Can someone with a working Cadet Sync Generator please do me a favor?

#1

I’ve been spending the last few weeks troubleshooting a non-working Cadet Sync Generator I built, and it seems like the ATMEGA88A I received from LZX either wasn’t programmed or got corrupted somehow. I’ve confirmed that the ATMEGA88A is getting supplied with the correct voltage, but the outputs seem messed up. If someone with a multimeter could measure the pins on their ATMEGA88A, it would be an enormous help in confirming the problem.

if you look at the ATMEGA88A with the notch at the top, the top left pin is #1, and the numbers run counter-clockwise until you reach #28 at the top right. Pin # 8 is the ground. If you don’t have time to read all the pins, if you could just measure the voltages on pins #12, #13, and #15–that would still help a lot. Thanks!

0 Likes

#2

On my sync generator (rev 1.1) the voltages read roughly as follows (nothing plugged into the front panel, switch set to NTSC).

Left voltage Left pin Right pin Right voltage
5V 1 28 2.0V
5V 2 27 2.1V
Toggling 0-5V 3 26 1.95V
0V 4 25 0.1V
0V 5 24 0.1V
0V 6 23 5V
5V 7 22 0V
GND 8 21 0V
2.5V 9 20 5V
0V 10 19 0V
1.2V 11 18 0V
2.5V 12 17 0V
0.1V 13 16 0.44V
2.5V 14 15 0.44V
1 Like

#3

I don’t think measuring the voltages are going to be very helpful, except to check that it’s getting power on the right pin (which you already did). It’s outputs are only digital, so they’ll either be a high (~5V) signal or a low signal (~0V), but since this should be generating sync signals, they’ll be switching between high and low really fast (too fast for most general multimeters to keep up). Now, if you have access to a decent oscilloscope (so probably not one of the $20-ish ones from ebay/aliexpress - but maybe one of those will show a signal enough to know if it’s doing anything?), you could use that to see the signals that are output. Or you could probably use a logic analyzer for the same thing, again as long as it’s fast enough.

I’m trying to think of an easy way you can check to see if any signals are being generated without a scope/analyzer, but I’m not coming up with anything good. Some of the outputs should be audio range, so maybe hook an amp and speaker up to an output and see if you hear a tone? (And for the ones that are faster, see if your dog can hear it? :grin: )

1 Like

#4

Thanks @csboling and @joem! I was wondering about the usefulness of a multimeter, but i’m getting some different numbers. Pins #12, #13, and #15, are ATMEGA88_ODDEVEN, ATMEGA88_VSYNC_INV, and ATMEGA88_HSYNC_INV, respectively. I don’t get any signal off those pins, so I don’t think they’re outputting any voltage at all. Those are the pins that feed the LM6172’s, and output to the front panel H-sync, V-sync, and Frame sync outputs–which are all dead, in my case.

0 Likes

#5

It’s possible your multimeter just isn’t seeing anything there due to the frequencies.

For fun and reference, here are the types of signals on the different pins of the atmega88a in this particular circuit:

Pin # Pin Name Function in the Cadet I
1 PC6 programming header
2 PD0 +5V
3 PD1 no connection
4 PD2 input from LM1881
5 PD3 output
6 PD4 input from LM1881
7 Vcc +5V
8 GND ground
9 PB6 input from crystal
10 PB7 no connection
11 PD5 output
12 PD6 output
13 PD7 output
14 PB0 output
15 PB1 output
16 PB2 output
17 PB3 programming header
18 PB4 programming header
19 PB5 programming header
20 AVcc +5V
21 AREF no connection
22 GND ground
23 PC0 either +5V or ground depending on the NTSC/PAL switch
24 PC1 no connection
25 PC2 no connection
26 PC3 no connection
27 PC4 no connection
28 PC5 no connection

All the outputs should have some sort of video rate sync output, all the programming headers don’t do much (or anything?) when you’re not actually programming it, and the inputs from the LM1881 I think will only have signal if you’ve got something connected to the sync input (and the LM1881 is working).

1 Like

#6

Good point—so the measurement differences may just be artifacts of the different multimeters.

0 Likes

#7

I would expect the voltage reading to be roughly the average voltage, so when measuring a pulse wave output it should be more or less linearly related to the duty cycle. This seems consistent with ODDEVEN and ODDEVEN_INV reading as 2.5V for instance, but this is certainly a rough ballpark. In case the issue is that your firmware hasn’t been flashed, there were some tips on doing that in this thread a little while back.

1 Like

#8

Thanks for posting that link. Flashing the ATMEGA looks like a headache, but I guess it’s my next step. It seems like the ATMEGA only needs a few inputs to function:

  • 5V input (working fine)
  • GND connection
  • Clock input (this looks ok on my build)
  • Pin 23 pulled up to 5V for NTSC
  • Optional input from LM1881 for genlock
0 Likes