Modification of old Yamaha piezo preamp for Fishman pickup - help needed

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
There is something weird here.
The Op Amp stage you show is working very well, as intended.
In fact, that 3V RMS value you show in the last line is maximum it can put out and is reaching clipping ... but the "preamp out" is 10X *smaller* or 20dB down ... why?
It´s useless to increase input stage gain, even reaching clipping, and then losing it in the output path.
Please post the *full* schematic, I guess there´s something very wrong there.
 
There is something weird here.
The Op Amp stage you show is working very well, as intended.
In fact, that 3V RMS value you show in the last line is maximum it can put out and is reaching clipping ... but the "preamp out" is 10X *smaller* or 20dB down ... why?
It´s useless to increase input stage gain, even reaching clipping, and then losing it in the output path.
Please post the *full* schematic, I guess there´s something very wrong there.

Hi JMFahey,
the whole preamp was intended for a piezo with much more output. Just comparing the dimensions of the pickups the old Yamaha pickup was about 3.4mm high and 5mm wide while the new Fishman is 1.35mm high and 3.2mm wide. So I'd say the issue should be solved in the gain stage of the preamp. There's a lot of EQing in the rest of the preamp circuit so this might be the part where the gain stage voltage is used up.
Sorry, can't provide the full schematic because there is none. I had to draw the schematic of the gain stage myself from the assembly

Yours
Rob
 
Administrator
Joined 2007
Paid Member
So 100pf. Normally such a cap would make little to no difference over the audio range, but here and given the possibly high value of source impedance feeding the preamp it does have an effect on the midband gain. Not a lot but noticeable.

Try it and see. If you right click the voltage source feeding the input then you will see you can add series resistance. I tried 500k.
 
C2 provides additional feedback at HF. As the frequency rises, the caps reactance or Xc (its 'resistance' to AC) falls. That extra impedance appears in parallel to the feedback resistor and so has the effect of reducing gain. For example, 150pf at 100kHz has a reactance of 10610 ohms. 10610 ohms in parallel with 22000 is 7158 ohms. So at 100kHz the feedback network is 'similar' to the equivalent of a 7158 ohm and the gain is reduced accordingly.
Slight correction here---a capacitor is a reactance, not a resistance, so its impedance is 90 degrees from the resistance---you can't calculate it like resistors in parallel. A 22KΩ resistor in parallel with a 150 pF capacitor will provide an impedance of 9.56KΩ AT 100 K Hz; as opposed to 21.5KΩ AT 10 K Hz. This, with the 22KΩ resistor to ground, will provide a gain of ~2 in the non-inverting circuit at 10 KHz, but only 1.43 at 100 KHz. This will roll off the 100 K Hz signal about -3 db compared to the audio-range frequencies.
 
Hi everyone,
after some more testing I think I figured out the best solution.

First I did some hearing tests comparing the loudness of my two Yamaha FG guitars (one newer and sounding great and the old one where I replaced the pickup) and tried to match the loudness by adjusting value of R4. At about 120k both guitars had a similar output volume. My former idea that I needed an amplification of 1.5 for the whole preamp came out to be a wrong assumption then.
Next step was entering that value into the LTspice simulation. As Mooly pointed out earlier, the C2 capacitor along with the new value of R4 changed the frequency response of the circuit quite a bit. So I run some simulations with other values for C2 and came to the result that a value of about 30pF resulted in a frequency response similar to the R4 22k C2 150pF combination.
@Mooly: would you be so kind to confirm those values in your simulation? I'm quite new in using LTspice so I'm not 100% sure whether I did everything right.

If Mooly can confirm my findings for the ideal value of C2 in combination with a 120k R4 then all that remains to be done is finding a capacitor of that value and soldering it in.

Would be great to be able to finish that work finally ...

Thanks to all hint givers and advisors

Rob
 
I have also played around with acoustic guitar under the saddle piezo pickup mod and installation. At first I thought you bought the fishman set that comes with an onboard amplifier. I have also used Artec piezo and hated it.

I would not expect a fishman piezo giving low output. Is the new pickup all sit in squarely with the saddle ? Is the bottom of the saddle exerting pressure on the new pickup evenly ? When I first started, I was assuming my saddle bottom is flat and square. Of course, turned out not flat and not exerting even pressure on the piezo. That affected string balance as well as volume output. I would recheck the mechanical aspect of the installation and make sure it is all good.

I have done quite a few installation on acoustics and have also built my own FET onboard preamp etc. Finally after years of trail and error, I settled on a Fishman presys 301 blend. Not expensive at all and it comes with sonicore as the pickup element. It has a condenser mic on the preamp therefore you can blend real sound with with the pickup. The sonicore installation is more forgiving then the standard piezo strip pickup element (that you need perfectly square and even pressure on the standard pickup). The scariest part of this installation is to cut a hole for the onboard preamp. But I have done it many times and even on an ovation.

I have a Yamaha F325 acoustic that I am planning to install the presys blend on it. Only reason I'm hesitate is the guitar itself sounds amazing acoustically and can compare to some Taylors but I'm just not sure how it will sound after adding a sonicore in the vibration path (string vibration -> saddle -> sonicore -> topboard).
 
Hi AC439,

thanks for sharing your experience with piezo setups. The presys blend might be an option for the future, but for the moment I think I'll stay with he piezo strip (Fishman acoustic matrix). I'm quite confident that saddle and saddle carving are quite even, had a guitar repair expert look over it.

And I wanted to stay with the old Yamaha preamp for several reasons the main being that I didn't want to cut any more holes in my guitar.

Yours
Rob
 
First I did some hearing tests comparing the loudness of my two Yamaha FG guitars (one newer and sounding great and the old one where I replaced the pickup) and tried to match the loudness by adjusting value of R4. At about 120k both guitars had a similar output volume........... a value of about 30pF resulted in a frequency response similar to the R4 22k C2 150pF combination.
So.....which guitar SOUNDED the best--the original or the one with the new pickup?
 
Administrator
Joined 2007
Paid Member
If Mooly can confirm my findings for the ideal value of C2 in combination with a 120k R4 then all that remains to be done is finding a capacitor of that value and soldering it in.

It looks a good fit. Also bear in mind that the TL062 is not as good a performer at HF as the model used here and so that will have an effect as well.

One thing thing we can do in LT is copy a circuit and have two instances running at the same time. Here we have the original circuit and the modified one. The output of the modified also has an attenuated output to match the level back to the original and so allow easy comparison.

I've changed the input to a 12.5 kHz square wave which is useful to check out the transient response and any ringing.
 

Attachments

  • Capture1.PNG
    Capture1.PNG
    61 KB · Views: 130
  • Capture2.PNG
    Capture2.PNG
    72.3 KB · Views: 129
  • 9 Volt Preamp revised.asc
    5 KB · Views: 46
So.....which guitar SOUNDED the best--the original or the one with the new pickup?

Hi dotneck335,
you're asking for an explanation how this project came to life ... ;-)

I'm owning 2 Yamaha FG guitars, one from about 1990 and one from about 2014. The guitars have a similar accoustic sound but while the newer sounds also great when amplified, the old one sounded horrible. As I told before: just mids, no bass, no treble. So after ruling out the preamp as a reason (see above), I decided to give a new pickup a try.

Now to your question: The old guitar sounds significantly better with the new Fishman piezo, more defined, more crispy, even warmer too.

Now that I managed with the help of you all to adapt the output volume of the old guitar with the new pickup to that of the newer FG, the old one sounds real promising.

I had to do all the testing with the innards of the poor guitar hanging out, so I'm looking forward to some more comfortable test situation when I'll have everything assembled and installed again.

Yours
Rob
 
It looks a good fit. Also bear in mind that the TL062 is not as good a performer at HF as the model used here and so that will have an effect as well.
Hi Mooly,
thanks again for your advice. So 33pF will be the value to go for, since the diagrams are looking good with it.
So I guess there is no real TL062 plugin for LT?

One thing thing we can do in LT is copy a circuit and have two instances running at the same time. Here we have the original circuit and the modified one. The output of the modified also has an attenuated output to match the level back to the original and so allow easy comparison.

I've changed the input to a 12.5 kHz square wave which is useful to check out the transient response and any ringing.

Thank you so much!
So the only thing that remains to be done is finding an online store that sells the capacitor without charging a fortune for postage. We used to have an electronics store in town, but no more.

Yours
Rob
 
Administrator
Joined 2007
Paid Member
The copy feature is really useful. Just click copy and drag and move the area you want duplicating. LT reallocates new reference numbers, all you need do is delete any duplicate .op commands and so on and relabel any nodes such as Vout etc so that there is only one of everything.

There probably is a more suitable model in there somewhere but I'm not familiar with a lot of the LT devices. You could search the Linear Tech site and try and find something more modest using the parametric search perhaps.

You might a suitable cap in some old scrap equipment or old circuit boards. Its always worth a look if you have any.
 
Hi Mooly,

I found a TL072 include for LTSpice which seems to work (just place the TL072.301 file in the same directory as the TL072_dual_supply.asc file (both files in the attached TL072_test.zip, that I found in the LTSpice yahoo group).

But I can't figure out how to transfer the working TL072 into my own project.

Maybe you can?

Yours
Rob
 

Attachments

  • TL072_test.zip
    1.5 KB · Views: 35
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.