Piezo / Preamp Noise Problem

[Mark Wybierala]

Every once in a while I'll get in an acoustic guitar that has what presents itself as a ground hum problem. These could be any typical manufactured acoustic with an undersaddle piezo and active preamp. The symptoms are excessive noise at higher signal levels and noise interrupt when you touch the metal backshell of the instrument cable and, the pickup is still actually functioning with a strong output (its understood that the cable and the amplifier has been ruled out). Sometimes its just a dirty/defective or miswired output jack and that is my first thought. But I have had instances where the cause has been a defective undersaddle piezo pickup. The fact that there is what seems to be ground problem and the noise is impacted by touching a finger to the backshell of the instrument cable causes me to not be so confident.

For clarification,... There are a number of piezo formats ranging from the simple piezo cable which is just a thin coaxial wire with the insulator made of piezo material to a number of other sandwich styles. The ones that I've had problems with are the very plain rigid strip that looks like it has a thin covering of foil and a typical example is the Martin Thinline. The current replacement from Fishman is called the AG Series.

I have personally accidentally damaged a piezo while rushing during a saddle height adjustment resulting in this noise problem and have successfully addressed this problem on at least two other occasions by replacing the piezo. Over the years I've purchased quite a few piezo pickups on ebay at dirt cheap prices in lots with a single unit cost of about $2 and have accumulated a wide range of sizes and thicknesses. The $2 piezos work great but it is rare to find them on ebay in the narrow width format (3/32 / 2.3mm) leading to the necessity of paying around $100 for a narrow format pickup from Fishman. Hey, if the client has a defective pickup and the cost is $100, that's okay but I don't want to spend 100 dollars of the client's money on a guess. Talking to Fishman support, they concur that the problem is likely a defective pickup but I feel so much more confident when I troubleshoot a problem and discover the cause and prove it through an actual measurement or visual confirmation. Swaptronics is a proven technique but I don't like it when its expensive.

The first thing is that I don't understand the theory or reason a piezo element would do this. My guess is a hairline fracture of the foil-like outer conductor. This foil is delicate and easily torn by sharp objects. I understand that this foil is the equivalent of the ground side of a magnetic pickup.

Can I replace the pickup with a 1 Meg resistor at the preamp and if the noise problem at high gain goes away be confident that the problem is actually the piezo? Any more thoughts on this are appreciated.

[Peter Wilcox]

Maybe I'm not understanding correctly, but could you just replace the pickup with another (presumably good one) at the preamp (size wouldn't matter since it's not in the guitar), and see if the noise is gone?

[Bob Gramann]

As I was reading, I thought “why not replace the pickup with a resister and see if the noise disappears?” Then, I got to your question in the original post. So, yes, I would consider that a good strategy as is Peter’s. If it doesn’t work, we all need to know because it just seems like the right thing to do. Replacing at the preamp doesn’t rule out a cause in the lead wire. I would also consider that the pickup could be wired backwards to the preamp, the foil side to the input rather than the ground. Or, there could be a cold solder joint at the ground connection. I bought a Fender electric bass some time ago that had a horrible hum. When I opened it, I found that the wiring to the pickups was not shielded cable, which I didn’t expect, but rather two single wires each. The wires to the pickups had not been twisted between the control panel and the pickups. I put a few turns on the wires and the hum disappeared.

[Mark Wybierala]

I was in this guitar a few weeks ago. Its a budget Martin with a Fishman preamp (OEM) setup. I replaced the output jack with no change. The pickup plugs into the preamp with a tiny jack and I found a small amount of corrosion on the jack. I removed the jack from the preamp circuit board and soldered the shielded wire from the pickup directly on to the circuit board. It seemed okay so I assumed the corrosion/compromised jack was the problem. The guitar is back again with the same noise issue. I actually did a thorough check and double check of the correct wiring of the output jack -- its the standard three conductor setup with battery/ground establishing a path via the stereo plug. I'm certain the pickup is wired properly too. I'm going to establish a few more observations before I proceed.

[David King]

If you suspect a grounding issue why not check continuity from the foil wrap of the pickup to the output jack sleeve/threaded barrel?

$100 seems like a lot to pay for a Fishman replacement. You might set yourself up as a Martin repair center and get the replacements from Martin at something closer to their cost. EMG makes a fine drop-in system that's also very reasonable as a dealer/luthier (though it might be easier to set something up with the local EMG dealer for the occasional need.)

I hate under-saddle systems so much I'd select something easy to manage, battery-free, cheaper and better sounding like K&K or equivalent and just replace every problematic one with a surface-mount transducer type.

[Mark Wybierala]

I spoke directly to Martin customer service about this instrument and they do not provide tech support for the electronics of these guitars unlike Taylor who have always offered outstanding support. I hoping that this was just a single Martin employee having a bad day because I would like to think highly of Martin. The preamp has the Fishman name on it and although the preamp is not a model of their current products catalog, generally Fishman will always spend the time on the phone talking with you and trying to help.

[Mark Wybierala]

I carefully searched through the text of ebay listings and found three listings that specified the 2.3mm thin format piezo and ordered a quantity of them cheap. Not a lot of investment if they turn out to be wrong. $99 is way too much for a piezo.

[David King]

Let us know how those inexpensive ones sound if you can hear any difference in signal level and tone.

[Chuck Tweedy]

Yes - please do continue the conversation. I actually have a very similar situation, but thankfully it is not a customer's instrument I'm working with.

[Mark Wybierala]

I ordered 3 ea. from three different ebay vendors in China. About $1.99 per unit average. Although each of these vendors were different in name and reviews, they all used the same pictures on their ebay listing. All three packages arrived about three weeks earlier than the estimated date of delivery in minimally protected packaging (a padded bag/envelope) but there was no damage. They are indeed the correct width and look very similar to the OEM pickup -- exact color (metallic Ruby Red), exact width, same length, same lead wire, but they are all slightly thicker in height by about 50% -- not really an issue -- about .030" and in my situation, the .030 resulting increase in saddle height was advantageous to the setup of this guitar. The noise problem is cured. I called the client two weeks after I finished the job and he is totally pleased with the performance and tonal response of the guitar.

Had the dimensions of the pickup been exact, I might be tempted to believe that I was getting the exact same $99 part sold by Fishman from the factory that supplies Fishman but this is not the case. They are an intentional copy/clone. But they work and seem to work well. It did address the noise problem.

[David King]

I'm willing to bet that Fishman still makes their own transducers as that's their main business and always has been. There are many variables to contend with regarding the material, it's formulation and how it's oriented.

[Rob Kidd]

Piezo pickups have huge impedance, and their signal and the high impedance input at the preamp needs very good shielding.

I would check the continuity of the copper shielding and the braid. I would also consider cutting the pickup's lead so that it's no longer than it has to be. Also consider replacing the lead itself. Some I've seen have a very loose and open weave for braiding. The lead should have tight and full coverage from the braided shield.

[Murray Leshner]

Rather than say, based on textbook theory, that replacing a piezo sensor with a resistor as a noise troubleshooting technique won't work, I'll instead share why I don't think it's that simple.

But it's been 4 years and it may have worked.

My thoughts on why it's not that simple. Results may vary depending on circumstances.

A piezo transducer is predominantly capacitive in addition to converting vibration (and/or sometimes pressure) to an output voltage.

Resistance is the parameter that thermal or Johnson noise depends on, along with temperature. Replacing an almost purely capacitive impedance with a large resistor could allow hiss type noise to be amplified by a high-input-impedance preamplifier. How would one select a resistor value to substitute for a capacitive piezo device? Probably by guessing. A low resistance compared to the piezo impedance, which is also a guesstimate, might minimize added thermal noise. My initial thought was that a short circuit across the input of the preamp would be helpful, but as I type, I think of reasons that could be misleading...it could minimize noise for different reasons.

If one is using a preamp optimized for a high-impedance piezo device, there is typically a 'noiseless bias' circuitry technique of powering the amplification component (I e., JFET or op amp) with a low resistance at DC to reduce noise from the bias technique but providing the designed input impedance for the audio function. If the input of such a preamp is direct-coupled (without an additional dc-blocking capacitor) based on assumption a piezo input device won't care about the bias voltage present, a low resistance would prevent the preamp from having proper voltage. A noiseless bias input typically has a high series resistance (1-10 megohms) and a voltage divider on the order of 10k to set the bias level. For some piezo preamps, a randomly selected resistor substituting for the piezo could reduce noise for unknown reason. Would it be because the bias was shifted too far and the JFET or op amp doesn't operate properly, or because a suitably low resistance that doesn't act as a Johnson noise sources was selected?

If a piezo is being used directly into a musical instrument amplifier without a preamp, the input impedance of the amplifier would be reduced by the substitute resistor being in parallel with it. This could reduce the hiss noise slightly, as well.

So, in summary, a resistor on the order of 1-10 M (Meg) ohms could cause hiss noise, and too low a resistance would probably misleadingly reduce noise depending on what kind of circuit it's interacting with.

Impedance is by definition a frequency-dependent phenomenon. An amplifier can be designed to have a fairly uniform input impedance over a wide frequency, but the impedance of a capacitor (piezo for example) is inversely frequency dependent, vs. an inductor's impedance. That is what causes low frequency loss with random piezo selection and random amp/preamp input impedance or the desire to avoid having a powered outboard preamp. Cable length with a capacitive sensor doesn't necessarily cause frequency response rolloff but can change the output level if a capacitive divider is formed unknowingly. Piezos typically need a high impedance load (or amplifier stage) unless they are relatively high capacitance.

A piezo feeding a predominantly resistive input amplifier forms a high-pass RC filter, and the bass response is dependent on impedance matching. An inductive pickup like typical guitar/bass non-piezo types are predominantly inductive. They do have significant resistance so people depend on measuring inductive pickup resistance for convenience. That serves a useful function but resistance does not equal impedance.

An inductive (magnetic) pickup feeding a resistive amp/preamp forms an RL low-pass filter, and the high frequency response is dependent on impedance matching. The resistance is significant but is typically still much lower than the AC impedance that actually determines the usability.

There is not an easily (for consumers) directly measurable resistive parameter with piezos.

Piezos also have different response along each axis of the element shape. You found say the same about a guitar pickup, but the physical characteristics prevent the latter from being installed in the wrong orientation.

Ignoring polymer piezos (and electrets)for a moment, ceramic piezos and formed of a variety of recipes in a variety of shapes. They can be optimized for purposes other than musical instruments where detection of vibration or pressure along a different axis is desired. So a square piezo shape may not have the same sensitivity on all sides. Attached wires or flat shapes may prevent incorrect installation. But surplus/repurposed piezo elements from other devices may do weird or unexpected things in a different (musical instrument) application.

If one reads patents, Fishman really stands out in having an astonishing number of patents and PhD's on staff.

One last remark on impedance vs. resistance. Thermal or a Johnson noise is a resistive phenomenon. Capacitors and inductors are generally not affected by this. I word it that way because there may be some arguable scenario where it is measurable.

The biggest YMMV factor with piezos is the obsession with selling them, or desiring them, to not need amplification or equalization. You can make them work, but they are not all the same.

Size and the ceramic recipe determine the amount of capacitance, which determines the relative impedance and what it will want to see connected as amplification.

Murray