SN76489 Hardware Question - Can't get this to work

[superjoebob]

Hey guys!
So far I've proven entirely inept when it comes to electronics. What I'm trying to do, is set up a simple test with my SN76489AN to get it to generate a tone. I tried hooking it up through a microcontroller, but I decided to step back and take a simpler route:



The buttons are hooked up to WE and CE on the chip, and the data pins are hooked up to the first 8 switches on the dip. READY is hooked up to an LED so I can visualize the state of the chip. I'm powering it with a 9-volt run through a regulator to bring it down to 5 volts. The problem I'm having is that I can't get the chip to do anything useful. I started by trying to get the ready light to come on in an un-clocked state by following the function table in the datasheet:


The chip next to the SN76489AN is an ATTiny85 running at 8mhz that I was using to give off a 4mhz square wave as a clock for the chip. Its output pin is currently disconnected, leaving the SN in an unclocked state.

I've tried every combination of holding/not holding buttons when I plug the chip in, and I've tried spamming buttons while it's plugged in but I just can't get the ready pin to light up. I've tried a few different chips to be sure it's not the chip, and I've tried basic data writing to the chip through a microcontroller with no results.

Am I doing something wrong with this?

[TmEE co.(TM)]

You need pull-ups on the switch lines, floating pin toggles betwees high asd low, not really being one or the other. The lines may need debouncing too to remove rapid toggle at momonts of button press/release.

But why it is not doing anything at all I am not sure why... clock you need in all cases.

[foobat]

Looks like you need to press the button on the right at 3.57mhz. You're probably not pressing it fast enough lol

edit: Aww it's upside down well that's a lot less funny now
I guess I can try to be helpful

absolute maximum rating for output current @5.5V is 2mA on pin 4 for the SN76489, is 2mA enough current to light that LED? (guessing no)

Maybe it lit it for a few nanoseconds and you didn't notice? And then your sn76489 promptly killed itself?

[superjoebob]

You need pull-ups on the switch lines, floating pin toggles betwees high asd low, not really being one or the other. The lines may need debouncing too to remove rapid toggle at momonts of button press/release.

Uhh, yeah, pull-ups, floating pins and debouncing! It's so obvious now how could I have missed it

Looks like you need to press the button on the right at 3.57mhz. You're probably not pressing it fast enough lol

edit: Aww it's upside down well that's a lot less funny now

Ahaha took me a while to figure out what you meant, I guess that button would be controlling the clock pin if I had it flipped the other way XD

absolute maximum rating for output current @5.5V is 2mA on pin 4 for the SN76489, is 2mA enough current to light that LED? (guessing no)

Hmm, never even thought of that. I figured in my ignorance that all the pins would output whatever voltage I put into the chip (if only). I'll bet that's it!

Since Tiido makes it sound like there's a lot more than meets the eye required to make my manual chip control method work, I guess I'll go back to trying to control it through my microcontroller. Would you guys mind giving me an idea of what I need to do with the chip to get a simple tone out of it? Reading the datasheet I have an idea, but I'm not completely sure. I think in code what I was doing is

CE = 1, WE = 1;
CE = 0, DATA = 10001110;
WE = 0; WE = 1;
DATA = 00001111;
WE = 0; WE = 1;
(thank you SMS Power for the data values)

After

CE = 1, WE = 1;
CE = 0, DATA = 10001110;
WE = 0;

I had code waiting for the ready pin to signal back on one of my inputs, but it seems it never did.

[Charles MacDonald]

READY is hooked up to an LED so I can visualize the state of the chip.

The READY pin is an open-collector output so it can only go low or tri-state. You'll need to add a 2K pull-up resistor to +5V if you want to see it toggle high and low. It can only sink 2mA and LEDs are in the 20mA range so you can't drive a LED with it.

The PSG needs a lot of time to accept data during a write, so it asserts READY for 32 clock cycles once it sees a write starting. This doesn't really indicate anything about the internal state of the chip, just that a write is in progress and the PSG needs more time to elapse. This happens for every write, so you'll see READY do the same thing each time.

Since you are using a MCU you can assert CE first, drive the data bus with the value you want to write, assert WE next, and then poll the READY pin state until it goes high. At this point 32 or more clock cycles have elapsed. Then negate WR and CE.

The PSG needs to be clocked at 4 MHz at all times for writes to work and for READY to show activity. Good luck!

[superjoebob]

Thanks dude. I tried hooking it all up again quickly but didn't have any luck, I think I'll give it a better try this weekend. For the pullup resistor, should I put it on the line between the MCU and the READY pin, or run the power through it?

[Chilly Willy]

Thanks dude. I tried hooking it all up again quickly but didn't have any luck, I think I'll give it a better try this weekend. For the pullup resistor, should I put it on the line between the MCU and the READY pin, or run the power through it?

PULL UP has a specific meaning in electronics - it means the resistor goes between the line and power (5V). By the same token, PULL DOWN means the resistor goes between the line and ground.