Boss DS1 mod for more volume, WITHOUT changing clipping diodes ?

[Xavier]

I am re-re-modding one of my Boss DS1's , and one of the things I did was to put back the back to back 1N4140 diodes...I really like it like this rather than with LED's

However, this means it again has the "barely less than unity gain" issue, where the pedal engaged sounds a bit quieter than disengaged.

I am looking at the output buffer here

I am wondering if the output buffer can be turned into a booster. By following an older post from user BAARON , I already changed R22 to 100R, C14 to 10mF and R23 to 10K, but no noticeable change in volume.

I was also wondering if a gain recovery stage with fixed gain (say 2 dB) could be inserted between Tone 2 and Level 3, before the switching, and what should that be ? Somehow I believe it would have to be integrated into the circuit, as part of it

Do you have any idea how this could be achieved ? Thanks !!

[stonerbox]

I would recommend you to read up on voltage dividers (R22 and R23). The changes you have done (100+10k) renders the exact same result as the 1k+100k.

[merlinb]

In an ideal world you would put the clipping opamp right after the first booster stage, then use the second opamp to add a little gain instead of being a useless buffer stage, but I guess you can't do this in an actual Boss build.

[Steben]

Quick question: do you prefer si diodes because of the higher amount of distortion?

[Xavier]

In an ideal world you would put the clipping opamp right after the first booster stage, then use the second opamp to add a little gain instead of being a useless buffer stage, but I guess you can't do this in an actual Boss build.

That's the key......converting this buffer stage into a booster stage, but I do not know if that would be possible....

[merlinb]

That's the key......converting this buffer stage into a booster stage, but I do not know if that would be possible....

Perhaps you could remove the opamp and replace it with an IC socket. Then make a little board that plugs into the socket, with the reconfigured opamp and extra gain components suitably fitted to this little board?

[Steben]

That's the key......converting this buffer stage into a booster stage, but I do not know if that would be possible....

Perhaps you could remove the opamp and replace it with an IC socket. Then make a little board that plugs into the socket, with the reconfigured opamp and extra gain components suitably fitted to this little board?

Not really, the diodes will still clip to the same headroom.
Why not adding extra low knee voltage diodes like Schottky's and raising the gain of the opamp stage a bit? This will add some level while retaining most of the character.
Another option is changing the BJT with a P channel JFET.

[mimmotronics]

What you have is a common-collector BJT configuration for the output buffer:
https://www.electronics-tutorials.ws/amplifier/common-collector-amplifier.html

Looking at the diagram, it seems like the most relevant thing to do is convert the Common-Collector configuration into a Common-Emitter configuration, which introduces the ability to add some gain to the signal:
https://www.electronics-tutorials.ws/amplifier/amp_2.html

I can't run the numbers this moment, but this would be the path I would start walking down first.

[merlinb]

Perhaps you could remove the opamp and replace it with an IC socket. Then make a little board that plugs into the socket, with the reconfigured opamp and extra gain components suitably fitted to this little board?

Not really, the diodes will still clip to the same headroom.

The gain stage would be moved *after* the clipping stage, as I explained:

In an ideal world you would put the clipping opamp right after the first booster stage, then use the second opamp to add a little gain

Why not adding extra low knee voltage diodes like Schottky's and raising the gain of the opamp stage a bit? This will add some level

Lower-knee diodes will *reduce* the level.

[Steben]

Perhaps you could remove the opamp and replace it with an IC socket. Then make a little board that plugs into the socket, with the reconfigured opamp and extra gain components suitably fitted to this little board?

Not really, the diodes will still clip to the same headroom.

Not if the gain stage is moved *after* the clipping stage, as I explained.

If one moves the gain stage after the clipping, the character changes entirely and you get lower diode clipping distortion. On top of that, you have a huge complex job in front of you.

Why not adding extra low knee voltage diodes like Schottky's and raising the gain of the opamp stage a bit? This will add some level

Lower-knee diodes will *reduce* the level.

Adding diodes in series is raising the total knee voltage.

[Steben]

What you have is a common-collector BJT configuration for the output buffer:
https://www.electronics-tutorials.ws/amplifier/common-collector-amplifier.html

Looking at the diagram, it seems like the most relevant thing to do is convert the Common-Collector configuration into a Common-Emitter configuration, which introduces the ability to add some gain to the signal:
https://www.electronics-tutorials.ws/amplifier/amp_2.html

I can't run the numbers this moment, but this would be the path I would start walking down first.

PNP tranny to the rescue! You need to add a resistor to control the subtle gain.

[merlinb]

If one moves the gain stage after the clipping, the character changes entirely and you get lower diode clipping distortion. On top of that, you have a huge complex job in front of you.

That opamp is currently a unity-gain buffer that serves no purpose. I'm saying he could put it between the clipping stage and the tone stack instead (and give it a little gain), which will result in no change in clipping and no noticeable affect on character.

Adding diodes in series is raising the total knee voltage.

Ah makes sense now.

[merlinb]

PNP tranny to the rescue! You need to add a resistor to control the subtle gain.

But isn't the OP modding a genuine Boss DS-1? In which case modifying that transistor will affect both the effect and the bypass signal equally. Maybe I misinterpreted the OP though.

[Steben]

PNP tranny to the rescue! You need to add a resistor to control the subtle gain.

But isn't the OP modding a genuine Boss DS-1? In which case modifying that transistor will affect both the effect and the bypass signal equally. Maybe I misinterpreted the OP though.

True, but it is the OP which mentioned the output buffer.

I would settle for adding schottky diodes to the silicon ones and adding gain to the gain stage.

[mimmotronics]

I would settle for adding schottky diodes to the silicon ones and adding gain to the gain stage.

What model Schottky's are the standard? For Si diodes I'm familiar with 1N4148's and 1N400x's, but what are the Schottky equivalents?

[Steben]

I would settle for adding schottky diodes to the silicon ones and adding gain to the gain stage.

What model Schottky's are the standard? For Si diodes I'm familiar with 1N4148's and 1N400x's, but what are the Schottky equivalents?

BAT85

[PRR]

> PNP tranny to the rescue!

As drawn, since the bias is V/2, it won't bias usefully except at gain less than unity.

[mimmotronics]

> PNP tranny to the rescue!

As drawn, since the bias is V/2, it won't bias usefully except at gain less than unity.

I don't see the need to sub in a PNP, my first approach would be to keep the NPN but put a resistor between +9V and the collector. Care to explain?

The trick to this approach is finding a way to modify the signal bias, and then balance that with the gain resistance so the volume can be brought up enough to where you want it... that sounds like a lot to go through though.

The schottky approach is much simpler to tackle first.

[Steben]

> PNP tranny to the rescue!

As drawn, since the bias is V/2, it won't bias usefully except at gain less than unity.

I don't see the need to sub in a PNP, my first approach would be to keep the NPN but put a resistor between +9V and the collector. Care to explain?

you wont get much gain if the signal comes from the emitter.

The trick to this approach is finding a way to modify the signal bias, and then balance that with the gain resistance so the volume can be brought up enough to where you want it... that sounds like a lot to go through though.

The schottky approach is much simpler to tackle first.

yes it is