D
Deleted member 550749
Recently made this schematic, and working properly, but my problem is without any capacitor when I connect from mobile out as a input direct there is no offset change or ZERO OFFSET WITH 1uma current, but I want to ad a opamp for that have input a cap but when I ad a cap offset goes 45mv , so do I need a servo circuit?? Or any circuit changes that will fix d problem 🙏
Are you sure it is offset and not noise/hum? Did you look at a scope?
That cap should do nothing to the offset, but it can be that the amp has an offset of itself which is nulled by an opposite offset from the phone.
Do you also get the offset with nothing connected and no cap?
What do you mean by " ZERO OFFSET WITH 1uma current"?
Jan
That cap should do nothing to the offset, but it can be that the amp has an offset of itself which is nulled by an opposite offset from the phone.
Do you also get the offset with nothing connected and no cap?
What do you mean by " ZERO OFFSET WITH 1uma current"?
Jan
D
Deleted member 550749
Bro I m not an engineer but make amp with some small knowledge and practice, yes without cap 0 offset but not checked without connecting audio source (phone output) I think u r right may the b input amp has minus offset thats why I m getting 0 offset volt, I want this same offset with input cap how can I achieve that ?? Or any servo circuit gonna give the same result?
4k7 emitter degeneration resistors on the input pair is way larger than normally used - the current mirror error will be amplified by those large resistances into a sizeable offset voltage on the emitters of the input pair, for instance 5uA difference would give 23mV of offset with 4k7 (but only 0.5mV for 100 ohms). Try something in the range 22 to 220 ohms for those degeneration resistors (although the compensation will probably have to change due to the increase in open-loop gain).
Then the input devices ought to be matching well with C1 present, since the bases see the same resistance to ground and output respectively of 22k, and the emitters are now close in voltage.
Or you can leave the 4k7 values as they are, and improve the current mirror to a more accurate one such as the standard 3 transistor design, and specify the 4k7's as 0.1%, in order to get good matching of emitter voltages. I don't think that's the way to go though.
Then the input devices ought to be matching well with C1 present, since the bases see the same resistance to ground and output respectively of 22k, and the emitters are now close in voltage.
Or you can leave the 4k7 values as they are, and improve the current mirror to a more accurate one such as the standard 3 transistor design, and specify the 4k7's as 0.1%, in order to get good matching of emitter voltages. I don't think that's the way to go though.
D
Deleted member 550749
Yes I think current mirror problem m tested without changing 4.7 but changed 2.1 with only 2k resistor and added 4148 diod at current mirror now the offset is 2.6mv when I try to lower down more offset then vas bias also goes down like 900mv form 1.2volt and sound distortion started, so that kind offset is not happening, 🥴
Sorry, you are cryptic. What current, where, into what?
From the output? If so, no worries about that, it will go to zero when offset goes to zero.
People are trying to find out what the cause is, then it will be easy to fix,
so try to answer the best you can and not randomly change the circuit while we are thinking ;-)
Jan
D
Deleted member 550749
Cryptic? May b, but I m learning, I m not a professional, perfectly understanding is not a bad thing , yes I have no special tools and knowledge like u that's why I m here, little bit of knowledge getting from here is not a bad thing also, hope u understand my feelings, this for my personal thing for my personal use but I love DIY that's why I m here 🥰
D
Deleted member 550749
The cap can appear to affect the offset if the source has a DC path to ground. So it looks like the amp has an inherent offset of 45mv and adding a parallel resistance (the audio source) across the 22k input resistance appears to reduce the offset... which it doesn't, it is just you are interpreting it as such.
Try altering the value of the 400 ohms in the current source (the parallel 800 ohm resistors) very slightly (I mean slightly, too much and you will get a latchup to the rails).
As @Mark Tillotson says, those resistor values (and the 2.1k's) are far to high by normal thinking,
Try altering the value of the 400 ohms in the current source (the parallel 800 ohm resistors) very slightly (I mean slightly, too much and you will get a latchup to the rails).
As @Mark Tillotson says, those resistor values (and the 2.1k's) are far to high by normal thinking,
D
Deleted member 550749
That's both 800 is offset pot parralel of 2resistor+ 2k potential resistor that I can easily make that 400ohm currently thats is 434ohm and offset is very good now currently I m playing music, circuit originally extracted from lm3886 , previously I made that using bootstrap, this time bias circuit changed that's why having this types of problems, by the way fixed that but still have another issue that's weird clipping, under 70watt I m happy , and I think about the compensation has to change because used Miller before that's sounds much better than this 😁
So is it all good now?
Another thing I spotted is that the current source supplying the VAS stage only looks to be set to around 1 milliamp or so which is imo far to low. The driver stage will struggle with only that current available.
D
Deleted member 550749
Actually thare is 1.05ma with 1.2volt , I think quiescent current is under 100ma , idle condition there very very low heating output transistor, and there is no distortion at all when I play music, clean sound , I have tda7265(pair) but diy amplifier sounds far better than that also frequency response is awsm and no noise at all,