Matrix M-Stage Class A Biasing

Some opamps come in DIP packages, some come in SOIC packages. The socket on the M-Stage is fitted for an 8-pin DIP package, so if you have a SOIC package opamp, you need to get it soldered to a SOIC to DIP adaptor like on my OPA627s. Here is the picture of the opamps, SOIC with bias resistors, SOIC without bias, DIP with bias resistors, DIP without bias.


It’s risky to perform the soldering with the opamps installed on the PCB, but I didn’t have a helping hand around, and this makes it easier as you don’t have to deal with having to hold the opamp. Just be careful on where you are pointing the solder, and take care not to drip any liquid tin on the PCB! Of course you can do the soldering away from the PCB, but you need a helping hand to help you hold the opamp.


Practice first with using cheap opamps before moving to the expensive opamps. Opamps are quite delicate and touching the leads with a hot solder iron for too long would break the internals.


After all is finished, plug the opamp back into the socket, maintaining a correct orientation of the pin (if you’re looking from the top side, the dot should be aligned to the top left hole on the socket. Or the U-shaped dip in the diagram should be facing the rear side of the amplifier. Turn the amplifier on, use a cheap headphone to test for sound, making sure there is no errors in the operation.


If all is well, you’re ready to enjoy your modded M-Stage. Plug in a good source and headphone and you’re ready to go!

Now, don’t expect a Beta22 beating amplifier with this simple mod. Rather, if you really like your M-Stage, you can push the performance further by doing this mod. The difference is quite pleasing to my ears. Without the Class-A biasing, I felt that the top treble frequencies are a bit restrained even with OPA627. With the Class-A biasing, I’m hearing a smoother and more linear treble extension, more refinement, and a more open sound.

I’m even more in love with the M-Stage now!

4.6/5 - (5 votes)
Facebooktwitterredditpinterestlinkedin

37 Comments

  • Reply August 31, 2010

    purrin

    🙂

  • Reply August 31, 2010

    coolfungadget

    great job, purrin & mike!

  • Reply August 31, 2010

    Professor00179

    Nice job.:D

    • Reply August 31, 2010

      Mike

      Thank you guys. 🙂

      @purrin: what current bias setting do you use? 5mA is running a little warm to hot on my set up. 🙂

  • Reply September 1, 2010

    Iyayy

    simple stupid question..

    how do i know what resistor value to start with on different opa?

    there should be something to refer to based on spec sheet right?

    is higher resistance = less current when biasing?

    thx.

  • Reply September 1, 2010

    Mike

    I was told that 5mA should be quite a good figure to aim for, as lower amperes may not have too much of an effect. Of course it can be different for different opamps, but I really haven't experimented much.

    Here is the basic formula:
    15 V (opamp running voltage, measure from pin 4 or 8 ) / 0.005 A (or whatever desired figure you want) = 3000 Ohms (the resistor you need to use). I use 2.7KOhms as a close approximation to 3000 Ohms.

  • Reply September 1, 2010

    Iyayy

    that would be very helpful..

    will probably try it out later,

    but im using LM4920NA, dont have 627.. i found it quite nice tho, except that the treble is slightly messy. hoping this biasing would fix the problem.

    thanks.

    • Reply September 2, 2010

      Mike

      Hmm… let me know how it works out.

  • Reply September 4, 2010

    wwenze

    Note that the pinout of single-channel and dual-channel op-amps are different, for the OPA2134 you connected both channels' output to V- which is correct, for the OPA627 you connected V+ to V- then to one of the offset-trim pins.

  • Reply September 4, 2010

    wwenze

    Unless yours is two OPA627s mounted onto a 2x single SOIC to dual DIP adapter, in which case it is correct. For single-channel op-amps (both DIP and single SOIC on DIP adapter it would be pin 6 to pin 4.

    • Reply September 6, 2010

      Mike

      Yes, wwenze, that is on a a 2 single SOIC to dual DIP adapter.

  • Reply September 8, 2010

    purrin

    Hey Mike,

    I'm currently using 3.3k right now which ~4.5mA. I may eventually settle on 3.9k. The chip gets warm but not hot – but I my ambient temps are around 70F. These op-amps don't like to take more than 5mA. You can push to 7mA, but that's really pushing it. I think as low as 2mA should get most of the good effects. I haven't done my math on the circuit, but 2mA should be enough, although what I've found is that different values tend to change the sound balance depending up on the chip. The LM4562 gets warmer (sound balance wise). LOL, I tried 2.2k for 6.8mA on LM4562 and it was too way warm/bassy. The BB chips seem to act differently though – I think they get brighter with more current.

    • Reply September 8, 2010

      Mike

      Wow, dude! Your findings are really in-depth! I really haven't experimented any further with the other op-amps, or changing the bias current. It's interesting to hear that the different current bias changes the balance of the sound.

  • Reply September 9, 2010

    purrin

    Yes, I read somewhere else that pushing the LM4562 more into class A warmed up the sound. I didn't really believe it or give much thought to it when I made my DIP class A adapter when I only had a spare 2.2k resister lying around (I was using 4.7k earlier before). After I put it in, I realized after some time "crap, this is way too warm". When my order of resistors came in, I promptly put in the 3.3k and thought that was more like it. I'm going to eventually settle with 3.9K probably for LM4562.

    For OPA2107, I think it gets more of that U shaped curve the more you push it, although some others have said the treble disappears when you go past 5mA to 7mA.

  • Reply September 10, 2010

    iyayy

    since i only have a pair of 10k resistor.. i end up using em with my LME 49860NA (sry for the wrong info previously)..

    apparently it does improve clarity by reducing treble smearing. i also got slight extra depth.. initial impression is very pleasing. though it does tilt the signature to warmth a bit.

    oh btw, im using lovely cube linear headphone amp, not the m-stage.. and, im not sure whether im more used to warm siggy, but i found it neutral at stock, and also with the un-biased 49860.

  • Reply November 7, 2010

    Arup

    3.3k is waaaayy too small. Been using a 1.5k with 12V supplies & its splendid (I actually mean it transforms the sound). Try it!

    • Reply November 8, 2010

      Mike

      Interesting. Thanks for the input Arup. What amp are you using?

  • Reply March 12, 2011

    cider-moe

    Can the M-Stage adequately drive the LCD-2 and sound good?

    • Reply March 12, 2011

      Anonymous

      Power level wise, yes. Sound quality is decent though.

  • Reply March 12, 2011

    cider-moe

    Can the M-Stage adequately drive the LCD-2 and sound good?

  • Reply October 6, 2011

    Ton luppenga

    Dear Sir(s),

    After reading the smart solution of biasing up to class A I read something about 15V (specified operating volt of the OPA627). Do you have changed thise power supply voltage to? I want to go from 12V regulator value up to 15V in relation to the OPA627.

    with my best regards, ton (the Netherlands)

    • Reply October 6, 2011

      Anonymous

      Ton,
      I don’t think you need to alter the voltage to 15 Volts as the OPA627 takes in a range of supply voltage from ±4.5 to ±18V as specified in the datasheet:
      http://www.ti.com/lit/ds/symlink/opa627.pdf

  • Reply October 26, 2011

    Alain

    Hi I don’t understand what I see and read … You write about a OPA627, but the picture shows a OPA2134PA, is it the same ?

    • Reply October 26, 2011

      Anonymous

      Hi Alain, 
      You’re right the main picture is an OPA2134 and it is not the same as the OPA627. 

      • Reply October 26, 2011

        Alain

        ok then, do I have to order a OPA627 ? and do I have to connect the resistors as the same place as on your OPA2134 picture ?… Sorry I’m a newbie, and english is not my mother tongue  🙂

        • Reply October 26, 2011

          Anonymous

          Alain,
          You can do the class-A biasing on either the OAP2134 or the OPA627. The OPA627 is a better quality opamp.
          You can buy the OPA627 already biased here: http://tamaudio.com/store/index.php?route=product/product&product_id=82

          • Reply October 28, 2011

            Alain

            Great !
            Thanks a lot, I’ll try this as soon as I receive it. Much easier for me 🙂

            • Reply October 28, 2011

              Anonymous

              Awesome.

          • Reply November 24, 2012

            thecansmancan

            so if i were to buy the after market op amp. I can open the m stage up just pull the stock op amp out (like gently take a pair of pliers to it?) And the aftermarket one just plugs right in? no soldering or otherwise special skills needed?

            • Reply November 29, 2012

              Mike

              No soldering, just plug in and out. Watch out for the opamp orientation. There is a dimple on the opamp and you have to pay attention to it for the orientation.

          • Reply August 3, 2014

            Traveller2225

            Hey so I know thi sfrom a while ago but basically with the pre soldered opa627 biased in the link all I would have to do is remove the 2134pa and put in the opa627 down in it’s place? just plug and play?

            Edit: I just read someone elses response sorry I did not see this prior. Do you know the difference from teh link you posted from the Burr Brown OPA627AP with Class A Biasing Mod ? it is 10$ more and was wondering if it is any good?

  • Reply July 27, 2012

    Alfani Gunawan

    how about OPA637AP?? Tamaudio says its better, but several pepaople says its cause overheating??

    • Reply July 27, 2012

      Mike

      I haven’t tried it with OPA637 to be honest.

  • Reply October 31, 2012

    George

    Would an OPA2227 be a good replacement for stock OPA2134?

    • Reply October 31, 2012

      Mike

      George,
      I haven’t quite tried that. I don’t see why it wouldn’t work. I liked the Matrix with the OPA627 since I haven’t yet “discovered” the OPA2227. Ever since I found the 2227, I’ve liked it better than the 627.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.