sage Posted March 30, 2013 Share Posted March 30, 2013 (edited) For those who care. Remark: Do not expect a clean square wave on the Lynx. The signal is (even for low frequencies) a clear RC charging signal. For very high frequencies the signals are terribly distorted. (I will post pics later). Different colors denote different prescale factors for the timers. Everything was measured with full-volume. What was a bit surprising: There are some combinations of prescaler and reload value which result in no signal. Edited March 30, 2013 by sage 1 Quote Link to comment Share on other sites More sharing options...
sage Posted March 30, 2013 Author Share Posted March 30, 2013 Illegal Timer combinations: On my Lynx II, the combination prescale 0 backup 1 or 0 and prescale 1 backup 0 do (somehow) disable the timer. I guess, because the frequency is too high. When I wanted to fix this in emulation, I noticed that there is a big mistake in all emulators: (at least handy 0.95 and everything which depends on it). backup values of 0 disables the timer (for optimization reason) as well as a volume of 0. Which is complete nonsense. Even if the volume is zero (no output) the timer might still drive a linked timer. the argument for disableing the timer for backup zero is even more wrong. even if the timer runs out of the audible spectrum, the waveform may not (eg noise). Funny enough, in most cases this is reproducing the hardware behaviour (for prescale 0 and 1). I prepared a patch, if interested, send me a PM. Quote Link to comment Share on other sites More sharing options...
sage Posted March 30, 2013 Author Share Posted March 30, 2013 Small correction: looks like i had not square but integrate mode enabled, i dont think that makes a big diference as the volume was at maximum. nevertheless, i will repeat the measurement for real square waves. The diff might be that i switched between x and 127+x instead of -127 and +127, thus only the scale is doubled. Quote Link to comment Share on other sites More sharing options...
GadgetUK Posted March 30, 2013 Share Posted March 30, 2013 Very interested in the patch, presumably a patch for Handy? I've suspected disabling the timer we wrong for backup value zero, I think he originally did it for 'performance reasons' but that zero value is why 'Gates of Zendecon' sample didnt work. I basically put a fudge fix in place which whilst it allows the sample probably makes the emulation even more inaccurate but it sounds OK at least. Quote Link to comment Share on other sites More sharing options...
sage Posted April 1, 2013 Author Share Posted April 1, 2013 Just a small remakr to the freq response plot: From the docs, it should "Finally there is a one pole RC filter in hardware on the output at about 4500 Hz. Future stereo versions anticipate a bass boost circui". This explains the rise towards low frequencies. Might be interesting to see how the Lynx I spectrum looks like. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.