New kind of Mosfet Boost - zBoost?

[DaveLT]

I haven't come up with a name but Mosfet and boost as an idea isn't new nor is the schematic, it is basically Nelson Pass's BoZ modified because there isn't a need for the linearity and 60V is something else to behold... I'm just having fun as I want to find a reason to play with power mosfets

But if you like a very clean boost try this one

(Ignore the P-channel mosfet, diy LC isn't terribly accurate with positive vs negative devices)

As explained in the BOZ notes : The gain of the circuit is the ratio of the output loading divided by the apparent resistance formed by the 330 resistor R1 and the inverse of the transconductance of the MOSFET. which is about 340 ohms

At audio frequencies the output load is the 10k of R3 in parallel with the combination of output potentiometer VR1 50K (the gain control) and the load. If you find the gain is too high and can't vary it much with the gain pot you can dial R3 down to 1K
Yes it uses 24V which is unlike anything that is ever used in pedals but there's reason for the madness : Clean boost and some of my projects use 24V and you can use a common 24v SMPS as a result

[antonis]

If you find the gain is too high and can't vary it much with the gain pot you can dial R3 down to 1K

Pardon me but I can't see any Gain pot.. 

IMHO, lowering R3 should upset Drain bias voltage which might result in signal negative waveform clipping (depending on signal peak & gain..) so you'll have to re-bias Q1 via VR1..
But I'm not sure if 9.6V on VR1 upper lug (due to R6/VR1 voltage divider) should be enough for bias trim with R3 down to 1k..
(you'll roughly need 10 times bigger Drain current to set Drain at the same quiescent voltage..)   

P.S.
Of course, well known MosFet boosters hopeless lovers (did I just hear blaring Sir Mike coming at furious gallop..??  ) could enlighten us more..

[DaveLT]

If you find the gain is too high and can't vary it much with the gain pot you can dial R3 down to 1K

Pardon me but I can't see any Gain pot.. 

IMHO, lowering R3 should upset Drain bias voltage which might result in signal negative waveform clipping (depending on signal peak & gain..) so you'll have to re-bias Q1 via VR1..
But I'm not sure if 9.6V on VR1 upper lug (due to R6/VR1 voltage divider) should be enough for bias trim with R3 down to 1k..
(you'll roughly need 10 times bigger Drain current to set Drain at the same quiescent voltage..)   

P.S.
Of course, well known MosFet boosters hopeless lovers (did I just hear blaring Sir Mike coming at furious gallop..??  ) could enlighten us more..

The output pot is THE gain pot  Actually the original preamp was design for 1k R3 and 100 ohm R1 so it doesn't matter much. As it is right now it running at 12mA bias and the bias from the divider just need to be set at 7.5V, more than enough for the Vgs

[antonis]

The output pot is THE gain pot 

Couldn't agree more if ONLY output pot was part of whatever comes after Booster..
(as it is, it's a Volume pot, in the mean of a humble voltage divider which always sets its "body" in parallel with R3 - at least for high impedance following stage or "open" output..) 

As it is right now it running at 12mA bias

If you mean with 10k Drain resistor, you have to place a decimal point between 1 & 2 .. 

[DaveLT]

The output pot is THE gain pot 

Couldn't agree more if ONLY output pot was part of whatever comes after Booster..
(as it is, it's a Volume pot, in the mean of a humble voltage divider which always sets its "body" in parallel with R3 - at least for high impedance following stage or "open" output..) 

Yes as I explained the gain of the circuit is the ratio of the output loading divided by the apparent resistance formed by the 330 resistor R1 and the inverse of the transconductance of the MOSFET.

As it is right now it running at 12mA bias

If you mean with 10k Drain resistor, you have to place a decimal point between 1 & 2 .. 

No, no the bias is set up by 4V running through the source resistor! 7.5V is to fully open up the gate to conduct. good catch, I didn't think of dividing the divider further. Change it 150K and 100K pot!

[phazon]

What would happen if you put another diode facing the opposite direction next to D1, like what you see on a tube screamer and various others?  Is that possible without blowing something up?

[antonis]

Somebody lost someone, here...

I'm trying to tell you that output pot, as it's wired, doesn't alter Drain total load despite its setting..
(you may consider it with its lower leg connected to power supply instead of GND, effectively setting it in parallel with R3 - with its wiper "open" or connected to a fair enough high impedance, you have a constant 8k3 Drain load and consequently a fixed Gain of 8333/670..)
It should work as Gain pot in case it was wired as variable resistor (in series with a cap) in parallel with R3.. 


>No, no the bias is set up by 4V running through the source resistor! 7.5V is to fully open up the gate to conduct.<

4V across 330R indeed sets Drain current at 12mA..!!
BUT..
for 10k Drain resistor it permanently bottoms Q1..!! 

With "placing decimal point" said above, I refered to 1.2mA which should be the correct working current for 10k Drain resistor..

[antonis]

What would happen if you put another diode facing the opposite direction next to D1, like what you see on a tube screamer and various others?

What you see on a tube screamer is a pair of diodes inside NFB loop..
(which would be the case here if they were placed between Drain & Gate..) 

D1 zener is placed here for Gate-Source over-voltage prevent.. 

[GibsonGM]

What would happen if you put another diode facing the opposite direction next to D1, like what you see on a tube screamer and various others?  Is that possible without blowing something up?

Nah. The purpose of D1 is to protect the gate from voltage spikes.  It normally does not conduct, but if you hit the gate with a voltage above the 20V zener voltage, it will send it to ground rather than thru (literally; pop) the gate.   The data sheet for the 610 says absolute max ratings for gate-source voltage is 20V.  The zener will hold it to this (ideally...).   I might choose a little lower zener just to be safe.  No input should be 20V! So something lower would be fine, 12V maybe...

You COULD try diodes from drain to gate for a soft-clipping feature.  Won't be much of a boost at 24v then, though....since this is meant to be VERY clean, I wouldn't go that route...I'd set it back to 9V and do it then.

Antonis - you know much more about the fine points of mosfets and so on than I do   I'm a mechanic, not a theorist!  Ha ha.

[antonis]

I didn't think of dividing the divider further. Change it 150K and 100K pot!

First in first, it isn't polite to edit posts when somebody is quoting them..!! 

Changing resistane values while keeping the same ratio doesn't further result to anything.. 
(except to alter divider's current consumption..)

I was talking about the possibility of not enough maximum VR1 voltage for new bias setting..
(which case doesn't stand for 7.5V Gate bias..) 

[antonis]

You COULD try diodes from drain to gate for a soft-clipping feature
....................................................
I'm a mechanic, not a theorist!  Ha ha.

Mechanics bother with clipping or not..
(soft-medium-hard etc are unknown words with no meaning to them..)

[rankot]

Maybe you shall take a look at this page: https://www.diystompboxes.com/smfforum/index.php?topic=119198.msg1111425#msg1111425

[DaveLT]

Somebody lost someone, here...

I'm trying to tell you that output pot, as it's wired, doesn't alter Drain total load despite its setting..
(you may consider it with its lower leg connected to power supply instead of GND, effectively setting it in parallel with R3 - with its wiper "open" or connected to a fair enough high impedance, you have a constant 8k3 Drain load and consequently a fixed Gain of 8333/670..)
It should work as Gain pot in case it was wired as variable resistor (in series with a cap) in parallel with R3.. 


>No, no the bias is set up by 4V running through the source resistor! 7.5V is to fully open up the gate to conduct.<

4V across 330R indeed sets Drain current at 12mA..!!
BUT..
for 10k Drain resistor it permanently bottoms Q1..!! 

With "placing decimal point" said above, I refered to 1.2mA which should be the correct working current for 10k Drain resistor..

Um... I left *the output pot which sets the gain* as is. Check the original schematic.

[DaveLT]

Somebody lost someone, here...

I'm trying to tell you that output pot, as it's wired, doesn't alter Drain total load despite its setting..
(you may consider it with its lower leg connected to power supply instead of GND, effectively setting it in parallel with R3 - with its wiper "open" or connected to a fair enough high impedance, you have a constant 8k3 Drain load and consequently a fixed Gain of 8333/670..)
It should work as Gain pot in case it was wired as variable resistor (in series with a cap) in parallel with R3.. 


>No, no the bias is set up by 4V running through the source resistor! 7.5V is to fully open up the gate to conduct.<

4V across 330R indeed sets Drain current at 12mA..!!
BUT..
for 10k Drain resistor it permanently bottoms Q1..!! 

With "placing decimal point" said above, I refered to 1.2mA which should be the correct working current for 10k Drain resistor..

I finally see your point. God knows what I was on when I spec'd 10K. But the issue with 1K and 100 ohms source resistor is that it sets gain to somewhat like 10...
More experimentation needed.