Problems with using 5V->3.3V translator in SA1 cart

Hi there, here I have the following problem- I want to use a M29W640 flash memory (64MBit 16bit wide memory) in an SA1 cart. To achieve that, I used a GTL2000 and designed a small adapter board for that. Unfortunately it doesn't work... I tried to check the connections and everything seemed fine. I also did a little test- I connected one of the addresses (A0) to GND and to 5V after that and it behaves in a strange way- while I connected one of the GTL2000 inputs to GND, the corresponding output gave something around 0.80V-1.80V and it changed all the time and when I connected the input to 5V, the output voltage was something like 2.0V-2.5V and it changed as well. The voltage on the board was 5.0V and it was stable and the voltage regulator gave 3.28V and it was stable as well. Here's the schematic of the connections on my board:



Datasheet: http://www.nxp.com/documents/data_sheet/GTL2000.pdf

Is there anything wrong?? Maybe the pullup resistor is too high or low?? Is there anything else that might be wrong??

Also, I've found this chip in my drawer as well: SN74TVC16222A. Will that work better instead?? Please for reply. Any kind of help or advices will be strongly appreciated. Many thanks in advance!!

There is nothing obviously wrong with your schematic.

That just sounds incorrect. Try instead measuring resistance across a source-drain pair? Maybe you somehow damaged it?

These voltage clamps should conduct as long as the source or drain pins are at a lower voltage than the Gref pin. Sometimes they do something a little odd if the higher side is pulling up and the lower side has no load at all but that's not what you're seeing.


FWIW, I was given a board identical to what you're describing you're making (SA-1, GTL2000) and it works fine in my SNES, so I'm not exactly certain what's going on.

I measured the resistance of the input and output (source-drain of one of the tested channels) and it displayed 99.0 value while I switched the ohmmeter to "200k" (the chip was off during the measuring). I don't remember how to calculate it though

Should I try a different GTL2000 and measure the voltage of input-output and measure resistance again??

That unfortunately doesn't really indicate anything; the resistance should be quite high if Gref isn't approximately at least a volt higher than at least one of the S or D pins you're measuring. (This is actually how its overvoltage protection works)

("99.0" when your ohmmeter is set to "200k" means "99.0k")

I tried measuring the voltage on another GTL chip and the results are a little better now but still not as it supposed to be I think. When I connect an input line to GND then the corresponding output gives something around 0.02V which is OK but when I connect the input to 5V then the output gives 2V (2.02V to be exact) but stays just like that. Maybe the pull up resistor is too high or maybe caps are bad?? What resistor is on your board??

EDIT: I've just changed the pull up resistor to 100k for the test purpose and the results start to get better: when I connect input to GND then the output throws 0.05V and when I connect input to 5V then the output gives something like 2.35V. So, was the pull up resistor actually too high?? What should I use then?? 10k, 20k, 50k or a lower/higher one??

P.S. I'm temporarily testing the board using Arduino Uno powered by USB cable right now and I connect the board to Arduino 5V and GND and the results are as follows. I'm not exactly sure how it will act on SNES though

EDIT 2: ahhh wrong wrong wrong. My device wasn't quite woring correctly because I realized that the pull up resistor wasn't actually connected to anything.... When I temporarily fixed that and measured the output while input was GND and 5V, the output gave 0.00V and 3.56V respectively and it was stable. I used 200k pull up again. So I guess it works good now?? I will try soldering it in SA1 cart in place and see what happens this time