An open source flicker fixer for Amiga 500/2000

[PurpleMelbourne]

Originally Posted by Gorf

How much is the fish chip?

its not available via mouser or digikey ... obsolete part?

I bought a dozen of them on ebay for about US$15 each. The Taiwanese manufacturer was still willing to sell them direct when I contacted them before Christmas.

A friend told me about the chip as he used it in a car video system a while back. But for Amiga purposes its entirely adequate. For more money they have a more advanced chip which also outputs in HDMI and DisplayPort.

I've heard that Analogue Devices have a similar chip too.

[Niklas]

Originally Posted by Niklas

Unfortunately the second version of the adapter PCB has some signal integrity issues, and I haven't taken the time to debug that yet, so that's where things stand.

Today I took the time to look into this a bit further.

In V1 of the adapter PCB I used a 74LVC16244 chip, which is a buffer/driver.
In V2 of the adapter PCB I used a couple of 74CBT16211 chips, which are FET bus switches.

The C7M and CDAC clock signals after the 74LVC16244 (V1) look like this: https://imgur.com/a/QS5rSpT.
And the same clock signals after the 74CBT16211 (V2) look like this: https://imgur.com/a/FJgYDsO.

It's quite a big difference. I then measured what the clock signals look like on Denise without any adapter PCB connected, and they looks exactly like the signals after the V2 adapter.

So the V2 adapter PCB lets the signals through without modification, while the buffer in V1 fixed the shape of the clock signals. So there is actually no signal integrity issues (like I thought it was), but what happened is that the timing of the signals changed going from V1 to V2, so that the HDL I had written for V1 didn't work for V2.

Regarding how to proceed I guess one possibility is to make a V3 PCB and add buffers. Another option is to continue using the V2 adapter PCB and hopefully work around the fact that the clock signals aren't exactly square waves. For now I will go with the later option and continue using the V2 PCB.

[Gorf]

Originally Posted by PurpleMelbourne

I bought a dozen of them on ebay for about US$15 each. The Taiwanese manufacturer was still willing to sell them direct when I contacted them before Christmas.

A friend told me about the chip as he used it in a car video system a while back. But for Amiga purposes its entirely adequate. For more money they have a more advanced chip which also outputs in HDMI and DisplayPort.

I've heard that Analogue Devices have a similar chip too.

Good to hear!
Yes I linked to Analogue Devices in one of my posts ... there might also be a suitable IC.

From my perspective hdmi or dvi would be the output of choice - I see no benefit in creating just an other soon obsolete analoge output. But besides that the chip you mentioned has all the criteria.

[Gorf]

Originally Posted by Niklas

Today I took the time to look into this a bit further.

In V1 of the adapter PCB I used a 74LVC16244 chip, which is a buffer/driver.
In V2 of the adapter PCB I used a couple of 74CBT16211 chips, which are FET bus switches.

The C7M and CDAC clock signals after the 74LVC16244 (V1) look like this: https://imgur.com/a/QS5rSpT.
And the same clock signals after the 74CBT16211 (V2) look like this: https://imgur.com/a/FJgYDsO.

It's quite a big difference. I then measured what the clock signals look like on Denise without any adapter PCB connected, and they looks exactly like the signals after the V2 adapter.

So the V2 adapter PCB lets the signals through without modification, while the buffer in V1 fixed the shape of the clock signals. So there is actually no signal integrity issues (like I thought it was), but what happened is that the timing of the signals changed going from V1 to V2, so that the HDL I had written for V1 didn't work for V2.

Regarding how to proceed I guess one possibility is to make a V3 PCB and add buffers. Another option is to continue using the V2 adapter PCB and hopefully work around the fact that the clock signals aren't exactly square waves. For now I will go with the later option and continue using the V2 PCB.

Yes - looks more like a sawtooth then square wave - but but to the RC less ripple. Keeping it this way would ensure to be able to use the original Denise in parallel, wouldn't it? Might be useful for testing...

[amigappc]

Hello, any news with this project?

[Niklas]

Originally Posted by amigappc

Hello, any news with this project?

No development has happened in a long while.

There are some options out there:

RGB2HDMI, https://github.com/c0pperdragon/Amiga-Digital-Video, seems like a good cheap solution if you don't mind having a raspberry pi in your Amiga, and keep the original Denise.

Indivision ECS V2, http://wiki.icomp.de/wiki/Indivision_ECS_V2, is a more expensive solution, but uses an FPGA and replaces Denise.

Deniser, https://github.com/endofexclusive/deniser, is an implementation of Denise in an FPGA, but has no flicker fixer/scan doubler functionality AFAIK.

It might be possible to combine Deniser with the open source flicker fixer, and use this chip to get DVI/HDMI output. It's not clear how large this board would become however, unless an FPGA in BGA package is used, which would make it harder to build for hobbyists.

I think the reason that the development of the project is stalled, is that I don't see which niche this design would fill. "An open source alternative to Indivision ECS" is the closest thing I can think of so far, which is nice in a way, but in practice I think there are very few persons who would take that open source design and develop it further. And if someone just wants to have HDMI output for their Amiga, then RGB2HDMI seems like a good and cheap option.

Maybe there is some way to think about this that I have missed?