Building BA-3 as analog output stage of noDac

R2 measures 47K both channels.

Junction of R1 and R2 8mV each channel.

Is the JFET's current close to 8ma? Using P3 I balanced the two JFets to have max. 6mA.

I don't know what's the wiper position of P1 & P2.

Are P1 and P2 shrting the C1 and C2 caps? I guess yes because I adjusted the mosfet bias to 1V.
 
I'd keep that R1 in place, especially with that low noDAc output impedance.

I am not sure why this is happening....
Why when connect the SG 1KHz sinuswave with 0,190Vrms when connect to BA-3 lower the voltage to 0,009Vrms?

The amp seems to be working correctly, as you stated here:

Seems BA-3 works ok, with 0,009Vrms gives 0,090Vrms.

I am thinking about this statement:

the SG 1KHz sinuswave with 0,190Vrms when connect to BA-3 lower the voltage to 0,009Vrms

How do you measure the voltage drop? Do you place the oscilloscope probe on SG output, and then connect the SG to the amp input, at which point you see the voltage drop on the oscilloscope screen?

On a side note, make sure the shunt regs can supply the required current... without collapsing.
 
Last edited:
The noDac it's an USB I2S output filtered with a 4th LPF, attached schematic, there isn't the output coupling capacitor that the value depend of input impedance that for BA-3 47K is 1,69uF but I installed 4,7uF that's the value that I have on hand..
 

Attachments

  • 02-05-2025 50 ohm characteristic impedance 4th order low pass filter.png
    02-05-2025 50 ohm characteristic impedance 4th order low pass filter.png
    6.2 KB · Views: 3
How do you measure the voltage drop? Do you place the oscilloscope probe on SG output, and then connect the SG to the amp input, at which point you see the voltage drop on the oscilloscope screen?
Measured with my true RMS voltimeter, yes probes direct to SG one measurement without nothing connected to set 1kHz 0,190Vrms, second measurement with BA-3 connected to Input + and GND probes connected also to to Input + and GND of BA-3 lowered to voltage to 0,009Vrms
On a side note, make sure the shunt regs can supply the required current... without collapsing.
Measured each reg is giving 200mA or a little more.
 
If you disconnect the BA-3 and play a sine wave through the JLSounds, do you get a sine wave at the 50R resistor? You would either need an oscilloscope for this or an RMS meter that can measure the frequency. 60 Hz is best. Trying to make sure that you are getting signal that you want from the JLSounds. If this works, then it is likely a question of loading.
 
Yes, otherwise didn't listen music. I listened several years only with Amanero USB-I2S + LPF + step-up 1:7 ratio without any problem but not reach the standard output of 2Vrms so when recordings are low level doesn't enjoy the same.
 
Last edited:
Hey looking for some more Jfets I have a couple of 2SK170V, one 11,66mA and the other 11,77mA so surely reach the 10mA with BA-3, also I have a couple of SJ74BL not so strong but can be good: one 10,73mA & the other 10,75mA, worth the change to increase the SQ?