LM3886 gainclone DC servo

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2007
Paid Member
The LF411A is the one with a typical 0.3mv offset voltage, the standard version is around 0.8mv still below the TL071 though but how much DC precision do you need for a servo ?

I just measured the offset on my amp which uses TL071's and the figures were 2.6mv on one channel and 0.9mv on the other... although as you may recall my amp actually derives its bias voltage from the servo output. In other words it is not simply a correction voltage but it biases the whole amp to the correct operating point.
 

ICG

Disabled Account
Joined 2007
I just want to remind that the DC offset is also dependent on the surrounding circuit, feedback and especally tolerances. That means, the offset can be compensated by the circuit/feedback or become worse by tolerances of the OPs and the resistors. That means, the 'off' isn't always the falut of the OP and before you claim OP xy isn't suitable/bad for the DC servo, you should probably check if something else is 'off'.

There are also schematics which correct the offset by using the 2nd OP (with assumed almost identical behaviour) by compensation by the 2nd OP of the same IC. If there's a remaining DC offset it might also be because of the following amplifier stage or amplifier chip. The final DC offset can be ignored in many cases but with PBTL or even multi-PBTL use the tolerances have to be much stricter to avoid power lossed due to the ICs compensating each other and much power potential is wasted by that, resulting in a much lower output and a lot higher heat dissipation or even burnt up components.

Is that a 6x LM3886 PBTL or 3x3886 PBTL amp board? Or 'just' a 6ch amp? If it is the former, that would mean, if you want to know if it works as intended, you have to measure each IC DC offset separately for each IC and if it's too far off, the DC servo of the particular amplifier IC might have to be modified or parts have to be switched. Just solder the parts in and/or substitute a few parts in such a DIY kit will probably not work up to the max specs if you leave out that part of testing, the more ICs are used parallel, the worse it gets.
 
Administrator
Joined 2007
Paid Member
Not necessarily although it would require some thought in the design.

Paralleling chips usually involves a low value series resistor on the output of each before summing to a single output. That prevents small differences in voltage from generating massive currents as the chips 'fight' each other trying to source and/or sink current between themselves.

If both amps to be paralled are designed correctly then I would say the offsets should be very similar and lend themselves to a single servo. A dual servo would also tend to cause massive currents to flow, particularly at power on due to slightly different time constants and so on.

Tom Christian is the one to ask on this as he has done more with the LM3886 than most.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.