AlienRelics

Member Since: July 31, 2013

Country: United States

  • OK, so… using “Send us your feedback” and “What’s on your mind?” both bounce.

  • Hello? Is anyone paying attention to this page any more?

  • I’d really like to hear the noise. Is there a video clip of this pedal somewhere?

  • Incorrect. Nyquist/Shannon states that there should be NO appreciable (-60dB) energy content at ½ the sample frequency. So it is dependent on the quality of your filtering, which usually isn’t much better than a 3rd or 4th order filter.

    In this case, the lowest filter puts the -3dB point at about 4MHz.

  • Nothing you said refutes anything that I said.

  • Regarding the noise floor:

    I see a noisy digital circuit suspended millimeters over a sensitive ADC/DAC. Use the header pin lengths and lift the Teensy much higher, and place a grounded shield between them. Make sure it isn’t shorting anything out. Even better, use a socket for the Teensy, that lifts it up and will help keep a ground shield from touching any pins. Tape, copper foil, tape.

    Use shielded wire for the Input and Output. Twisted pair works, too.

    Put 0.1uF bypass capacitors on the LM7805.

    As much as possible, keep analog wiring away from digital circuits. Separate the digital and analog grounds as much as possible. I’d prefer seeing separate voltage regulators for digital and analog circuits.

    Make that input buffer into an amplifier with gain control. You can add a resistor and anti-parallel LEDs to the output of that Op Amp to ensure it doesn’t go above 3Vpp. Then you can adjust the gain to just under the point where the LEDs blink while playing. Get rid of that ridiculous attenuator/voltage divider of R3 and R4. You want the maximum signal without clipping going into the ADC.

    In fact, I see a microphone input on the Audio Board. Have you tried using that? I’d remove the 2.2k MICBIAS resistor and maybe put a larger capacitor (10uF?) across the 0.1uF capacitor going to pin 15 of the SGTL5000. A 500k audio taper volume control, and adjust to just under clipping. That removes one Op Amp from the circuit entirely.

    You should have antialias filtering on the input. Otherwise, any noise above 22kHz gets translated down into the audio band. Use that Op Amp for good - turn it into a dual Op Amp lowpass filter, -3dB about 12 to 16kHz.

    Don’t amplify the output, just buffer it. Why did you amplify it? A guitar amp is expecting from 100mV to 1.5V peak to peak, ie, guitar to line level. Not 9Vpp.

  • The analog bandwidth is not listed, not even in the Wiki. Based on an analysis of the circuit, the analog channels are 4.4MHz -3dB bandwidth. In addition, there are a lot of these scopes out there with an incorrectly specified TVS protection diode on the digital channels that limits their bandwidth to under 1MHz.

    Also, this is NOT 72Msps per channel, It is dual 36Msps ADC converters.

    Nyquist has nothing to say about bandwidth of a scope. Clearly, you cannot represent a sine wave with only two samples. It takes more like 8 or 10. That fits with a 4.4MHz bandwidth with 36Msps.

    In addition, square waves require at least the 5th harmonic to approximate, more like 9th harmonic if you want any reasonable representation of the rise time. One of the requirements of a square wave is rise time. The rise time of the scope must be several times faster (shorter) than the rise time of the square wave, or it affects how the actual rise time is displayed.

    There is another requirement with square waves: You are often using a scope on them to look for glitches, overshoot, and ringing. Those things will have a spectral bandwidth many times greater than that of the square wave. For that, the two digital channels are useless.

    I consider a 4.4MHz scope barely adequate for audio work and sub100kHz digital signals.

    The function generator is primitive. Sine, triangle, and sawtooth waveforms look like they are generated from only 3 or 4 bits of information, with big noisy steps between levels. They are stairstepped, is what I’m saying.

    Over 20 years an Electronics Technician, and an Electronics Engineering Technician.

  • Yes, because it is embarrassingly low. Based on an analysis of the circuit, the analog channels are 4.4MHz -3dB bandwidth. In addition, there are a lot of these scopes out there with an incorrectly specified TVS protection diode on the digital channels that limits their bandwidth to under 1MHz.

    Also, this is NOT 72Msps per channel, It is dual 36Msps ADC converters.

    Nyquist has nothing to say about bandwidth of a scope. Clearly, you cannot represent a sine wave with only two samples. It takes more like 8 or 10. That fits with a 4.4MHz bandwidth with 36Msps.

    In addition, square waves require at least the 5th harmonic to approximate, more like 9th harmonic if you want any reasonable representation of the rise time. One of the requirements of a square wave is rise time. The rise time of the scope must be several times faster (shorter) than the rise time of the square wave, or it affects how the actual rise time is displayed.

    There is another requirement with square waves: You are often using a scope on them to look for glitches, overshoot, and ringing. Those things will have a spectral bandwidth many times greater than that of the square wave. For that, the two digital channels are useless.

    I consider a 4.4MHz scope barely adequate for audio work and sub100kHz digital signals.

    The function generator is primitive. Sine, triangle, and sawtooth waveforms look like they are generated from only 3 or 4 bits of information, with big noisy steps between levels. They are stairstepped, is what I’m saying.

    Over 20 years an Electronics Technician, and an Electronics Engineering Technician.

  • They still require a resistor. If you are running them off of batteries without a resistor, then you are relying on the internal resistance of the batteries and the LED to limit current, a dangerous way to run LEDs.

    So they may work on a Lithium coin cell, or two AA batteries, but burn out on a Lithium Ion rechargeable or a couple of C cells.

  • I know this is old, but for those who may find this:

    The input offset voltage is amplified along with the signal, that is why you will see a large output with no input.

    A 741 has offset null terminals. Connect the outer terminals of a 10k trim potentiometer from pin 1 to pin 5. Connect the center slider terminal of the trim pot to the negative supply voltage, or to ground. You really need to use both positive and negative supply voltages with the 741, unless the signal you are amplifying is a couple of volts above ground already. The 741 was not designed to work well with a single supply.

    In any case, for amplifying a signal this small, you will really want to use either a newer Op Amp with input offset voltage in the range of nV, find one with offset adjust terminals, figure out how to add offset compensation externally, or use a chopper amplifier.

No public wish lists :(