Ampeg Based Piezo Preamp Build

[blown240]

I am on vacation for a week and I am starting to plan my next build.  I play an upright bass that has a dual piezo pickup.  It came with a preamp that sounded so bad I literally threw it away.  What I want to build is an Ampeg b15n based dual channel preamp that I can use to plug into house sound, my regular bass amp, or use to record direct.  

My bass has a stereo jack on it, so the preamp will have a single stereo input.  It will have separate volume, treble, and bass controls for each channel, and I will use push/pull tone pots for the bass and treble boost.

I want to make the preamp circuits as close as possible to the original B15n but obviously setup for a piezo.  Also, when playing at louder volumes, or when the acoustics of a room are terrible, my bass feedsback at about 100hz.  So I want to do a notch filter on each channel that drops 100hz, but leaves everything above about 120hz and below 80hz. I will have a separate notch for each channel that is switched with a small micro switch.

Here is a link to the pickup I am using:
http://www.kksound.com/bassmasterrb.html

Here are some questions I have:

1.  Input Resistor – the pickups input resistance is 1m per channel.  Should I match that, or should I got with something like a 4.7m input resistor like a lot of the peizo preamps I see online.

2.  Rectifier – I want to use a tube rectifier, but I am not sure if the tranny I am using will let me, or what rectifier tube to use. Here is a pic of the small trans that I will be using (tranny on the left):

 

3.  Notch Filter – No idea how to do this.  Any ideas?

4.  Is there anything else I need to do to the basic b15n preamp circuit to better work with a peizo?

5.  I want to make sure I have loads of clean headroom, so anything that will help that...


Thanks for any input!

[blown240]

Here is a schematic for a b15n for reference:

[birt]

if you make only one input for each channel (the one with the 100K for example) then you don't have the 47K to ground from the other input. this means the 5M6 resistor is more or less the input impedance. good for a piezo.
this will also boost the signal a bit when compared to the original circuit. you might use a lower gain tube so it will be cleaner. it doesn't matter for the volume because the FOH mixer has loads of volume for you ;-)

can you post what is written on top of that transformer that you have?

you might want to look into good passive DI transformers so you can have a balanced XLR output without having to use a separate DI. add a jack output too and you can plug into a poweramp on stage at the same time.

[blown240]

I am only going to have 1 input total.  It will be stereo. So I guess the 47k to ground wont be an issue.

I will post whats on top of the trans in a week when I get home.  I got it a few years ago from Hoffman. Its just a basic preamp transformer.  What is a FOH mixer?  Sorry, but I am still pretty new at this.

[birt]

FOH = Front Of House
you plug your preamp into the system of the house so they will provide all the volume you need in monitors and PA for the audience.

i'm so used to the term FOH from my job that i forget to write it out on a forum like this.

[blown240]

Thanks!

Ya I really want something that will bring out the sound of my upright bass, and its hard to go wrong with the portoflex sound.  So this preamp hopefully will get me that sound no matter what I am plugged into.  Plus it will give me the impedance I need to make the piezo sound better and louder.

[blown240]

Here is a schematic I did to start with.  I still need to make sure the input impedance is correct for a piezo.  And I need to figure out where to put the notch filter(s) and exactly how to do them.

[PRR]

The two 270K mix network is just strong enough to run inside the box, maybe thru a short cord. 

You need more beef to drive modern power amps and long runs to sound system.

Don't over-think the input impedance. After all, you can always change it on prototype.

[blown240]

I never even considered that.  What should they be?  100k?

[sluckey]

Changing the mix resistors is not the answer. Put a cathode follower between the mix resistors and the output jack.

[blown240]

Ok, Thats easy enough.  Something like this:



What value should the load resistor be?  And what tube should I use that will color the sound the least.  And does it matter what filter cap to get I get the B+ off of?  I can add another since I was going to get cap cans anyway and would have an extra.

[sluckey]

That's basically it, but you need a wee bit more to bias the CF and block the dc voltages. This should work. The input cap can be a small value .001µF to .047µF) since the bootstrapped grid circuit is very high Z. The output cap should be larger (0.1µF to 0.25µF) since it's in a much lower Z circuit. Caps are easy to play with if you don't like those values.

Some of the other guys can put you on a tried and true CF circuit. You may even consider a MOSFET source follower.

[blown240]

Thats looks pretty good.  I will add that to my schematic and get a few extra caps when I order my parts.  I dont really like the idea of a mosfet when its so easy to do it with a tube.

It seems that some like the 12AU7 for this...


What are the thoughts on a "twin T filter" to give me the 100hz cut? I still need to figure that part out.

[blown240]

Here is an updated schematic:




I was looking at a pic I have of the trans that I going to use.  There are only 4 wires coming out of the secondary side.  1 pair is the heaters.  I am not sure on the power of the 2nd pair.  I may have to go with a SS rectifier because of this.  Which is fine I guess.

So I still need to figure out:

1.  Notch filter at 100k
2.  What tube to use for the CF
3.  If my trans will work with a tube rectifier

[HotBluePlates]

Don't over-think the input impedance. After all, you can always change it on prototype.

Don't we need to over-think it just a little. I don't think he can just copy the B15 input circuit because...

I play an upright bass that has a dual piezo pickup. 

He may need some kinda small buffer amp right at the instrument, to give a ~4-5M load impedance for the piezo. No bass might be a bad thing for an upright bass pickup.

[blown240]

Wouldn't I have 5.6m input impedance with the schematic I have?  Thats what I got from the post from birt. 

But maybe there is more too it than that.  I would prefer not to have anything else mounted on my bass, but instead to integrate it into this preamp.  Thats half the reason I am building it...

Would a 15-20 foot cable really change things as far as having the buffer right at the bass or first thing in the preamp?

[HotBluePlates]

Wouldn't I have 5.6m input impedance with the schematic I have?  Thats what I got from the post from birt.

Yes, but...

But maybe there is more too it than that.  I would prefer not to have anything else mounted on my bass, but instead to integrate it into this preamp. 

Would a 15-20 foot cable really change things as far as having the buffer right at the bass or first thing in the preamp?

Have you noticed how almost every guitar using a piezo pickup has an onboard preamp? It's only partly about convenience. A piezo element might as well be a capacitor; in fact, the mechanics that make it like a capacitor are the same that allow it to produce an output signal.

Have you ever used a condenser mic? They need phantom power because there is a preamp internal to the mic. The old-school name for "capacitor" was "condenser" and this type of mic operates on the same basic principle as a piezo element. Anyway, for several reasons, the pre-preamp is mounted in the mic, and presents an impedance of more than a gigohm.

Anyway, a classical-style guitar from the late-60's that introduced the first piezo pickup (the Baldwin Prismatone pickup), and considered to have one of the best available sounds, has a small internal preamp mounted directly on the jack in the guitar. There can't be much to it except a couple of FETs and a couple of resistors, because it's encased in a glob of potting material maybe 1-2" long. It uses a stereo jack, so that power for the FETs is carried on a different conductor than the signal, and is powered by an external power supply.

So it can be done. I don't know anything specific about your pickup. I can't tell you if it will work fine with the preamp you have in mind. If it were my instrument, I'd look to have some kind of simple onboard preamp, perhaps even one small enough to mount on the jack.

[blown240]

Perhaps I should build this preamp, then see how it is. 

Then if I need to I can build one of these:
http://www.diyguitarist.com/Guitars/PiezoBuffer.htm

Either way, this preamp will get me alot of what I am looking for.

[HotBluePlates]

Sounds like a plan.

[blown240]

I still need to see what exact trans I have to determine if I can go with a tube rectifier or not. 

Plus I need to figure out the notch filter and see what tube to use for the CF.

[sluckey]

There's no advantage when using a tube rectifier for this type circuit.

I don't have any ideas on a notch filter.

Wire the CF for a 12AU7. If you don't like it, swap in a 12AX7, 12AT7, 5751, etc.

[blown240]

Sweet.  Thanks!

Now I only have to figure out the notch filter.

[blown240]

Here is the updated schematic:




Couldn't I just use a cap in parallel with a resistor to make a high/low pass?  Use the cap to cut out everything below 120hz and the resistor to cut everything above 80hz.  It seems to me that in the end there would be a cut from 80-120hz. Maybe put this in series between the 270k resistors and the CF, and use a double pole, double throw switch to take it in/out of the circuit.

[birt]

take a look on how the filters in a graphic EQ work. i'm sure you can find schematics for that. there's also stompboxes lie the Boss GE-7 that you can look at.

and don't forget you will need a DI when you plug this into the sound system of a venue. you might as well add a transformer and XLR output to your preamp.

[terminalgs]

Couldn't I just use a cap in parallel with a resistor to make a high/low pass? 

for hi-pass 'yes', for low-pass, no.  for low-pass, you can use a cap, or a set of parallel caps and resistors to ground.


BIRT has a good suggestion about the XLR balanced output. 

I've built a few tube pre-amps,  here are some thoughts:

1.) have a ground lift switch,  if you plug this into an amp with unbalanced signal (1/4 inst. cable)  and both the preamp and the amp have 3 prong plugs, you can get ground loop noise.  the shield on the inst. cable will provide path to ground from with the chassis.   This is different than a ground lift switch on a XLR/balanced output box, but you can have a ground lift there as well (for pin 1).

2.)  re-evaluate the tube rectifier.   pre-amp tubes won't draw enough current to ever get a tube rectifier to cause sag, and I'm not sure there is any reason beyond that to use them.  If you are modeling your engineering of Ampeg, they switched to SS as soon as they could.

3.) Triode transformers makes a nice Dynaco pre-amp PA211/PA522 that is  a 330-0-330 and 12.6v/900ma filament transformer that works well for this application, tho its 12.6v, not 6.3v and I see you'd like to use 6SL7s.   If you reconsider and decide (1) use 12SL7s or (2)  decide to use 9A based tubes (12ax7/12au7 etc), this is a perfect tranny.  I've used the International version PA522 (which has 120v and 240v windings) by using the 240v windings with 120v USA wall power to produce 165-0-165 and 6.3v secondaries.

4.) you've got stereo inputs, you could add extra switched output jacks to have optional stereo outs,  pre-mixing resistors...

5.) you've got a 12au7 enlisted for the CF, you've got another section there for another CF out (maybe for the other channel,  or to drive a transformer XLR/balanced out.   

6.)  that preamp has enough gain to hit a power-amp section, you might want to tame it down with lower Mu-triodes (6SN7s, or 6072As, 6211s, 12AU7s).   If you go with 6SL7s (or 12ax7s), your usable gain will be between 0 and 3 on your volume knobs,  I suspect). 

7.)  9A based tubes will give you a lot tuning of the overall gain,  with an 8BD pinout, you are limited to 6SL7s and 6SN7s, with 9A's, you can you use 12au7s and 12BH7As all the way up to 12AX7s (there are about 10 different Mu tubes in between).

Piezoelectric pickups in acoustics produce sound a lot different than magnetic pickups. Besides the much much high output impedance (which you might have to phtuz with on your 1st stage to match),  you might find that the tone stack in a traditional amp (like the one you are looking at) change the sound of your Piezos differently than magnetic pickup guitars/basses.   Be prepared to have to change the tonestack a bit.    also look at the bass tone stack on an SVT.   

[blown240]

Thanks for the reply terminalgs!  You have some good ideas.

1.  Ground switch - will do, thats a no brainer.

2.  Going to use a solid state rectifier.

3.  Transformer - Hopefully the one I have will work.  If not my local amp guy has loads of them that he sells CHEAP to his buddies.

4.  Stereo out - I may do this after the initial build is done.

5.  How could I use the other half of the 12AX7 to drive a transformer XLR out?  What would the circuit look like?

6.  I will see how it sounds with the tubes and switch them out of need be.  Another advantage of my local amp guy!

7.  I really want to try the 6SL7, if that sucks, then I may switch to 9 pin.

[PRR]

http://www.national.com/ms/LB/LB-5.pdf

[RicharD]

>ground lift switch

Last several pres I've built, I've blown off the switch and just simply wired the output lifted.  You run it grounded unless it it buzzes, then you lift it.... which never buzzes.  I 100% think you should conjure up a post EQ direct out.  Your instrument, you're the player, you should be in control.  I have a tried and tested bass preamp circuit that everybody who has used it loves it. Drop me a line if interested.

[blown240]

http://www.national.com/ms/LB/LB-5.pdf

Thanks!  Thats exactly what I am looking for, except how can I get it from 60hz to 100hz?  I assume that it would sit between the 270k resistors and the CF, is that right?

[RicharD]

I got your PM and I have attached the last (I think) schematic.  This circuit is loosely based on the STV Channel B preamp.  This is the side without all the cool inductors which in 2011 are unobtainable. (L to R)  V1b was changed to a plate follower to reduce gain and distortion.  S1 inserts a RC circuit that cuts the lows @ 20ish Hz.  It also loads V1b but doesn't really cut volume all that much.  It's the magic button when feeding a Big BooM PA.  R6 through V2b are a direct copy of the SVT sans the summing network R19 & R14.  V2b is the HiZ output (J2).  I added V2a to drive a DI transformer.  Yup, I know you don't wanna add another transformer but life is cruel that way in the world of tubes.  It's designed around a cool old Peerless 15095 transformer.  15k:600 ohm plate load.  We had a few of them and they're too cool not to use.  The Edcor XSM10k/600 will do the job just fine.  http://www.edcorusa.com/products/160-xsm10k-600.aspx  It's $15.00.  The B+ is pretty straight forward SS full wave recto.  I did 12VDC regulated filaments.  The 555 timer is simply a 60ish second delay that keeps the outputs disconnected (via a very poorly drawn relay) until the filaments are hot and the circuit is quiescent.  The little relay and timer thang is good to have.  This circuit goes NASTY for about the first 5 to 10 seconds.  It's all food for thought. 

[xm52]

This might be of interest to you.

http://personalpages.tds.net/~fdeck/bass/hpfpre.htm

[blown240]

Thats an interesting circuit.  There are some aspects of it I really like.  It definately gives me some food for thought.

Also, I was looking at the schematic for the Boss EQ pedal.  There is an EQ circuit for 100hz that I may try to incorporate right before the CF in my preamp.


The like for the peizo buffer cool as well.  I will have to go back to that once I see how the preamp effects my overall setup.

Thanks again guys!

[bigsbybender]

I use that same K&K preamp and actually like it.  It's beat everything else I've used to date.

[blown240]

The best pickup I have ever used was a King Doublebass Doll Head.  It was AMAZING.  But with them being out of biz, and the fact that so few were sold, I was offered a TON of money for it, so it went... 

[HotBluePlates]

http://www.national.com/ms/LB/LB-5.pdf

Thanks!  Thats exactly what I am looking for, except how can I get it from 60hz to 100hz?  I assume that it would sit between the 270k resistors and the CF, is that right?

Figure 1 of the sheet tells ya. It's all based on R*C.

But since you'd like to know now, leave the resistors alone, change C1 & C2 to 150pF each, and change C3 to 300pF (not a standard value; use two 150pF caps in parallel). Those values land on 106Hz, but the original values land on ~58Hz, so I'm close enough.

You likely want to build the variable-Q version.

You could trim the resistor values down and cap values up, but if you build what the sheet says, you know it'll work right. Plus, larger-valued caps cost more.

[blown240]

HotBluePlates - thanks for the reply.  Thats exactly what I needed to know.  I appreciate it!

[blown240]

OK so here is where I am at.  I have added the balanced DI out, the ground switch and the variable HZ cut.



What do you guys recommend for:

1.  The first resister on the XLR DI out.  It can be 5k to 100k.  What should I start with?
2.  will the SPDT switch work for the notch filter, or should I use a DPDT to completely remove it from the circuit?

Thanks!

[HotBluePlates]

1.  The firs resister on the XLR DI out.  It can be 5k to 100k.  What should I start with?

Probably a 50k trimpot in series with your actual level pot.

I don't know what signal level your line out needs to supply. We can guess your preamp should output 1v or less of the normal signal, and if we knew the required output for the XLR, we could figure a ratio.

But it's easier to just add a trimpot, turn your level full up, then adjust the trim for the max signal you'll likely need. Button up, and your done.

2.  will the SPDT switch work for the notch filter, or should I use a DPDT to completely remove it from the circuit?

DPDT.

The variable Q pot of the notch is a 50k load that is fine for the opamps, but too low for the tube stage ahead of it. You want to be able to completely bypass the notch filter when it's off.

And a question you didn't ask:
Don't do the ground-lift the way you've drawn. The 3rd prong should always be attached to the chassis of any amp/preamp. If B+ is shorted to the case, you want a breaker to trip, not you to trip after you get zapped touching the case. I've grabbed ahold of 450vdc before... it's not fun.

If you really want to incorporate a ground lift that works, but is safe, you need to use isolated jacks, and have a ground buss on the circuit board for everything with a ground symbol. 3rd wire of the power cord gets connected to the case. Connect a 100pF cap and 50k resistor in series.  Connect this resistor/cap combo across a SPST switch, which has wires running to your wired ground system and the chassis (close to the 3rd prong wire).

It works like this:
If the switch is closed, the circuit ground is connected to chassis ground, which is hardwired to the 3rd prong.
If the switch is open, circuit ground is disconnected from chassis ground. 3rd prong wire provides shock safety. 100pF/50k combo connects the circuit ground to chassis ground at radio frequency, which helps keep RF buzz/noise out of your circuit. This particular combination is -3dB down at about 31kHz.

[blown240]

So a 50k trim pot, sort of like a bias pot.  What does that resistor actually do? is it even necessary with the level pot?

SPDT - will do, makes sense.

Ground Switch.  Is it even worth the hassle? Seems like if it had a ground loop thru a cable to a power amp, it wouldnt matter if there was a chassis ground thru the power cord.  It would ground to the power amp thru the cable.

[birt]

don't rely on a cable between devices to provide a ground path like that. (it can be helpfull in some situations but since you are designing something you better go for the safer way)

[blown240]

Good Point.  I suppose an easier way to do it would to use a non conducting chassis.  I had planned on a 16aug steel chassis, because I have a ton of 16aug sheet metal.   OR I can isolate the pots and jacks from the chassis.  I guess that isnt too bad either, just a bunch of nylon washers...

[blown240]

Here is where I am at now.  I have isolated all the grounds after the standby switch and have them going thru a ground switch.  Also added a DPDT switch for the notch, and a 50k trim pot for the DI.

Here is the schematic.  The isolated ground path is in green:

[jjasilli]

Re input impedance matching:  I'm not sure what your input impedance is.  It looks like the incoming signal is split so there's about 2X 6M in parallel = 3M.  Piezo's usually have about 6M output impedance, so 3M input impedance is rather low.  It might work well enough though. 

FWIW Merlin recommends a 12AX7 cathode follower (not preceded by a gain stage) as the input stage for a piezo PU.  It has an effective AC impedance of 6M.  Designing Tube Preamps for Guitar & Bass.  This equals the input impedance of hi-end SS acoustic guitar preamps. 

Another alternative is an outboard SS preamp/buffer designed for piezo's, in front of the amp.

[PRR]

> It looks like the incoming signal is split so there's about 2X 6M in parallel = 3M.

No, he's got _2_ piezos (bridge and fingerboard), wired with one 3-conductor cable and jack. Each piezo sees 5.6Meg.

He can readily try 2Meg or 22Meg just by changing resistors.... no need to fret about it before the Build.

[RicharD]

You're missing the boat on the ground lift.  I don't think it's safe IMO to float a HV circuit.  All you lift is pin one of the XLR output.  This prevents the possibility of a loop through other equipment down stream.

[blown240]

OH YEAH, I missed that one for sure.  Well thats a whole lot easier!

Thanks Richard!

Something else I am considering:

If I wanted a 1/4 inch output jack on the front AND back of the preamp, can I just run 2 jacks in parallel?  One on front (easier when plugging into some amps), and one on back (for running to house sound/recording).

[RicharD]

I'd use the other 1/2 of the 12AU7 for the DI  A phase switch is awfully good to have.  If you're gonna run with the 6SL7, why not use a 6SN7 in place of the AU?  How are you powering the OpAmp?  Can it run on 12V?  Tis a good excuse to go with DC filaments which puts you back in the Ax -n- AU tubes.

[blown240]

What is the advantage of using the other half of the 12AU7 for the DI?  I am totally for it, just curious if there is an advantage.

6SN7 - I am not agianst using it, I have no idea what it is.  I only picked the 12AU7 because I have seen others use them.  I do want to stick with the 6SL7.

I didnt know I had to power the OpAmp, I know absolutely ZERO about them.  Apparently they need 12-15v

[RicharD]

The advantage of using separate triodes for the 2 outputs is so that the 2 loads don't affect each other.  You could easily stick a pot and another iso cap in front of the 2nd cathode follower to get a separate level control for the DI.  You've got the triode, why not use it.  The 6SN7 is for the most part the octal equivalent of the 12AU7. 

Yes the opamp has to have juice.  If you use the PA211 DynaClone PT, you've got 12.6 VAC so you could copy my regulated filament supply.  Stack the 2 SL filaments in series and stick with the 12AU7 and wire it for 12V filaments.   

[jjasilli]

> It looks like the incoming signal is split so there's about 2X 6M in parallel = 3M.

No, he's got _2_ piezos (bridge and fingerboard), wired with one 3-conductor cable and jack. Each piezo sees 5.6Meg.

He can readily try 2Meg or 22Meg just by changing resistors.... no need to fret about it before the Build.

Ok.  Then that's as good as it gets!