Is the 60 Hz in Fig. 2 with the 7 kHz hum from the powersupply? Terrible.... Or is that your signal?
I do not expect that proper boards and proper resistors will help this.
The design is not good I think.
Try the design of Dr. Frost before buying new stuff. It is just eliminating most components and adding one 4 a 6 Ohm resistor.
I do not expect that proper boards and proper resistors will help this.
The design is not good I think.
Try the design of Dr. Frost before buying new stuff. It is just eliminating most components and adding one 4 a 6 Ohm resistor.
fig 2 is rightmarks imd test,so i dont know what freq it uses . this was tested on laptop with laptop psu connected, with windows output adjusted to 6 percent and using cheap usb soundcard as input/output. 100ohms in paralell with input and pot in series for adjust level.
why would i steal other complete design and not try at all to optimise what i think should work? i dont care even if it dont end good.
i am actualy using 2200uf/rail couse i am cheap basterd so i use as low as i can get away with.
it was intended to use more as evident by bigger print on pcb(4700uf /rail) and as it is only one channel here i use half.
edit: sorry for mixing up different results ,
in resistor measurment it goes like this:
fr,imd,thd,noise.imdsweep,dynamic range.
why would i steal other complete design and not try at all to optimise what i think should work? i dont care even if it dont end good.
i am actualy using 2200uf/rail couse i am cheap basterd so i use as low as i can get away with.
it was intended to use more as evident by bigger print on pcb(4700uf /rail) and as it is only one channel here i use half.
edit: sorry for mixing up different results ,
in resistor measurment it goes like this:
fr,imd,thd,noise.imdsweep,dynamic range.
Last edited:
I don't know what I saw really but you are right Andrew, we talk about distortion levels of approx. 1%. The reason for this is the output stage but also a poor layout. I don't want to be negative but it's crucial to get the pcb layout right because you will have substantial voltage drops if we want to pick up uV in error voltages. 0.005% distortion requires a pcb layout in Spitzenklasse.
i belive second posted pcb would work much better as i try to make all traces as short,all traces are higher width , there is ground plane on both top and bottom layer and there is no vias, while there is lot of vias on original pcb i am measuring now,and also not thickest copper my board house allows.
i dont understand why you say this work in class b when there are bias diodes and resistors and this should work class ab with 6k8
i tested it today with 2ohm speaker on for few hours on almost full volume ,it got realy hot but it survived.
i dont understand why you say this work in class b when there are bias diodes and resistors and this should work class ab with 6k8
i tested it today with 2ohm speaker on for few hours on almost full volume ,it got realy hot but it survived.
there is little current it is litle bit conservative and as low as possible,but there was overlap in simulator. only i am not sure how good output models for spice from internet are.
can you suggest ,do i put multiturn pot in final board for easy adjustment and adjust by hand or simply put little lower value resistor?
can you suggest ,do i put multiturn pot in final board for easy adjustment and adjust by hand or simply put little lower value resistor?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.