I have a bigger problem, what makes the distortion voltage? The voltage across the wire is ~0.1% of the voltage in the loop (if the AP is 600 Ohms) which increases what you are looking at by 60dB. At higher impedances the distortion of the voltage across the wire as a percentage gets enormous. What you are looking for is just not well posed.
Hum.
According we were talking about power amps, what kind of noise ?
Ripple ?
Diode switching noise VS frequency leakage, and what about their frequencies ?
AC isolation (currents in ground connections, specially when asymetrical) ?
If you cannot measure differences in distortion neither hear any noise out of the amps in absence of signal, what is the point ?
That is a good question! The distortion is as mentioned -110 or 50 dB below that which would leave the mechanism at .01% into a 10k load.
Do you have any thin magnet wire to try it yourself?
I'm thinking of it being used clamped between two metal plates with say a cm or two of surface area. A ring of resistive wire may work, but nichrome isn't resistive enough, and I don't know about tin or steel.
I'm thinking about velostat:
Pressure-Sensitive Conductive Sheet (Velostat/Linqstat) ID: 1361 - $3.95 : Adafruit Industries, Unique & fun DIY electronics and kits
But I'm not sure how to calculate the resistance. It would be ideal if I could adjust to the resistance I needed just by increasing pressure, but I'm not sure how to calculate the resistance I need and how pressure would affect it. What I want to know is how much surface area would I need to get 5R between two plates, and could I adjust this within the right range using pressure?
I have a spool of #30 and a drill and jn's current sensing resistor, I'll see if I can try this weekend I have an idea to improve the sensitivity. Let's keep this fun.
Still no response on my thought about two ways of twisting? Imagine the insulation having an identifying stripe running along it, which always has the same orientation along its length, ie., doesn't spiral around the conductor: in twist 1 the stripe always stays, say, on the outside of the twisted pair, is always visible - this is what will nominally happen using a drill; in twist 2 the stripe spirals to the inside and outside of centre of the twisted pair, but from a distance the stripe constantly faces a certain direction.
The 'stress' on the cable is different in the two cases - does this make a difference, in any audible area?
The 'stress' on the cable is different in the two cases - does this make a difference, in any audible area?
Last edited:
If you put a compliant conductive layer between the resistive sheet and the electrodes and the resistive sheet is essentially incompressible (i.e., hard plastic, not rubber), you can minimize the resistance change with pressure. Otherwise you'll have a surface area effect and a volumetric effect.
To calculate resistance, you need the resistivity in ohm-cm. Multiply by thickness, divide by area.
Here, some measurements L.C did with the Connex that Scott talked about.
http://www.diyaudio.com/forums/vend...e-mosfet-amplifier-module-40.html#post3882608
L.C First One (A simple current feedback amp, and, as it, with less PSRR than a long tailed pair) has something like < 0,0016 % distortion (at 1W) and >110dB signal noise with Hypex Smps.
Often, commercial amps with SMPS are cheap, may-be not a good idea to generalize
It would be interesting John, if you can compare your listening impressions with your favorite amp with an Hypex SMPS VS Linear...
Last edited:
Bruce Hofer has developed a measurement filter for class D (and by implication SMPS) to avoid this kind of issues. It should be on the AP website. They sell them at AP prices but should be possible to home-brew.
Jan
Yeah, right .. if the equipment can't meet the specs, change the test conditions
You know this equipment? Can it not meet its specs? What spec are you referring to? I'm in the market for something like that so am interested.
Jan
- Status
- Not open for further replies.