HPF HPF-Pre 1 Series User manual

HPF-Pre Series 1 and 2 Technical Manual
Francis Deck, Mar. 15, 2012
HPF Technology, LL
Overview
HPF-Pre is a preamp with adjustable high pass filter and phase switch for use with piezoelectric pickups on
the double bass. It is intended to address the primary issues faced by bassists when playing through
mainstream amps and PA systems.
Foremost HPF-Pre is a preamp. Its high input impedance ensures proper buffering of the signal from piezo
pickups. It is battery powered and housed in a tough plastic enclosure with internal shielding.
An adjustable high pass filter tunes out the subharmonic “thump” generated by the plucked bass which
otherwise can drive ported bass speakers to over-excursion. It is also a useful tool for coping with adverse
acoustical environments and PA systems with excessive subwoofer support.
A phase (polarity reversal) switch is provided to help control feedback problems. It could also help when
setting up systems with multiple pickups or a pickup and a microphone.
Series 2 of the preamp adds an output volume control and an LED battery status indicator. The volume
control is handy when your stage setup requires your amp to be out of arm's reach.
HPF-Pre is an "open" design. I sell the HPF-Pre at a modest price from my part-time home business.
Otherwise you are welcome to build your own.
Specifications
Here are some preliminary specifications for the HPF-Pre. This report goes into considerable detail on how
these parameters are measured. The results are from initial builds that I expect to be representative.
•Input resistance / capacitance = 10 MegOhms / 220 pF
•Overall gain = +2 dB
•High pass filter corner = Adjustable from 35 to 140 Hz
•High pass filter slope = 12 dB/octave with minimal “hump”
•Phase switch = Applies phase inversion to signal
•Connectors =1/4 inch input and output power switched by output jack
•Distortion = 0.1% THD at 1 V p-p input
•Noise = -107 dB referred to 1 V p-p input
•Input level = Up to 4 V p-p with tolerable distortion
•Power requirement = Standard 9 V battery
•Design topology = Solid state Class-A circuit

User's guide
There isn't much to the HPF-Pre. Put in a
battery. There is no reason to use an
expensive battery. You should turn the
volume down on your amp when flipping
the phase switch until you are certain
that there won't be a loud "pop" under
any conditions.
Now you hook it up. In and Out are self-
explanatory. The knob adjusts the cutoff
frequency. Plugging into the Output jack
connects the battery. Make sure it is not a
stereo plug because the HPF-Pre uses
the solid sleeve of a mono plug to
connect the battery. Unplug it when you
are done. As with all audio devices it is
best to turn your amp volume down
while connecting or disconnecting to the
HPF-Pre.
Series 2 of the preamp has a battery status indicator LED just below the phase switch. The LED will flash
briefly if the battery is OK. The indication is sufficient to provide you with at least a gig's worth of power.
For convenience if your amp already has a high-impedance input you can use the HPF-Pre in the effects
loop of most amps.
I think that for best results start with your amp EQ controls flat. Many electric bass amps are "voiced"
meaning that they produce a non-flat response curve when the control knobs are centered. It is my opinion
that a flat voicing works better for upright bass – at least as a starting point. If your amp is designed for
electric bass don't be surprised if you end up with seemingly extreme EQ settings. You are probably
dialing your amp to a more flat response curve.
Start with the cutoff frequency at 35 Hz. Increase it until you are happy. If possible find a favorable setting
of the HPF-Pre before adjusting the EQ on your amp. Having the HPF-Pre doing its job liberates your EQ
controls to deal with feedback. Use your ears and not your eyes. If you get good tone at a relatively high
frequency setting don't worry about what is happening to the low frequency content of your bass. It's there
but is under control.
What can I say about the phase switch? It is not a panacea but can possibly give you a few dB of added
gain before the onset of unacceptable feedback. What does it do? Perhaps "phase" is a misnomer and
"polarity" would be a better term. The sound coming from your amp is coupled into your bass and feeds
back into the pickup. The amount of feedback coupling depends on your bass and pickup amp gain and
tone settings and acoustical surroundings.
In simplistic terms if that coupling is additive then you get worse feedback than if the coupling is
subtractive. The phase switch lets you find the polarity that gives you less feedback.
To make it complicated the coupling is not described by a simple phase relationship but by frequency
dependent phase shifts. The practical effect is that the phase switch can often help cope with feedback at
moderate volume levels but the amount of control is unpredictable and limited in effectiveness at high
volume. For these reasons don't expect miracles.

My experience has been that I can set the phase switch in the following way. Gradually raise your amp
volume until feedback just begins to be noticeable. Now flip the phase switch. The feedback will get worse
or better. You want it to be better. Done. Now get out there and play!

Design description
Let’s start with background. The HPF-Pre uses design elements that are well known but combined in a
novel way. Here are some online sources that deserve particular credit:
http://en.wikipedia.org/wiki/Butterworth_filter: Design concept for active Butterworth filter
with equal resistors.
http://www.jensen-transformers.com/as/as098.pdf: This is the “starting point” for my active
filter design a nice 2-transistor source feedback preamp.
http://www.ciphersbyritter.com/RADELE T/PREJFET/JFETPRE.HTM: More information
about simple discrete preamp circuits including source feedback.
The preamp circuit is based on a classic active filter design using discrete transistors instead of op amps.
While I have nothing against op amps the discrete design turned out to be simpler and more efficient. With
half the transistors compared to op amps current consumption is cut in half and noise is reduced by 3 dB.
Also a Class-A design with minimal open loop gain will exhibit asymmetrical soft clipping – a musically
forgiving overload response behavior that is shared with tube preamps. (Note that I am stopping short of
claiming “tube sound.”) Here is a schematic of the circuit.
Q1 is a classic JFET phase inverter circuit. Since negligible current flows through the gate of Q1 equal
currents flow through R3 and R4. The signals at the source and drain of Q1 must be of equal amplitude but
opposite phase. A SPDT switch selects one of these signals – it is the phase switch. C2 and R5 hold the DC
offset of these signals equal so there is minimal "pop" when you flip the switch.
While C1 and R2 form a highpass filter the cutoff is 0.5 Hz not musically useful. I wanted a high value of
C1 to allow for the lowest possible noise with pickups that have high capacitance. R1 limits current into Q1
if the input is temporarily overloaded thus providing a little bit of protection.
A shunt capacitor at the front end of an audio preamp is unorthodox. However the circuit is designed for
piezo pickups. C2 provides some measure of protection for the gate of Q1 eating voltage spikes. Its low
capacitance means that it has a negligible effect on the response of capacitive pickups.
The second active stage is the classic Butterworth active filter design. (Compare to the Wikipedia
reference). Q2 and Q3 form a two-transistor source feedback amplifier – think of it as a crude op amp. The

gate and source of Q2 are the positive and negative inputs. R14 and R15 set the gain and also determine
the bias points of the transistors. With gain greater than unity in this stage it is possible for the filter to
have equal resistors – critical for using two stages of a dual pot. An audio taper pot is used and it is wired
"backwards" with increasing corner frequency as the knob is turned counterclockwise. This trick makes use
of a widely available audio taper pot to provide a roughly linear frequency scale. The analogous op amp
circuit is shown here for reference:
The source feedback amplifier is a clever general-purpose gain stage. The gain is set by R14 and R15.
Variations on this theme abound. It is shown as a "circuit idea" in The Art of Electronics by Horowitz and
Hill (Cambridge University Press 1980) and of course in the preamp design on the Jensen website. Tube
power amps often use an analogous cathode feedback scheme. I have used it as a low noise preamp with
gains up to 20 dB when working with a homemade magnetic pickup with unusually low sensitivity.
The negative terminal of the battery is connected to the "ring" terminal of the output jack. Inserting a mono
plug into this jack grounds the ring terminal and applies power to the circuit. Guitar effects pedals have
used this arrangement since the dawn of civilization. The only drawback is that you really don't want to
unplug the output cable while your amp is turned on.
Series 2 of the preamp has an output volume control. This component is absent in Series 1.
The battery indicator deserves a bit of attention. A schematic is shown below. The circuit starts out with
C21 discharged so when the circuit starts out current will flow through the D21 if the battery voltage is
sufficient to overcome the Zener breakdown voltage. Drawing some of this current through the base of Q4
will cause current to flow through D22 flashing the LED. As current flows into C21 it charges up until the
Zener shuts off.
Turning off the preamp alows C21 to discharge through R21 resetting it for the next time.
There's enough variation in Zener diodes that some of them are unsuitable. I have a test jig with a 1k
resistor in series with the Zener and measure the forward voltage of each diode. With the 1N5237B diodes
most come in around 7.8 V in my test and I throw away the ones that are much higher than that. This
appears to result in a “flash” threshold of around 8.5 V which seems like a decent level. Since 9-V alkaline

batteries start out around 9.5 V you get a reasonable lifetime with a 1-V drop. My hope is that if you see a
“flash ” it means that you have enough juice for one more gig.
Test results
Frequency response: I used my FreeSA spectrum analyzer program to measure response curves at three
positions of the frequency knob. (Note that this program is obsolete due to “progress” in Microsoft
Windows but the free program Visual Analyzer works nicely). The result shows that I nailed the
Butterworth response function pretty well.
HPF-Pre frequency response curves
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
10 100 1000 10000
Frequency (Hz)
Response ( B)
CW
Mi le
CCW
Noise: By necessity my noise measurement process is a bit convoluted as I am not using commercial
calibrated measurement gear. The first edition of The Art of Electronics by Horowitz and Hill suggests
comparing the noise from a grounded input and from a calibrated noise source. My noise source is a 100
kΩ resistor which generates thermal noise at a level of 40 nV/√Hz at room temperature or 5.7 µV RMS
over a 20 kHz bandwidth. My setup is:
Resistor  HPF Preamp  Another Preamp  PC sound card input
I used my own FFT spectrum analyzer program to collect spectra of the resistor the grounded input of the
preamp and the grounded input of “another preamp ” which boosts the signal above the noise floor of my
PC sound card. Subtracting the first two measurements (using quadrature subtraction meaning the square
root of the difference of the squares) gave me a reference that corresponds to 40 nV/√Hz. Subtracting the
second two measurements in the same way gave me the noise level of the HPF-Pre. Scaling this noise level
to the 40 nV/√Hz reference resulted in a measured input noise for the HPF-Pre of 9 nV/√Hz. Over 20 kHz
bandwidth and flat weighting this is a noise level of –107 dB referred to 1 V p-p. Some premium JFET op
amps have lower noise but draw much more current.
Distortion: My test setup looks like:
PC sound card  Buffer with gain voltmeter at output  HPF preamp  PC sound card
First I took the HPF preamp out of the system to check the inherent distortion of my measurement system.
I used my own software and simulated distortion waveforms by adding a small amount of the second
harmonic. Here are the results:

HPF-Pre istortion plot
-100
-80
-60
-40
-20
0
20
0 5000 10000 15000 20000 25000
Frequency (Hz)
Amplitu e ( B)
1 V p-p
5 V p-p
.
At 1 V p-p The second harmonic is at ≈ -55 dB and the third harmonic at ≈ -78 dB. The rest of the
harmonics are consistent with the distortion floor of my measurement system. Pure second harmonic
distortion at -55 dB corresponds to a THD of 0.17% which is quite respectable for a Class-A circuit. This
is with an input voltage of 1 V p-p.
At a higher input voltage you see a slew of harmonics. That's serious distortion.
To estimate the dynamic range of the filter I measured the total harmonic distortion at different input
voltage levels using Virtual Analyzer. Here is the plot:
HPF-Pre istortion versus input voltage
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 1 2 3 4 5 6
Input voltage (V p-p)
THD (%)
Distortion rises with increasing amplitude. But unlike an op amp circuit the dynamic range is not a hard
upper limit. You can push beyond this range at least on the transient at the front end of a plucked note
with minimal audible effect. I expect this circuit to be quite forgiving.
Acknowledgement
I am indebted to drurb from the TalkBass forum for trying my first prototype for many valuable
comments and suggestions and for prodding me to put the preamp on the market.