Atari 800 clock signals scope capture

[ijor]

Anyone has a high quality scope capture of the PHI2 and PHI0 clock signals on an 800 cpu board (not XL, not Sally) ?

 

[cwilbar]

@ijor, I don't know when I'd have the time, but I should have a non Sally 800, and I have a Rigol scope that can save to USB, so I should be able to capture this.  But I'm not sure when I'd have the time (not to mention needing to clean up bench space to set it up (benches are a disaster right now)).  But, if nobody else comes up with one, I can try to carve some time (and space ? ) out to do it sometime, but it might not be for a bit.

 

[ijor]

Hi cwibar,

 

Thanks a lot for the offer. Note that I don't really need a file capture. A good screenshot is good enough as long as you measure the distance from PHI0 to PHI2 (ideally from PHI0 raising edge, please). Please take the board level signals at ANTIC pins. Don't use the PHI0 and PHI2 signals at the CPU, which have the same name, but they are actually different signals.

 

Thanks a lot again, and I do understand it could take some time.

[reifsnyderb]

Phi2 on bottom, Phi1 on top. 

 

Signals taken from:

    Phi2 -- RAM Module Slot 3, Pin 14

    Phi1 -- CPU Board, Pin 22

 

 

 

Edit:  From Phi1, not Phi0

 

 

Edited June 26, 2022 by reifsnyderb

[reifsnyderb]

While I am at it:

 

R/W Early on top, Phi2 on bottom

 

    R/W Early from ROM Slot, Pin U

 

[reifsnyderb]

R/W Late from RAM Slot 3, Pin 13

 

R/W Late on top, Phi2 on bottom:

 

[reifsnyderb]

RASTIME on top, Phi2 on bottom

 

RASTIME from ROM slot pin V

 

[ijor]

41 minutes ago, reifsnyderb said:

Phi2 on bottom, Phi0 on top. 

 

Signals taken from:

    Phi2 -- RAM Module Slot 3, Pin 14

    Phi0 -- CPU Board, Pin 22

 

Thanks a lot. But that's PHI1, not PHI0. I need PHI0 from ANTIC (or from the 7474 input). Also, can you please zoom a little more so that it would be possible to measure the time distance between PHI0 and PHI2 more accurately?

 

Thanks!

Edited June 26, 2022 by ijor

[reifsnyderb]

2 minutes ago, ijor said:

 

Thanks a lot. But that's PHI1, not PHI0. I need PHI0 from ANTIC (or from the 7474 input). Also, can you please zoom a little more so that I would be possible to measure the time distance between PHI0 and PHI2 more accurately?

 

Thanks!

Oops.  Thanks for catching that.  I just edited it.  I'll snag Phi0 while I have the 800 apart.

[reifsnyderb]

Phi0 from Antic Pin 34.

 

Phi0 on top, Phi2 on bottom.

 

[reifsnyderb]

It looks like Phi2 follows Phi0 by 40 nanoseconds, give or take.

 

Out of curiosity, why do you want this reading?

 

 

Edited June 26, 2022 by reifsnyderb

[+DrVenkman]

I have a number of similar scope readings, but most (all?) of the ones I can recall making are from SALLY on one of my 1200XL’s when I got my then-new Siglent scope. Any particular reason to think SALLY and 6502B would vary in this regard?

[Rybags]

Sally has some of the otherwise external logic built in - maybe the propogation delays become much shorter?

[ijor]

5 hours ago, DrVenkman said:

I have a number of similar scope readings, but most (all?) of the ones I can recall making are from SALLY on one of my 1200XL’s when I got my then-new Siglent scope. Any particular reason to think SALLY and 6502B would vary in this regard?

 

This is not something related exactly to Sally vs 6502. The CPU doesn't generate PHI2 on the 800. If you look at the schematics, the PHI2 and PHI1 CPU pins are unconnected. These signals are generated externally at the board level on the 800.

[ijor]

18 hours ago, reifsnyderb said:

Phi0 from Antic Pin 34.

Phi0 on top, Phi2 on bottom.

 

Thanks a lot, I appreciate. Unfortunately the bandwidth of that scope is not good enough for an accurate measurement.

 

So I would still like a high quality scope of these signals. Again, must be on an 800 with a true 6502, not Sally. The signals should be ideally probed at the ANTIC pins. I need to measure the delay from PHI0 to PHI2 as accurately as possible. Probably a minimum bandwidth of 200 MHz is required, if the bandwidth is higher, even better.

[reifsnyderb]

2 minutes ago, ijor said:

 

Thanks a lot, I appreciate. Unfortunately the bandwidth of that scope is not good enough for an accurate measurement.

 

So I would still like a high quality scope of these signals. Again, must be on an 800 with a true 6502, not Sally. The signals should be ideally probed at the ANTIC pins. I need to measure the delay from PHI0 to PHI2 as accurately as possible. Probably a minimum bandwidth of 200 MHz is required, if the bandwidth is higher, even better.

What are you looking to do with this information?

[+warerat]

23 minutes ago, ijor said:

 

Thanks a lot, I appreciate. Unfortunately the bandwidth of that scope is not good enough for an accurate measurement.

 

So I would still like a high quality scope of these signals. Again, must be on an 800 with a true 6502, not Sally. The signals should be ideally probed at the ANTIC pins. I need to measure the delay from PHI0 to PHI2 as accurately as possible. Probably a minimum bandwidth of 200 MHz is required, if the bandwidth is higher, even better.

Green is PHI0, Yellow is PHI2 obtained directly from ANTIC pins.  Collected with my 500MHz, 2GSa/s Agilent scope.

 

T2-T1 (rising edge) ~22ns

G2-G1 (falling edge) ~26ns

[ijor]

11 minutes ago, warerat said:

Green is PHI0, Yellow is PHI2 obtained directly from ANTIC pins.  Collected with my 500MHz, 2GSa/s Agilent scope.

 

T2-T1 (rising edge) ~22ns

G2-G1 (falling edge) ~26ns

 

Fabulous @warerat Thanks a lot! As expected, significantly smaller delay than the one produced by Sally.

 

37 minutes ago, reifsnyderb said:

What are you looking to do with this information?

For the time being I am just performing a timing analysis of the clocking system. Can't say yet if it would be used for something specific.

[+DrVenkman]

59 minutes ago, ijor said:

 

This is not something related exactly to Sally vs 6502. The CPU doesn't generate PHI2 on the 800. If you look at the schematics, the PHI2 and PHI1 CPU pins are unconnected. These signals are generated externally at the board level on the 800.

Per this 1976 MOS datasheet for the 6502, it is still fully capable of generating Phi2 on pin 39 just like SALLY. I wonder why Atari would have chosen to generate it externally?

 

 

mos_6500_mpu_preliminary_may_1976.pdf

[reifsnyderb]

12 minutes ago, DrVenkman said:

Per this 1976 MOS datasheet for the 6502, it is still fully capable of generating Phi2 on pin 39 just like SALLY. I wonder why Atari would have chosen to generate it externally?

 

 

mos_6500_mpu_preliminary_may_1976.pdf 6.59 MB · 0 downloads

My theory is that Phi2 is generated externally due to the timing and use of the address buffers.  (Maybe someone else can confirm this.)

 

 

 

[ijor]

17 minutes ago, DrVenkman said:

Per this 1976 MOS datasheet for the 6502, it is still fully capable of generating Phi2 on pin 39 just like SALLY. I wonder why Atari would have chosen to generate it externally?

 

The 800 can't use the 6502 clocks because during DMA all CPU clocks are stopped.