Jawed
1000+ Head-Fier
- Joined
- Dec 7, 2014
- Posts
- 1,287
- Likes
- 739
@Rob Watts In DAVE the pulse array has 20 elements and therefore, as I understand it, an element can be switched on for between 0 and 20 clocks, and each higher-numbered element switches on for the same duration, but offset by one clock. e.g. this is showing three elements (out of 20), and each element switches on for 6 clocks, followed by being off for a further 14 clocks:
So all 20 elements are jointly outputting the code for "6", which is effectively -14/20 (-0.7) as a delta, since the pulse array is generating changes, and when less than half the clock cycles are switched on, the elements are signalling a negative delta. If each element was on for 14 clocks that would be a positive 0.7 delta.
Now, if we compare this with say Hugo TT2, which has 10 elements, we can see there's only 10 clocks maximum time for an element to be switched on and the delta is a multiple of one-tenths instead of one-twentieths. So the first implication is that there's more "roughness" in the deltas being produced by TT2, because there's less choice of deltas available.
But in both DAVE and TT2, the pulse array is clocked at 104.25MHz. So, over the same period of time as DAVE (20 clocks) TT2 can produce two coded outputs. For example TT2 can output "3" two times in a row:
In both cases, over the duration of 20 clocks, each element is on for 6 clocks. Now it might be preferable not to output the same code twice in a row, so maybe 2 followed by 4 or 4 followed by 2 would work too, depending on the overall gradient (short section of a curve) that's being constructed.
Is this right? What is the downside for TT2's pulse array arrangement? It seems to me that while the "roughness" in TT2 is greater than DAVE since there's half the codes available to use, because it runs "twice as fast" as DAVE, it can output twice as many codes in the same time as DAVE. In other words, the available codes in TT2, when considered over 20 cycles, match those produced by DAVE.
Code:
____|------|______________
_____|------|______________
______|------|______________So all 20 elements are jointly outputting the code for "6", which is effectively -14/20 (-0.7) as a delta, since the pulse array is generating changes, and when less than half the clock cycles are switched on, the elements are signalling a negative delta. If each element was on for 14 clocks that would be a positive 0.7 delta.
Now, if we compare this with say Hugo TT2, which has 10 elements, we can see there's only 10 clocks maximum time for an element to be switched on and the delta is a multiple of one-tenths instead of one-twentieths. So the first implication is that there's more "roughness" in the deltas being produced by TT2, because there's less choice of deltas available.
But in both DAVE and TT2, the pulse array is clocked at 104.25MHz. So, over the same period of time as DAVE (20 clocks) TT2 can produce two coded outputs. For example TT2 can output "3" two times in a row:
Code:
____|---|_______|---|_______
_____|---|_______|---|_______
______|---|_______|---|_______In both cases, over the duration of 20 clocks, each element is on for 6 clocks. Now it might be preferable not to output the same code twice in a row, so maybe 2 followed by 4 or 4 followed by 2 would work too, depending on the overall gradient (short section of a curve) that's being constructed.
Is this right? What is the downside for TT2's pulse array arrangement? It seems to me that while the "roughness" in TT2 is greater than DAVE since there's half the codes available to use, because it runs "twice as fast" as DAVE, it can output twice as many codes in the same time as DAVE. In other words, the available codes in TT2, when considered over 20 cycles, match those produced by DAVE.
, really?
