You haven't discovered the secret of putting slots in the board for them?
Then they don't roll away.
Never actually used them on a PCB I designed, just a board I assembled. Nice tip, but I'm far too lazy to create a special footprint with slots cut in it 😉. Plus I figured there would be less issues in pick-and-place if going the automated route with standard rectangular resistors.
A touch of flux will keep them from rolling around. I usually hold them with tweezers and tack solder one end, then solder the other and go back to the first and do a good job. I haven't had too many issues with the MELF components. Having colour codes on resistors is nice.
-Chris
-Chris
Too bad the Twin-T doesn't have a differential realization. Seems a bit unfortunate to measure gear with differential outputs with the notch limiting you to single ended, and then the ADCs all want differential inputs again.
If you consider a balanced source as 'signal + reference', and the tee input as 'signal + reference', would that not work? Also on output side.
That is how Bruno showed a single pot controlling a different signal in his G-word article.
Jan
That is how Bruno showed a single pot controlling a different signal in his G-word article.
Jan
Unfortunately, I don't have access to the instrumentation capable of such measurements.
I understand Samuel did some measurements so I'm relying on his design.
Regards,
Braca
This works on the output, but on the input you're going to still need to convert to single-ended in front of the filter unless you want to just ignore one of the outputs, no?
You can use two identical Twin T filters for the balanced signal. Then would be the symmetry and the similar performance.
And an even bigger tuning challenge but its the right solution. I'm not sure how best to wrap feedback around it if needed.
Have tested KEMET 200V 0.01uF 1206 COGs. Very good performance. Practically the same as the Wima FKP2 (in post #5199). Needs to compare these capacitors in better conditions 🙂.
Film capacitors also have lower distortion with higher voltage ratings. I suspect you may be getting to the point where the capacitor distortion is not the limiting factor.
The other possible improvement is to use four capacitors in series parallel to keep the value the same as a single. Then with an "H" layout you can minimize both vibration induced changes and lower the inductance a bit.
What tests did you perform for the comparison of the two types of capacitors?
Cheers,
Bob
If you can try the series parallel thing that would be an interesting result.
It was suggested to me that the variation of results from polypropylene testing is due in part to
the purity of the polymer. Is ceramic process easier to control?
It was suggested to me that the variation of results from polypropylene testing is due in part to
the purity of the polymer. Is ceramic process easier to control?
Last edited:
There are some possible phase cancellation issues that should at least be considered. Maybe nothing, but I never looked at how the harmonics add up doing this.
The last comparison was made by replacing all three Wien bridge capacitors (two of them in parallel) in my 1kHz oscillator board - I have one old semi-experimental board for these purposes. Oscillator output ~2,7V RMS.
Yes, would be interesting... But for the further experiments I need better conditions - needs to replace LME49720 with some better 🙂
A very simple experiment is to make your measurements with the card flat and then on edge. See if that changes the LF noise level.
What's better?
LT1468 maybe but the Vn is higher. However you won't get the low frequency junk.