2600 cartridge port - how much power can you draw?

[rolandg]

Hi,

 

I'm half way through building an eeprom cart for my 2600 (one of the later slim black ones) and I was going to use a Xicor x28c256 32kx8bit EEPROM so I could hold several images.

 

However (after ordering) I notice that despite being listed as 'low power' these devices can draw up to 60ma at 5Mhz! Now I know its not running that fast and the typical value listed is 30ma - but this is still I understand rather more that the PROM or EPROM usually used.

 

Anybody care to comment on whether this is too much? Assuming the whole circuit takes 30-40ma - am I drawing too much current?

 

Thanks

 

Roland.

[rolandg]

I did a bit more testing tonight. I wired up a power extension cable (3.5mm socket to jack) that let me hook my multimeter into the 9v line (its a 2600 jr type). Results were interesting and a little surprising.

 

No cartridge - power drawn was 240ma.

With a cartridge inserted power draw varied from 260ma to 300ma. COMBAT was the most power hungry and the Atari 32 in 1 was one of the least power hungry. I expected the opposite.

TV on/off - didn't seem to make any difference.

 

This is by no means conclusive as it still doesn't tell me anything about the cartridge port itself. However as the PSU is rated at 500ma at 9V - so that bit at least will cope.

 

I'm still hoping the x28c256 will draw considerably less than 30ma (at 5v) due to running at only 1Mhz instead of 5Mhz but I will just have to wait and see.

 

Roland.

Edited February 21, 2013 by rolandg

[DanBoris]

I did a bit more testing tonight. I wired up a power extension cable (3.5mm socket to jack) that let me hook my multimeter into the 9v line (its a 2600 jr type). Results were interesting and a little surprising.

 

No cartridge - power drawn was 240ma.

With a cartridge inserted power draw varied from 260ma to 300ma. COMBAT was the most power hungry and the Atari 32 in 1 was one of the least power hungry. I expected the opposite.

TV on/off - didn't seem to make any difference.

 

This is by no means conclusive as it still doesn't tell me anything about the cartridge port itself. However as the PSU is rated at 500ma at 9V - so that bit at least will cope.

 

I'm still hoping the x28c256 will draw considerably less than 30ma (at 5v) due to running at only 1Mhz instead of 5Mhz but I will just have to wait and see.

 

Roland.

 

The difference between Combat and 32 in 1 is not surprising since Combat was an early cartridge and the 32 in 1 was much later. Chips generally get more power efficient over time.

 

Besides the PSU the other thing you need to look at is how much current the voltage regulator in the system can handle.

[rolandg]

Hi,

 

Thanks for the comment :-)

 

I agree about the differnce between Combat and 32 in 1. I had assummed the big diffrence in capacity (2K vs 64K) - would make more of a difference to the power consumption - but obviously not.

 

Good point on the voltage regulator too. Based on the circuit diagram the 2600 jr uses a 7805:

 

http://www.atariage....Junior_PAL.html

 

These I believe are good for 1000ma (with a heatsink) - although I can't currently check the power draw on this - it should cope.

 

I also had a look at a few pictures of circult boards (I've not opened my 2600 jr up yet):

 

http://www.the-liber...-1983-rev-f.htm

(note - this link seems to bounce you back to the front page of the site - but the 2600 jr pages are easy enough to find)

 

These seem to agree with the circuit diagram to the extent that a 7805 is used and it is fed directly to the cartridge port through PCB tracks which are significantly wider than the data ones.

 

On my new board I will be able to check the power draw from the cartridge port. I can't yet do that for ordinary cartridges - though my EEPROM programer (based on an Arduino Mega 2560) lets me read original cartridges and I could check the relative power consumption from there. (only relative as the Mega 2560 won't be reading at full speed). Maybe I'll build a cart extender :-)

 

I also now realise that the 9v input supply is used in a video circuitry - so presumably some of the power draw when I insert a cart is due to that and not the cart itself.

 

Anyway - my EEPROM board is now nearly completely wired - though I have to connect it up to the original cartridge PCB - and the biggest problem at the moment is getting it all to fit back into the cart case! :-)

 

Thanks

 

Roland.

Edited February 24, 2013 by rolandg

[rolandg]

The design I'm using has a 74LS04 on the cart board and another one on the programmer board. The cart one is to invert the chip enable (CE) and the programmer board one is for safety to invert the Output Enable (OE) and Write Enable (WE) - both active low. When the Arduino starts up its pins may appear low which unless inverted and causes an accidental write to the EEPROM. I'm basically following the design at:

 

http://z80dave.blogs...programmer.html

 

but with the Arduino connected to a fake 2600 cart port so I can hopefully reprogram the cart without having to take it to pieces.

 

I know the fake cart port works as I can dump an original cart and get the image to work on Stella. However the real reason I'm trying all this is to be able to code and run my own Batari basic games on a real Atari. Running them under emulation is great - but I won't be satisfied until I've run them for real (just need to stop messing with electronics and get back to coding).

 

I only have one 74LS04 so I'm using a 74LS00 also - and tonight I wired them up to check if they still worked (probably date to the 1980s) and see how much power they draw.

 

Anyway - under a simple test (not at speed) they draw around 4ma each. A little more than expected - but to be honest I'm just glad they worked :-)

 

A bit more testing and I'll risk an EEPROM chip

 

Roland.

Edited March 2, 2013 by rolandg

[rolandg]

Well the good news is that it works! :-)

 

I found one dry joint during testing and I had to add a resistor to pull the WE line low as it went into the 74LS00 (used as an inverter). This is so that the 74LS00 would always deliver a high signal to the active low WE pin on the EEPROM to stop accidental writes when the Arduino was reset. I also had a fair bit of trouble getting the image correctly onto the EEPROM. In the end I used the C++ rom sending program found at:

 

http://arduino.cc/fo...?topic=129567.0

 

I also changed the Arduino script so that it set the address and data lines up before enabling the write. This seems to work more reliably - though I have to admit I've only tried one image so far and the software may need more work.

 

As I wanted to try a known good image - I used a PAL copy of COMBAT which I had extracted off a cartridge - as I knew for sure this would work on the 2600.

 

I rigged up the multimeter again and it shows about 240ma to 250ma current draw for the whole 2600 - even when playing the game. So it seems my initial fears about the EEPROM being power hungary were unfounded.

 

 

Here is the programmer rig based around an Arduino Mega2560 (actually a Sainsmart clone - which I can recommend).

 

 

Here is the cart itself. Its built to go back into the case. The 8pin socket is for a 12F675 PIC chip to select the images. Currently I have a jumper in the socket which just grounds the 3 output pins connected to the high address lines on the EEPROM. The 3.5mm socket on the wires carries the active low WE and OE lines from the Arduino to the cart as these are only used for programming. In normal use they are just pulled up and down respectively by resistors.

 

 

:-)

 

Roland.

Edited March 9, 2013 by rolandg

[keoni29]

Very interesting. I made something similar and it looks almost the same as yours exept I used Sram and a battery backup instead of an eeprom. The eeproms I got are pin compatible, so I could swap it out for an eeprom if I wanted to.

Can you fit this back in the original case? If so: would you like a label for it? I designed mine a while back (when I was using an eprom) and perhaps with some modifications it can make your eeprom cartridge look cool too!

Edited March 12, 2013 by keoni29

[rolandg]

Hi, thanks for the post - much appreciated. I did see an SRAM cart board in my search for information - maybe it was yours? Anyway - I liked the idea and did consider it - however as I would still had used an Arduino to load it - once I'd found a suitable EEPROM chip - that seemed a simpler way to go. I've done various very small electronics projects over the years - but this is my first attempt at interfacing a memory chip.

 

I'm going to shelve the idea of using a PIC chip for the moment as I'm not sure it will power up and set the address lines fast enough. If the 2600 had any sort of loader or bios that ran before the cartridge was accessed - I wouldn't worry - but its going to try to access the cart immediately and I'm not sure this would work. Also the PIC can only change the selected image while powered up - but there is no obvious way of resetting the 2600 to pick this up - other than power cycling it. That just leads back to the first problem of is the PIC fast enough (4Mhz clock - 1MiP). Whats not a problem is the PIC remembering the image selected as it has its own built in EEPROM.

 

Anyway - for the moment I'm going to use a 3 (actually 4) block of DIL switches (like many other people who have built these) - so I can select each of the 8 images (32k EEPROM - 4K images). I think there should be a way to use the spare 4th switch to split the 4K bank down so that I can pack two 2K images into 1 bank - and still have the ability to use 4K images. Eg switch off - A11 (if I remember correctly) under control of the 2600. Switch on - A11 set to high to force the use of the top half of the bank. Its worth a try.

 

Having said all that - the board is designed to fit back into the case. I've had to remove much of the internal plastic supports and will have to drill a hole for the 3.5mm panel mount socket I will use (on order). The DIL switch will fit externally on the end.

 

Thanks also very much for the cart label :-) I'll certainly need one - and that looks like a great start for me!

 

:-)

 

Roland.

[rolandg]

Well its done! Rewired for dip switches and squeezed back into the cartridge case. I had to remove most of the internal plastic so the stripboard is held in place just by the mass of wires and the original board is held down by a stack of double sided foam pads between it and the back of the case! There is also no longer a screw holding the 2 sides together. I'm just relying on the side clips. Its hardly high tech - but it works.

 

Thanks again Keoni for the sample label. I used the controller text from it for mine - just what I needed. In the end I quickly knocked up a label using

 

http://www.labelmaker2600.com/

 

I didn't have a suitable picture to hand - so I just used a chunk of the datasheet for the EEPROM. Its a bit simple looking - but does make the cart look 'finished'. The label is strictly speaking upside down - but as the cart won't sit the other way around due to the switches - it works for me.

 

I've still got a little work to do on the programmer board. Currently I have to reload the Arduino with a reader or writer script - depending on what I'm doing. As the writer script can't verify - I keep swapping between the 2. I just need to wire up one of the i/o pins to a switch which I can check within the script to do the right thing. Same sort of process as the 2600 uses for its switches.

 

Also I had 4 LEDs on a breadboard which I was using to track the download. they were only meant to be for debug - but they have proved very useful - so I've built a little stripboard version which can sit permanmently on the Arduino.

 

I probably won't do any more current tests - as I'm happy the cart is low powered. However one final note is that when active on the Arduino it draws around 10-15ma.

 

Hope that helps anybody else working with EEPROMs. A few pictures:

 

 

 

 

 

 

Roland.