Does anyone know the thought process behind the HP 41 team's choice of the built-in tone frequencies? It has often seemed so odd to me.
Why not choose frequencies that were actual musical note frequencies for a bit more than an octave? Is the bender not able to drive those? Seems as if that would have made using them much "easier" for music.
Any ideas ?
(02-18-2015 01:09 PM)Gene Wrote: [ -> ]Does anyone know the thought process behind the HP 41 team's choice of the built-in tone frequencies? It has often seemed so odd to me.
Why not choose frequencies that were actual musical note frequencies for a bit more than an octave? Is the bender not able to drive those? Seems as if that would have made using them much "easier" for music.
Any ideas ?
FWIW, there are a few MCODE functions in the SVROM to play the musical note frquencies "DO, RE, MI, FA, SOL, LA, SI" (i.e. french for C - D- E- F- G- A- B-).
As to why they chose those TONES, my non-informed guess is they were easier to implement?
One saturday back in '86 I assigned the ten notes on the PPC rom to wrap around the right side of my 41 and took it over to my bud Jordan's place. I thought it would impress him, being a piano player and all. He touched all the keys once, thought a bit, and busted out with a rousing version of “Mary Had A Little Lamb”. I turned out to be the impressed one. He told me that he'd have to think of a better song for it because it sort-of approximated three notes per octave like Chinese classical. Jordan sure knew a lot. The PPC ROM "telephone book" didn't even say that.
The biggest problem is the fact that the high or low time of the signal driving the piezo element has to be a multiple of the instruction cycle time. This cycle time is nominally 155.6uS. So, for example TONE 9 has a three-instruction low and high time, giving a frequency of 1071Hz. TONE 8 has a four-instruction low and high time, giving a frequency of 803Hz. TONE 7 has a five-instruction low and high time, giving a frequency of 643Hz. These tones are individually coded. The remainder of the tones use a common routine to save code space. This common routine is 6+n instruction time long (for each phase of the piezo drive). And n is set by the TONE number as follows: TONE 6 has n=2, TONE 5 has n=4, and so on, down to TONE 0 with n=14. So, you could get better control at the low end of the frequency range, but it would take more code space. I guess that what they came up with was a reasonable compromise.
For 41CL users, try loading YFNX and then executing "41CL" Is what I was trying to do recognizable?
Monte