Boss HM2 analysis HM-2

[Renegadrian]

I was experimenting again with the HM-2, and studying its schem to understand what does what.
http://www.hobby-hour.com/electronics/s/schematics/boss-hm2-heavy-metal-schematic.png

now we have a typical input buffer at q1, q4 is part of the switching, then q6 is a npn gain stage (the same you find in the DS-1) then q7 going to IC1b (ic stage with soft clipping diodes and cap+res to ground via the gain pot.) then "brake" diodes (which I removed) then hard clipping stage, then IC buffer then tone and gain recover?! now that is quite clear to me. (hope I didn't mess anything, correct me if I am wrong)

now, what's the use of q7?! can it be ditched altogether?! also, what's the point of r49/c31 to gain pot?! can I take them off to have a "regular" IC gain pot?! Does it make sense to mod it to get closer to a DS-1!? compare the schem, DS-1 has the same npn gain stage (here q6) then IC (without soft clipping) going to the tone section.
any ideas?!

[Transmogrifox]

Q7 is extra gain.  The purpose of HM-2 and what makes it "HM" is lots of gain.

R49/c31 pre-attenuates the signal, which cancels the added gain of Q7 at lower gain settings.  With the pot turned down to a lower gain setting it approaches something much more like a DS-1.

By turning the pot clockwise the gain going to the input of the op amp is simultaneously increased with the gain of the op amp stage itself.   The end gain after the op amp looks like it maxes out around 2000, but it gets there by cascading moderate stages: gain of 20 followed by a gain of 20 followed by a gain of 5 (op amp).  Cascaded gain stages generally are used in order to optimize each stage for noise performance.

My guess without analyzing it is the multi-gain stage strategy is used to get a lot of gain with minimum noise. and in my memory the HM-2 is a surprisingly quiet pedal considering the amount of gain it has.

For example, if you do away with both Q6 and Q7 stages and put a 100 ohm resistor in the feedback then you will get the gain but it will probably have a lot more audible hiss.

Also the pre-gain stages give some drive above the nominal op-amp clipping threshold to drive the following grounded diode pair into more hard-clipping.

I wouldn't recommend changing Q7, R49, C31 unless you think it has too much gain and want to tame it a little bit.

It would make sense to remove all of these and connect output of Q6 to op amp if you just want to have a tamer distortion (make the HM-2 something other than HM).

[Quackzed]

check out these links for info on the eq section of it. gyrators man.
http://www.geofex.com/article_folders/eqs/paramet.htm

and heres a gyrator calculator thingie, very helpful if you want to mess w/ the tonestack gyrators.
http://www.muzique.com/lab/gyrator.htm

gyrators man, gyrators.

[Renegadrian]

Yeah I knew about gyrators man!!!  Tried to put a mid pot, but didn't like it so much...
here are my mod so far.

psa mod (removed d1 e r2)

d6 d7 removed & jumper
r30  removed & jumper
c28 removed
c30 47n
c4 1µ non el.
c12 1µ non el.
c15 1µ non el.
c32 1µ non el.

buffer out removed (wire 6 to q10/r48)
c7 c3 removed

tried leds and jumper on d8-d9 - didn't like it
tried bypassing pnp gain stage. so so, yeah you have less gain, but gain pot still is a on/off kinda thing.

[Transmogrifox]

If you want more useful range out of the gain pot you might do better to short gain pot pins 2 and 3 so U1B has a fixed gain. Change the gain pot to something between 10k to 50k and tweak R20 to reduce to the maximum desired gain.

If you don't want to change the gain pot (for example if you don't want to search for a footprint-compatible replacement) you can also tweak resistance values around Q6&Q7 to make the input and output impedances more comparable to 250k so you notice the difference more as you roll off the pot.

For example, multiply resistor values by 10:
R19, R16, R18, R21, R17, R22, R26, R29, R28, R27

Except R16, R21, R29, R26 ratios will work out differently in order to bias it correctly.  With those you keep the thevenin resistance looking into the base of the transistor (parallel equivalent) needs to be close to 1Meg.  Most likely you can eliminate R21 and R26 and then adjust R16 and R29 until the bias is correct.

Then of course you need to tweak the surrounding capacitor values to get back to a useful frequency response.

[Renegadrian]

wow that's a great reply! thx man!!!

[Fancy Lime]

Out of curiosity: Does someone have a graph of the frequency response of the tone stack with maxed controls? I'm wondering if the tone shaping couldn't be approximated in a much simpler way since the bass and treble knobs are always at max for most people anyway.

Thanks,
Andy

[BuryMeInSmoke]

@Fancy Lime

Don't know about the first part of your question but I came across these graphs of the FR

https://atomiumamps.tumblr.com/post/139197356031/boss-hm-2-analysis