High offset drift

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi
I would like to gather your opinions please.
The problem is my amplifier is low feedback (<30dB) and has no huge cap in feedback divider, neither DC servomechanism. Only DC trimmer.
As a result I can trim the DC offset to <20mV, but after 15 minutes or so it becomes a pretty random value in range of , say, -150mV<offset<150mV.
I know what is cone excursion caused by DC, why it should and how it could be avoided, but...

...My question is: do I REALLY have to worry unless I sell anything?

waiting for comments

best regards
Adam
 
Administrator
Joined 2004
Paid Member
Hi Adam,
Your DC offset is not harmful at this point, just annoying. It may do something you don't expect because it doesn't go to the same polarity even. Interesting.

Can you post or PM the diagram so we can get an idea what's going on?

-Chris
 
OK
seriously:
After several minutes of warm-up I trimmed offset and it stayed below 40mV most of time. After I heated up the heatsink with a hairdryer it went up to ~130mV but quickly diminished to <80mV. I think the source of this high sub-1Hz noise has something to do with complementary Vbe multiplier close to output devices.
By the way bias thermal stability is very good. After switch on the idle current goes from ~200mA slowly to settle at ~120mA.
Thanks for interest.
regards
 
Hi, Adam,

David, do you think my circuit 'could' become a component of the month then? Maybe it is fashionable to deny the importance of yet another parameter?

Maybe? What are you building?

OK now, stop joking :D

I've noticed this. CCT's that has "not stable" or quite high DC offset tends to sound better. I ask question to myself, why? Why CCT's like NP's SuSy (patent version) or DartZeel (both do not have R to ground at input transistors) sounds better than average other topologies?

I don't know exactly why, but in "common" topologies (like 3 stages amp), 0 DC offset is "set up" by R to ground (input impedance R), usually at input base (or base is grounded in inverting mode). The problem is, is the so called "ground" is clean or not? If not, this ground R could be injecting noise at the same time it sets up 0 DC point. Maybe this "input impedance" resistor is quite high in value, like 22k or 47k, still it doesn't stop this R from tranfering "ground garbage" while setting up 0 DC offset.

The best one is like the 2 example above, no input impedance R, making sure that the base of the input transistor is only accepting music signal, nothing else.

Drawback = you will have crazy DC offset if you don't use smart topologies.
 
I had many trouble about that, when i repaired VHF radios

I had a very old and experienced boss, he is now in the highest floor of the skies.... maybe the man is asking him.

- Well, i made this world, and it is full of problems...what is happening there?..... the anwer may be the same i receive when worked with him...was an unforgeatable honor to me.

I used to ask him why those floating voltages in biasings...if this is a result fo oscilation, some low frequency motor boating applied to radio frequency circuits?...I always iasked him... a monster of knowledge, in many aspects of the life...85 years of deep know how and intense life...a big man, a very respectable man.

- What is this?

He also answered me:

- A deffect!....meaning that "it is defective, damaged!"

I did not felt very good with that "small help"...and one day i asked him directly:

- What a hell!, you do not answer my questions, always telling me
a defect!, a defect!...a defect!

And he told me:

- Because we never know, everyday new defects, new things happening, new parts and new circuits operating differently, we never know...the only thing we really know, for sure!...is that it is a Defect!

Here is the sittuation.... the man want some help and is asking for some help...i cannot help him.... all informs he is telling me ,do not allow me to discover what is the circuit's problem.... all i can tell him....sorry....

-This is a defect!

regards,

Carlos
 
The Electrocompaniet Ampliwire has a single diff pair, and will drift 700mV within the first 20 minutes or so.

The diff pair is even in one can.

I've had problems with quad diff input drift. Haflers are particularly bad for this. Ripping out the CCS and replacing with resistors and zeners (Leach design) really helps, sounds better too.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.