diyAudio (
-   Chip Amps (
-   -   Buffer/Gain stage suggestions for LM3886 (

zakman35 13th August 2012 07:47 PM

Buffer/Gain stage suggestions for LM3886
As the title says, I need a buffer/gain stage for my 8 channel LM3886 amp which till now was driven by a behringer CX3400 active crossover.

My amp is based on Linkwitz Riley desing at
This amp has an input sensitivity of 1,1V.

Recently I replace behringer with a miniDSP 2x4 and although I am very happy with the sound of miniDSP I have a amplification problem since miniDSP 2x4 has an output of 0.9V max.

My problem is that miniDSP output (0,9V) cannot drive the amp to full power and this is very annoying especially with low level recording even with the volume at full the sound is not loud enough for me. I assume there is also an additional voltage loss from the long interconnects (2 meter) that I have to use since the amp and minidsp are not located close to each other.

My thoughts are to amplify miniDSPís output to 6db or more by adding a buffer/ gain stage which is supposed to be beneficial for gainclones either way.

So, I would like your suggestions in order to add a 4 channel buffer/gain stage between these two units.

My thoughts are:
A) To add an opamp buffer/gain stage with a 6db gain as the first stage described at ESP preamp project 88.

Discrete buffers as described at Decibel Dungeon pages the problem with this is I donít know if they can modified to provide a 6 db gain or not.

Not the best but probably the easiest to by two readymade preamps like this from Ebay and remove the pots.

I forgot to mention that I am using miniDSPís digital volume control, and I already have an adjustable symmetrical psu with maximum output of +-14V that I would like to use for this project.
The buffer/ gain stage is going to be placed inside the miniDSP box.

Thanks in advance.

Mooly 13th August 2012 07:55 PM

Nothing wrong with an opamp gain stage. It's just the buffer with two feedback resistors. It really is the classic text book implementation.

An alternative idea. Modify the LM3886 stage. Reducing the 1K feedback return would increase the gain.

AndrewT 13th August 2012 08:03 PM

A 3886 has a maximum gain of 10times (+20dB). But at this gain the phase margin is quite low.
The 3886 performs better at reproducing the input signal when the gain is increased to about 28 to 30times. (+28dB to +30dB).
The 3886 will still work well with the gain up at 40 times (+32dB).

Only when more gain than +32dB is required would an extra gain stage be worth fitting.
That extra gain stage could be in the Pre-Amp. There it could be made switchable, so that higher level sources could pass straight through and thus avoid excess gain + excess attenuation.

A non-inverting 3886 does not need a buffer or preconditioning of the signal.
An inverting 3886 does benefit from a stage that preconditions the line to allow quality signal into the low impedance presented by the inverting input.
I would prefer to put that pre-conditioning stage back at the source (the Pre-Amp) and call it a low impedance buffer stage.

zakman35 13th August 2012 08:11 PM

Thanks for your fast response.

I would prefer to build a buffer instead of modifying gainclone's gain stage because i am afraid that with the long interconnects i use and low level recordings miniDSP will still be incapable to fully drive the amp.

zakman35 13th August 2012 08:13 PM

What about the 3 transistor buffer, described at Decibel Dungeon pages.

Can this stage modified to provide amplification of 6 db or more?

Mooly 13th August 2012 08:14 PM

Well a buffer is really easy... but you want a gain stage :)

I can probably find lots more worked examples but this was first out of the bag that I remembered,

read the whole thread though.

Mooly 13th August 2012 08:22 PM

If you want discrete then something like this is excellent although originally designed for single rail,

tbh though you will not really better an opamp stage using something like the OPA2134 dual opamp.

zakman35 13th August 2012 09:05 PM

I guess you have wright, I had good experiences in the past with opamps like OPA2134, LM4562 & LT1028, why not now.
Discrete stages could wait for the moment.


zakman35 14th August 2012 08:45 AM

Thinking again the option to increase the gain of the amp and according to my calculations the current gain of the amp is 20,6.

So if i replace the feedback resistor of 1k with a 687 ohm, the gain will be increase to 29,5 which i think will be enough.
Is that correct?

Mooly 14th August 2012 10:29 AM

That's correct. The voltage gain is 29.5 with the new value.
On a practical level don't get hung up on exact resistor values for gain setting. 687 ohm or 620 or 560. The difference in perceived level won't be as great as the numbers suggests.

Also from a theory point. You have attenuator in the 1K and 19.6K input resistors. Again in actual use its a non issue.

Best way is to try values around the 680 ohm mark and take it from there.

Have you got Ci fitted that is on your circuit ?

All times are GMT. The time now is 06:56 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio