If you have two wire mains connections ( Class II ), you must much more better fixed mains conductors ! Fixing by screw is not sufficient, says bloody norm 😀 and mains conductors must have double insulation ( inside chassis ) ! Unforunately norms must be keeped in DIY constructions too.
Franz G said:
The TO220 chips closer to the buffer/line preamp are regulators for the ICs? Are you using just a single chip as buffer or a two chip combo?
Which yellow/green are you talking about? The one coming out from the transformer on the right?
Carlos
It is the regulated version copied from CarlosFM, with two discrete OPA627. The TO220 chips are 7818 and 7918 regulators.
There was one yellow/green conductor from the center grounding point to both speaker connections.
I hope, this answers your question.
Franz
There was one yellow/green conductor from the center grounding point to both speaker connections.
I hope, this answers your question.
Franz
carlosfm said:
Hi carlosfm!
Is it also possible to bias 637 in class A? I need a gain of about 7 so the OPA627 is not applicable for me.
Do you have a schematic to start with OPA637?
Thanks.
aHobbit, the OPA637 needs to be run with gain to keep it stable. The OPA627 does not but it can still be run with gain (6 or otherwise). 😉
From its datasheet, its gain should not over 5 for non-inverting type and 4 for inverting type if you want it stably operate. 😀
Nuuk said:
Hi nick,
OK, gain of 6 for 627 is fine, but can you biased it class A, just as suggested by carlos?
With 637, my intent is really to have it with a gain (of 7), can it also be biased class A, carlosfm style?
I think the both can be biased into class A.
I will know how it is good when opamp is biased into class A after I get the PCB. 😀
I will know how it is good when opamp is biased into class A after I get the PCB. 😀
Sawadee
imho, this is a misinterpretation of the datasheet.
Correct:
OPA637 needs gain over 5 (or 4 for inverting) and has better values in this region. It will not work stable below this gain.
OPA627 can be driven stable between gain = 1 and open loop gain, but dont forget the high frequencies (thats why a 4pF cap is needed in the feedback path, setting high frequencies to unity gain).
And dont forget: the less gain, the more feedback, therefore some circuits can be instable driven below spec gain.
Forget about "class A" in this case. Definitively!
A true class a amp is designed to be a class a amp. Everything else is cheating yourself, by a way.
How do you set the perfect operation point of the residing +transistor at the output? Is it working linearly at this point? I am just waiting for people, who tell, it is better to use the negative transistor, therefore short circuit the positive one 🙂
It IS not a class a opamp. Setting it to "class A" is like you remove one left wheel from your car. This way, you can drive lefthand turns much faster with less resistance...
Franz
imho, this is a misinterpretation of the datasheet.
Correct:
OPA637 needs gain over 5 (or 4 for inverting) and has better values in this region. It will not work stable below this gain.
OPA627 can be driven stable between gain = 1 and open loop gain, but dont forget the high frequencies (thats why a 4pF cap is needed in the feedback path, setting high frequencies to unity gain).
And dont forget: the less gain, the more feedback, therefore some circuits can be instable driven below spec gain.
Forget about "class A" in this case. Definitively!
A true class a amp is designed to be a class a amp. Everything else is cheating yourself, by a way.
How do you set the perfect operation point of the residing +transistor at the output? Is it working linearly at this point? I am just waiting for people, who tell, it is better to use the negative transistor, therefore short circuit the positive one 🙂
It IS not a class a opamp. Setting it to "class A" is like you remove one left wheel from your car. This way, you can drive lefthand turns much faster with less resistance...
Franz
That's correct! This is what I was told recently by a highly repected hi-fi designer.
So if you want to stop worrying over which is the best opamp, or which is the best circuit, etc click Here, scroll down and you'll find a gem of a circuit that will take only a few minutes to solder up. Do yourself a favour and compare it against any of your opamp circuits! 😉
If I would have believed everything what respected people have told me I'd be a rich man by now 😉
Seriously, this current mirrors at the output of opamps can't be called class A. They just pull current at only one output transistor and heat up the chip at a constant level which makes their sound somewhat better. Class A is superior to class B or AB. When one wishes true class A operation it is better to build a discrete circuit or to go for class A opamps like some types Harris produced.
Nuuk said:
Mmm... but I have heard from another respected designer that biasing op amps into class A is a well known benificial technique.
and I have also heard this from my ears...😉
so as ever, there seem to be a range of opinions on this one
why am I not suprised ? ....🙄
No conflict there Mike!
You can tune a Mini and make it faster but it won't be as fast as an F1 Ferrari. Likewise, you can make an opamp sound better by biasing toward class A but it still isn't operating in true class A.
Anyway, the proof of the pudding is in the eating! 😉
Nuuk said:
🙄
I'm not that, but I can tell you this about your buffer schematic (on your link):
1. Remove those 0.22R resistors, like Carlos Martinez (carlmart) says.
Instead, change them for 10~22R resistors😱, before both PSU pins (and psu caps) on both op-amps. This helps isolating each op-amp from each other, as the regs are common for both.
2. Remove those OsCons
, and change them for something around 22~100uf (yes, these op-amps like capacitance, big caps).3. Remove the input cap, you have one on the output.
4. Put a 0.1~0.22uf poly cap between + and - rails on each op-amp, under the circuit, directly on the pins.
5. Disconnect those transistors and try the simple resistor, with the value that I suggest.
6. Put the OPA627 back in.
7. Check my next post.
8. Maby it's faster to make another buffer circuit with my suggestions than make these mods.😀
jean-paul said:
Oh, you can go a little higher, to 220nf, if it fits.
jean-paul said:
Amen.
jean-paul said:
Amen.
It shows that you know the "animal".😀 😉
I apologize for my mistake. Yes, OPA637 needs gain at least 5 for stable operation. Thanks Franz G. 😀
I know opamp is designed by class B or AB. We cannot do it to be true class A. Anyway, if the biasing the opamp make me get the good sound, I will try it.
Thanks for your pointers Carlos. I will give them a try with an open mind although I do feel that the discrete buffer is probably the better solution for this particular job! 😉
Oscon SG capacitors are equivalent to much higher values in other capacitors so those 15 uF Oscons are more like 150 uF. However, I will try some Panasonic FC's in yet another new circuit!
I note that some people put the electrolytic decoupling caps nearer the regulators and have smaller film caps (0.1 uF) right on the pins of the opamp. Do you recommend that approach?
I am a little puzzled as I have separate regulators for each opamp circuit. Did you realise that or am I missing something else?
Oscon SG capacitors are equivalent to much higher values in other capacitors so those 15 uF Oscons are more like 150 uF. However, I will try some Panasonic FC's in yet another new circuit!
I note that some people put the electrolytic decoupling caps nearer the regulators and have smaller film caps (0.1 uF) right on the pins of the opamp. Do you recommend that approach?
I am a little puzzled as I have separate regulators for each opamp circuit. Did you realise that or am I missing something else?
- Status
- Not open for further replies.