Interesting.
AD1862 is a 20-bit DAC with +/-1mA full scale. So 1 LSB is ~2nA in current.
The Sparklos has single-ended BJT inputs with a bias current of ~5µA, in additional to back to back protection diodes.
No idea how low-leakage those diodes are, but probably well below 5µA, unless you start exposing them to temperature.
Then they can also jump to µA level very quickly.
https://sparkoslabs.com/wp-content/uploads/2018/03/SS3601_SS3602.pdf
How much is 5µA then in the AD1862 ?
About 11-bits.
I hope this input bias current is dead constant and does not change with signal.
The current mirror IV also has BJT input, not ?
Yes, BUT it has complementary BJT input pair with matched hfe, so that the base current largely cancels out.
AND those chosen BJTs has very good hfe linearity.
BUT, they also cannot beat JFET inputs for distortion.
You cannot change physics.
Patrick
AD1862 is a 20-bit DAC with +/-1mA full scale. So 1 LSB is ~2nA in current.
The Sparklos has single-ended BJT inputs with a bias current of ~5µA, in additional to back to back protection diodes.
No idea how low-leakage those diodes are, but probably well below 5µA, unless you start exposing them to temperature.
Then they can also jump to µA level very quickly.
https://sparkoslabs.com/wp-content/uploads/2018/03/SS3601_SS3602.pdf
How much is 5µA then in the AD1862 ?
About 11-bits.
I hope this input bias current is dead constant and does not change with signal.
The current mirror IV also has BJT input, not ?
Yes, BUT it has complementary BJT input pair with matched hfe, so that the base current largely cancels out.
AND those chosen BJTs has very good hfe linearity.
BUT, they also cannot beat JFET inputs for distortion.
You cannot change physics.
Patrick
dont calculate too much, i think many members here to buy a lot pcb from group buy or diy alot dac, but this ad1865 pcb, dont need any expensive parts, you can put any parts what you have, i tried tx2575, rhopoint on feedback 1k4 position and output resistor 100r, trial different opamp, but the funny thing is whatever you replace it, the sound is beautiful, now the opamp is ne5534, sound beautfiul........ i dont know why....... the magic sound easy to come out......
You should try the voltage output with it...it may sound even better, ahhaa...joking of course. Notice it means you like the Nichicon Muse at analog decoupling or it simply marry good with the rest of your hifi chain. Me having also an ad1865k I have not better result VS ad1862. Migth be a subjective preference as usual.
I would try it with an op1678... 5V rails.
@ Zarkstic : check the code color rings of the resistors story to see if you made not an error there.
Last edited:
Check LDOs - whether they are cold or hot. If they are cold as ice - mouser again.
Did someone ordered a good case for 1862 Dac?
I have the case Breeze hifi but he is small.
https://m.aliexpress.ru/item/1000004643167.html?sku_id=65088120641
Please write example.
I found another Dac on PCM1702u and I want to copy him.
http://pavouk.org/hw/audiosystem20/pcm1702dac.html
Does anyone know this project?
I have the case Breeze hifi but he is small.
https://m.aliexpress.ru/item/1000004643167.html?sku_id=65088120641
Please write example.
I found another Dac on PCM1702u and I want to copy him.
http://pavouk.org/hw/audiosystem20/pcm1702dac.html
Does anyone know this project?
A friend got it for me. From Farnell I think.@skrstic Where did you buy the LDO regulators?
It should not be the output capacitor, but try to remove the electrolyt from the output and measure the voltage only with foil capacitor. Otherwise the components seems be installed good.