• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

TB11 Preamp Project

#61
But I ordered a run of everything I didn't already have laying around, so I will have a proto ready next week
Click to expand...

That's awesome I can't wait to hear some feedback before my batch of parts comes in! One thing that was omitted (on purpose) was pole compensation and DC blocking stages. I've found that keeping the phase shift to a minimum tends to 'sound better', and makes applying global negative feedback (not used in this design) easier to implement. I'm guessing that with low GBW op-amps that are low passed to around 40kHz that oscillations aren't going to be a problem, but we shall see in the prototypes.

I'm working on the analysis of the parametric mid-control. It's going to be somewhat tricky to implement as it needs another op-amp stage and I wanted to stick with four. I might be able to get rid of the blend section if I can compensate for some signal loss. Another option is to use common source JFET buffers for the pickups to save two op-amp stages, but that's the next design. ;)

One thing that comes to mind: if I implemented a sweepable 'mid' that spanned from about 300Hz (assuming a Q of 1) that went up to around 4 kHz, could the bass and treble controls be removed? Do people actively boost or cut multiple bands?

I ask because I had a 3-band EQ on my Carvin fretless (and still have one on my Ibanez SRX505) but I found I rarely touched anything but the mid control (in terms of applying boost).

If anyone cares to share the Digikey/Mouser part numbers they used I can make a 'recommended order list' to add to the schematics.
 
#62
Why not make the mid control a modular thing (like a seperate board, buffered on its own) so its possible to add more than 1? :)

I would love to have a 5-band with 3 sweepable mid sections on my Yamaha... :)
 
#64
Here is the parametric mid control:

Invalid Link Removed

If you take a look at this circuit at DC and high frequency, you can get a good idea of how it works without resorting to Laplace transforms.

Start at DC: the caps look like opens, in which case the non-inverting input looks like a resistor to ground (which is the same as just saying ground). The non-inverting input is set up with equal feedback resistors R5 and R6 which means the stage looks like a unity gain inverting amplifier (no cut or boost).

At high frequencies, C1 and C2 look like wires, and again the non-inverting input looks like ground (through C1), again which means that the circuit looks like an inverting unity gain buffer.

So obviously the interesting things happen in the middle frequencies. ;)

Taking the first case again: as we move up from DC, at a certain point there will be a signal through R3 and C2 (and R1). At middle frequencies, the input to the non-inverting section of the op-amp looks like a complex voltage divider between R3 and C2 and R2 and C1.

The gain will be maximum (maximum in the boost or cut sense) at a frequency where the voltage divider equation gives maximum signal to the non-inverting input. The +-3dB points are governed by the fundamental equations:

f_turn_low = 1/2*pi*R3*C2

and

f_turn_high = 1/2*pi*R2*C1

As the frequency pot is swept, R3 and R2 increase which causes the curve to shift downward in frequency.

Finding a transfer function involves writing the node equations, converting the circuit to it's Laplace representation, the solving for Vo / Vin. You then take the first derivitave, set equal to zero, and solve for s to find the local frequency maxima. However that's overkill for this, as the above two equations will tell you the 3db points and all you have to do to find the center frequency is multiple them together then take a square root (called the geometric mean).

Using my values:
C2 = 4.7nF
C1 = 2.2nF
R2 = R3 = 13k + 100k (pot on full = lowest frequency)
f3_low = 299 Hz
f3_high = 640 Hz
f_center = 438 Hz

R2 = R3 = 13k (pot on min = highest frequency)
f3_low = 2.6 kHz
f3_high = 5.5 kHz
f_center = 3.8 kHz

This circuit gives roughly 15dB of boost/cut from 438 Hz to 3.8 kHz. For those that are interested, the Q is equal to the 1.414 (the square root of 2), or stated another way the bandwidth of the filter is one octave (underdamped Butterworth response).

The SPICE agrees pretty well with the equations.

Invalid Link Removed
 
#65
the free OPA2277's just arrived from TI :hyper: this is going to a challenge, never built a preamp like this on a blank PCB. can somebody explain what the bold and thin red & green lines represent? I am assuming one is the signal path and the other is the ground path

Invalid Link Removed
 
#66
Four layer pcb with thru hole? I'd guess from top layer to bottom: red, light red, light green, green.
If you bought the development kit, that'd be the board that came with it.

Almost always worthwhile to spring for the kit IMHO unless you like to make your own pcbs.
They're usually nicely done and well suited to the part.
 
#68
Actually I think this board is just two layers, but the different widths of traces make the colors appear different to your eyes.


Already started on mine, about half way through at the 1.5 hour mark, going to pace myself on this... I might even side track a bit first and build the para-mid circuit (if we can find out what value the level pot is there on the far left ;)) just to see what thats like.. I also plan on building another one with Tillman discrete FET buffers for each pickup.


Made a few changes, but kept mostly to the prelim layout just to make life easier on the next step.. however my interconnects will be as quick/short as possible, not even going to worry about following them on a perfboard proto..

4809472931_33e98ed641.png


Obviously it wont fit in a bass like that, but thats not the point yet. Seeing if it works, with room to wiggle is key right now..
 
#70
can somebody explain what the bold and thin red & green lines represent?
Click to expand...

It's a two layer PCB.

Don't follow the layout to wire the prototype: follow the schematic posted above.

I'll see if I can publish a netlist that can be used for wiring to make it easier to follow and double-check the connections.
 
#72
I got my parts too! I'll be traveling over the next week so I'll try to assemble one when I get back.

I also plan on building another one with Tillman discrete FET buffers for each pickup.
Click to expand...

That's a very clean proto-board JackANSI. I like how you used DIP pin sockets so you can easily swap out part values.

I also found a few more op-amp candidates that might work: I can't see them being better than the OPA2277, but those might not be available to everyone:

OPA1641
OPA1642
MAX4126
 
#73
MglMatador said:
It's a two layer PCB.

Don't follow the layout to wire the prototype: follow the schematic posted above.

I'll see if I can publish a netlist that can be used for wiring to make it easier to follow and double-check the connections.
Click to expand...

would it be too much trouble to make a tagboard or vero layout of the circuit? that would make it a lot easier for intermediate noobs like myself to make the preamp and prevent time consuming mistakes :hiding:
 
#76
Only problem I'm running into so far is the V+/- rails are dropping too far for my taste during long low notes. I switched the 100K's (R1 - R2) out for 33K's and so far so good.
Click to expand...

Interesting....there must be more total ground current than the spice simulations predicted.

Did you play around with C1/C2 as well? Like I said previous, 10uF (or even 1uF) is probably enough and saves a lot of room on the board.
 
#77
Here are the nets (the first one means "the net is called X3.1, and it connects to pin 2 of R8, pin 2 of R7, pin 1 of R9, and pin 2 of U2").

I would suggest that all of the components be placed vertically, and call the top pin #1 and increase the numbering going down. Most components are two pin (resistors, caps) so the top is #1 and the bottom is #2.

Code: 
*NET X3.1
*NET X3.1 R8. "2" R7. "2" R9. "1" U2. "2" 

*NET 10
*NET 10 R8. "1" J9. "2" 

*NET 9
*NET 9 R7. "1" J8. "2" 

*NET X3.2
*NET X3.2 R9. "2" R10. "1" R13. "1" R15. "1" U2. "1" 

*NET X3.12
*NET X3.12 R12. "2" C7. "N" U2. "6" C8. "N" 

*NET X3.5
*NET X3.5 R12. "1" J11. "2" 

*NET X3.4
*NET X3.4 J11. "1" R11. "1" C5. "N" 

*NET X3.3
*NET X3.3 J11. "3" R10. "2" C5. "P" 

*NET 11
*NET 11 R11. "2" R14. "2" R16. "2" U2. "7" J10. "1" 

*NET X3.7
*NET X3.7 C6. "N" J12. "1" R14. "1" 

*NET X3.6
*NET X3.6 C6. "P" J12. "3" R13. "2" 

*NET X3.8
*NET X3.8 C7. "P" J12. "2" 

*NET X3.10
*NET X3.10 J13. "1" R16. "1" 

*NET X3.11
*NET X3.11 J13. "2" C8. "P" 

*NET X3.9
*NET X3.9 J13. "3" R15. "2" 

*NET 0
*NET 0 U2. "5" U2. "3" R4. "2" R3. "2" R5. "2" R6. "2" C3. "N" C4. "N"  
R1. "2" R2. "1" C1. "2" C2. "1" J6. "2" J8. "3" J9. "3" J7. "2"  
J10. "2" 

*NET 2
*NET 2 U2. "4" U1. "4" J1. "2" R2. "2" C2. "2" 

*NET 1
*NET 1 U2. "8" U1. "8" J1. "1" R1. "1" C1. "1" 

*NET 3
*NET 3 J3. "1" R4. "1" J2. "1" U1. "3" J6. "1" 

*NET X2.4
*NET X2.4 J3. "2" C3. "P" 

*NET X2.2
*NET X2.2 J2. "2" R3. "1" 

*NET X2.1
*NET X2.1 R5. "1" J4. "2" 

*NET 4
*NET 4 J4. "1" R6. "1" J5. "1" U1. "5" J7. "1" 

*NET X2.3
*NET X2.3 J5. "2" C4. "P" 

*NET 7
*NET 7 U1. "6" U1. "7" J9. "1" 

*NET 5
*NET 5 U1. "2" U1. "1" J8. "1"

If I get time I'll try and make a Veroboard diagram.
 
#80
filter-style preamp
Click to expand...

What does this mean exactly?

I got the buffers on my prototype board wired up, and they...well, buffer. ;) What's cool is that you can swap in a 250k for R4 and R6 independently to re-voice each pickup in isolation. I'm not liking the low-Z mode but that's subject to personal preference. It also looks like the 910pF might be a bit too high but again it's personal preference.

I'll see if I can figure out a way to record a few samples of passive vs. active that's aren't too noisy (right now, the pickup leads are extended from the bass to my prototype board so it's picking up a lot of noise). Perhaps I'll run a shielded cable from the back of the pickups all the way to my proto board to improve it.
 

Latest posts

Top Bottom
Back