[spendorspain]

Hi all.
I’ve moved this post to the DIY section of the forum. My apologies for the previous mistake.

I’m going to build an AMB M3 Headphone amplifier, which probably will serve also as preamplifier section for a LM3886-based loudspeaker amp. I’m currently using another headphone amp with a gain of 4.5x, that is rather high for my Grado HP-2 (40 ohms), so I plan to use the M3 at only about 2x gain.

One of the best things about the M3 is the possibility of customize its sound through the selection of the op-amps (BBs, more mellow, Ads, more bright, or even one of those GD discrete-opamps) and the bass-boost option. If Cbb is socketed, it is possible to test different capacitors (for instance, in the Wima MKP10 series there are several values from 0.01 to 0.1uF, all with the same LS=15mm) and to use some high-grade caps when a definitive value is chosen.

Using the online “Bass Boost Calculator”, I’ve calculated the predicted response with the default values (Cbb=0.22uF/50k log pot as Rbb) and with R3=3.32k/R4=10k (gain=4) and R3=R4=3.32k (gain=2).

I’ve draw, in one graphic for each gain, the results for several potentiometer positions (different Rbb values) and the curves for gain=2 are a little puzzling. Please, see attached images. The curves 1 to 4 (Rbb=50k, 20k, 10k and 5kohms, respectively) show the expected bass boost from the flat response (red line 5 with Rbb=0ohms), but the curves 6 to 9 (Rbb=1k, 500, 100 and 10ohms, respectively) show a depression in lower frequencies, even below the 6dB gain baseline. Does this mean that, using these low Rbb settings, the response is "boosted" (relative to bass) in the high frequencies? If this is correct, when a recording is too bright, adjust Rbb to 10k or 20k values with the Bass Boost pot (or switched discrete resistors); but if it is dull, adjust Rbb to 1k or 500ohm values. This would be a convenient and easy “tilt adjustment” to compensate less than ideal recordings or different-sounding sources. Is this correct? Am I missing something? Why this bass cut is so pronounced in the gain=2 setting and less at gain=4??

Thanks in advance for your help.

Jose

 

[amb]

You're running into some boundary conditions in the graphing function as Rbb approaches zero. In reality this bass boost circuit will never cause the response to dip bellow the basic gain of the amp anywhere. I'll try to improve the graphing function to handle this better when I find some time.

 

[rhester]

I am a lil confused on the varibale bass boost on the M3. If I want to have true variable boost controlled by pot, what do I do with Rbb and Cbb?

 

[amb]

If you want variable bass boost, install the bass boost pot, but don't install Rbb. This is spelled out in the "Parts list" section of the M³ website.

 

[spendorspain]

Thanks for your response, Ti.

Perhaps using in Cbb a very small cap, combined with the 50k log pot as variable Rbb and a gain=2x or 4x, would put the point “Fc”, from which the bass boost takes place, higher in frequency, to around 1k-3kHz. This way the behaviour of the filter is over a broader range and not limited to bass. Moreover, high quality caps (Solen, Mundorf, ICW…) are usually very big and probably only a very small value would fit, perhaps with some lead bending or even under the M3 pcb. Bass boost circuit is in the feedback loop and I think every part in it must be of the highest possible quality.

I’ll try different cap values to listen to the differences in sound, socketing the 15mm-spacing holes in Cbb. I think there is no risk in using these low value caps (like 0.047 to 0.01uF) in Cbb, am I right?

Thanks again for your help

Jose

 

[amb]

Quote:

Originally Posted by spendorspain /img/forum/go_quote.gif II think there is no risk in using these low value caps (like 0.047 to 0.01uF) in Cbb, am I right?

There is no risk of damage, if that's what you're worried about.