ZZ9000 Compendium

More than a modern Amiga graphics card
Compiled by a ZZ9000 owner and enthusiast

Updated: March 25, 2023 | ZZ9000 is © 2019-2023 MNT Research GmbH

Source(s): gdonner

Status: Unknown

Note(s): On my Amiga 4000, this occurs when using Amiga-M to cycle between RTG and NTSC screens (video examples below):

ZZ9000 RTG <-> NTSC screen switching bug 1 (SysInfo v4.4)

ZZ9000 RTG <-> NTSC screen switching bug 2 (Asimware AudioThunder v1.1)

ZZ9000 RTG <-> NTSC screen switching bug 3 (DiskMaster 2.1)

ZZ9000 RTG <-> NTSC screen switching bug 4 (Workbench 3.2 screen 'compound fracturing')

Source(s): my_pc_is_amiga (Original MNT Community post.)

Status: Unknown

Note(s): Possible cause from Chain|Q on IRC: "Based on the hit, and the offset, I think SetGC() is called with NULL ModeInfo in this case, so when the driver tries to read mode_info->Width, it hits. (0x12 offset is Width in the ModeInfo struct, and that's the first read there, also A1 is NULL in your hit regdump) Now, in turn SetSwitch() (gonna assume only called with DISPLAYCHAIN=NO) in the driver calls SetGC(), with b->ModeInfo as argument, but in turn SetGC() b->ModeInfo is set inside SetGC() first... So if that's some first "init" call, and SetSwitch is called first, ModeInfo could be NULL there, and there's your hit. This is entirely just by reading the code, but I'm pretty confident something like this happens."

Source(s): dschallock (English Amiga Board post) ^{[archived image]}

Status: Confirmed by Lukas as a bug.

Note(s): Occurs when running programs that use native screen modes like DeluxePaint IV. Thomas Richter notes: "This is *not* a P96 bug, but a ZZ9000 firmware bug that was confirmed by Lukas. P96 cannot control what the ZZ9000 does to native screens, and the native output is fine."

Source(s): my_pc_is_amiga (IRC/#mnt-amiga)

Status: Unknown

Note(s): my_pc_is_amiga: "I have [firmware] 1.9.1 and see this split screen on the flickerfixer side when switching between RTG and the flickerfixer. This is seen in PDraw but also have seen it when in Scala as well, when changing between RTG and Amiga. I have [Tool Type] DISPLAYCHAIN=YES. The native Amiga screenmode is NTSC Hires.

Source(s): ugreen (IRC/#mnt-amiga)

Status: Unknown

Note(s): Passive (non-powered) HDMI to VGA converters don't work. However, ugreen reports that at least one active/powered one works (Claude tested this).

Issue(s):

Source(s): punk_ (IRC/#mnt-amiga)

Status: Unknown

Note(s): A phenomenon called the "red or blue screen." At the time of writing (Sep. 20, 2019), Lukas had not yet been able to trigger it.

Source(s): FranciscoJr (English Amiga Board)

Status: Unknown

Note(s): "Accepts only 4 GB. Could not transfer 128 MB 32-bits Fast RAM from ZZ9000 to Amiga. This could be set manually or automatically, like as done in the Elbox Mediator. (ZZ9000 RAM -> Amiga is no longer an issue) Waiting for a solution from MNT." ^{[source link]}

Source(s): gdonner

Status: Unknown

Note(s):

Amiga 3000 Dell U2211H TFT Horz: 30-83 kHz Vert: 56-75 Hz		NTSC: Low Res Laced **Broken**
Amiga 4000 Dell P2319H IPS LED Horz: 30-83 kHz Vert: 56-76 Hz		NTSC: Low Res Laced **Broken**		NTSC: High Res Laced **Broken**

Source(s): multiple (including a1k.org)

Status: Fixed as of firmware/drivers v1.11.

Note(s): There appears to be a general issue with ZZ9000 Zorro register access on both 68040 and 68060 Zorro III-based Amigas, making some of the communication unreliable, probably including some of the important ZZ9000AX register access. In addition, the ZZ9000AX's AHI driver source code appears to be incompatible with AHI high-level access—which is used by both AMPlifier and AmigaAMP.

In short, both the ZZ9000 and ZZ9000AX firmware, and the AHI driver code need fixing and thorough testing.

SongPlayer might work on some systems (as it uses low-level register access), but there's no guarantee.

Source(s): my_pc_is_amiga (IRC/#mnt-amiga)

Status: Fixed as of firmware/drivers v1.10.

Note(s):

From Thomas Richter:

"The above problem stems from the InitBoard function of the rtg.library, namely

Code:

sprintf(bi->VBIName, "VBInt %.20s (%ld)", card->Name, bi->BoardNum);

That is, "card->Name" is not initialized properly.

This bug report goes please to mntmm (i.e. Lukas). I cannot fix anything on this from my end."

Source(s): gdonner

Status: Fixed in v1.10 firmware (a known issue in firmware v1.8).

Note(s): Can occur when trying to view native Amiga screen modes. The screen may also temporarily open, but seconds later, return to out of sync:

Source(s): _Bnu (IRC/#mnt-amiga)

Status: Fixed.

Note(s): _Bnu: "It would need a 640x400 or 720x400 mode over HDMI to scale cleanly to. The card is aware, so it could probably be handled on the ZZ9000 side; the question is just how cooperative monitors would be with those resolutions. They're both somewhat standard DOS resolutions. I usually end up with really crappy cropping when I use those resolutions over VGA, but it might be fine over DVI."

Source(s): _Bnu, gdonner (IRC/#mnt-amiga); ProfPlum (English Amiga Board)

Status: Fixed.

Note(s): There appears to be some kind of issue(s) related to 1.3c firmware and NTSC mode (PAL modes don't seem to be affected). Some still occurs during bootup with firmware/drivers 1.5.1 (Amiga 4000). Amiga passthrough video (e.g. Amiga Early Startup) comes through, but may be vertically squashed a bit. Amiga 2000 NTSC screens are also fuzzy on the HDMI output.

Lukas: "The garbage at the bottom is leftover content in the framebuffer that doesn't get cleared. The NTSC image has less lines than the actual screen resolution; the NTSC screens are completely captured, but the screen just doesn't "end" at the right line."

Source(s): tryphoncosinus (IRC/#mnt-amiga)

Status: Confirmed as a fix for A2000 systems with a rev. 6.2 motherboard (and original PSU).

Note(s): If you have an Amiga 2000 with a rev. 6.2 motherboard and an original PSU, and your system appears to be switching back and forth between PAL and NTSC, try moving jumper J300 (just above the Paula chip) from pins 2-3 to pins 1-2.

Note that this is NOT a solution if you have an ATX PSU.

a1k.org user A10001986 sheds some light on this issue: "J300 switches between tick from power supply vs. VSYNC as clock source. Has nothing to do with PAL/NTSC."

_Bnu adds: "It keeps switching back and forth between ~50 and ~60 Hz on the video output. ... It's definitely the computer being unable to decide if it should keep the video output at PAL or NTSC refresh rate, for some reason. Knowing the root cause as to why one of the two tick sources don't work would be infinitely more interesting."

Installing a tick generator may resolve this.

Source(s): Ravi Abbott (YouTube), gdonner (IRC/#mnt-amiga)

Status: Fixed as of firmware/drivers v1.5

Note(s): Lukas: "There is a "swapped Hi-res pixels" bug in the scandoubler that happens for about 10% of cold starts."

Source(s): Lukas, various batch 1 owners

Status: Fixed as of batch 2 (and later) cards

Note(s): Batch 1 ZZ9000 cards may be a tight fit in your Amiga 4000 (noticeable card bowing in above photo), but they should go in; just be as careful as you can. Lukas has noted that this issue was resolved in batch 2 (and later) cards ("...and will have a slightly improved mechanical fit for easier installation"), and will fit (albeit tightly) in the Checkmate A1500 plus case.

Source(s): trixster (IRC/#mnt-amiga)

Status: Fixed/Workaround

Note(s): Possible fix/workaround for Prometheus owners: Having a second Zorro card installed in any Zorro III slot may well 'fix' the issue. Any other Zorro card in combination with the ZZ9000 makes—or should make—it work in an Amiga 4000 with a Prometheus PCI Zorro III card.

Note that the Prometheus has a "Swap Config" jumper that determines the order of memory allocation for Zorro and PCI cards, so its open or closed state might affect the Mediator's interaction with the ZZ9000 and other Zorro cards (or the absence of other Zorro cards).

Source(s): Jope, FranciscoJr (English Amiga Board)

Status: Fixed/Addressed as of firmware v1.6.2/drivers v1.6

Lukas (via 1.6.2 firmware change log): "Fixes to network driver including the possibility of switching to INT2 instead of INT6 in the event you have hardware conflicts. To do this, set the interrupt jumper on ZZ9000 to INT2 and create an empty file ENVARC:ZZ9K_INT2."

Jope: "...the X-Surf 100 + RapidRoad combo can't coexist with the ZZ9000 LAN. There is an interrupt conflict." ^{[source link]} ... It shows in the way that the zz9000net will not really receive packets when the X-Surf 100 is plugged in to the same machine. I didn't bother checking whether it was the X-Surf or the RapidRoad that was conflicting, I just pulled the X-Surf out and zz9000net started working for me."" ^{[source link]}

FranciscoJr: "ZZ9000.net only works if the X-Surf 100 is not plugged in to the same machine. But X-Surf 100 works fine with ZZ9000 plugged. Using Roadshow TCP/IP stack. This is a known issue, and perhaps it will be fixed next firmware, as informed by Lukas Hartmann."

Source(s): hUMUNGUs (IRC/#mnt-amiga), ProfPlum (English Amiga Board)

Status: Partially fixed in v1.4b; pixel trails on the mouse are gone, but there are still pixels on Workbench windows and dropdowns.

Note(s): Using Zorro III firmware leaves a trail of pixels all over the place when moving the mouse. The same thing happens with drop down menus—most of the pixels from the window stay behind. The pixel trail issue also affects the A2000 with TF534 and hangs the system. Temporary (albeit slow) workaround: Use the Zorro II firmware.

Lukas: "This is a bus timing/loading issue, might be '030 related, might not be. This will take a bit of time as I have to repair my A3000, etc."

Source(s): my_pc_is_amiga (IRC/#mnt-amiga)

Status: Possible Explanation

Note(s): my_pc_is_amiga: "I took out the ZZ9000 and put in a PicassoIV card to compare. PicassoIV does work with the Video Toaster. However, I have to start the Toaster in Amiga video mode for it to come up properly and not corrupting the Workbench screen. If in RTG mode (P96) on Workbench, the screen gets corrupted. This is a different effect than what I saw in the ZZ9000 of the split screen. The split screen seemed to happen even when I was in Amiga graphic mode in Workbench on the ZZ9000.

Source(s): Francois (twitter)

Status: Workaround as of firmware/drivers v1.7.0.

Note(s): Francois: "It happens after a soft reset." ^{[Original link]}

Source(s): dev74.com (website)

Status: How-To

Note(s): How to use the ZZ9000 on systems running AmigaOS 4.1 F.E.: ZZ9000 works fine also on Amiga OS4.1 FE with PowerPC !