Tillman Preamp

[Moosehead]

I was wondering if there's a way to optimize this circuit for high output pickups? I've read that raising the supply voltage from 9v to 18v makes a difference, but is there anything else I should do to make sure the preamp stays clean without overdriving?

Thanks.

[petemoore]

  Running @ higher voltage will raise the headroom, whether you notice a difference depends on if there was a lack of it to begin with, using your input source.
  "nothing I can think of' [in answer to your question].

[Jeremy]

You can make a buffer instead of a preamp, with nearly the same components.  The preamp has voltage gain, which would make clipping a little more likely; a simple buffer does not have voltage gain.  To change the existing circuit into a buffer, connect the output cap to the source of the jfet instead of the drain.  You can then omit the 6k8 resistor if you like, and connect the jfet drain directly to +9 volts.

Or, you can use an 18 volt power supply, as you mentioned.

Or you can keep your guitar turned down. 

[Jeremy]

On second thought, the type of jfet you use might cause my idea to distort more, if it doesn't bias right.  But it was a thought.

[Gus]

This one?
http://www.till.com/articles/GuitarPreamp/

If so you will need to adjust values for higher voltage.  Start with the drain resistor first adjust the resistor so the drain is at about 1/3 to 1/2 of the supply voltage you want

also a FET has a Max Id so the circuit will need a fet with a higher IDSS rating than the Id used you might need to increase the drain R to something like a 10k or...

IDSS is the max current a jfet will conduct with gate to source connected.

.  The use of the j201 is good because it has tighter specs than some other fets you see used in effects and guitar circuits.

The tillman uses a 6.8k drain R so Id at Q would be 6.8k/voltage drop across the resistor say 4.5 about .67ma

If the pickup has a high output maybe use a SF circuit.

[davebungo]

It is a little more complicated than just raising the supply voltage and here's why:

The bias is set by a combination of the FET characteristic and the 2K2 source resistor.  Looking at the graphs for the J201 and picking an average characteristic it would be reasonable to pick an ID of say 300uA.  This will produce a VGS of -0.66V and a VS of 0.66V since the gate is grounded i.e. 300uA * 2200.  Whatever this voltage is, the drain voltage will be Vsupply - 6800/2200 * VS because the same current flows through both drain and source resistors.  For 9V this will be 6.96V.  Remember these are just guesses from the datasheet but they won't be too far out.  VDS will be 6.96 - 0.66 = 6.3V.  The original circuit probably would have had the drain a bit lower but this does not matter in this case.  The + swing will be limited to +2V maximum but remembering that this stage has a gain of about 3 (6800/2200), this translates to only +0.67V at the input. 

So what happens if you raise Vsupply to say 18V?  Well VD will simply rise to the occasion i.e. to 18 - 6800/2200 * VS = 15.96V.  i.e. you still have the same + swing even though the - swing i.e. VDS has increased by 9V to about 15.3V!

So what if we just raised Rdrain to lower VD and increase the + swing headroom?  Well this would work but would also increase the gain of the circuit correspondingly.

Swings and roundabouts spring to mind.

One idea I have toy'd with is to increase Rdrain but put a bypass cap and resistor across it.  Note, *not* across Rsource.  If you adjust Rdrain to give you a more symetrical potential swing and select a parallel R which would give roughly the original Rdrain as far as AC is concerned then this should work.  The C would have to be quite large >1uF.

Using the example figures, if you set Rdrain to 13K and put in parallel with this another 13K resistor and a 10uF cap, then this should give you a more symmetrical swing.  You can use the same calculations to predict roughly the right starting values for an 18V supply.

The thing to remember here though is that although your guitar plus the combination of a super pre-amp with large a swing will be big and clean, it probably won't be after it hits the next effect in the chain or your amp.

Ultimately, you may need to but a volume control somewhere or use your guitar volume to limit the levels.

[Moosehead]

Wow, thanks a lot guys. I wasn't expecting so many replies so soon!!

Gus, what's an SF circuit?

[Gus]

source follower.  Gain of just under 1.

As davebungo posted there is a bit more to things when changing the circuit for higher voltages with jfets.

One is jfets are all over the place with the specs that count.

If I was to do a circuit that wanted a controlled gain but high input z I would use a source follower to a vol control to a BJT stage with the BJT part of the circuit designed for a certain gain easy to do.  With a Source follower you can do some different things gate to ground or gate to a split source R bootstapping the input for even higher input Z.


Did you build the circuit and have a clipping problem with a piezo?

[bassmeister]

According to the specs at http://www.till.com/articles/GuitarPreamp/, the output impedance is 6kOhm. Calculating... R3//R4 = R3*R4/(R3+R4) = 6. This is just a shot in the dark, but the result is right. Is this how output impedance is calculated for a common source amplifier stage?

Also, the gain is stated to be ~ 3 dB. I thought the voltage gain for CS amp was calculated as Rdrain/Rsource = 6.8/2.2 = 3.1 => gain dB = 20*log(3.1) = 9.8 dB. I suspect I'm missing something, but what? I'm seriously trying to learn, not just ranting about mathematics 

[davebungo]

According to the specs at http://www.till.com/articles/GuitarPreamp/, the output impedance is 6kOhm. Calculating... R3//R4 = R3*R4/(R3+R4) = 6. This is just a shot in the dark, but the result is right. Is this how output impedance is calculated for a common source amplifier stage?

The output impedance is roughly Rd.  It is actually Rd//ro where ro is the dynamic (i.e. for small signal variation about the operating point of the circuit) output impedance of the JFET but this is probably an order of magnitude higher than the 6K8 or so of the resistor in use.

Also, the gain is stated to be ~ 3 dB. I thought the voltage gain for CS amp was calculated as Rdrain/Rsource = 6.8/2.2 = 3.1 => gain dB = 20*log(3.1) = 9.8 dB. I suspect I'm missing something, but what? I'm seriously trying to learn, not just ranting about mathematics 

I believe the original article is incorrect.  I think he confused 3dB with a gain of 3.  The gain of this configuration is usually taken as -Rd/Rs, but in actual fact it is slightly less - just as a source follower has a gain of slightly less than unity.

Also, as the circuit has a built in 51K resistor to tie down the other end of the output cap, the level will be further reduced by about 1dB i.e. 51K / (6K + 51K).  Any further loading (volume control etc.) will reduce the level even further.

[Mark Hammer]

However you do it, unless your master plan is to always run your guitar straight into an amp, I would advise against adding any gain if using high output pickups.  There are just too many pedals that are predicated on seeing an input signal of a certain amplitude range in order to work properly.  Yes, you end up with a nice hot output signal that preserves bandwidth.  But the downside is that you impede yourself from being able to get a vast range of pedals to work properly for you unless the signal gets padded down.

[Moosehead]

This may be a long shot, but is there a preamp circuit other than the Tillman that would give more headroom for a high output pickup?

Also, I'm not sure if this has any relevance whatsoever but I looked up some stats for my pickup, the DiMarzio Norton. The output is 352 milivolts.

[Mark Hammer]

That spec depends on context and what was strummed.  I'm not disputing the spec but it may be based on circumstances you personally do not produce.

Again, if you DO have higher output pickups, though, just exactly where are you sending this signal that you need the cleanliness of a higher headroom preamp?  Unless you are sending it to a power amp directly (and I've done that myself), or to an amp designed expressly for a pedal steel (and you thought YOUR pickup was hot?), there are a whole whack of places along the way where your clean headroom turns into audio grime further along the signal path.

What you want most from that first stage is really and truly impedance reduction; high-in, low-out.  Even if there is no trace of gain added, it will sound like there was because you'll get back some of the top end and cutting power you might have lost to cable capacitance.  Any 9v battery should provide all the headroom you need to do that, even IF those pickups regularly put out 350mv.

[Gus]

google source follower.

setup right you will have good input headroom.  set it up for 10meg or more input Z. 

[Gilles C]

This may be a long shot, but is there a preamp circuit other than the Tillman that would give more headroom for a high output pickup?

Also, I'm not sure if this has any relevance whatsoever but I looked up some stats for my pickup, the DiMarzio Norton. The output is 352 milivolts.

This one maybe.

http://www.uwm.edu/People/msw/ActivePlug/BootStrap.html

I  also remember I experimented with this circuit from Jensen, but I can't remember the result...

http://www.jensentransformers.com/as/as004.pdf

[Gus]

Gilles C

Cool link.  A source follower to a EF with bootstrapping to make a quasi constant current load for the fet source leg.

[lovekraft0]

This is what I've been using for an onboard buffer:

Simple, cheap, quiet, low current, and while it's not quite unity gain, the sonic impression is one of a slight boost. No audible distortion, even with Invaders (but then, who can really tell with Invaders?). And it works with a 25K volume pot, so it'll work with EMGs (which was the whole point of the original idea - it also works great with standard volume pots).

[cd]

Yes, you end up with a nice hot output signal that preserves bandwidth.  But the downside is that you impede yourself from being able to get a vast range of pedals to work properly for you unless the signal gets padded down.

Well... that's what the volume control is supposed to be used for, right?

The nice thing with a built in preamp, provided you have it mounted in the right place, is your volume control will actually act like a volume control, and not an additional volume and high sucking control.  With some gain, you can leave your volume control turned down most of the time, then have a built in solo/whatever boost when you need it.  Of course, this goes completely against how most folks use their volume control (turned up full all the time) so will probably be cumbersome/counter-intuitive to use.

[Mark Hammer]

With some gain, you can leave your volume control turned down most of the time, then have a built in solo/whatever boost when you need it.  Of course, this goes completely against how most folks use their volume control (turned up full all the time) so will probably be cumbersome/counter-intuitive to use.

Yeah.  Which is more or less why I advised against adding gain.  Besides, maybe the solo involves use of a pedal which has limited headroom (oh, say, a FET-based phaser) and gets whupped HARD with an ugly stick when that headroom is exceeded.  In which case the user has to be fussier about how they dial in any additional volume on the guitar, and that can be a royal pain.

It is easy to imagine a whole host of circumstances where your suggestion to simply turn down makes absolutely eminent sense.  At the same time, this is a forum for folks who are more than likely to have stuff in between the guitar and amp, right?  My gut response is to suggest maintaining a reasonable guitar output which will not tax anything in that path, even at full tilt.  If the user wants to push the signal level beyond that, then that can be nicely done in a pedal, where your fingers can keep doing what they do to the strings, and your feet can substitute for what you'd otherwise have to do with your hands and guitar volume control. 

Boosting you can do at a lot of points.  Buffering against that stupid cord hanging off the guitar can only be done at one point.  So there is every reason to want to go active for buffering purposes, but many fewer reasons to make added gain part of it at that point.  As with anything, though, that's merely an opinion.  YMMV.