PCB. It makes removing and reinstalling them MUCH easier, and the connections are much more secure.
Two replacement microdrive PCBs are part of the layout.
Redesigned 4 layer PCB, other changes include giving clean 5v instead of 9v so no regulator on board - keeping your fragile microdrives away from heat sources.
DB9 serial/joystick.
Can't get the old sockets. Wouldn't use them if I could. Joystick would be Kempston compatible layout, unless feedback tells me to do different. Serial, I can either follow the German QL pinout or standard 9-pin PC serial layout.
ROMOE jumper.
Select if using Sinclair ROMs or an EPROM - inverts the chip select signal.
Reset button.
I couldn't find a good source for the QL reset button, so I have replaced it with an upright tactile switch.
Serial headers.
You can disconnect internal serial from the DB9s and hook up the SuperHermes serial. Maintains the clean lines of your QL without bits hanging out of it.
Fits in existing case.
All socket positions, mounting holes, etc. are in the same place. Put it in your QL case and nobody will know. Unless you're english, in which case you will need to remove some plastic at the back of the case for the DB9 sockets - then it will look exactly like a German or US QL.

Added:
New GAL replaces IC38 HAL
Revised logic, primarily focused on improved handling of /DTACK. Replaces 8301 decode functionality. Also, a bit of future proofing.

Deleted:
MC1377, UHF modulator and related circuitry.
Who uses UHF any more? It is quite hard to find all the parts now. I decided to leave some space free to have the option of adding a SCART connector and/or S-Video/composite. All video signals are buffered and shock protected, so no more blown up 8301s.

Not changed:
Expansion port, ROM port, RGB functionality, microdrive capability, QLNET, sound (now buffered), keyboard, board shape/size, microdrive expansion port, component positions. All previous internal expansions will fit, including (Super)Hermes. Included headers so you can disconnect internal serial from the DB9s and hook up the SuperHermes serial.

I can see anyone using an issue 5 board buying this as an upgrade, and anyone with a bad issue 6 or 7 board might buy this as a service/replacement part. It's ideally suited to large systems that are heavily expanded with (S)GC, etc. etc.. Just provide a 3A 9V supply.

I do have options for future revisions. However, this board isn't about big changes. It's about refining the current design in a 100% compatible way. I have been incredibly slow about releasing anything. I think this board being in people's hands will maybe change some minds and I might get a couple more offers of help with other projects that need coding support. (See separate post)

So, bear in mind the character of the board - it is very definitely an Issue 7-like board, with just modernised component selections and fabrication technologies. Think of it as just a new 100% QL board.

With that in mind, the engineers, developers, etc... What signals would you like brought out to test points?


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

Re: QL test points...

Post by Dave »

I'll be assembling most of these so they can be fully tested before shipping, but I will make a few available with only the SMD components mounted and the rest of the components in a kit bag with instructions.

ALSO! Any place where a component is the same or equivalent to one from the Sinclair designs, it has the exact same part number. So all the components that have matching numbers are the exact components in a US.5 QL.

Also, you're going to LOVE the documentation that comes with it.

Side question... Since I am in the US, and there's no US Sinclair trademark registration, do you think I'd be ok putting a small sinclair logo on them, like the originals?

Side observation... I still haven't figured out how the 8301 generates NTSC timings. It has to be configured by the JSU ROm version, because replacing it with a Minerva or plain JS ROM stops NTSC, but I don't know of a register in the 8301 that sets this value.


User avatar
mk79
QL Wafer Drive
Posts: 1349
Joined: Sun Feb 02, 2014 10:54 am
Location: Esslingen/Germany
Contact:

Re: QL test points...

Post by mk79 »

Dave wrote:Side observation... I still haven't figured out how the 8301 generates NTSC timings. It has to be configured by the JSU ROm version, because replacing it with a Minerva or plain JS ROM stops NTSC, but I don't know of a register in the 8301 that sets this value.

Code: Select all

mc_stat equ     $18063  status register address
*               bit     clear   set
mc..blnk equ    1       normal  blank   if set, the rest are irrelevant!
mc..m256 equ    3       512/4   256/8   pixels per line / colours per pixel
mc..ntsc equ    6       PAL625  NTSC495 scan lines (512/384 displayed)
mc..scrn equ    7       $20000  $28000  screen base address (scr0/scr1)


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

Re: QL test points...

Post by Dave »

Thank you, Marcel.

:D


Nasta
Gold Card
Posts: 443
Joined: Sun Feb 12, 2012 2:02 am
Location: Zapresic, Croatia

Re: QL test points...

Post by Nasta »

mk79 wrote:
Dave wrote:Side observation... I still haven't figured out how the 8301 generates NTSC timings. It has to be configured by the JSU ROm version, because replacing it with a Minerva or plain JS ROM stops NTSC, but I don't know of a register in the 8301 that sets this value.

Code: Select all

mc_stat equ     $18063  status register address
*               bit     clear   set
mc..blnk equ    1       normal  blank   if set, the rest are irrelevant!
mc..m256 equ    3       512/4   256/8   pixels per line / colours per pixel
mc..ntsc equ    6       PAL625  NTSC495 scan lines (512/384 displayed)
mc..scrn equ    7       $20000  $28000  screen base address (scr0/scr1)
Interesting, I had no idea they included bit 6, i tried it on my 2 QLs a long time ago and it did nothing.
This suggests it's not in all 8301 versions?


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

Re: QL test points...

Post by Dave »

ALL my US QLs have exactly the same components/markings, as follows:

68008P8 or 68008L8, clocked at 7.552445 MHz
ZX8301 marked CLA2345 with red dot
ZX8302 marked NCR0371725 F813816 G 8442
8049 marked MAB 8049H A038 DSD436V4Y: (C)SZX83V0.7

So, the only irregularities are the 15.10489 MHz crystal and they all have the later version 8301.

Since you're here, Nasta, would you maybe like to post in a new thread the equivalent register details for the Aurora?

It would be very enlightening. :D


User avatar
mk79
QL Wafer Drive
Posts: 1349
Joined: Sun Feb 02, 2014 10:54 am
Location: Esslingen/Germany
Contact:

Re: QL test points...

Post by mk79 »

Nasta wrote:Interesting, I had no idea they included bit 6, i tried it on my 2 QLs a long time ago and it did nothing.
This suggests it's not in all 8301 versions?
No idea, I just took this out of the Minerva sources. But looking at SMS.DMOD one also sees

Code: Select all

D2.B key: -1 read display D2.B Display type
0 monitor
1 625-line TV
2 525-line TV
and disassembling the JSU-ROM we get

Code: Select all

move.b d2,sys_dtyp(a6)		; Display TYPe (0=normal, 1=TV 625, 2=TV 525)
lsl.b  #5,d2
andi.b #$40,d2		; Bit 6 set if value was 2
andi.b #$bf,d0
or.b   d2,d0
move.b d0,$00018063
As far as I can see with a short look, D2 is ignored on normal JS-Roms.

Marcel


User avatar
mk79
QL Wafer Drive
Posts: 1349
Joined: Sun Feb 02, 2014 10:54 am
Location: Esslingen/Germany
Contact:

Re: QL test points...

Post by mk79 »

Dave wrote:Since you're here, Nasta, would you maybe like to post in a new thread the equivalent register details for the Aurora?
The Aurora registers are already well documented, what's missing?

Marcel


Post Reply