yea pretty much, im just worried with grounding boxes that the difference people hear come atleast partially from this antenna effect...
Reading with interest. This is the kind of DIY I like to see - someone working on a solution to what is sometimes a real problem.
It would be good if you had a way to visualize the problem being remedied. When you connect two pieces of equipment together you could record the output with an audio interface or sound card and visualize the waveforms in a DAW program like Adobe Audition.
It would be good if you had a way to visualize the problem being remedied. When you connect two pieces of equipment together you could record the output with an audio interface or sound card and visualize the waveforms in a DAW program like Adobe Audition.
the best way is probably to measure some exact things, like how are the common mode chokes on the data lines performing... or what does the LP nature of a supercap do to the powerrail
tho im wondering since i have now the BLA interface if i could route audio from DAC to ADC and send some testsignals and show them, tho i never played around so far with measurements
the BLA isnt good performing measurement wise, around 10db higher noisefloor than other dacs/adc but it might be still interesting to look out for some minor changes, specially since the interface is usb powered, or to atleast verify that nothing got consistently worse
hmmm my first ADC measurement... tho i did soom roomacoustic measurements before...
but im unsure if i got things right, the results look different from amirs
this is -106db noisefloor , i set the generator to -5db and didnt had the volume knob turned all the way up, i just hooked up a TRS to XLR cable from output to input on black lion audio revolution 2x2 with zero gain on the input, so with generator volume and volume knob compensated we probably land at -115db
Do i have to set the generator to generate harmonics or something? or the noisefloor might be to high, i actually see a blip at 2khz
edit: instrument mode measurement was with filter too
Last edited:
and yea... basicly no difference... maybe under the rather high noisefloor but i think i can just verify that nothing got worse (above the noisefloor)
but it actually showed me that amir's measurement of the noisefloor looks far worse from a scale perspective than mine, this was one thing i was worried about but 100khz still sits at -70-80 db...
i actually think that a higher noisefloor might sound better if there is no "crap" sticking out
but it actually showed me that amir's measurement of the noisefloor looks far worse from a scale perspective than mine, this was one thing i was worried about but 100khz still sits at -70-80 db...
i actually think that a higher noisefloor might sound better if there is no "crap" sticking out
Last edited:
Did you make the recording with RTA? Also choose dBFS (or dBc/dBV) for the vertical axis.
yes
hmm where do i find the setting for it?
i think i just set the generator to 24bit dither instead of 16 bit, all other settings were basicly defaults
i dont know the noisefloor doesnt get shifted if i try to calibrate on the top of the RTA window... tho in the end it doesnt matter much, i just wanted to show differences, which are nonexistent, the db/spl values still seem applicable
im using linux, REW should open the interface with ALSA drivers
here i did a measurement with generator set to -4db and volume knob on max... this should give 8.224 dbu (since the dac output is able to produce 12.224dBu, where the max mic input at min gain is 8.6dbu
i calibrated the generator to 8.224 dBu (if this matters)
H3 is around -102db
Attachments
im pretty sure REW on linux is playing "around" pipewire and uses alsa directly but i could be wrong
im also not able to play sounds trough pipewire once the rew generator is running
well since its only a scale issue i dont think its that crucial to do some quick and dirty measurements
but i actually noticed too low levels too with room accoustic measurements with a Umik-1 since i use linux... i dont know what goes wrong here but since i didnt use the measurements "directly" but just for doublechecking where roommodes lie and what Q i need i didnt persuit to troubleshoot this crap...
i have a laptop, i might setup windows on it for it (i kinda completely banned windows...)
i played a bit more around changed my setup up so i could remove the 2m extension...
this setup
50cm usb A to usb B cable -> ground filter (ground passtrough) -> 50cm usb A to usb B cable -> ground filter (333ohm ground) with ext power supply -> 50cm usb A to usb B cable
doesnt work either, the ifi isilencer works in addition to one ground filter no problem, so my guess is that 4 common chokes are just too much, i actually think about staying with 2, this stills allows me to filter input and output seperately with one common choke and i dont risk degrading the usb signal "too much" which is definitely a thing with common mode chokes from what i read, even two common mode chokes sound already very good
if i switch both ground filters to ground passtrough it doesnt work either
and interesting thing... im not sure if i heared for the first time a ground loop but after playing around i forgot to switch the single ground filter to 333ohm again and it was immediatly audible as something was wrong, ground loop would be definitely a thing since i removed the topping hs01
what i heared was that with ground connected it kinda sounded irretating in the highs and overall a bit worse, i havent listen for long but its also possible to switch the ground passtrough dipswitch in a A/B comparison while the dac is running which is definitely interesting to test, clear win for ground over resistor imo
__________
i also read a bit about current limiting resistors on supercaps
10ohm, 2,5W resistor would be needed for 500mA of standard usb ports
Can someone help me out here, i dont quite seem to get wether the resistor consumes 2,5W itself and i would need 750mA to charge the supercap with 500mA or if it just lets 500mA trough and essentially heats up "because" of the 500mA... can someone clarify here?
if this works this would definitely be the easiest and possibly best route because completely passive and the resistor makes the DC filter property of the supercap even more effective because its a RC filter
i could also easly mod the current pcb to take an additional resistor, then i feel a bit safer using them
this setup
50cm usb A to usb B cable -> ground filter (ground passtrough) -> 50cm usb A to usb B cable -> ground filter (333ohm ground) with ext power supply -> 50cm usb A to usb B cable
doesnt work either, the ifi isilencer works in addition to one ground filter no problem, so my guess is that 4 common chokes are just too much, i actually think about staying with 2, this stills allows me to filter input and output seperately with one common choke and i dont risk degrading the usb signal "too much" which is definitely a thing with common mode chokes from what i read, even two common mode chokes sound already very good
if i switch both ground filters to ground passtrough it doesnt work either
and interesting thing... im not sure if i heared for the first time a ground loop but after playing around i forgot to switch the single ground filter to 333ohm again and it was immediatly audible as something was wrong, ground loop would be definitely a thing since i removed the topping hs01
what i heared was that with ground connected it kinda sounded irretating in the highs and overall a bit worse, i havent listen for long but its also possible to switch the ground passtrough dipswitch in a A/B comparison while the dac is running which is definitely interesting to test, clear win for ground over resistor imo
__________
i also read a bit about current limiting resistors on supercaps
10ohm, 2,5W resistor would be needed for 500mA of standard usb ports
Can someone help me out here, i dont quite seem to get wether the resistor consumes 2,5W itself and i would need 750mA to charge the supercap with 500mA or if it just lets 500mA trough and essentially heats up "because" of the 500mA... can someone clarify here?
if this works this would definitely be the easiest and possibly best route because completely passive and the resistor makes the DC filter property of the supercap even more effective because its a RC filter
i could also easly mod the current pcb to take an additional resistor, then i feel a bit safer using them