Oakley Flanger User manual

Oakley Sound Systems
5U Oakley Modular Series
Flanger
PCB Issue 2
User Manual
V2.4
Tony Allgood
Oakley Sound Systems
CARLISLE
United Kingdom

The Flanger module as a 1U wide MOTM format module.
2

Introduction
This is the User Manual or the issue 2 5U Flanger module rom Oakley Sound. This
document contains an introduction to the module, some in ormation on how to make the best
use o your module and the calibration procedure.
For the latest Project Builder's Guide, which also contains the parts list or the components
needed to populate the boards and a list o the various interconnections, please visit the main
project webpage at:
http://www.oakleysound.com/ langer.htm
For general in ormation regarding where to get parts and suggested part numbers please see
our use ul Parts Guide at the project webpage or http://www.oakleysound.com/parts.pd .
For general in ormation on how to build our modules, including circuit board population,
mounting ront panel components and making up board interconnects please see our generic
Construction Guide at the project webpage or http://www.oakleysound.com/construct.pd .
3

The Oakley Flanger Module
The issue 2 Oakley Flanger as a single width MOTM format module in a natural finish Schaeffer panel.
The Oakley Flanger module is an analogue delay line capable o delaying audio signals rom
0.5mS to 15mS. In conjunction with an external LFO module this unit can be used to produce
real time vibrato, chorus and langing e ects. With the eedback control set to maximum the
module can be set to sel -oscillate.
The delayed audio output is available rom the DLY OUT socket. While the MIX OUT socket
has a equal mix o input signal and delayed signal. The phase o the delayed signal that is
mixed may be selected to be inverted (NEG) or non-inverted (POS) by the use o the MIX
switch. This is particularly use ul when used or chorus or langing as the two modes can
create completely di erent e ects.
A eedback path rom the delayed output back to the input is available and you have control
over both the amount, via the FEEDBACK pot, and phase o the signal, via the FBK switch.
Inserting a jack plug into the FBK IN socket will break the internal eedback path and allow
another signal to be sent into the delay line. This could allow additional processing o the raw
delayed signal, rom the DLY OUT socket, prior to being sent back into the delay line again.
For example, using a VCA to process the eedback signal would allow CV control o the
eedback level.
There are two control voltage (CV) inputs. CV1 has a ixed sensitivity while CV2 has its own
sensitivity control. CV2's control pot is a reversible attenuator, so you are able to adjust not
only the depth o the modulation but also the polarity. Although not speci ically designed to
track to 1V/octave, using the CV1 input, the module can be used as tuned delay line over a
use ul proportion o the ull operating range.
4

The delay line itsel is based around two 3207 bucket brigade delay (BBD) devices controlled
by a voltage controlled oscillator (VCO) running rom around 33kHz to over 1MHz. Audio
bandwidth o the delayed signal extends to 14kHz and several unusual design eatures ensure
that the module is relatively quiet compared to other units o its type.
Two LEDs provide visual indication o the signal level being sent to the BBD devices. The
DRIVE control should be adjusted to ensure that the green LED remains lit or optimum
signal level. The red LED will light when the signal level is close to, or is, being overdriven.
The module will not be harmed when the signal is overdriven and there is a so t clipping
circuit in place to ensure the overdriven sound is musically use ul. The DRIVE control allows
a variety o di erent signal levels to be used with the module although the expected signal
level would be expected to be between 0.5V(peak) and 8V(peak).
5

The Delay Control
The Oakley Flanger uses a high requency voltage controlled oscillator (HFVCO) to clock the
BBD devices. The aster the clock the shorter the delay time. The clock runs rom 33kHz to
just 1MHz. The clock can be controlled by the ront panel delay control pot and control
voltages sent to the CV1 and CV2 sockets. Increasing positive voltages will decrease the
delay time.
The delay control pot will create the longest delays at its minimum setting and shortest delays
at its maximum setting. This may seem counter intuitive i you think about the module as a
simple delay line – most delay lines would have a delay control that would increase the delay
as it is turned up. However, the langer creates its distinctive sound by a orm o notch
iltering whereby some requencies in the audio spectrum are cut and some are boosted.
Shortening the delay time gives the impression that the iltering e ect is rising in tone. It's
actually more complex than that but having the delay pot work in a reverse ashion does seem
to sound more natural.
The delay time goes rom approximately 500uS (0.5mS) to approximately 15mS. Control
voltages applied to either CV1 or CV2 will extend this slightly but the module has limiters in
place to prevent the module rom going too ar away rom its intended range. With the delay
pot in its middle position the delay time is around 3mS.
For langing e ects you can use the ull sweep o delay times rom 0.5mS and 15mS. For
chorus and vibrato type one would normally use delay times o around 5mS to 15mS. All o
these e ects require the use o some orm o modulation. That is, the delay time is dynamically
altered to create movement in the sound. Modulation sources normally take the orm o a
connected LFO with triangle or sine wave orms available. However, you can also use a
sequencer or envelope generator. With an LFO as the modulation source the delay pot is used
to set the mid point by which the modulation shi ts the delay time. The e ect heard varies
greatly with delay time, modulation depth, modulation wave orm and modulation requency.
Vibrato is best heard rom the raw delay output and using a sine LFO o around 6Hz to gently
modulate the delay time. Chorus is best done with a relatively slow moving triangle wave and
using the mix output. The pitch shi ting e ect created by modulating the delay time is more
noticeable at longer delay times even with only small amounts o modulation.
Flanging requires the use o eedback while both chorus and vibrato e ects are done with no
eedback. The greater the amount o eedback the more pronounced the langing e ect.
Feedback
There is a eedback input socket, a eedback pot and a eedback switch. For the switch and
input socket, eedback is abbreviated to FBK.
The FBK IN socket can be used as an additional input which will add to the signal going into
the INPUT socket. The Feedback control pot will then control the level o the FBK IN signal.
I there is no signal plugged into the INPUT socket and the input is being applied solely to the
6

FBK IN socket then only the delayed version o this signal will heard at both the DLY OUT
and MIX OUT sockets.
When the FBK switch is in POS it has no e ect on the signal passing into the FBK IN socket.
In NEG mode the module inverts the signal so the audio going into the delay circuitry is now
completely out o phase with the FBK IN input signal.
With no jack plug inserted into the FBK IN socket the delayed audio signal is automatically
passed to the eedback control pot. When the control pot is at its minimum then no signal is
passed rom the delay output back into the input. In this case the position o the eedback
switch has no e ect as there is no signal to invert.
At high levels o eedback the module may sel -oscillate – although this will depend on how
the unit has been calibrated. The requency o sel -oscillation will depend on a variety o things
but the key actor is the delay time. Feedback is stronger at longer delay times, that is, when
the delay control pot is at its lowest settings. It also slightly stronger when the FBK switch is
in NEG mode. I the unit has been calibrated to sel -oscillate then sel -oscillation will happen
more readily in NEG mode and at longer delay times. The output signal level at high eedback
levels can get very loud and be somewhat unpredictable.
For the classic langer sounds both switches would normally be set to the same mode.
The Drive Control
Whilst having dedicated input level and output level controls is more common in rack e ects
the Oakley Flanger eatures a single Drive control. This combines the actions o both the input
and output level controls in one control. Turning up the Drive increases the signal level sent to
the BBD devices, while at the same time reducing the signal level rom the BBD devices. This
way input sensitivity can be increased without any change in output level. This is very use ul
when using the Flanger module to process external instruments or indeed the wildly varying
signal levels o a modular.
For clean signal processing the green LED should be mostly lit and the red LED lit only very
occasionally. Allowing the red LED to activate shows that the signal is now being overdriven
producing a more distorted sound. I the red is lit even with the Drive pot at its minimum this
indicates that the input signal level is too large. This will not damage the unit but you should
either turn down the signal level via the sending instrument's volume control or use another
module (like the Oakley Multimix) to attenuate the signal a little.
7

O Signal Levels and Insertion Loss
The signal levels o both outputs are approximately 8dB lower in signal level than the input.
This means a 5V peak input signal will be lowered to around 2V peak. This reduction allows
the output audio signal more space to rapidly increase in volume, which it may do under
certain conditions, without any signi icant risk o clipping distortion.
When the module's eedback control is at its minimum value the delayed output's signal level
pretty much ollows that o the input. There are no surprises. However, increasing amounts o
eedback, both negative and positive, can produce massive changes in output signal level. The
actual peak output level is hard to predict as it can change with the harmonic structure o the
input signal, the undamental requency o the input signal, the delay time selected and the
modulation signals used. This rapid change in volume at certain requencies is a distinctive
part o the langing sound.
Any 5U modular system is powered rom +/-15V so the maximum signal output o any
module cannot be more than +/-15V. In practice it actually works out somewhat less than this
and outputs o over +/-13V are unlikely. I the output wants to be above this value, but the
module cannot produce it, then the output clips. This means the module does what it is
supposed to do up to +/-13V and then any desired output greater than that is clipped at +/-
13V and goes no urther. This abrupt lattening o the top and bottom o wave orms creates
distortion and is generally not wanted.
The Flanger module is quite capable, when eedback is applied, to produce signal levels ar
greater than what goes in. I the output matched the input per ectly, ie. 5V peak ended up as
5V peak with no eedback, then when eedback was applied the output could easily try to
exceed the +/-13V. By lowering the nominal output level to approximately 2V peak the output
can rise by over six times be ore clipping sets in.
The two signal activity LEDs both monitor the signal level going into the BBD devices. The
module is calibrated so that the red LED will only light when the BBD audio driver circuitry
starts to overdrive. Keeping a signal under this level will help keep the signal as clean as
possible. It is, however, still possible, under certain conditions, or the output stages o the
module to distort. This should be obvious to the user as output stage clipping can normally be
heard. Clipping does not damage the module and some users may ind the sound appealing.
When the delay time is very short the signal running through the BBD devices is not passed
through as e iciently as it is when the delay times are longer. Engineers say that the insertion
loss through the BBD increases at higher clock requencies. The upshot o this is that you may
notice a small volume drop as delay times are shortened. While this has no real detrimental
e ect, especially or chorus and pitch shi ting e ects, it does reduce the signal level available
or eedback. At very high levels o eedback, that is, close to or at oscillation, and with very
short delay times you will notice that the langing e ect is not as pronounced. In practice this
means that the position o the eedback control is important to getting the right sound at the
desired delay time.
8

Noise
BBD devices are noisy in comparison to other electronic devices. They hiss and the clock
signals that control the delay o ten sneak into the audio pathway. It is use ul there ore to keep
the audio running through the module as high as possible but not high enough to cause audible
distortion – unless you actually like the distortion.
The Oakley Flanger does not use the noise reduction circuitry that some other devices use.
There are two types o noise reduction techniques typically used in BBD based langer and
chorus units; pre-emphasis/de-emphasis and compression/expansion. Both types do indeed
reduce the overall level o hiss and grunge. But they also cause other sonic e ects that in my
view detract rom the sound. The best sounding analogue langers in my opinion are the ones
that have no noise reduction circuitry.
That said, to reduce CV and clock breakthrough, the Oakley Flanger uses two langers
running in parallel. Each langer circuit, based around one MN3207 BBD, is controlled by the
same clock signal so the delay produced by each o them is identical. Also the audio signal
passing through each one is the same but with one important di erence – one o them is the
inverse o the other. This is a di erential audio signal and when one wave orm is going up, the
other one is going down. I you were to add the two signals together then they would
completely cancel each other out. Here though, our two delayed, but out o phase, signals are
subtracted rom each other. This reproduces the original signal at twice the size but also any
similar noise that both signals would have gained rom going through the BBDs is cancelled
out. This technique is particularly e ective at reducing CV breakthrough – where the CV that
controls the delay time sneaks into the audio path producing unwanted thumps and wheezes –
but it also improves signal quality too.
I you do ind the level o noise objectionable in certain patches try using the langer module
be ore your inal VCA. This way the audio signal, including any added noise, rom the langer
will only be heard when the VCA is opened. The audio passing through the langer will usually
drown out the relatively small amount o noise produced by the BBDs.
9

Calibration
There are seven trimmers, or presets as we used to call them in the UK, on the printed circuit
board (PCB). An oscilloscope is required to set several o the trimmers to their optimal
positions.
You should use a proper trimmer tool or a ine blade jeweller's screwdriver or adjusting the
two multiturn trimmers. Vishay, and others, make trimmer adjusters which are available or
very little. The ive single turn trimmers need a small electrician's screwdriver.
Switch the module on and allow to warm up or i teen minutes. Trimmers should be adjusted
in the order as presented here.
NU 1 and NU 2
Turn the delay control to its minimum position. Place a scope probe on pin 6 o U7. Adjust
the multiturn trimmer TUNE so that the requency o the wave orm seen at pin 6 is 33kHz.
With your oscilloscope probe monitor the voltage on the right hand side pad o the sur ace
mount resistor R22. Set your probe to AC input, 200mV/division and 5uS/division. You
should see a wave orm with alternate spikes o di erent heights at 33kHz. Adjust NULL1
until the wave orm's peak amplitude is minimised. Essentially we are making the alternating
spikes the same size as each other.
Now put your probe on the right hand side o R24. Adjust NULL2 until the wave orm seen is
minimised.
OFF1 and OFF2
Connect a 5V peak 220Hz (the A below middle C) sawtooth to the input. Adjust the Drive
control until the red LED just comes on. The green LED should also be lit.
Set your scope or DC input, 500mV/division and 1mS/division. Probe the bottom pin o R45.
A sawtooth wave orm with slightly rounded edges should be seen i OFF1 is set correctly.
Adjust OFF1 so that the wave orm shows no hard clipping on the lower part o the wave orm.
The best place or OFF1 is just backed o rom point hard clipping starts to occur.
Probe the top pad o R32 and repeat the above or OFF2.
FBK
This one can be set to taste and it sets the maximum internal eedback level. The way I do it is
as ollows. Set both switches to NEG. Set the Feedback pot to its maximum setting and set
the Delay pot and the Drive pots to their minimum. Connect an ampli ier up to the DLY OUT
socket. Adjust FBK so that the module starts to oscillate and then gently back it o so that it
is sitting on the point that is just, but not quite, about to oscillate.
10

I you want your module to oscillate then you should know that the two modes o the FBK
switch behave slightly di erently. The amount o eedback required to sel -oscillate in NEG
mode is a little less than that required in POS mode. As such, i you want your module to sel -
oscillate in both modes, then you should trim FBK so the module just starts to sel -oscillate in
POS mode.
Remember, the classic langing e ect does not require either mode to oscillate and some olk
will ind the unstable nature o an oscillating langer disturbing. So eel ree to set FBK as you
wish. Whatever position that FBK is set, the maximum eedback levels are always obtained at
the longer delays thanks to the inherent gain drop (insertion loss) o the 3207 BBDs as the
clock requency rises.
SCA E and TUNE
I you have no need or Karplus-Strong physical modelling techniques or other tuned delay
e ects you can skip this bit entirely and just leave the scale trimmer in its de ault position and
adjust the TUNE trimmer to set pin 6 o U7 to 33kHz when the delay pot is at its minimum
position.
To set SCALE more accurately involves looking at the output o the high requency voltage
controlled oscillator (HFVCO) directly and using an oscilloscope or requency counter to
measure its requency. The HFVCO output is available on pin 6 o U7. Applying a 1V signal
to the CV1 socket should double the requency seen at this point. The easiest way to do this is
with your midi-CV convertor and then playing octaves on your midi controller. The 1V step
change when you play between a C on one octave and then a C an octave higher is ideal.
Set the delay pot to its mid point. With no CV input applied check the voltage on top solder
pad o D9. Fine tune the delay pot until it gets as close to 0.00V as you can get it. Using an
oscilloscope or requency counter measure the requency at pin 6 o U7. Adjust TUNE until it
reads 60.0kHz. Set your midi-CV convertor, or octave selector on your controller, to produce
0V when you play a particular C note on your controller. It should then produce +1V the
octave above that and +2V the octave above that.
Apply the midi-CV's output to CV1. Pressing the lowest C should not move the requency o
pin 6 U6 rom 60kHz. Play the C two octaves above that and adjust SCALE until it reads
240kHz. Play the bottom C and it should still be at 60kHz. The C up rom that should produce
120kHz and top one should again be at 240kHz. SCALE is now calibrated.
Now set the delay control to its minimum and remove the CV input. Re-adjust TUNE so that
pin 6 o U7 is 33kHz +/-0.5kHz. Note that the SCALE trimmer also a ects this inal value so
TUNE should always be re-adjusted a ter any re-calibration o SCALE.
To test the accuracy o the scaling you can 'ping' the langer module by sending a very short
AD envelope to the audio input. With the eedback control set to just shy o sel -oscillating,
the short sharp trigger pulse will tickle the langer to producing a lovely plucked string noise
which can be played via a keyboard using the CV1 input. The eedback control adjusts the
decay time o the note heard although higher notes will decay aster anyway.
11

Remember that the Flanger will not be able to be made to track per ectly over a very wide
range. As such, even i you have achieved a per ect octave spread with your two C notes, you
won't be able, or example, to have a per ect octave spread rom two similarly spaced C notes
a couple o octaves above that. I managed to get just under three octaves o playable tracking
using the ping method described above.
Final Comments
I hope you enjoy using the Oakley Flanger module.
I you have any problems with the module, an excellent source o support is the Oakley Sound
Forum at Mu wiggler.com. I am on this group, as well as many other users and builders o
Oakley modules.
I you have a comment about this user manual, or have a ound a mistake in it, then please do
let me know.
Last but not least, can I say a big thank you to all o you who helped and inspired me. Thanks
especially to all those nice people on the Synth-DIY and Analogue Heaven mailing lists, and
those at Mu wiggler.com.
Tony Allgood at Oakley Sound
Cumbria, UK
© May 2017 – updated November 2018
No part o this document may be copied by whatever means without my permission.
12