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

[eh la bas ma]

Hello,

My last build is a BD-2 with Galaxie mod :

https://aionfx.com/app/files/docs/sapphire_documentation.pdf

As explained in instructions :

"This modification—or more accurately, major surgery—changes out about half of the components to completely restructure
the pedal. It lowers the gain and fully eliminates the fizz while improving the tone control’s usefulness."

I was a bit surprised by the result, the circuit acted like a volume pedal with a tone and a bass control. No gain, no compression at all. I can't see the interest, so I think there's probably an issue with my build...but i am not sure.

I can hear that it's not behaving like the exemples I could find on youtube, there is much less gain on my build :

https://www.youtube.com/watch?v=ua4fvQSzCtw
https://www.youtube.com/watch?v=5-etn9qaxFE

I decided to put back a couple of diodes (D6 & D7) and i got a little bit of compression with Gain fully CW, but not really enough.
Then I added a switch with the remaining diodes (D2, D3, D4, D5) and the gain knob is finally doing something useful.

I suspect something is wrong. There are 5 smd transistors, I did a continuity check following schematics and all looks good.

I am using a RC4558P instead of JRC4558D, maybe that's the problem ? I found this thread, but even these wise forumites don't really answer the question unfortunatly...

https://www.diystompboxes.com/smfforum/index.php?topic=110891.0

Here are my readings:

DC 9V+ = 9.45 V

1= 4.22 V   8= 8.45 V
2= 4.22 V   7= 4.30 V
3= 4.22 V   6= 4.30 V
4= 0 V        5= 4.095 V

There are 10 transistors, I can give my readings if required.

Some pictures, in case I missed something obvious :












Is there really an issue with my build ?

Is the Galaxie mod from Aion supposed to decrease the gain untill there is almost none ?

If we take every diodes out of the modded BD-2 circuit, how is the saturation created ? is it the op-amp we are hearing in the utube exemples ?

[idy]

It is going to be hard to provide real deep analysis because there isn't a schematic for the version you are using; many changes from the BD-2, hard to follow without printing out the schematic and then penciling in the changes. That being said, fools rush in...

It is very unlikely that the opamp (which has good voltages) is a problem, or that some subtle difference between batches will account for less gain.. it is being used as a VB buffer and as the bass control with a transistor gyrator.

It is also unlikely that the opamp is supposed to be a distortion element at all.

The likely "distorters" are the gain stages, each made of two JFETs and a BJT. The JFETs are called a differential pair and with the BJT they make an opamp. The best read I found is at
https://www.analogisnotdead.com/article25/circuit-analysis-the-boss-bd2

The part where it says

most of the hard clipping occuring in fact comes from the opamps

means the discrete opamps. So there. They should be clipping.

The likely way to find your issue is with a signal trace. Are those two gain stages boosting?

[antonis]

Try to lower R7 & R18 resistor values..

P.S.
I'd delete C26 (or put a 100R resistor between IC1_A out and C26) and alter Q1 and Q9 bias configuration, if I were a snoopy guy..

[eh la bas ma]

Thanks for your replies !

So, I guess I should audio probe the gain stages, and if everything sounds ok, I 'll try Antonis suggestion.

Just to make sure : According to Aion's schematics, I should probe Q4 on the R10 side, then Q7 on R21 side ? These are the outputs of the two "discrete op-amp stages" connected to the gain pot ?

[idy]

Those are the outputs, yes. First probe the inputs so you can compare volume. Gates of Q2, Q5.

[eh la bas ma]

Here are my results :

Q4 output (collector) is much louder than the middle leg (input, Base), same for Q7.

Q4 output is a bit louder and a bit more distorted than Q7 output.

The distorted signal is rough on the edges, and quite sputtery.

The signal is dirty and crackling on both inputs (Base), it gets cleaner if I decrease the Gain, but I am not sure, hard to hear as the volume gets quite low. Q4, Q7 input volume is increasing, if I turn the Gain CW.

I also notice some background noises on both inputs, the volume of the background noise varies gradually and apparently randomly, untill it totally disappears.

Keeping the same settings, Gain fully CW, the signal is much cleaner on the circuit output compared to Q4 or Q7 output.

I am not sure what to think about these results. Do you think there is something wrong ?

[eh la bas ma]

First probe the inputs so you can compare volume. Gates of Q2, Q5.

Oh... I just noticed my mistake, I probed the wrong inputs.

I probed again and I have differents results than yesterday, I don't know why, today there is no distortion on inputs and outputs with gain at max. A mystery to me, I hope I'll understand why later.

Edit : I was using hearphones yesterday, and not today, maybe that's it...earphones were saturating or something...

I recorded a loop just like yesterday, the signal is distorted if I try an other gain effect...

On the first gain stage, Q4 ouput is clearly louder than Q2 input. About roughly + 8 or 10 db, I would guess.

But  Q7 output is at the same volume than Q5 input, maybe 1 or 2 db louder but hardly noticeable. I tried to change Q7 2N3906, and I have the same result.

I guess I will try to reflow some pads now...see if I can get Q7 output louder than Q5 input.

Edit 2 : it's not the earphones that produced the distortion yesterday, I just tried them and there is still no distortion.

I notice Q2 input is at the same volume level than the jack IN. Q5 input is louder than Q2 but a bit quieter than Q4 output.

[idy]

You are starting to make sense of it.
You expect Q2 in to be the same as inject: Q1 is a buffer
Good you are hearing a louder signal out of Q4.
It is normal that

Q5 input is louder than Q2 but a bit quieter than Q4 output.

There is a lot of filtering going on in between.

Now you need to get Q7 out louder than Q5 in. How is the second stage different from the first? It should work the same way!

I know there are two Rs Antonis posted to, the mods changed those to 10k... I think those are feedback loop elements, "ground legs."

[eh la bas ma]

I reflowed some pads around Q5 Q6 and Q7, the signal is a bit louder on Q7 output than on Q5 input. But the difference in volume is really lower than the difference on the first gain stage, between Q2 in and Q4 out.

Isn't it because Q5 input is already louder than Q2 input ?

Just to make sure, should be obvious but G on these smd transistors is the middle pin, the one alone on one side ? The BOM on musikding is unavailable, should be 2SK209-GR, but I bought a kit and I have no way to look for the pin out on musikding. This is all I can find :

https://eu.mouser.com/datasheet/2/408/2SK209_datasheet_en_20140301-1594295.pdf

my readings on transistors :

Q5

G = 4.17 V
S = 4.60
D = 7.79 V

Q6

G = 4.28 V
S= 4.60 V
D= 8.46 V

Q7

E = 8.46 V
B = 7.80 V
C = 4.29 V

Readings on Q2 and Q3 are the same than Q5 and Q6. Same voltages too for Q4 and Q7.

[eh la bas ma]

I circled in red the points where I am probing, gain stages inputs and outputs... in case I am doing it wrong...

[eh la bas ma]

I tried the first part of Antonis 's idea.

I have soldered a 4k7 at R7 and a 5k6 at R18 (standard BD-2 uses 1k5 and 2k2). No obvious change on the gain stages or on the Gain control. Do you think I should try the BD-2 values ?


The circuit can produce some distortion with the signal if I increase the guitar volume (active hambucker pickups) or if I increase the input volume somehow. But with my usual volume setting, volume around 12' on my guitar, it's very clean and the circuit can only be used as a weak booster.

It gets much closer to a BD-2 and to the youtube exemples only when the diodes are back in the signal path.

These diodes are located between the two gain stages, maybe that's where the lack of gain is coming from...maybe I made a mistake  wiring the diodes switch ?

Does it look correct to you ? 

[antonis]

I reflowed some pads around Q5 Q6 and Q7, the signal is a bit louder on Q7 output than on Q5 input. But the difference in volume is really lower than the difference on the first gain stage, between Q2 in and Q4 out.

Isn't it because Q5 input is already louder than Q2 input ?

Quite right..
Q4 is the VAS (Voltage Amplifier Stage) following Q2/Q3 differential pair stage, so Q4 output (Collector) should hit supply rails..
(distortion dependent on input signal amplitude..)
Q5/Q6 stage input amplitude should be about 1.3V due to hard clipping diodes pair and Q7 output should be louder that Q5 Collector (Q7 Base) but in a SMALLER proportion compared to Q2/Q4 one..
[ 4.5V/1.3V compared to 4.5V/0.5V(say)]

Your measurements are good enough, indicating a healthy good-working circuit..
In case you indeed want more gain, there are 2 options:
Lower Q3 and/or Q6 NFB amount (via the way previously suggested) or lower C5 and/or C12 Miller feedback capacitors values..
(both have their pros and cons but such an analysis should be beyond the scope of present thread..)

[antonis]

I tried the first part of Antonis 's idea.
I have soldered a 4k7 at R7 and a 5k6 at R18 (standard BD-2 uses 1k5 and 2k2).

I think, Antonis told you to LOWER R7 & R18..!! 

The above mentioned resistors are the lower part of NFB respective voltage dividers..
By raising their values, you raise the amount of NFB applied on Q3/Q5 Gates, hence you lower stages gain..

[eh la bas ma]

but... but I did lower the resistors ? 

I applied the Galaxie mod to the circuit, and both resistors were 10k, according to the instructions.

You think I should lower them again, back to stock 1k5 and 2k2 or even lower ?

I just wanted to know if there was something wrong with my build and clarify all doubts.

I don't think I will modify the modification for more gain if everything is as it should be, because I have the diodes switch which works well.

I still have doubts about how I wired it though... a mistake here could probably impact the following gain stage. Can you please confirm the wiring is correct on the switch (see reply #10) ?

[antonis]

I have to admit that I don't understand your diode switch wiring..

Exactly what do you wish to do..??

edit: If you want to bypass them, just implement a SPDT switch lifting D3 cathode & D4 anode grounding..

[eh la bas ma]

I wanted to have on a switch exactly the same thing happening on the board with the diodes.

I noticed they are connected to ground, so I tried to recreate that somehow.

Green and orange wires from the diodes goes to the pads connected to ground (D3- D4+).

When I want to disengage the diodes, the signal coming from R13 goes to D2+ and D5- which aren't connected to ground. And the signal goes its way, without going through diodes ?

...but it does look weird...there's probably something wrong...

[antonis]

In case all diodes are soldered on board and there is no trace cut to respective GND or D2/D5 joint, switch is useless..

[eh la bas ma]

In case all diodes are soldered on board and there is no trace cut to respective GND or D2/D5 joint, switch is useless..

you mean with an SPDT as you suggested ? it wouldn't work without a cut ?

Right now with this wiring, it is really effective ...even if it's suspicious, I can't see how it's wrong ?

I would prefer not to cut traces on the board.

Edit: but I will do it if there is no other way. Aion's board are sacred, aren't they ?

[antonis]

You can try this: 

[matopotato]

@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