Let's design a new video mod

[bircoe]

Are you sure I can't convince you to sell me a PCB??? Happy to cover full expenses for the boards.

 

I just created a batchpcb account and uploaded your zip file, the boards are pretty cheap, $5 each, but postage is a killer to Australia...

 

 

Subtotal $10.00

Shipping $52.38

Handling $10.00

Grand Total $72.38

 

I'm not paying $72.38 for a pair of PCB's!

 

I just purchased an Atari Flashback 2 which was shipped from the USA and shipping was only $16... a Flashback is larger and heavier than 2 PCB's!

[Game-Tech.US]

I still have a few unused boards from my first two prototype runs.

Batari a/v, Batari a/v and pause, LHE a/v, LHE a/v and pause, and separate pause boards.

PM me.

[TheShanMan]

This project is very cool. I'm looking forward to the conclusion. I should be getting a 2600 shortly from a friend to add to my game room (I'm mostly a collector of classic arcade games), and I've been trying to figure out which video mod to use. Sounds like this will be worth the wait. Any sense of how far off you guys are from calling it done?

[+batari]

Are you sure I can't convince you to sell me a PCB??? Happy to cover full expenses for the boards.

 

I just created a batchpcb account and uploaded your zip file, the boards are pretty cheap, $5 each, but postage is a killer to Australia...

 

 

Subtotal $10.00

Shipping $52.38

Handling $10.00

Grand Total $72.38

 

I'm not paying $72.38 for a pair of PCB's!

 

I just purchased an Atari Flashback 2 which was shipped from the USA and shipping was only $16... a Flashback is larger and heavier than 2 PCB's!

There might be a cheaper shipping option (I know in the USA, it defaults to "priority" but you can select "first class" for much less.)

[bircoe]

I wish that were the case (see attachment), in any case I've PM'd akaviolence.

 

Please keep up the good work and keep us up to date when boards become available to purchase!

Edited April 22, 2011 by bircoe

[axh174]

Fantastic work! Really looking forward to this! Please keep us informed!

[Game-Tech.US]

I've been chatting with Batari and decided to post my testing results here as well.

He was kind enough to send me one of his boards AND a telegames pcb with a colorbar generator ROM on it!

First round of tests:

Ok, I put the board together and as I did I noticed an error on my board. I've been using a 6.2K in place of the 8.2K resistor, more on that later. With the proper resistor and 2K pots in place I hooked it up to another JR and got the same bars on the svid to the LCD tv that I got with my board and the LHE board. Now its either the LCDTV or possibly the way I hook up the board, i've been cutting resistors instead of lifting pins, are you actually lifting pins 6, 9, 12, and 13? Again something I want to test and compare. Also just as you had found, the pots are best at zero and 200ohms, so i've already removed them from the batari board, though this had no effect on picture quality. I just don't like pots.

I also went ahead and used 1.5K resistors as audio pull ups, just in case your board somehow had better quality and lower volume than my design had, it didn't, still too loud and poor audio quality, with PFII I might add. Still need to test with lifted pins vs cut resistors. Another test would be to hook up another audio source to my tv and switch between that and the atari to get a good idea where the volume needs to be. Then i'd like to have a test ROM for audio, know of any good ones? I don't think there is really much we can do about it though, and even if there were i'm certainly no audiophile and have no ear for good sound...

Second round of tests:

Now more testing results, i'm still getting the bars on svid, but i've pretty well eliminated most possible problems and am starting to suspect the lcd tv. If it weren't for my projector, my only other svid capable... I just remembered my capture card has an svid input, maybe i'll try it out tomorrow and get a better pic of the symptom bars. Anyway, the projector does not have the bars, but the black square of the colorbar pattern is green, btw the bars are greenish on the lcdtv colorbar pattern as well. The projector doesn't handle color too well either, both blue seem to be the same hue.

So here's what I did, still using JRs.

I cut traces leading away from pins 6 and 9, no better.

Lifted pins 6 and 9, no better.

Separated 6 and 9 wire leads from the ribbon cable I use, no better.

Separate all leads of the ribbon cable, no better.

My only working svid cable is 50' and coiled up behind the lcdtv, uncoiled it across the floor, no better, while uncoiled it was also hooked up to the projector.

I plan to get another svid cable to try asap as well.

Also replaced the 8.2K with two 5K pots in series to see if my using a 6.2K was causing any symptoms, only change was brightness, lower value resistance gave a brighter picture, not necessarily a good thing actually, but even 5K ohms looked good, may go back and get a better read on that.

I also hooked c49 and r31 back up, pin 6 traces lead to these components, and didn't seem to have any effect on video output. Have you even not lifted pin 6?

Pin 9 definitely needs cut or lifted.

More info:

Bought a new 6 foot svid cord, no better.

My capture card was acting weird and I wasn't able to get a screen shot of the svid quality yet.

Should hear from Bircoe in a few weeks about how my board works for his PAL console.

[Game-Tech.US]

Ok, that was old news, here some new info.

I've been working on an all SMT Batari board. I've had the layout done for a while, but wanted to be sure the SSOP16 4050 would work as good as the DIP16 IC we've been using before I have a board fabbed. I got one soldered to an adapter board and wired into one of my batari boards and hooked up to an atari. This time I used a 4 switch to if just maybe it was the JR causing the vertical bars i'm haveing trouble with.

Good news/Bad news: The smt 4050 works as well as the dip16 4050, but still has the bars.

 

 

You think the fms soic8 chips are hard to solder? Nothing compared to ssop16!

Notice the solder bridges? This was my first attempt, with no stencil its hard to judge how much to spread on with a toothpick.

A little solder wicking should fix this issue.

[Uzumaki]

Geez I hope you're not planning to make an assemble yourself kit with those tiny SMDs Those can be hard to solder on if you don't have experience with any SMD 9surface mounted devices, which includes the above tiny SOIC chip) and don't have a tiny pointed soldering iron. Give me wire wrapping vacuum tubes over any of these new fangled SMD any day.

 

SMD resistors and some caps however can make the board a lot smaller since those audio and video generally don't need much watt's going through the resistors. However I'd suggest making these tiny all SMD boards ready to install over assemble-it-yourself version.

[Game-Tech.US]

I don't have plans for the smt board either way yet, just sharing info that the ssop16 chip works the same as the DIP16 chip.

I don't even really want to assemble an smt board for myself until I get a toaster oven modded for reflow work and a solder paste stencil.

[bircoe]

Those can be hard to solder on if you don't have experience with any SMD 9surface mounted devices

 

In my past experience I found the best way to solder SMD is with solder paste (you lay it on the board with a syringe) and a SMD hot air solder rework station...

 

Bit pricey but the results look like they've been done by machines.

 

Something like this:

http://www.jaycar.com.au/productView.asp?ID=TS1574

 

They blow hot air rather than using an iron to transfer heat, works a treat for desoldering as well... heat it up, remove the device the wick up the left over solder.

[bircoe]

Duplicate post.

Edited April 28, 2011 by bircoe

[bircoe]

Should hear from Bircoe in a few weeks about how my board works for his PAL console.

 

I have a pair of Dell UltraSharp 2005FPW's as my desktop monitors which have SVideo input I can use for testing, will let ya know. The bits to finish my LHE boards should arrive tomorrow, so can hopefully have some results by the weekend.

Edited April 28, 2011 by bircoe

[Game-Tech.US]

I am using a hot air work station, but the syringe that came with my paste is huge and does not apply evenly or in small enough quantity to prevent similar problems.

If I ever want to do boards in quantity and get an oven modded, i'll most likely order a stencil from Pololu.

[bircoe]

I am using a hot air work station, but the syringe that came with my paste is huge and does not apply evenly or in small enough quantity to prevent similar problems.

If I ever want to do boards in quantity and get an oven modded, i'll most likely order a stencil from Pololu.

 

That's pretty groovy, if i did more SMD work that would definitely be on my buy list.

[bircoe]

akaviolence: the boards you sent me arrived today (so did my flashback 2), I'd be soldering the resistors on at the moment but I have a sick 12 month old girl in my arms...

 

I also scored a Nintendo 64 with Golden Eye for free today!

Edited May 4, 2011 by bircoe

[wildstar87]

I think it was a FMS6406. I have a half-built board that uses a ML6428 but have been a bit busy lately.

 

The 6400 is interchangeable with the 6428. The 6406 and 6410 are the same except for pin 2. I put a jumper on my board to allow either chip.

 

Eventually I do plan to make some boards available but for now I only have a few on hand. I will repost when that day comes.

 

I've been looking at this thread for the last few days, I'm looking at probably building this or the LHE mod. Would I be better off with an FMS6400, or FMS6406, or does it really matter at all?

[+batari]

I think it was a FMS6406. I have a half-built board that uses a ML6428 but have been a bit busy lately.

 

The 6400 is interchangeable with the 6428. The 6406 and 6410 are the same except for pin 2. I put a jumper on my board to allow either chip.

 

Eventually I do plan to make some boards available but for now I only have a few on hand. I will repost when that day comes.

 

I've been looking at this thread for the last few days, I'm looking at probably building this or the LHE mod. Would I be better off with an FMS6400, or FMS6406, or does it really matter at all?

I didn't see any difference between the 6400 and 6406 chips. However, I thought the ML6428 looked a little bit better, though this was on the breadboard. I still haven't finished with my 6428 on the actual board, but if I do, I'll post pics if it still looks better.

[bircoe]

Alrighty folks, Heres some pics from a PAL 4 Switch Vadar with a akaviolence mod: batari edition.

 

The below pics are taken with a completed (mostly... sans audio as my test monitors have no audio input) batari mod, the board was kindly supplied by akaviolence and well... the results speak for themselves really.

 

 

I have a pair of Dell UltraSharp 2005FPW's (20" Widescreen LCD's they're a bit old but they look good and I got them for free) on my desktop, each of them have a composite and svideo input, which came in really handy for doing these side by side pics, the left monitor was quite obviously set to svideo, while the right was composite, the batari mod (thanks to the FMS6400) is able to drive composite and svideo simultaneously).

 

Some games exhibit a slight blue bleed (Berzerk and Space Invaders), others look spot on, overall I think this is a great mod without going to extremes to get every last ounce of quality.

 

After weeks (possibly months) of messing around with different mods including my own designs I've come to the conclusion that composite video is just a massive compromise and should be ignored where possible, the results just aren't nice. Composite results in a washed out mess of an image while svideo is sharp clear and the colours look about right. The only time composite looks acceptable is when connected to a CRT TV, LCD/Plasma's just don't like Atari+Composite.

 

On to the in game pics...

 

Now if you'll excuse me I see a pair of Sega's that look like they could do with some svideo lovin...

Edited May 7, 2011 by bircoe

[keilbaca]

Is there a way to purchase a kit yet? I just don't have parts laying around to make one of these, and I'm sure the cost of me ordering just one set of these will be more expensive than some of you would charge for a set of components.

 

I also have various different consoles I can test it out with, including the Sears video arcade II.

 

If anyone would like to help out, let me know. I can solder very well, so even the SMT 4050 isn't a challenge for me.

[axh174]

Just wondering what the latest word is on this mod.

 

Did the ML6428 end up looking just as good on the printed board as it did on the bread board?

[supercat]

I just came upon this thread. One thing I've been pondering for awhile is whether it would be practical to build a 480P component video mod using a CPLD and a fast cheap micro to act as a scan doubler? The basic idea would be a board that would sit between the 6507 and the TIA, intercepting pins D1-D2, so that they would be passed through in all cases except when writing to COLUxx; on COLUPx writes, the board would latch the value of A0-A1 and D1-D7, and would force D1-D2 to match A1-A2. The board could then use the two lower luminance outputs to detect which of the four colors the TIA is outputting.

 

I don't think a cheap ARM would be quite fast enough to do everything, since it would be necessary to output 7 million dots per second. It may be possible for an FPGA to do everything, but CPLDs don't have RAM and adding a fast cheap micro would probably be cheaper than adding 160 bytes of RAM.

[+batari]

I just came upon this thread. One thing I've been pondering for awhile is whether it would be practical to build a 480P component video mod using a CPLD and a fast cheap micro to act as a scan doubler? The basic idea would be a board that would sit between the 6507 and the TIA, intercepting pins D1-D2, so that they would be passed through in all cases except when writing to COLUxx; on COLUPx writes, the board would latch the value of A0-A1 and D1-D7, and would force D1-D2 to match A1-A2. The board could then use the two lower luminance outputs to detect which of the four colors the TIA is outputting.

 

I don't think a cheap ARM would be quite fast enough to do everything, since it would be necessary to output 7 million dots per second. It may be possible for an FPGA to do everything, but CPLDs don't have RAM and adding a fast cheap micro would probably be cheaper than adding 160 bytes of RAM.

It is possible:

http://www.atariage.com/forums/topic/145950-call-for-beta-testers-cybertech-2600-vga-rgb-board/

 

I can't think of any microcontroller that would be fast enough. It looks like he used an FPGA and DAC.

[supercat]

I can't think of any microcontroller that would be fast enough. It looks like he used an FPGA and DAC.

 

How many cycles does the ARM require to read or write a port? Is it a 3-cycle LDR or two cycle STR?

 

My estimate would be that every four chroma clocks would require four port LDR (12 cycles), eight port STR (16 cycles), one memory LDRM of two registers (4 cycles), one memory STR (2 cycles), and nine shifts (9 cycles). A total of 12+16+4+2+9 = 43 cycles. I don't know if the internal PLL of the ARM could easily run at Chroma*12, but if my estimates are right an ARM like you use for Harmony would be plenty fast to do the data buffering if it didn't have to do anything else. Adding 8-to-16-bit translation would add an extra 4 memory LDR for color translation (+12 cycles) and add two registers to the buffer LDRM (+2 cycles). The extra cycles for the buffer STRM would be offset by eliminated shifts; the net would be 57 cycles every four chroma clocks (probably use chroma*16).

[+batari]

I can't think of any microcontroller that would be fast enough. It looks like he used an FPGA and DAC.

 

How many cycles does the ARM require to read or write a port? Is it a 3-cycle LDR or two cycle STR?

 

My estimate would be that every four chroma clocks would require four port LDR (12 cycles), eight port STR (16 cycles), one memory LDRM of two registers (4 cycles), one memory STR (2 cycles), and nine shifts (9 cycles). A total of 12+16+4+2+9 = 43 cycles. I don't know if the internal PLL of the ARM could easily run at Chroma*12, but if my estimates are right an ARM like you use for Harmony would be plenty fast to do the data buffering if it didn't have to do anything else. Adding 8-to-16-bit translation would add an extra 4 memory LDR for color translation (+12 cycles) and add two registers to the buffer LDRM (+2 cycles). The extra cycles for the buffer STRM would be offset by eliminated shifts; the net would be 57 cycles every four chroma clocks (probably use chroma*16).

The issue I think would be timing. It needs to read and write from ADC/DAC ports at the same time and (particularly with writing) do it at precise intervals, then in the middle of the screen it somehow needs to send blanking and HSYNC pulses to the output without missing a beat. I'd be surprised if it could be pulled off effectively with a single MCU.