Confirm max instruction counts for the modes

[dihm]

We should confirm (update?) what the actual maximum instruction number is for each mode.

Based on the logic used in the firmware already, we should have way more instructions than what Ethan documented, so we might need to do some empirical testing to see what is going on. Pretty sure the current instruction array size is set by the biggest number that still compiles. I suspect we hit stack space issues in practice.

[Json-To-String]

Current:

Table Mode	1 Channel	2 Channel	3 Channel	4 Channel
Single Stepping (Ext Timer)	16656	8615	5810	4383
Sweep Mode (Ext Timer)	8327	4234	2838	2135
Single Stepping (Int Timer)	5000	5000	5000	4032
Sweep Mode (Int Timer)	5000	3895	2611	1964

What I find:
The values in parentheses are the ones that actually sent all the triggers, whereas the non-parentheses values are the maximum that would fit in memory without complaining. The xs I haven't tried yet

Table Mode	1 Channel	2 Channel	3 Channel	4 Channel
Single Stepping (Ext Timer)	16383 (16352)	8473 (8464)	5714 (5710)	x
Sweep Mode (Ext Timer)	8472 (8464)	x	x	x
Single Stepping (Int Timer)	x	x	x	x
Sweep Mode (Int Timer)	x	x	x	x

[Json-To-String]

(misclick on the pin) I think this https://petewarden.com/2024/01/16/understanding-the-raspberry-pi-picos-memory-layout/ may cover some of the things we talked about earlier re: pico memory behavior. Putting it here for a good reference

[Json-To-String]

Updating the table:

Table Mode	1 Channel	2 Channel	3 Channel	4 Channel
Single Stepping (Ext Timer)	16383 (16352)	8473 (8464)	5714 (5710)	4311 (4311)
Sweep Mode (Ext Timer)	8472 (8464)	4311 (4302)	2890 (2880)	2174 (2169)
Single Stepping (Int Timer)	4998	4998	4998	3960
Sweep Mode (Int Timer)	4998	3960	2653	1995

After figuring out the problem (I'm dumb and wasn't including timing in the dtype like I should've been) internal timing was pretty straightforward with no discrepancies between what triggered and the max size that would fit.

Also helpful are the following from external timing:

Maximum values that fit in memory:

The amount of values that actually trigger:

The difference between the two (Max - Actual)

Note that we were modifying MAX_SIZE to be less than 245760 a bit today, but these values are found with that.