Hello, I've been trying to gather information on this Alembic style preamp I've built.
It works, but, I wanted to get some expertise opinion or suggestions for three points that are lacking on this project.
First: the filter resonance goes from 300 hz to 4.8k approx.I wanted to get it to go a bit lower, perhaps 100/150 hz.
Second: from what I've gathered, this circuit consumes a bit more power than usual, and I'm worried since I will be using two of these on my bass with a single battery.Is there anything it can be done besides choosing a low power IC?
Third:On the schematic, there is a final Volume(10k)pot.I've made the circuit without that pot, just wired it straight to the output jack and it works great.But, doesn't the schematic say that there should be a 10k resistor always connecting from output to ground!?
Could anyone take a look on what could be done here?
Thanks millions
Schematic: http://forum.musikding.de/cpg/displayimage.php?pid=871&fullsize=1
One thing you could do to reduce power consumption is replace the Vbias circuit with a TLE2426. Since you have two units involved, you could run them both from the same TLE2426.
Going to a low power op-amp could save a ton of current.
I haven't looked at the specifics of this filter, but typically there are some tradeoffs made in the frequency range it can do. You could probably adjust some values to lower the minimum freq, but you'd probably end up either losing the max. freq. or having a lower, broader peak that didn't work all that well.
On the output, are you passively summing the two filter sections at this point?
Ei RobA.
Well, the ic being uses is a tl074, so I'm not sure How that would work.
Regarding the mixing, I haven't decided yet If I'll do it passively or actively.
But before that, I can't figure If that last supposed 10k res is actually part of the circuit.For now I want to work the circuit individualy.
I might forget the freq range issue.
The last 10k resistor is going to have some influence on the circuit, but mainly in its interaction with the cable and other downstream components. The output from the final op-amp stage of the SVF is a reasonable low-Z buffer, so that part is OK. For me, even if I were going to have a volume pot at the end of the chain, I'd put an output buffer after the volume pot to drive the output and help avoid cable impedance and capacitance issues.
If you are going to use an active summing circuit, you can change the output cap and that resistor to whatever the input of the summing circuit needs and then use the 10µF cap on the output of the summing buffer.
I'm not sure what you mean by you're not sure if it will work with the TL074, but the TLE2426 will work in any circuit where you need a reference voltage at half the supply voltage as long as the total current it needs to source or sink is less than about 20mA. I think you'll be below that for this circuit even if you add in the output buffer and power both of them from the same TLE2426 reference point.
The spec sheet on the TL064 versus the TL074 indicates a pretty big current savings there. It might be worth doing that change if the current draw ends up being too high.
If you do use an active summer and you end up having a leftover op-amp, you could use that to buffer a single Vbias for both filters. You could move to higher value resistors in the divider and save some current there. The savings using this method though will depend on what type of op-amp you are using.
As the frequency is determined by 1/(2 pi R C) with R=R8=R9 and C=C4=C5 you can halve the frequencies by doubling either the values of C or R. Using C4=C5=22n should shift the frequency range to 136Hz...2.2kHz.
So, regarding that 10k final resistor that shows off as a pot on the schematic, How should it be wired If I use a fixes resistor.should it be in series, or should it go parallel from output to GROUND!?
Quote from: RobA on March 06, 2014, 03:35:19 PM
The last 10k resistor is going to have some influence on the circuit, but mainly in its interaction with the cable and other downstream components. The output from the final op-amp stage of the SVF is a reasonable low-Z buffer, so that part is OK. For me, even if I were going to have a volume pot at the end of the chain, I'd put an output buffer after the volume pot to drive the output and help avoid cable impedance and capacitance issues.
If you are going to use an active summing circuit, you can change the output cap and that resistor to whatever the input of the summing circuit needs and then use the 10µF cap on the output of the summing buffer.
I'm not sure what you mean by you're not sure if it will work with the TL074, but the TLE2426 will work in any circuit where you need a reference voltage at half the supply voltage as long as the total current it needs to source or sink is less than about 20mA. I think you'll be below that for this circuit even if you add in the output buffer and power both of them from the same TLE2426 reference point.
The spec sheet on the TL064 versus the TL074 indicates a pretty big current savings there. It might be worth doing that change if the current draw ends up being too high.
If you do use an active summer and you end up having a leftover op-amp, you could use that to buffer a single Vbias for both filters. You could move to higher value resistors in the divider and save some current there. The savings using this method though will depend on what type of op-amp you are using.
Is it not possible to shift Only the lower frquencies?
Quote from: kothoma on March 06, 2014, 04:03:49 PM
As the frequency is determined by 1/(2 pi R C) with R=R8=R9 and C=C4=C5 you can halve the frequencies by doubling either the values of C or R. Using C4=C5=22n should shift the frequency range to 136Hz...2.2kHz.
Quote from: garfo on March 06, 2014, 04:30:06 PM
Is it not possible to shift Only the lower frquencies?Quote from: kothoma on March 06, 2014, 04:03:49 PM
As the frequency is determined by 1/(2 pi R C) with R=R8=R9 and C=C4=C5 you can halve the frequencies by doubling either the values of C or R. Using C4=C5=22n should shift the frequency range to 136Hz...2.2kHz.
Oops, need to be make a little correction: R=P2+R8=P3+R9, of course.
Using 50k+3k3 and 10nF as in the schematic:
R=53.3*10^3 and C=10*10^-9 gives 298.6Hz
R=3.3*10^3 and C=10*10^-9 gives 4823Hz
Using 50k+1k5 and 22nF for example:
R=51.1*10^3 and C=22*10^-9 gives 140.5Hz
R=1.5*10^3 and C=22*10^-9 gives 4823Hz
You should get the idea...
But now your freq pot could be a bit harder to dial in as it now sweeps over 5 octaves instead of 4.
Quote from: garfo on March 06, 2014, 04:25:56 PM
So, regarding that 10k final resistor that shows off as a pot on the schematic, How should it be wired If I use a fixes resistor.should it be in series, or should it go parallel from output to GROUND!?
You would put the 10k to ground, but what you want to do with it really depends on what comes after, like active or passive summing. What you want to use there depends on the input to the next stage of the circuit.
Thanks guys.Well, I might Check the frequencies.
I think I'll just use an mn tapper blend pot I already have on my bass.so, each circuit will Go straight to the blend pot, in the end I will use a master volume pot, I'm not Aurélio there will be any need for buffering the output.
Also, the tl064 is not jfet input, or is it!?what other good opamps besides the 064 would work Well and push less power?
The TL064 has J-FET inputs like the TL074.
Less power consumption, but a bit more noise.
RobA, If I decide to Go all passive after, the 10ufcap and the 10k should be as in the schemattic, right!?
Would I Only remove them if I want to use a buffered summing(blend) and a normal volume pot; or should I Also remove them using a passive blend and then a buffered volume and place them at the buffer output!?
https://docs.google.com/viewer?url=http://diy.musikding.de/wp-content/uploads/2013/06/bufferschalt.pdf
I have that buffered laying around at Home.would it work well after the volume pot on my bass?it is the jfet version.
Let me understand what you want to do now. You have two pickups with each pickup through an independent SVF. Then the SVF output is passively summed and then run to a single volume pot and then into the output buffer. Are you going to have a pickup selector toggle in there somewhere too?
That is precisely it.No, there is absolutely no need for a selector when I have the Blend Pot.It does what the selector switch does, plus more.I'm using a Bourns MN taper blend pot.It shous almost no resistence at the center.Much better than the other Blend pots where I would always find a volume loss and tone sucking at center position.
I forgot about the blend when I was thinking about how you have this set up. There are a couple of ways to use the blend pots. There's one that replaces the toggle and goes into a volume. But, there's another way to do it that leaves them more independent and still uses the toggle. It essentially works out to be like the standard Gibson LP circuit, but with both volume pots replaced by the blend pot. I have the Bourns blend on one of my guitars now using the first type of circuit, but I'm going to be moving it to the second and putting the toggle back in.
Since you are going to be putting some resistive elements after the SVF's, I don't think you will need or want the 10k to ground. The 10µF cap is probably going to be bigger than you need, but it won't hurt. You could probably replace it with a 1µF film cap though. You do need a cap there though on the outputs of each of the SVF's.
The JFET buffer looks like a good choice for the output. Since you don't need the high input impedance of the JFET, the BJT version of the same circuit might actually be a bit better of a choice, but the JFET should be fine.
What is the bjt version???Sorry for my ignorance ::)
The one with the Si-NPN Transistor...
because it has a lower impedance input!?I could build it with the 2n5088.Or is there a better silicon choice!?
Yes, mainly because of the lower R1/R2.
A 2n5088 is a good choice, as is a BC549C/BC549B/BC550 (turned 180°). Or a MPAS18.
I've wired it as it is in a box.Works fine, will build the second svf with precision resistors, will try 064 and 074 opamps and decide for one before I decide to build it as you suggested with another ic making the bias.
Well, by the way, What values in terms of pots should I use.For Blend I can get 250k and 500k only, but what about volume pot.Should I go with 500k or should I go under that with this build in mind as it is!?
Quote from: garfo on March 07, 2014, 01:49:19 PM
Well, by the way, What values in terms of pots should I use.For Blend I can get 250k and 500k only, but what about volume pot.Should I go with 500k or should I go under that with this build in mind as it is!?
It depends a bit on which blend circuit you are using with the MN pot. One of the circuits I've seen commonly used effectively halves the impedance of the pot, so the 250k ends up looking like a 125k to the output of the previous stage. That's still going to be OK here really because you have the active buffers from the SVT, but you do want to match your volume pot to be about the same level as the impedance from the blend circuit. If you are using this circuit with the 250k MN pot, then I'd think that a typical 100kA volume pot would be a good choice.
I do wish that Bourns would come out with some 50k and 25k versions of the blend pots to use with active pickups.
The other reason the BJT version of the output buffer is a better choice than the JFET version is that BJT's typically have better current drive than JFET's. This allows them to better overcome the capacitive load that the guitar cable puts on the signal.
Yeah, i know, ideally we should keep impedance low when mixing the two signals, and I could do that latter on, and a smaller value mn taper blend pot would be great.Unfortunately they only have those two.
I'm using a 500k Bourns Blend Pot, and at the moment I actually have a CTS 500k volume pot that I was intending to use here(It's the one on my bass).Also, I have wired the blend pot leaving the ground connection off, don't know how that affects impedance, but it works passively on my bass.
So,impedance wise, if I use the 500k blend, should I go 250k on the volume and then buffer it!? :P
I'm curious about this because I'm not sure how this specific pot will interact with the volume.
Wouldn't it be 250k only when at center position?I believe that these pots in particular due to its taper have a different interaction, correct me if I'm wrong.
I take it that you are doing the wiring method like the way I currently have it in my guitar since you mention leaving off the ground connection. (I do have the ground connection on mine.) I'd have to think about what leaving off the ground does impedance wise in the circuit. But, when the ground connection is in place, the resistance from signal to ground is 250k for the 500k MN pot no matter what the position of the pot is. It's a bit weird, but it happens because of the way two sections are interconnected. You can just measure the resistance with a DMM with the jumpers in place to see what it is on the way you have it wired.
After saying all that, it would be more ideal if you could match the low output impedances of the active electronics, but I doubt if it is going to cause you a great deal of problem if you can't. I guess I'd try to use the 500K pot you have and see how it goes. If it causes any tone loss issues, you could always buffer the blend output with a transistor buffer stage like the output buffer you have planned. You could get both buffers and the volume pot on one small board if you needed to.
For now I will be doing the blend/master volume passively and only buffer the output.If I wire the ground of the blend and end up using a 250k volume pot at least they would be matching and the output buffer would see that load which would more or less match it's input load.
I will get together with a friend that has more experience designing boards and we will together try to develop this a bit further, but for now it's just me and I'll use what I have.
Nevertheless all the suggestions here are awesome and I'll try to use some of those ideas.
If I leave the 3.3k resistor as it is and change only the cap to lets say 12n I can get a minimum of 248.84hz and a max of 4019.06khz.It goes a bit lower, and the highs play in a better ball park for the bass.
Quote from: kothoma on March 06, 2014, 05:05:10 PM
Quote from: garfo on March 06, 2014, 04:30:06 PM
Is it not possible to shift Only the lower frquencies?Quote from: kothoma on March 06, 2014, 04:03:49 PM
As the frequency is determined by 1/(2 pi R C) with R=R8=R9 and C=C4=C5 you can halve the frequencies by doubling either the values of C or R. Using C4=C5=22n should shift the frequency range to 136Hz...2.2kHz.
Oops, need to be make a little correction: R=P2+R8=P3+R9, of course.
Using 50k+3k3 and 10nF as in the schematic:
R=53.3*10^3 and C=10*10^-9 gives 298.6Hz
R=3.3*10^3 and C=10*10^-9 gives 4823Hz
Using 50k+1k5 and 22nF for example:
R=51.1*10^3 and C=22*10^-9 gives 140.5Hz
R=1.5*10^3 and C=22*10^-9 gives 4823Hz
You should get the idea...
But now your freq pot could be a bit harder to dial in as it now sweeps over 5 octaves instead of 4.
Quote from: garfo on March 08, 2014, 10:51:03 AM
If I leave the 3.3k resistor as it is and change only the cap to lets say 12n I can get a minimum of 248.84hz and a max of 4019.06khz.It goes a bit lower, and the highs play in a better ball park for the bass.Quote from: kothoma on March 06, 2014, 05:05:10 PM
Quote from: garfo on March 06, 2014, 04:30:06 PM
Is it not possible to shift Only the lower frquencies?Quote from: kothoma on March 06, 2014, 04:03:49 PM
As the frequency is determined by 1/(2 pi R C) with R=R8=R9 and C=C4=C5 you can halve the frequencies by doubling either the values of C or R. Using C4=C5=22n should shift the frequency range to 136Hz...2.2kHz.
Oops, need to be make a little correction: R=P2+R8=P3+R9, of course.
Using 50k+3k3 and 10nF as in the schematic:
R=53.3*10^3 and C=10*10^-9 gives 298.6Hz
R=3.3*10^3 and C=10*10^-9 gives 4823Hz
Using 50k+1k5 and 22nF for example:
R=51.1*10^3 and C=22*10^-9 gives 140.5Hz
R=1.5*10^3 and C=22*10^-9 gives 4823Hz
You should get the idea...
But now your freq pot could be a bit harder to dial in as it now sweeps over 5 octaves instead of 4.
Right, then the range shifts (down) by a factor of 10/12.
I just remembered that I have a CTS 250k Volume pot that I've scrapped the final taper and it offers infinite resistance(No Load), maybe I could use it after the blend pot, in those terms the load before the last buffer would be The blend's Load.
I took the measurements of the 500k Blend pot and it is as you say, around 230k load.
Quote from: RobA on March 08, 2014, 02:04:43 AM
I take it that you are doing the wiring method like the way I currently have it in my guitar since you mention leaving off the ground connection. (I do have the ground connection on mine.) I'd have to think about what leaving off the ground does impedance wise in the circuit. But, when the ground connection is in place, the resistance from signal to ground is 250k for the 500k MN pot no matter what the position of the pot is. It's a bit weird, but it happens because of the way two sections are interconnected. You can just measure the resistance with a DMM with the jumpers in place to see what it is on the way you have it wired.
After saying all that, it would be more ideal if you could match the low output impedances of the active electronics, but I doubt if it is going to cause you a great deal of problem if you can't. I guess I'd try to use the 500K pot you have and see how it goes. If it causes any tone loss issues, you could always buffer the blend output with a transistor buffer stage like the output buffer you have planned. You could get both buffers and the volume pot on one small board if you needed to.
I have built everything and it is working.I had to put two 10k resistors in series with the output of each preamp because they were fighting each other when mixed in the middle position.Someone told me to do this and said that 10k resistors should work.It feels like I've lost a bit of the high end when I did so.
Is 10k enough, or should I increase the resistors value in order to get the most treble out of it?