Prop as a special signal generator for tube amp tests. Using C with CMM and fcache.
Christof Eb.
Posts: 1,201
This is just an experiment using the propeller as a special signal generator.
Idea:
Guitar players are rather happy with tube amplifiers. It is astonishing, that even with clean amp settings -no distortion- a tube amp sounds much better compared to a high quality transistor amp. So why?? One explanations lies in the much higher output impedance. Damping is much less. This allows the speaker cabinet to resonate like a room or just like an acoustic guitar body, which has been stolen from an electric guitar.
I wanted to measure this with an (Parallax USB) oscilloscope and needed a signal generator, that has to output a number of sine cycles and to make a pause then.
Implementation:
Propeller demo board.
One cog, that updates its counter A as a DAC to form a couple of sine waves followed by a pause. A second cog forms an ultra simple user interface to vary the frequency.
Propgcc, the new C-Compiler has two very interesting features:
1. The compact memory model CMM makes it possible to write larger programs. This one is actually very small but uses floating point maths (sine).
2. Fcache loads smaller loops into a fast cache of cog RAM. Here they can be executed like very fast assembler routines. You don't have to be an assembler wizzard to write very fast code. In this program the sine dac can be updated 325000 times every second!!
Results with a guitar amp (Tubemeister 18 combo):
At 145Hz you can see one additional half cycle at the output (blue) to the input (red, which is the output of the demoboard):
At 6155Hz you can see, that there are very vivid oscillations with about 22kHz:
The code uses math.lib
Have fun Christof
Idea:
Guitar players are rather happy with tube amplifiers. It is astonishing, that even with clean amp settings -no distortion- a tube amp sounds much better compared to a high quality transistor amp. So why?? One explanations lies in the much higher output impedance. Damping is much less. This allows the speaker cabinet to resonate like a room or just like an acoustic guitar body, which has been stolen from an electric guitar.
I wanted to measure this with an (Parallax USB) oscilloscope and needed a signal generator, that has to output a number of sine cycles and to make a pause then.
Implementation:
Propeller demo board.
One cog, that updates its counter A as a DAC to form a couple of sine waves followed by a pause. A second cog forms an ultra simple user interface to vary the frequency.
Propgcc, the new C-Compiler has two very interesting features:
1. The compact memory model CMM makes it possible to write larger programs. This one is actually very small but uses floating point maths (sine).
2. Fcache loads smaller loops into a fast cache of cog RAM. Here they can be executed like very fast assembler routines. You don't have to be an assembler wizzard to write very fast code. In this program the sine dac can be updated 325000 times every second!!
Results with a guitar amp (Tubemeister 18 combo):
At 145Hz you can see one additional half cycle at the output (blue) to the input (red, which is the output of the demoboard):
At 6155Hz you can see, that there are very vivid oscillations with about 22kHz:
The code uses math.lib
Have fun Christof
Comments
Interesting project. I haven't done any C programming on the prop yet (except once to try the compiler) but hopefully I can get to it this winter. Just last week I started to research and collect parts to build a tube amp for my Pedal Steel Guitar. I may not agree with your (or the theory) theory of why tubes sound better but I do agree that they do sound better and I will take time to go through the info you provided before I comment
Bob
now I had the chance/time to measure a very small ("bedroom amp") self-built tube amp called PP-micro connected to a small single speaker cabinet. It is easyly (I still have the ability to hear after the tests...) driven to limits where special tube amp effects can be seen in the graphics. The output power is only about 0,05Watts with clearly beginning distortion. Bigger amps will show these effects only when playing them very much louder.
1. "Sag", the first part of the burst is louder than the follwing parts. (444Hz)
2. Asymectric distortion. The positive half-waves are "rounder" than the negative ones.
3. Resonance of this speaker cabinet around 175Hz. It is interesting, that the speaker driven with 365Hz will resonate still at about 175Hz directly after the burst. If the input is 174Hz then the voltage will be enlarged after the first waves.
Good sounds for you,
Christof