Hi and thanks for a fun video again! It would be interesting to se the difference of the measurement results using a standard 100MHz probe as a comparison to the active one. /Tomas
I love this home brew stuff! Couple of comments, the original circuit has 2 caps in parallel to have the resonances spread over a wider frequency range and the signal is taken off a 47 ohm resistor which provides a better match to the 50 ohm coax. These are not criticisms just observations. Most 1 uF surface mount ceramic caps have already dipped and are the inductive side long before 1GHz. What did you use for the output cap?
I tried using multiple caps in parallel to feed the output signal but did not notice any considerable difference; I guess my layout was already adding far more inductance than the capacitors, so the change did not make a difference; regarding the low side resistor - the initial circuit was designed for 5V supply, but I used 9V and the static current trough the circuit is dependent on this low side resistor; at least in the simulation, I had to increase the resistor to reduce the static current to a point where I was not exceeding the datasheet recommendation. I'm not sure what do you mean by "output capacitor"?
Nice video FesZ, with plenty of brain food. There looked like there was a delay line in the schematic, is that right? Was that modelling your test lead length? I noticed a small phase difference between reference and probe output, and the output was lagging Was that due to inductance, or cable delay, or the complex response of this active 'filter'? (everything is a filter)
I added the transmission line to highlight any impedance mismatch related issues - to see if the response gets affected; but that did not seem to be the case. Part of the delay is coming from the cable - above 10MHz the signal after the cable starts lagging (this is visible in the AC simulation).
Considering the simplicity of the device, just by distance you can ensure low capacitance. And since it works below 1GHz, you should not need special PCB materials.
Indeed, there is no DC component; I drew it this way just so the signal does not overlap the Xaxis. I know its common to consider the intersection of the axis as the 0 point, but this point can be placed anywhere - I should have clearly marked the "0" on the Y axis where the signal starts.
Nice job encapsulating this task with motivation and objective, an affordable approach, and then satisfying observations and result!
Another excellent video. Thank you for the great info!
Hi and thanks for a fun video again! It would be interesting to se the difference of the measurement results using a standard 100MHz probe as a comparison to the active one. /Tomas
I agree that would be very interesting
nice explanation
Very good job that you are doing, thank you very much!
I love this home brew stuff! Couple of comments, the original circuit has 2 caps in parallel to have the resonances spread over a wider frequency range and the signal is taken off a 47 ohm resistor which provides a better match to the 50 ohm coax. These are not criticisms just observations. Most 1 uF surface mount ceramic caps have already dipped and are the inductive side long before 1GHz. What did you use for the output cap?
I tried using multiple caps in parallel to feed the output signal but did not notice any considerable difference; I guess my layout was already adding far more inductance than the capacitors, so the change did not make a difference; regarding the low side resistor - the initial circuit was designed for 5V supply, but I used 9V and the static current trough the circuit is dependent on this low side resistor; at least in the simulation, I had to increase the resistor to reduce the static current to a point where I was not exceeding the datasheet recommendation. I'm not sure what do you mean by "output capacitor"?
Nice video FesZ, with plenty of brain food. There looked like there was a delay line in the schematic, is that right? Was that modelling your test lead length? I noticed a small phase difference between reference and probe output, and the output was lagging Was that due to inductance, or cable delay, or the complex response of this active 'filter'? (everything is a filter)
I added the transmission line to highlight any impedance mismatch related issues - to see if the response gets affected; but that did not seem to be the case. Part of the delay is coming from the cable - above 10MHz the signal after the cable starts lagging (this is visible in the AC simulation).
Hi FesZ. How did you handle the stray capacitance in this PCB? Just by trace distance or would use PCB material other than FR-4? Any coating?
Considering the simplicity of the device, just by distance you can ensure low capacitance. And since it works below 1GHz, you should not need special PCB materials.
👍❤️
The AC plot (2:58) should obviously not contain a constant component.
Indeed, there is no DC component; I drew it this way just so the signal does not overlap the Xaxis. I know its common to consider the intersection of the axis as the 0 point, but this point can be placed anywhere - I should have clearly marked the "0" on the Y axis where the signal starts.
compared to other videos from other channels, your audio level is a bit too low :)