Coupling capacitor dac.

Jump to Latest
#22
20 to 20 was no theory, just someone tossing out numbers. It is not based on either music or human hearing, let alone the environment. When you learn more about actual listening environment, you will understand my comment better.

It is often confused between amplifier BW and open loop slew rate. The only thing above about 18K is harmonic distortion.
 
#23
Huhuh, it was just tossed up 😀 It originally was 40/50 Hz to 12,5 kHz. The DIN stated that human hearing was from 20 Hz to 20 kHz in most optimal case. It all does not matter as people that listen to electronic music will like their set to be able to play 20 Hz. That is felt not heard 😀
 
Last edited:
#26
jean-paul said:
Make it simpler then. If you use 25 kOhm as load then Fc = 1(2 x pi x R x C) can be written as 1/(157050 x C). You want the outcome to be around 10 Hz. Now there is only 1 variable and that is the cap.... A V-cap ODAM of course 🙂 -> with a 2.2 µF cap this is 1/(157050 x 0.0000022). Fc = 1/0.34551 = 3 Hz. The formula uses resistance in Ohms and capacitance in Farad not µF. This makes the - 3dB point around 3 Hz. That is fine with modern source devices. However in some cap loving circles the mantra is that large caps are worse sounding than small caps (never say the terrible words that the best cap is no cap!). There is some truth in that at the extreme side of things like 100 µF film caps but even if it would be slightly suboptimal it still is way better than randomly chosen too small caps that indeed lead to low roll off.

If you are worried that a 2.2 µF V-cap ODAM sounds worse than a 1 µF version then recalculate the - 3 dB point with 1 µF. Simple. If it is still below/around 10 Hz it is also fine.

Trick/tip: if any of the devices one has is for example 10 kOhm then it makes sense to adjust the previous device (preamp) to a 10 kOhm load. Then you have guaranteed results with all devices regardless if they're 10 or 50 kOhm. Nothing worse than the tube world practice of needing to "marry" devices with eachother.
Click to expand...
Hi again, if I want 10Hz cutoff frequency then I get 0.68uF from calculator with 25K. Or accurate 9.36Hz, that was what you suggested right? Or am I missing something, I can try with that? Frank
 
Last edited:
#27
Yes but I would use 1 µF as that is a common value. BTW I "forgot" to mention this: since we calculate worst case we should take part tolerances also into account. Yeah I know, boring. A 1 µF with 10% tolerance can be 0.9 µF, the 680 nF cap 10% can be 612 nF. The old rule of thumb to play safe was to choose the next standard value in the E6 range. Better a bit too large (no penalties) than a bit too small.

ha lykke til!
 
Last edited:
#29
1 or 1.2 µF will both work OK. Coupling capacitors will "charge" at power on and they will do this via both the hopefully existent very high resistance output resistor and the following device. This causes severe plops in the following devices. This is what comes from asymmetric PSU's and coupling caps. There is an excellent invention called "relay muting" for that and then especially the version that only shorts outputs to GND at power on/off. If you are into asymmetric PSU devices chockfull of coupling caps you might want to look into that. Silence at power on/off for only a few Euro extra is a blessing not a luxury.
 
Last edited:
You must log in or register to reply here.
Top Bottom