the electrostatic shield/screen wires should each be short and each be connected to the chassis where they come out of the transformer insulation. These sheild wires pass interference from the mains to the chassis. To do that at high frequencies they MUST BE LOW INDUCTANCE. They must be short ! Otherwise there is no point in wasting money specifying a screen if you use a high impedance connection to the Chassis.
the green or green/yellow PE wire should be permanently and mechanically fixed to the chassis. Not shared with other connections, that may be dismantled at some later time. It is worth making this close to the IEC socket rather than trailing across the Chassis.
I am not talking about the long coaxial cable. I am referring to the connection at the input terminal. The core/Hot has been exposed when you stripped off the screen/shield. The exposed core is not protected from interference. It must be short, very short. The screen has been twisted to form what we colloquially call a pig-tail. That too should be short. Now look at the gap between the pigtail and the hot connection. That gap creates an area. The bigger the area the more interference that is picked up. You must make the loop area small. Insulate the core/hot with a bit of tape. Push the pigtail right up next to the tape. That reduces your loop area. Using the screwed on tag makes the loop area bigger. Instead you can solder the very short pigtail to the end of threaded collar around the hot pin. This minimises the loop area. Do this for every low level socket.
Hoprefully explained clearly above.
Last edited:
I like those links they look proportional in that board space reserved for Rz/J
They are fine ultrafast diodes with soft recovery too. They are a logical choice I would say.
Offset +-0.5mV maximum that I measured over several minutes.
Also rails voltage sag about 0.1V when I connected the amplifier.
Andrew, I know, inputs will be replaced with shielded wire next time I open the lid. I just got fed up stripping cables! Although there is no ill effect I can detect, even on FFT, at least in my environment.
Also will try a 10K pot, as I have a DACT one to spare. For the time I have a 50K Tocos.
Yes, from TME so not fake, but I have a sample from ebay and they do not look very similar. How can you tell?
Talking about AD823
Talking about AD823
The originals should be having weak brownish but sharp probably laser etched logo & characters that tend to come out in much starker relief if you shine a torch on them at an angle
The right one must be fake. Mine have the same lettering and orientation edge dent like the one on the left.
I haven't looked much into it as if it comes in various legit looks from different production facilities or something but I also got OP275 here that looks like your left one again.
The Chinese fake on the right is clearly a relabeled NE5532 or some other T.I. chip that has the large flat pin 1 dot.
Is the OP275 suitable for the servo ?
I am having difficulties finding AD823 in my usual supplier