Decided to measure my Dual² Sapphire and its understandable. The high noise floor is coming from unshielded wires, in a plastic box without a faraday cage and its cable is running across every power cable under my desk.The 3KHz spike is coming from the soudcard so it can be ignored. LTSpice showed the same distortion profile..
@marin-cornel
Nice work. The only thing i don't understand what is the arduino doing there.
@marin-cornel
Nice work. The only thing i don't understand what is the arduino doing there.
That arduino takes care of the spectrum analyzer displayed on the 16x2 display.
A kind of vu-meter.
A kind of vu-meter.
Hi ,
So, I've completed my build with Standard BOM, checked the values of all the components and they measure same,
I don't know why the voltage across the R17-18 is around 500mv , Also the output is coming muffled (I've chosen 15db fixed gain)
Kindly help
So, I've completed my build with Standard BOM, checked the values of all the components and they measure same,
I don't know why the voltage across the R17-18 is around 500mv , Also the output is coming muffled (I've chosen 15db fixed gain)
Kindly help
Last edited:
Lets start with the basics.
You are missing 3 heatsinks.
Check the resistor values again, especially the bias and gain resistors
check the dc offset on the output.
Tidy build but as @kazuviking says " missing 3 heatsinks" and it appears "visually' that transistors are compromised by being placed on shared heatsink.
Explain ?
Explain ?
Now i got a really really weird issue after adding a SSR for the psu. If i plug my beyers into the amp it gives a square wave sound then the amps DC offset goes to 600mV and the proptection board kicks in, this is on repeat. This only happens IF i have both rca(regardless of source) plugged into the amp. If i plug my other headphone or only one rca regardless of side in nothing happens.
I removed the SSR and still happening. If i add 200 ohm resistors inline with my 32ohm headphone it doesn't happen. My beyer works without issues on the portable amp, phone and pc output so that is not the issue.
I'm currently at loss on what to do.
I removed the SSR and still happening. If i add 200 ohm resistors inline with my 32ohm headphone it doesn't happen. My beyer works without issues on the portable amp, phone and pc output so that is not the issue.
I'm currently at loss on what to do.
Found out that i shuffled q5-q7 with q6 q8 amp is working now but new problem
After rectifying that everything works fine , but after 5min random cracks and pops started to build on one channel (and the volume of noise was initially less and after 10min it nearly quadrupled)
After testing all small npn and pnp found
Q7 is dead as no readings on multimeter on any legs , changed that and amp is alive but again after 10mins pops n all started to appear again
Also i’ve checked that noise is not under the signal path , even after disconnecting the signal input , random static is there and they are very loud (not in a pattern, random)
I'm waiting for a pair of PCB's for this(42.b version).
I've ordered 200ea of BC327 and BC337, 50ea of BD139 and BD140.
I also ordered something I've wanted for some time, a Peak Atlas DCA75 Pro to match the transistors.
I don't know how closely they need to be matched?
Looking forward to this build 🙂
Edit:
My headphones are:
Beyer Dynamics DT 1990 Pro
Beyer Dynamics DT 990 Pro
Beyer Dynamics DT 770 Pro
Shure SRH940
HIFIMAN Sundara
HIFIMAN H400se
Philips Fidelio X2
AKG K702
I've ordered 200ea of BC327 and BC337, 50ea of BD139 and BD140.
I also ordered something I've wanted for some time, a Peak Atlas DCA75 Pro to match the transistors.
I don't know how closely they need to be matched?
Looking forward to this build 🙂
Edit:
My headphones are:
Beyer Dynamics DT 1990 Pro
Beyer Dynamics DT 990 Pro
Beyer Dynamics DT 770 Pro
Shure SRH940
HIFIMAN Sundara
HIFIMAN H400se
Philips Fidelio X2
AKG K702
I admit not having read the entire thread...
Should i match for anything else than the Hfe stated in the BOM?
Should i match for anything else than the Hfe stated in the BOM?
I'm guessing you'll have a more uniform spread if temperatures across the heatsinks, I've noticed a slight difference between mine but it's not much. Just a hunch...
Without matching the DC offset is gonna be a bit higher in one channel. For me one channel stays between +-12mV while the other channel +-48mV.
For me adjusting R6 nulled the DC offset to zero on both boards once they were up to operating temperature.
I gave up on my mistery issue. Built two more dual boards and it worked for a few days then the issue started to happen regardless. I even added buffer before and after the relay volume control but no use. Gave the 3 boards to my friends and done with it.
@arfadow For me the DC offset is never 0, even if i null it while its at operational temperature it drifts to +-5.4mV and +-9.7mV. Added a DC servo to deal with it and even with that the left channel is stuck at +-3.4mV while the other shows 0.00mV
@arfadow For me the DC offset is never 0, even if i null it while its at operational temperature it drifts to +-5.4mV and +-9.7mV. Added a DC servo to deal with it and even with that the left channel is stuck at +-3.4mV while the other shows 0.00mV
Yours looks to be a much more complex build with a single board combining both Sapphires, a different PSU setup and the addition of line amps? I wonder if your mystery issue and large DC offsets (+/-48mV) would improve if you experimented with two standard sapphire boards built as per original BOM. That would be quite an undertaking though....Apologies if I've misinterpreted anything.
At first i was using the original dual mono boards but switched to a combined board for size as the relay volume wouldn't fit inside the box(when i did the layout in cad and forgot that i can stack boards). The headphone driver board uses the same parts as the first dual mono ones and it have no issues. I combined two sapphine boards because i don't have good enough equipment to work on tight parts and the headphone drivers gain is 8dB while the preamp sapphire is 12dB. Swapping the bias resistors took more time than soldering an entire new board.
For peace of mind i desoldered the umblical cords ground an ran a super short direct connection to the psu, same thing.
How the board was wired and working for a month after i added the second sapphire board at first. For the second run i connected the ground coming from the volume controller but it did nothing. The layout is not the best and could be refined but there is no audible 50hz hum otherwise. Tts at -90dB from the measurement and i can only hear -42dB down if the gpu/cpu's fan going.
Do I really need two dual (0-12vac - 0-12vac) Transformers, or can I use (12-0-12vac) Transformers?
I have several of these in various VA ratings. See attached picture. Since I have several of them, it would be good if I didn't have to buy two new transformers.
Thanks
I have several of these in various VA ratings. See attached picture. Since I have several of them, it would be good if I didn't have to buy two new transformers.
Thanks
