endofexclusive / deniser Goto Github PK

Currently I have had much luck and the board behaves nicely.
Most Errors are dividable in three schemes:

1. Sprites are shown/repeated above/beneath the display window
2. Offset Failure (kickstart 3.1 screen animation is cut off a few pixels from the left and repeated on the right side.)
3. Jitter/Display Errors (Looks much like the RGB2HDMI calibration was missing but only in some places)

Currently some tests show failures, I'll try to compile a list:

Turrican 3: Error 1 in Intro
Agony: Same as above but Ingame (Sprites near Bottom)
AIBB: Error 3 in some Visual Benchs
SWIV: Error 3 in top Score Lines
Battle Squadron: Error 3 (Some Sprites are drawn wrong)
CopperMaster Demo: Most likely something like Error 2 but also different parts are drawn outside the display area

I'll go one with the list. System is an A500+ with a 68ec020 Accelerator running at 35Mhz. 2MB Chip, 8MB Fast, 1.5MB Ranger Mem. Kickstart&OS3.1.4 from CF, WHDLoad.

I'd be very interested in a Super Denise version...ideally one with NTSC compatibility. Any plans to do so?

Hi,
I am new to this sort of thing, and found this project and felt it has promise to help me with one of the AMIGA500s I have to repair.
I would like to know how do I order a pre-fabricated PCB to build a Deniser.
Or if someone has a Pre-etched board they are willing to sell I am happy to buy it off you, and pay for shipping.

Hi, would it be possible to provide JEDs since I had issues building it in my Environment. (WIn10)
I haven't used the machx03 series before and Diamond doesn't recognize the device itself nor does it offer the exact board. Could you elaborate on programming?

Amiga's power is notoriously bad given that they're 35-year-old machines (or more). The Denise socket is close enough to the mains that this might not be worth it, but a cheap, drop-in 5V boost like the TPS81256SIPT might be a good option to ensure a stable 5V.

In this use case, they're functionally similar, but the CBTD:

• has no direction (save drd_rl_to_fpga)
• has a much shorter propagation delay (150ps versus ~2.5ns)
• has slightly more IO per 48 TSSOP (20 channels versus 16)
• use virtually no power themselves (current is drawn thru the CBTD, not from it)
• if space permits, two CBT16211's could replace the three LVCH162245's

The CBT's are more common and cheaper than the CBTD's. TI makes a specific 5V to 4.3V LDO for this (since the CBT's clip another 1V off that for 3.3V output): TLV70243DSER.

You would then be able to omit the 3.3V LDO by taking the 4.3V power meant for the CBTs to some of the spare pins and using the other side pins as a 3.3V power source. Then you're just swapping the existing 3.3V LDO for the 4.3V LDO.