Aion's Sapphire Galaxie mod question (BD-2 with mod)

[eh la bas ma]

@eh la bas ma: I did not follow all the details yet, but I am interested in this same kit with the mod. Did you manage to get it working before changing various components?
Cheers

Sorry I didn't see your question sooner.

I don't know if something was wrong on my build, it turned the circuit into a sophisticated boost pedal. On its own it could have been useful for clean tones. As a booster it didn't even have much volume on tap.

There was much less dirt on my build than on the demo I can find on Youtube. None in fact : clean tone through all the Gain rotation, even fully CW, unless I put back some diodes in there. I could hear the signal was colored a bit, EQ worked fine,  that's all, no dirt.

I am glad I changed it back to the classic BD-2 with some useful mods. I am using it much more often. Bass and Treble controls are great compared to the standard BD-2. I am really happy with the mods I found.

You can check musikding's Sapphire reviews to find what I've done.

I never had any answer from Aion. On musikding's forum I've been asked to post the full modded schematic from Aion, because there is none in the instruction files. No help without it, and I must admit I wasn't curious enough about the mod to do that. It seemed easier to unmod the circuit.

The most interesting feature for me about the mod is eliminating the fizz. I am guessing it depends also on the variety of transistors selected. I went with the 2SK209 from musikding and it's really good. My other BD-2 vero build has 2SK117 and it's great too.
I did as suggested in this document :

https://lectric-fx.com/wp-content/uploads/2016/04/Bloozehound.pdf

Based on Lectric fx notes, I wonder if the fizz isn't produced by J201, anyway I never felt the need to do something about some "fizz" on my builds.

Be that as it may, for the sake of science and the common good, I invite you to try it and report back your findings. I am very curious to know if other builders get to the same results.

[matopotato]

No worries, I did not receive the kit yet 

Did you do everything/most from the premierguitar mod (https://www.premierguitar.com/gear/boss-bd-2-mods)?

From Das Musikding, I figured out:
Resinsert some diodes. Now the BOM says 1N914, but you tried and liked better Ge diodes, right?
So the D7 and D2, did you put (D9k + 1N4001) for each of them? ¤ (Edit: 4 diodes...) diodes in total?

I assume this is an DPDT so 1M and 15k in the one setting and 100R and 8k2 in the other flip?
R13 = 1M resistor/100R on a toggleswitch
R14= 15k/8k2 on the same switch

I was thinking of breadboarding to check the mod, but not sure how to deal with the 2SK209 assuming they are SMDs?

For sure I will let you know how things go 

[eh la bas ma]

Resinsert some diodes. Now the BOM says 1N914, but you tried and liked better Ge diodes, right?
So the D7 and D2, did you put (D9k + 1N4001) for each of them? ¤ diodes in total?

I assume this is an DPDT so 1M and 15k in the one setting and 100R and 8k2 in the other flip?
R13 = 1M resistor/100R on a toggleswitch
R14= 15k/8k2 on the same switch

I was thinking of breadboarding to check the mod, but not sure how to deal with the 2SK209 assuming they are SMDs?

For sure I will let you know how things go 

I think both 1n914 and ge diodes sound good. The ge diodes give something a bit different, I'd say it helps smooth the distortion around the edges. I already built a BD-2 on vero and I wanted something a bit different. Could be less fizzy with ge diodes too.

Edit : it was in june, so I don't clearly remember. But knowing myself, I surely spent a lot of time trying all the combinations to make sure I find the best possible settings.

D2/D7 : OA1182 in series with 1N4001. So 4 diodes in total for D3 and D7. The article mention D3 instead of D2, but it's the same i guess.
D3,4,5,6 are D9K.

2SK209 are smd. You can start by soldering some wires at the key locations for the mods, instead of fully populating the board. Being able to decrease a bit of low ends and to toggle back to the original sound is really useful. 

So, on the DPDT, I did the logical thing : one position is standard values, other position toggles two mods at once, two values are changed as described.

A breadboard would have been nice to test all the possibilities, the article is full of great ideas. The problem is smd, but it's a simple thing to solder them on adaptors :

https://www.musikding.de/Adapter-SMD-TO23-RM254_1

[matopotato]

Sorry for the delay on this. I finally managed to breadboard the Sapphire. And yes, it is breaking up a bit with high gain and high tone settings. It seems also overall very loud compared to other builds, but that might change once it is finished in my experience.
So I tried the Galaxie mod and it did take away that breaking up ("fizz"?) sound, but at the expense of the overall character and treble IMHO. Might be a future project, but not for "my BD-2".
While I did not get your "no gain at all" feeling, I still had to set gain to 75% and higher to get anything out of it. The stock version had useful gain from 0 to 75% so in the end the choice is easy.
Then I tried your diode mods. And that got a bit complicated. I had only 4 D9Ks, but seems you had OA1182 for two places so that should work. I had no 1N4001 so I used 1N4002. As protection diodes in pedals I understand they are interchangeable, but as clipping diodes there might be a small difference. I figure I will not be able to tell though.
So D3 or D4 changed to OA1182+1N4002 in series, same for D6 or D7. I am guessing it would not matter that much where the asymmetry lies or is it important that it is D3 and D7 that get the combo?
The other clipping diodes were all D9Ks, with their backward marking, making the whole operation a bit of a challenge.
I think the diodes took away some of the "fizz" for sure, but I plan to make some more tests. I would like to keep stock BD-2 character as much as I can and at the same time lose some of the "fizz" to make the pots useful throughout their rotation.
Why the asymmetrical diode setup? Why not "4 of the same" as the 1N914 were doing in the stock setup?

I also tried the R13 and R14 changes you did, but for me there was not a huge difference and if anything some bass was lost, which I don't mind having anyway, so I will probably let those stay stock.

SMDs was a new experience, and the adapters works fine so far.

[eh la bas ma]

So I tried the Galaxie mod and it did take away that breaking up ("fizz"?) sound, but at the expense of the overall character and treble IMHO. Might be a future project, but not for "my BD-2".
While I did not get your "no gain at all" feeling, I still had to set gain to 75% and higher to get anything out of it. The stock version had useful gain from 0 to 75% so in the end the choice is easy.

I always felt like I failed my Galaxie Mod attempt, and missed the Ultimate BD-2 Mod, with all the horrific consequencies.

According to your experience and despite your lack of good Morals (you said it yourself), standard version is better, so it's a relief. Thanks !

is it important that it is D3 and D7 that get the combo?

I guess so,  but not really sure. You breadborded the circuit : you tell me?

Why the asymmetrical diode setup? Why not "4 of the same" as the 1N914 were doing in the stock setup?

Depending on the context, sym & asymetrical settings don't sound the same. Probably something about odd and even order harmonics more or less present in the signal ? According to B.Wampler, author of the article, it's supposed to sound more "woolly", old Fender Tweed style.

In my case, when building the Sapphire, I already had a BD-2 built on perfboard with standard 1n914, so I wanted to try something different. Build report :

https://www.diystompboxes.com/smfforum/index.php?topic=36392.msg1239191#msg1239191

You breadborded the circuit : which diode config do you prefer ?

I also tried the R13 and R14 changes you did, but for me there was not a huge difference and if anything some bass was lost, which I don't mind having anyway, so I will probably let those stay stock.

Decreasing some bass and some mids to lighten the distortion was what i intended, based on premierguitar's article :

"After this first gain stage, we go through what first looks like an odd tone filtering stage. It is actually a Fender-type 3-band tonestack with fixed values (with the treble on 0 and the bass and mid on 10). This is a really cool thing to mess with if you want to go hog wild, because you can add trim pots in place of R37 (use a 250k trim pot for treble), R50 (use a 1M trim pot for bass), and R51 (use a 25k trim pot for mids)."

R50 and R51 on the article's schematics are R13 and R14 on Sapphire's.

On my build, both values can be toggled at the same time, from custom to stock values, and the change is quite noticeable.

Compared to most Overdrive/Distortion circuits, BD-2 is quite heavy on bass. These values were the most interesting I could find to get the BD-2 closer to the "average" overdrives : decreasing the bass presence and keeping the BD-2 sound character from the famous discrete op-amp stages. I do like the Bass-heavy original character too, though.

There should be less compression and less bass when you use 100R & 8k2 at R13 & R14. Depending on how you use the circuit, this mode can be useful : to boost an other gain effect without overloading it with low frequencies, or with modulations... especially with modulations... like the amazing Flintlock flanger, for exemple.

SMDs was a new experience, and the adapters works fine so far.

New experience ? Do you mean you never built any PedalPCB FV-1 circuit yet ? Don't you have the incredible Unison chorus ?  What about the fabulous Hydra Delay ?

[matopotato]

So I tried the Galaxie mod and it did take away that breaking up ("fizz"?) sound, but at the expense of the overall character and treble IMHO. Might be a future project, but not for "my BD-2".
While I did not get your "no gain at all" feeling, I still had to set gain to 75% and higher to get anything out of it. The stock version had useful gain from 0 to 75% so in the end the choice is easy.

I always felt like I failed my Galaxie Mod attempt, and missed the Ultimate BD-2 Mod, with all the horrific consequencies.

According to your experience and despite your lack of good Morals (you said it yourself), standard version is better, so it's a relief. Thanks !

Morals are continously improving 
I would say it comes down to personal taste. Some might even like the "fizzyness" of the stock build, some might not care that Galaxie mod changes it to something almost different.
But as far as taste goes, I agree with yours in this case.

is it important that it is D3 and D7 that get the combo?

I guess so,  but not really sure. You breadborded the circuit : you tell me?

I will do some more testing, but I compared Brian Wamplers mod in premier guitar article https://www.premierguitar.com/gear/boss-bd-2-mods with Aion schematics and I would say that Brian's D3 nd D7 would correspond to the Aion D6 and D2 respectively (after trying to sort out all the ends). Not sure if that makes much difference circuit and music wise though, but no harm in doing the same as he suggested.

Why the asymmetrical diode setup? Why not "4 of the same" as the 1N914 were doing in the stock setup?

Depending on the context, sym & asymetrical settings don't sound the same. Probably something about odd and even order harmonics more or less present in the signal ? According to B.Wampler, author of the article, it's supposed to sound more "woolly", old Fender Tweed style.

In my case, when building the Sapphire, I already had a BD-2 built on perfboard with standard 1n914, so I wanted to try something different. Build report :

https://www.diystompboxes.com/smfforum/index.php?topic=36392.msg1239191#msg1239191

You breadborded the circuit : which diode config do you prefer ?

So far the Ge diodes took some of the shrill out so I might go with that. Will rebuild the breadboard to allow A/B testing. But would probably not go as far as putting in a switch for them.

I also tried the R13 and R14 changes you did, but for me there was not a huge difference and if anything some bass was lost, which I don't mind having anyway, so I will probably let those stay stock.

Decreasing some bass and some mids to lighten the distortion was what i intended, based on premierguitar's article :

"After this first gain stage, we go through what first looks like an odd tone filtering stage. It is actually a Fender-type 3-band tonestack with fixed values (with the treble on 0 and the bass and mid on 10). This is a really cool thing to mess with if you want to go hog wild, because you can add trim pots in place of R37 (use a 250k trim pot for treble), R50 (use a 1M trim pot for bass), and R51 (use a 25k trim pot for mids)."

R50 and R51 on the article's schematics are R13 and R14 on Sapphire's.

On my build, both values can be toggled at the same time, from custom to stock values, and the change is quite noticeable.

It would be a bit tight to add trimmers. I might test some more with this, but at some point I probably like to stay close to original and sort the worst shrill out, hopefully with diode changes.

Compared to most Overdrive/Distortion circuits, BD-2 is quite heavy on bass. These values were the most interesting I could find to get the BD-2 closer to the "average" overdrives : decreasing the bass presence and keeping the BD-2 sound character from the famous discrete op-amp stages. I do like the Bass-heavy original character too, though.

There should be less compression and less bass when you use 100R & 8k2 at R13 & R14. Depending on how you use the circuit, this mode can be useful : to boost an other gain effect without overloading it with low frequencies, or with modulations...

I must have mixed things up since I felt the opposite. Will test this some more.

especially with modulations... like the amazing Flintlock flanger, for exemple.

Hmm, Flintlock? Sounds interesting? Might try that 
I read somewhere they can be tricky to get cyclic thump out of the mix though.

SMDs was a new experience, and the adapters works fine so far.

New experience ? Do you mean you never built any PedalPCB FV-1 circuit yet ? Don't you have the incredible Unison chorus ?  What about the fabulous Hydra Delay ?

Correct. I have looked at pedalPCB Deflector, but will try Don't tell Ray (Adineko) https://scientificguitarist.wixsite.com/home/don-t-tell-ray and pedalPCB Gravitation first.

[matopotato]

I have tested a few options:

The diode settings with 1N914 on one side and the D9K and OA1182+1N4002 on the other.
The germaniums take down the shrill somewhat, so I like them better. But while playing I also felt the 1N914s serve a purpose. I admit that the Gain and Tone on 100% with single coils *is* a bit shrikey, but other settings were useful, IMHO.
So either keep both on a switch or just the Ge. As I mentioned earlier I am quite fond having stock as an option.
It should be OK I think to wire from the PCB over a DPDT on-on and to ground for the 4-5 group and the 2-3 constellation to the pins 5 or 6 of the OpAmp. I beeped around a bit so it seems straight forward. (Famous last...)

R13 and R14 as 1M and 15k vs 100R and 8k2 was difficult for me to tell the difference so I had to bring them to the side and use a one wire swap setup for faster A/B testing before I forgot the previous tone. I did not detect much difference to be honest. The stock (1M and 15k) took down the intensity a notch and with it some shrill (but it is a very close call), so I think I will stick with stock.

I also experimented with R18 and C11 (2k2 and 1uF stock) and used the Galaxie variants (10k with 100n, 82n or 68n). I felt earlier that the 1uF was a bit too much, loud and fizzwise, but now the Galaxie variants felt a bit flat. The 100n better and 68n variant most flat. They did kill quite a bit of fizz though, but a bit of character as well.
So either keep stock here or put the 10k+100n on an SPDT switch but most likely keep stock settings.

[eh la bas ma]

R13 and R14 as 1M and 15k vs 100R and 8k2 was difficult for me to tell the difference so I had to bring them to the side and use a one wire swap setup for faster A/B testing before I forgot the previous tone. I did not detect much difference to be honest. The stock (1M and 15k) took down the intensity a notch and with it some shrill (but it is a very close call), so I think I will stick with stock.

That's a surprise. Now I wonder if I am mistaking... Let's check this now :

If I read you correctly, you're saying that decreasing R13 & R14 values result with few changes, taking "the intensity" up a notch, like some kind of a slight boost. Toggling back to stock values quiet things down a bit.

I guess the question is : In the BD-2 Tonestack, if we decrease the values at R13 & R14 (R50 and R51 on the article schematics), does it do a bass and mid boost, or a bass and mid cut ?



Edit : " a Fender-type 3-band tonestack with fixed values (with the treble on 0 and the bass and mid on 10)"  quoting from https://www.premierguitar.com/gear/boss-bd-2-mods

This legendary article says that with fixed values, (assuming "10" means fully clockwise) it's like having bass at max with 1M, Mids at max with 15k.

15k and 1M pots can only go one way : toward 0k.

So, I would have thought that if we use any values under 1M and 15k, it would be like turning these pots down, CCW : resulting by some bass & mids cut, right ?

From 1M to 100R for the bass resistor value, I'd say you should easily hear the difference, with a huge noticeable bass cut ?

Sorry if i am raising doubts over nothing here, but it feels like I should hold my position on this....

Edit 2 :  I found a similar article on the same topic, I built a similar mod based on this, on my first perfboard BD-2, years ago : i guess it checks out.

https://www.analogisnotdead.com/article25/circuit-analysis-the-boss-bd2

"Bass is maxed out, while Mid and Treble are kept at their minimum value"

Analogisnotdead article is contradicting premierguitar article about the mid setting !
This is getting confusing... I think BD-2 is quite neutral or flat on the mids, so i would guess analogisnotdead got it right ? It probably depends on which fender tonestack they took as comparison...

Let's notice that analogisnotdead article does also say : "Lowering R12 will lead to less bass frequencies, (...) You can also increase the mid frequencies by increasing R11." (R12 being 1M, R11 being 15k).

So i still think i am right about the bass and mid cut...  sorry to tell you this so bluntly (especially with your current Flintlock situation), it means that you have a critical issue with your BD-2 circuit.


What's your project ? Will it be a perfboard build, true bypass, without the buffers ? Will you get Aion's board, to get the buffers and 2-knobs Eq controls ?

[matopotato]

R13 and R14 as 1M and 15k vs 100R and 8k2 was difficult for me to tell the difference so I had to bring them to the side and use a one wire swap setup for faster A/B testing before I forgot the previous tone. I did not detect much difference to be honest. The stock (1M and 15k) took down the intensity a notch and with it some shrill (but it is a very close call), so I think I will stick with stock.

That's a surprise. Now I wonder if I am mistaking... Let's check this now :

If I read you correctly, you're saying that decreasing R13 & R14 values result with few changes, taking "the intensity" up a notch, like some kind of a slight boost. Toggling back to stock values quiet things down a bit.

Yes, more or less. In my limited world the difference is a bit subtle.

I guess the question is : In the BD-2 Tonestack, if we decrease the values at R13 & R14 (R50 and R51 on the article schematics), does it do a bass and mid boost, or a bass and mid cut ?

I did not feel much impact on the tone to be honest.

Edit : " a Fender-type 3-band tonestack with fixed values (with the treble on 0 and the bass and mid on 10)"  quoting from https://www.premierguitar.com/gear/boss-bd-2-mods

This legendary article says that with fixed values, (assuming "10" means fully clockwise) it's like having bass at max with 1M, Mids at max with 15k.

15k and 1M pots can only go one way : toward 0k.

So, I would have thought that if we use any values under 1M and 15k, it would be like turning these pots down, CCW : resulting by some bass & mids cut, right ?

From 1M to 100R for the bass resistor value, I'd say you should easily hear the difference, with a huge noticeable bass cut ?

Maybe so, but I didn't. On the other hand I was testing with Tone pot high and Bass pot really low. Should try again but with much more Bass "on" from the start.

Sorry if i am raising doubts over nothing here, but it feels like I should hold my position on this....

No problem. I think you should.
Maybe we experience this build differently.
I have it on breadboard with many wires and components picking up surrounding disturbances and what not. Perhaps there are some not 100% perfect connections as well. Breabdoarding to me feels less like some exact science and more of prototyping to test things so you can make soldering decisions.

Edit 2 :  I found a similar article on the same topic, I built a similar mod based on this, on my first perfboard BD-2, years ago : i guess it checks out.

https://www.analogisnotdead.com/article25/circuit-analysis-the-boss-bd2

"Bass is maxed out, while Mid and Treble are kept at their minimum value"

Analogisnotdead article is contradicting premierguitar article about the mid setting !
This is getting confusing... I think BD-2 is quite neutral or flat on the mids, so i would guess analogisnotdead got it right ? It probably depends on which fender tonestack they took as comparison...

Let's notice that analogisnotdead article does also say : "Lowering R12 will lead to less bass frequencies, (...) You can also increase the mid frequencies by increasing R11." (R12 being 1M, R11 being 15k).

So i still think i am right about the bass and mid cut...  sorry to tell you this so bluntly (especially with your current Flintlock situation), it means that you have a critical issue with your BD-2 circuit.

Thanks, it was a nice article.
Maybe I have some issue with my build, but will be hard to tell until it is done. Now on breadboard it is a bit "open to suggestions" from both me and apparently from the build itself as well.
I am merely trying to figure out if the Galaxie mod was for me, and I decided "No".
Next, is the "fizz" together with limited usefulness of pots (as a consequence) annoying enough to do something: "Yes, at least try. If not then keep stock"
And now I am at finding out what makes a useful and wanted difference in my case.
Btw, all pots contribute with what they should.
Diode swaps between Si and Ge give a clear (well, distorted, but...) difference for the 4 (or 5) diodes between C7 & C9. The 2 (or 3) ones at C17, C18, C19 area not really. But Brian Wampler wrote that the effect was subtle and more of a protection for the OpAmp clipping, and he kept the change so why not be consistent?
Did you do his C22 and C101 change as well? Seems his C22, our C3 is the same anyway, although the C101 suggested by him as 100nF is our C16 at 18n. That would make the diode switch a bit more complicated, so I hope I can stay with only switching the diodes.

I am wondering what R18 and C11 actually do. Strangely, if I remove either, the sound stops, so I might have done something backwards here. Since they only go to ground I do not see how they should cut the signal. Chnage it, yes, but not cut.
Maybe that is my "critical issue". If so I will have to try things out again.
Anyway, I first thought they were a Low Pass filter, but that would have taken the signal onward from between R18 and C11 and not from the "top" side of R18. Do you know what this use of R&C does? Also comparing with Brian's schematic, they are swapped, and if they were a low pass filter that would have had different effect, so I doubt they are.
"There are two of these types of gain stages in the BD-2, controlled by a dual gang 250k pot wired as a variable resistor. Just like the IC opamp circuits, there is a resistor/capacitor pair going to ground that will also help set a frequency to clip. This pair will also help set the gain, though they are fixed values in the BD-2. R31 and C22 are the pair for the first stage, while R15 and C9 are the pair for the second stage."
So some kind of filter frequency definition I guess...

What's your project ? Will it be a perfboard build, true bypass, without the buffers ? Will you get Aion's board, to get the buffers and 2-knobs Eq controls ?

It is the Sapphire by Aion FX (topic title is a bit of a hint    )
No perf board.
Yes I have Aion's board with Tone and Bass controls.
As mentioned earlier, I was focused on the (eternal) "to fizz or not to fizz" question, so I wanted to have low bass, high treble and high gain.
And then just kept on trying things out in that setting. (I have begun to get a bit used to it as well...)
But I will check out R18 connection and also with more bass and less treble to see if also the R13 and R14 variants have some impact in my case.
Then when I go PCB later on, I suppose things might sound a little bit different. We'll see.

[eh la bas ma]

Did you do his C22 and C101 change as well? Seems his C22, our C3 is the same anyway, although the C101 suggested by him as 100nF is our C16 at 18n. That would make the diode switch a bit more complicated, so I hope I can stay with only switching the diodes.

I didn't change the caps values, I can't remember if i tried a 100n at C16, but I kept the 18n greenie cap. It probaly changes the tone control action for even more bass range ?

The change about C3 isn't exactly the same : same value but different kind of cap. From what I could find on musikding, metal film cap are those big brown cap, with a vintage look.
It could do a slight difference in sound, probably unoticeable on its own, but many tiny details can make a noticeable difference when they're adding up.

If i understand correctly, Boss often choose their parts depending on what they have available in large quantity, and also for economic reasons. When a modification on a Boss circuit involves a change of cap's category, it's often an interesting upgrade, in my opinion, even if there isn't a direct noticeable change. I will try this metal film cap at C3...

[matopotato]

I did some more testing yesterday.

The R18 and C11 was not breaking the signal, just making it very much weaker when removed, so I suspect nothing wrong in the circuit there.
I think I will put a switch to select 2k2 and 1uF on one setting and 10k and 100nF (from Galaxie) on the other. It does dampen the fizzy gain a bit and might come in useful although I might not use it all the time.

The diodes I already checked and just have to set them up in a DPDT for the two locations. It will probably be a bit of a messy nest, but some shrink tubing or tape should cover that.
So two new holes to drill.

I also tried the R13 and R14 variants and this time with lots of bass dialed in from the start. Yes there is a small difference in my case, but not enough to construct a third switch. I do not think my circuit is necessarily wrong, but we might have other differences that make it more pronounced for you than it is for me.
I hope I will be happy with my version, right now it is quite loud and somewhat strained at max Gain and Tone, but I think I will find good use of it.

[matopotato]

I ended up with an SPDT to switch between 2k2/1uF on one setting and 10k/100nF. I called it EQ but it is just as much boost or "anti-boost".
The build with original values 2k2 and 1uF turned out to be really loud, so that 10k/100n left from Galaxie is really useful. It makes the knobs contribute full turn, instead of having to keep level and gain quite low.
The Ge and Si goes between stock and Wampler's diode suggestions mentioned earlier. Not a huge impact on the overall tone, but at least I like the knowledge of having Ge in there. Not sure why...

All in all an nice build, sounds close to the original, but the taming of the fizz to some extent with the minimal mod above was needed in my case.