the official Channel F thread!

[tnt23]

1 hour ago, e5frog said:

Yes, it worked very well on the Channel F II, but not on the Luxor V.E.S. Just got a B/W and rolling image there. Will try bypassing chroma and add it in after. 

Schematics for Luxor differ in couple of moments, firstly Luxor appears to be PAL version and features dedicated delay line, and secondly, they have omitted LS07 gates after 74153 mux for CHROMA and BURST signals.

Luxor also swapped another LS07 gate in Sync circuit for open collector BJT, but this really should not matter.

[e5frog]

Yes, it's PAL, I'm guessing I haven't found a place with the proper level. In my earliest first attempts I isolated the signals from the RF modulator, adding in the chroma after the amplifier helped getting a color image then but the signal was weak and dark. 

[tnt23]

On 7/31/2019 at 10:05 AM, tnt23 said:

bit of tech whinning on THS7316 inside

Could've saved me some efforts if only I haven't overlooked a THS7314, same as THS7316 but for SDTV (with 8MHz filters) LOL!

[ubersaurus]

New Easter egg discovery! It’s Michael Glass at it again. https://selectbutton.net/t/i-found-an-easter-egg-from-1977/8828

 

I should note that with Spitfire having come out in April 1977 based on newspaper ads, it beats the previously known first video game easter egg (Atari's Star Ship 1 arcade game) by a few months!

Edited August 2, 2019 by ubersaurus

[+atari2600land]

How was this not noticed even when the game's code was dumped?

http://channelf.se/veswiki/index.php?title=Disassembly:Videocart_4

[e5frog]

Wow - how cool!!

 

Did such things on the Amiga a few times, have never considered it for Channel F data.

Did he do anything else we can look closer on? Maybe all roms should be re-examined. 

The ROM-disassembly in the Wiki has only be run through a disassembler and hasn't been checked or commented.

Shouldn't be too hard to find the right spot, now that we know where to look. 

 

Here's the last data in the rom - which means he had over 188 bytes free in the ROM before adding the Easter Egg.

 

000007D0   04 08 04 02 08 02 01 02  08 01 04 04 02 04 01 08   ................
000007E0   04 02 01 08 02 04 01 08  04 02 02 01 04 02 04 08   ................
000007F0   02 04 04 08 01 01 01 02  02 04 04 02 08 08 08 04   ................

 

Here's the button data:

%00000001 button 1 (1)

%00000010 button 2 (2)

%00000100 button 3 (4)

%00001000 button 4 (8)

 

Which turns into these button presses (as per description in the article): 343 242 124 133 231 432 142 314 322 132 342 334 111 223 324 443

Perhaps it's musical notes... 

I wonder if there's a time constraint...  probably.


EDIT:
I checked the disassembly and added some comments. 
http://channelf.se/veswiki/index.php?title=Disassembly:Videocart_4

Quick patch: It's a lot easier if you only have to press 1 (after first pressing 3)

04_press3then11111etc.bin

Edited August 3, 2019 by e5frog

[e5frog]

Has anyone managed to show the Easter Egg on a real machine?

I have tried it on a Channel F II, ITT Tele-Match and a Luxor V.E.S. when the last button is pressed nothing happens, no resets happens in the sequence so it should still be in the loop. 
 

[e5frog]

Past free edit time... 

So, I tried my patched version (press button 1 only) and that worked, you can just hold button 1 for about 33 seconds (on a 2MHz console). So question is if something works differently on the real cartridge. A second between keypresses should be a long enough wait IMHO. 
I just stuffed 32 copies in a row on an eprom hooked to a SMI. One official difference between the SMI and a PSU is that the former has 2 data counters, which could explain the use of H and Q in the code - address could just have been backed up in DC1. Maybe something happens when the PSU(s) data counter rolls over, maybe it doesn't turn to $1000 but $0000 instead...?

 

... if so, perhaps it works as it was meant to on the Multi-Cart...

 

 

[ubersaurus]

I wasn’t able to get it up on my system, it’ll just go into the game after so many presses. 

[e5frog]

Got it showing from the Multi-Cart, press a key, wait a second and repeat.

[e5frog]

Still no luck with an original cart, I'm not sure data counter in the PSU(s) rolls over to $1000, perhaps they just used the 12 bits that was needed.

 

I was following the example and noticed #15 seem to have a complete alphabet (4x5 font) but as far as I know it only spells "MEMORY-MATCH", 10 different letters? You can play the game with numbers or symbols, symbols version doesn't use letters. So... would they stuff the entire alphabet in there just for fun?

But then again, #18 is full of letters and it's not easy to spot anything in that one... there are so many ways to hide away Easter Eggs.  
Seems if Michael K Glass programmed them, his name is somewhere in it.  

 

 

 

[tnt23]

Lost my +5V rail again, don't know was it due to short circuit or overload. Not wanting to fiddle with another L7805 in TO-3 case (which are of limited availability locally, and rather pricey), decided to go for DCDC regulator instead.

Got a 5V/3A assembled kit for about $3.

 

That one came on aluminium PCB to be conveniently installed on heatsinks. Fits pretty well with some wires:

 

 

And stays cool after 40 minutes work, worth the hassle.

Edited August 5, 2019 by tnt23
typo

[ubersaurus]

16 hours ago, e5frog said:

Still no luck with an original cart, I'm not sure data counter in the PSU(s) rolls over to $1000, perhaps they just used the 12 bits that was needed.

 

I was following the example and noticed #15 seem to have a complete alphabet (4x5 font) but as far as I know it only spells "MEMORY-MATCH", 10 different letters? You can play the game with numbers or symbols, symbols version doesn't use letters. So... would they stuff the entire alphabet in there just for fun?

But then again, #18 is full of letters and it's not easy to spot anything in that one... there are so many ways to hide away Easter Eggs.  
Seems if Michael K Glass programmed them, his name is somewhere in it.  

 

 

 

Definitely post up if you find his name in any more programs! At this point the only games we know for sure he wrote are Spitfire and the Democart. His son has posted up on that original forum post announcing the discovery that his father passed away back in 2005 and since they weren't close he's not too sure what all he worked on.

 

Also, has anyone ever tried a TSR 1-2450 as a replacement for a 7805 in a channel f? Curious if that's a viable drop-in replacement.

Edited August 5, 2019 by ubersaurus

[tnt23]

On 8/5/2019 at 4:07 PM, ubersaurus said:

Also, has anyone ever tried a TSR 1-2450 as a replacement for a 7805 in a channel f? Curious if that's a viable drop-in replacement.

Nice and small one. Alas, according to the datasheet its maximum current is 1A, three times less than that of original LM7803 in TO-3 casing.

[RT-55J]

On 8/4/2019 at 1:15 PM, e5frog said:

Still no luck with an original cart, I'm not sure data counter in the PSU(s) rolls over to $1000, perhaps they just used the 12 bits that was needed

Hi. It's me, the fellow who found this thing.

 

So, to make sure I'm up to speed: the easter egg on an emulator, and on the original hardware with the multicart and with a flash cart, but apparently not on the original cart. Is that correct?

 

I know dumping Channel F games requires dedicated hardware, but I think someone with the original cart should test it and redump it just to make sure there isn't an undocumented revisional difference in the ROM causing the issue.

 

Truncating the bit-width of the data counter register definitely sounds like a plausible cost-cutting measure a semiconductor company would do (fewer transistors -> lower failure rate in manufacturing). IIRC, didn't Sean Riddle decap some Channel F PSUs? That could be one way of verifying if they actually did that.

 

(Note: I don't have any Channel F hardware myself.)

[e5frog]

Nice find, nice to see you're here. I have a dumper and 11 #4 carts/loose pcb:s. They all dump the same, same data that works when in an eprom/flashrom connected to a memory interface circuit. I can probably take the time and try all of the originals in case there's a difference in the chip construction. 

Sean have decapped a few chips, maybe it's possible to find the data counter and check the number of bits. There's perhaps no reason to have a counter that goes past the available memory size. The memory interface allows the entire 16-bit range. 

Doesn't really matter though, the code is there... and it was an exciting find. 

[RT-55J]

Testing the different carts to see if there's a hardware difference would be nice. I doubt they would have changed any of the logic in the PSU between different production runs of the cart if they kept the ROM the same. Still, it would be nice to know for sure. 

 

As far as examining Sean's decapping photos, it looks like all of them lead to 404s now. Bummer.

 

In any case, I'm glad that I was able to find the easter egg, and that Michael Glass Jr. ended up finding out about it at an opportune time in his life. I'm just slightly embarrassed to have shared unintentionally inaccurate information on the subject.

[RT-55J]

I wrote a post explaining the probable issue with the original cart that I'm going to add to the original topic I made on that other forum. If anyone has any complaints about it's accuracy or clarity, I'll wait before posting it there.

 

https://pastebin.com/v7swQL0i

 

Thanks.

[e5frog]

It's just a theory about the Data Counter but it seems the condition to jump to plotting the text doesn't happen on the real cartridge and it doesn't reset until the series is written in and a key after that is punched. If it has rolled over to $0000 instead it expects $70, which you can't feed it since buttons only affect the lower nibble. Upper nibble isn't even fully used - and even if you could insert all bit patterns you'd go through the entire BIOS and cartridge again without being able to get that roll over. 
As the Data Counter is inside the chip there's not really a way to check it to see what happens or manipulate it. 

I guess if he had checked the lower byte for $00 instead it would have solved such a problem but as a programmer it's quite likely he didn't know all the details about the hardware - or they changed it. 

Speculating further, Mr Glass probably put Easter Eggs in everything he wrote.  

Brad Reid-Selth wrote something about being vain in his youth, perhaps Mr Glass got over such "foolish" things later on in life. Personally I feel it's a nice thing to sign your work of art. 

@RT-55J I read the text and I'm not sure why it's embarrassing - because we haven't been able to trigger it from an original cart?
It's true the ROMC control signal was most likely a cost reducing measure. I suspect Fairchild already had a platform for 40 pin IC:s and wanted to use that so they divided their CPU system into several chips (you needed at least two at the start) with ROMC to sync the parts. Pretty clever IMHO. As far as I understand they made their CPU before the introduction of the Fairchild Video Entertainment System so I wouldn't say they used the method to save pins in the cart, they used it to save pins on the chips themselves, number of cart pins was just. 

Yes, the Easter Egg would most likely be accessible on a development cart, as these have "plain" ROMs and a memory interface and the memory interface has a Data Counter that covers $0000-$FFFF.

 

Yes indeed, the Easter Egg became free!

 

[+atari2600land]

I don't care how "rare" something is, a few slips of paper are not worth $70.

eBay #223604825181

[TheCoolDave]

1 hour ago, atari2600land said:

I don't care how "rare" something is, a few slips of paper are not worth $70.

eBay #223604825181

Yep, expensive in my eyes as well. As they say, something is only worth what someone is willing to pay.  No harm in them trying I guess...

 

Odds are it will be sitting for a while....

[Ashevent]

I paid $20 for a signed letter from Fairchild (pre-zircon) thanking some lady for her inquiry so I guess I can't say much but $70 for that does seem steep. 

 

I don't understand why such an obscure system that what..5 of us collect for?..lol. still holds such high prices for a few carts as well. I'm 3 away from the whole library and i'm saddened to realize i'll never finish the first consoles library I had as a child. I've seen high numbered carts on eBay for YEARS, way to high priced, never sell and...the vendors just keep them on at inflated prices.oh well hopefully one day. 

[Mikebloke]

2 hours ago, Ashevent said:

 

I don't understand why such an obscure system that what..5 of us collect for?..lol. still holds such high prices for a few carts as well. I'm 3 away from the whole library and i'm saddened to realize i'll never finish the first consoles library I had as a child. I've seen high numbered carts on eBay for YEARS, way to high priced, never sell and...the vendors just keep them on at inflated prices.oh well hopefully one day. 

Are we all looking for slot machine, poker and checkers???

[e5frog]

Seems to me they usually start way too high and then lower until they get sold - some merchants keep the same prices over and over again but I guess it doesn't cost them anything.

 

If we all offer around $10, I bet they'll lower the price... I'll offer $9 now.

 

 

EDIT: 
$9 offered.

BTW, no I've already got those carts, chances are boxes of them will pop up, like some that are still available on eBay at rocket high prices. #17 Zircon release, white box IIRC.

Edited August 8, 2019 by e5frog

[e5frog]

I succeeded in composite modding my Luxor VES, I think the birds-nest-prototype-pcb-thingy I made first was wired wrong, potentiometer wasn't working as it was supposed to. But the small ones I ordered from Aisler - no problems. Forgot to buy those expensive SMD potentiometers though, so this one has legs. 

 

Had a pretty good image with my proper "75 Ohm F-connector with adapters"-cable but waves in the image and distorsion on the white zeroes. Completely gone on the composite, nice clear image, like you took a cleaner and wiped all the interference away. It's of course not emulator sharp, it still has the normal composite video trades, checkerboard flicker along the sides etc.