Man O' War Deluxe

FX TYPE: DELAY
11.5.19 - See an important update on the last page to
Enclosure Size: 1590BB or 125BB
Based on the Maxon® AD-900™ address an issue with the modulation!
© 2019 madbeanpedals

Overview

The Maxon AD series is sometimes overlooked by guitar players in search of a good analog delay and has most definitely
been neglected in the DIY pedal world. While it shares a lot of likeness to the more popular DM-2™, the AD-900™ has its
own unique voice. It’s neither as percussive as the DM-2™ nor so warmly filtered as the Memory Man™. The AD-900 is
neither too bright nor too dark and has perhaps more low end than the DM-2. Overall, it’s an excellent in-between of the
pillars of DM-2 and DMM.

Design-wise it parts most noticeably from the DM-2™ in the way the BBDs are arranged. It uses one main clock and one
slave clock as drivers. My guess is that this was done for maximum efficiency and lowest noise floor (one MN3101 can
drive up to two MN3005 easily so it is not a matter of economics). But, I wouldn’t necessarily call this a pristine analog
delay either. It can, and does, get pretty gritty after a few repeats. But, that’s not a bad thing for an analog delay.

The Man O’ War takes one further step and adds two bypasses: Regular and Tails. The regular bypass is just like
every other true bypass build. The Tails allows you to spill over the delay repeats when you turn the delay off. Both are
footswitches so you can change the bypass method on the fly.

The Man O’ War Deluxe is the same exact delay circuit as the Man O’ War with optical modulation added. It is a
more difficult and slightly more costly build but if you like modulation with your delay (and who doesn’t) it is a
great addition!

This is not a build for the novice. You should have some experience building pedals and also have a testing/prototyping
rig as well as an audio probe.

Controls

• DELAY: Sets delay time (max delay time will be between 500 and 550ms).
• FDBK: Number of delay repeats from 1 to “infinity”.
• MIX: Delay level mix.
• BAL1, BAL2: Sets the balance between the two outputs on each BBD.
• BIAS1, BIAS2: Used to calibrate the input bias on each BBD.
• CLOCK: Sets the correct clock range for the min and max delay times.
• LVL: Sets the output of BBD1 for cleanest delay signal.
• T1: Adjusts the point at which the FDBK control goes into “infinite” repeats. The ManOWars do not do self-oscillation.
• RATE: Modulation rate from slow to fast.
• DEPTH: Modulation depth from min to max.
• T/O/S: Triangle, Off, Square modulation. Setting this switch to the center position turns modulation off. Switch UP is
the triangle position.

Like the AD-900™, the Man O’ War runs on 12v DC power. But it can also run on 9v. What’s the difference? About 3v,
dawg! Also, 12v operation has a bit more output and overall sounds better to me. If you do run it at 9v change R55 and
R56 from 470k to 220k for the LFO.

Terms of Use: You are free to use purchased ManOWarDX circuit boards for both DIY and small commercial operations. You may not
offer ManOWarDX PCBs for resale or as part of a “kit” in a commercial fashion. Peer to peer re-sale is fine, though.

Technical assistance for your build(s) is available via the madbeanpedals forum. Please go there rather than emailing me for
assistance on builds. This is because (1) I’m not always available to respond via email in a timely and continuous manner, and (2)
posting technical problems and solutions in the forum creates a record from which other members may benefit.
Layout ManOWarDX

madbeanpedals
Traces ManOWarDX

madbeanpedals
B.O.M. ManOWarDX
Resistors Resistors Caps Caps ICs
R1 1M R34 12k C1 47n C35 4u7 IC1 4558
R2 1k R35 10k C2 1uF C36 4u7 IC2 NE570
R3 510k R36 10k C3 33n C37 1uF IC3 MN3005
R4 10k R37 10k C4 470pF C38 10uF IC4 MN3101
R5 10k R38 10k C5 4u7 C39 100n IC5 MN3005
R6 2k R39 10k C6 470pF C40 47pF IC6 MN3101
R7 3k3 R40 10k C7 33n C41 10uF IC7 TL062
R8 10k R41 10k C8 1uF C42 10n Photocell
R9 10k R42 33k C9 4u7 C43 22uF LDR1 9203
R10 3k3 R43 15k C10 4u7 C44 220uF Switches
R11 2k R44 10k C11 220n C45 100n TAILS 3PDT
R12 10k R45 2k C12 100pF C46 47uF T/O/S On/Off/On
R13 470R R46 10R C13 10uF Diodes Trimmers
R14 100k R47 1M C14 10uF D1 1n914 CLOCK 2k
R15 100k R48 2M2 C15 10n D2 1n914 BAL1 10k
R16 100k R49 8k1 C16 56n D3 8.2v BAL2 10k
R17 10k R50 10k C17 330pF D4 LED BIAS1 10k
R18 10k R51 10k C18 10uF BP D5 1N5817 BIAS2 10k
R19 10k R52 330R C19 1uF LED3 LED T1 20k
R20 10k R53 33k C20 1uF Transistors LVL 250k
R21 10k R54 100R C21 10uF Q1 - Q4 Si NPN Pots
R22 10k R55 470k C22 1uF Q5 2N5088 MIX 10kA
R23 5k1 R56 470k C23 1uF Q6 2N5088 DELAY 10kC
R24 5k1 R57 220k C24 1uF Q7 2N5087 FDBK 20kB
R25 100k R58 220k C25 1uF DEPTH 50kB
R26 130k R59 1k C26 10uF RATE 100kC
R27 130k R60 4k7 C27 4n
R28 12k R61 4k7 C28 39n
R29 5k1 R62 4k7 C29 820pF
R30 5k1 R63 4k7 C30 27n
R31 100k R64 4k7 C31 470pF
R32 130k R65 4k7 C32 10uF BP
R33 130k R66 47k C33 220n
R67 47k C34 100pF

The transistors used for Q1-Q4 in the stock unit were 2SC1815 but the exact component type
doesn’t really matter much. A lot of NPN will work fine (pinout on the board is C-B-E). I suggest
MPSA18, 2n3904 or 2n5088. You could also use BC550 but the pinout is reversed on those
(E-B-C). The 2SC1815 has a different pinout, too: B-C-E!

madbeanpedals
Shopping List ManOWarDX
Value QTY Type Rating Value QTY Type Rating
10R 1 Carbon / Metal Film 1/4W 100n 2 Film 25v Min.
100R 1 Carbon / Metal Film 1/4W 220n 2 Film 25v Min.
330R 1 Carbon / Metal Film 1/8W 1uF 1 Film 25v Min.
470R 1 Carbon / Metal Film 1/8W 1uF 8 Electrolytic 25v Min.
1k 2 Carbon / Metal Film 1/8W 4u7 5 Electrolytic 25v Min.
2k 3 Carbon / Metal Film 1/8W 10uF 6 Electrolytic 25v Min.
3k3 2 Carbon / Metal Film 1/8W 10uF BP 2 Electrolytic - BiPolar 25v Min.
4k7 6 Carbon / Metal Film 1/8W 22uF 1 Electrolytic 25v Min.
5k1 4 Carbon / Metal Film 1/8W 47uF 1 Electrolytic 25v Min.
8k1 1 Carbon / Metal Film 1/8W 220uF 1 Electrolytic 25v Min.
10k 21 Carbon / Metal Film 1/8W 1n914 2
12k 2 Carbon / Metal Film 1/8W 8.2v 1 Zener
15k 1 Carbon / Metal Film 1/8W LED 2 Red, Diffused 3mm
33k 2 Carbon / Metal Film 1/8W 1N5817 1
47k 2 Carbon / Metal Film 1/8W Si NPN 4 MPSA18, 2n5088 or 2n3904
100k 5 Carbon / Metal Film 1/8W 2N5088 2
130k 4 Carbon / Metal Film 1/8W 2N5087 1
220k 2 Carbon / Metal Film 1/8W 4558 1
470k 2 Carbon / Metal Film 1/8W NE570 1 or, V571
510k 1 Carbon / Metal Film 1/8W MN3005 2
1M 2 Carbon / Metal Film 1/8W MN3101 2
2M2 1 Carbon / Metal Film 1/8W TL062 1
47pF 1 Ceramic / MLCC 25v Min. 9203 1 Photocell
100pF 2 Ceramic / MLCC 25v Min. 3PDT 1 or, DPDT (footswitch)
330pF 1 Ceramic / MLCC 25v Min. On/Off/On 1 SPDT, Pin Mount
470pF 3 Ceramic / MLCC 25v Min. 2k 1 Bourns 3362p
820pF 1 Ceramic / MLCC 25v Min. 10k 4 Bourns 3362p
4n 1 Film 25v Min. 20k 1 Bourns 3362p
10n 2 Film 25v Min. 250k 1 Bourns 3362p
27n 1 Film 25v Min. 10kA 1 PC Mount, Right Angle 16mm
33n 2 Film 25v Min. 10kC 1 PC Mount, Right Angle 16mm
39n 1 Film 25v Min. 20kB 1 PC Mount, Right Angle 16mm
47n 1 Film 25v Min. 50kB 1 PC Mount, Plastic Shaft 9mm
56n 1 Film 25v Min. 100kC 1 PC Mount, Plastic Shaft 9mm

madbeanpedals
Parts Guide ManOWarDX
10uF Bi-Polar cap:
http://www.mouser.com/Search/ProductDetail.aspx?R=ECE-A1EN100Uvirtualkey66720000virtualk
ey667-ECE-A1EN100U

8.2v Zener:
http://smallbear-electronics.mybigcommerce.com/diode-zener-1n4738a/

NE570:
http://smallbear-electronics.mybigcommerce.com/ic-ne570/

V571 (sub for NE570):

http://smallbear-electronics.mybigcommerce.com/ic-v571d/

Xvive MN3005:
http://smallbear-electronics.mybigcommerce.com/mn3005-re-makes-xvive-audio/

MN3101:
http://smallbear-electronics.mybigcommerce.com/ic-mn3101/

Bourns 3362p 22k:

https://www.mouser.com/ProductDetail/Bourns/3362P-1-223LF?qs=sGAEpiMZZMvygUB3GLcD7v%2
F2K2JTtKgbVPDHLENkzyQ%3D

Bourns 3362p 10k:

https://www.mouser.com/ProductDetail/Bourns/3362P-1-103LF?qs=sGAEpiMZZMvygUB3GLcD7k%2
52Bod3ZqvEIQboRRPdOKB6M%3D

Bourns 3362p 1M:

https://www.mouser.com/ProductDetail/Bourns/3362P-1-105LF?qs=sGAEpiMZZMvygUB3GLcD7kdd
hVJPyV2kST8Lo8Gl%252B%2F8%3D

16mm Right Angle PC-Mount:

http://smallbear-electronics.mybigcommerce.com/alpha-single-gang-16mm-right-angle-pc-mount/

9mm Right Angle Plastic Shaft:

http://smallbear-electronics.mybigcommerce.com/alpha-single-gang-9mm-right-angle-pc-mount-w-
knurled-plastic-shaft/

On/Off/On SPDT:
http://smallbear-electronics.mybigcommerce.com/spdt-center-off-short-lever-pc-mount/

9203 photocell:
http://smallbear-electronics.mybigcommerce.com/photocells-cds-5mm-diameter/

madbeanpedals
Notes ManOWarDX
Calibration (by ear)

This procedure should be done in a testing environment before boxing up the pedal. You do not need
to fully wire up the jacks and switches to do it. You’ll need wires for 12v, GND, IN, OUT, and the two
wires for the TAILS connection. An audio probe is required.

Set pots and trimmers as follows (make sure the TAILS wires are connected first):

• DELAY: 12 o’clock
• FDBK, MIX, RATE and DEPTH: Min
• T/O/S: Off (center position)
• LVL: A little less than half-way up
• All remaining trimpots to 12 o’clock

1. For this step, remove IC5 from its socket on the PCB. Connect power.
2. Use an audio probe to probe pin7 of IC3 to verify that you have signal to the input of the first BBD.
If you do not, check pin7 of IC2a and the emitter of Q2 for output. Debug as necessary.
3. Using the audio probe, probe either pin 3 or 4 of IC3. Adjust BIAS1 until you get the cleanest
sounding delay.
4. Disconnect power and insert IC5. Reconnect power.
5. Audio probe pin7 of IC5 for input. Adjust the LVL trimmer so the volume at pin7 is more or less
equal to the outputs of pins 3 or 4 of IC3.
6. Adjust BAL1 left and right to listen for any improvement in the delay output of IC3. If none, leave it
in the center.
7. Probe pin 3 or 4 of IC5 and adjust BIAS2 to get the cleanest sounding delay output.
8. Probe R35 and adjust BAL2 left and right to see if it improves delay output at all. If not, leave it in
the center.
9. Set the Delay pot to max and FDBK to 12 o’clock.
10. Probe pin10 or 11 of IC2b. Adjust the clock trimmer clockwise for the most delay time possible
without any clock noise (whine) in the signal.

After these steps, disconnect your audio probe and listen to the actual output of the pedal. Turn
FDBK all the way up. While listening to the output, adjust T1 left to increase the maximum number of
repeats to the desired amount. You can adjust the LVL trimpot up to increase both the total FDBK and
MIX output. These two trimpots are interactive and I advise against adjust the LVL too high (shoot for
a setting between 1/3 and 2/3 up) or it may start to distort the delays. The Man O’ War doesn’t really
do self-oscillating feedback. At least not without adjusting the LVL so high as to make the volume of
feedback get too loud. Shoot for as close to infinite repeats as possible when making your T1 and
LVL adjustments.

Bypass Operation

• True bypass operation: Leave the Tails switch on and use the Byp switch for on/off. Delay
repeats will cut off when the effect is bypassed.

• Tails bypass operation: Leave the Byp switch on and use the Tails switch to toggle the effect on
and off. Delay repeats will continue after the effect is turned off. In this state the effect is not true
bypass.
madbeanpedals
Mods ManOWarDX
As mentioned, the Man O’ War(s) don’t do self-oscillating feedback. IMO, this is due to the low output
of the first half of the Compander (NE570). On a DM-2 this is typically about 3v on a 9v supply. Here
it is about 3v (pin7) on a 12v supply. There may be a good reason it was done this way. Perhaps the
designer thought it was better to hit the input of the first BBD with a lower amplitude signal and then
use the LVL trimmer to make up volume at its output to keep the delays as clean as possible. And,
yet, the Sallen-Key style filter directly after the output of the compressor portion would benefit from a
higher bias voltage. But, these are guesses. I did not spend any time working through this “problem”
since I like the effect as designed.

Point being, this is an area where you could try to mod the Man O’ War(s) for self-oscillating
feedback: by increasing the Compander output. To do this, change the values of R17 and R18.
Probably 15k, 18k or 20k for both would be the starting point. Of course, you will want to socket those
two resistors if you do this. It might require a re-bias on the BBDs if you calibrate it first with the stock
10k resistors.

And, since someone will ask: can you run the Man O’ War(s) on 15v? You should be able to. Again,
I have not done it but there is no reason it cannot be done. You’ll want to set that compander output
(pin7) for somewhere between 5v and 7v, I think. Additionally, you should increase R62 and R63 to
either 8k2 or 10k. Keep in mind that it would have to be a regulated 15v and you need to take into
consideration how to get that. The best way would be to use an 18v supply and make a little breakout
board with a LM78L15 or LM7815 regulator plus bypass caps on the 18v input then jumper through
D5.

I would not advise using a charge pump. Even though the total current draw of the effect is pretty
low, it already has two clocks in it. Adding a charge pump increases the chance of clock noise and
heterodyne. YMMV.

madbeanpedals
Wiring Diagram ManOWarDX

madbeanpedals
1590BB Drill Guide ManOWarDX
Note: Drill Guides are approximate and may require tweaking depending on the types of jacks, switches and pots you use.

The two LED bypass indicators (effect bypass and Tails bypass) as well as the modulation
LED (LED3) should be soldered directly to the PCB. Place them loose in their pads. After
mounting the PCB in your enclosure, move the LEDs into place in their respective holes and
solder in place.

madbeanpedals
Voltages ManOWarDX
Q1 Si NPN IC1 4558 IC4 MN3101 Delay Time Freq (CP1) mA
C 11.71 1 5.84 1 11.04 min 150kHz 33mA
B 5.32 2 5.86 2 5.54 max 7.6kHz 20mA
E 5 3 5.84 3 0
4 0 4 5.51
Q2 Si NPN 5 5.83 5 10.81
C 11.71 6 5.85 6 1.7
B 2.96 7 5.85 7 varies
E 2.39 8 11.71 8 0.745

Q3 Si NPN IC2 NE570 IC5 MN3005

C 11.71 1 0.88 1 11.71
B 6.13 2 1.77 2 5.58
E 5.55 3 1.77 3 6.18
4 0 4 6.16
Q4 Si NPN 5 1.77 5 0
C 11.71 6 1.77 6 5.58
B 5.53 7 2.97 7 5.96
E 4.94 8 1.77 8 0.745
9 1.77
Q5 2n5088 10 3.98 IC6 MN3101
C 0.472 11 3.98 1 11.11
B 0.65 12 1.77 2 5.54
E 0 13 11.71 3 0
14 1.77 4 5.57
Q6 2n5088 15 1.77 5 11.04
C 9.3 16 0.832 6 1.69
B 0.47 7 9.38
E 0 IC3 MN3005 8 0.755
1 11.71
Q7 2n5087 2 5.54 IC7 TL062
C 9.4 3 6.04 1 varies
B 11.58 4 6.04 2 varies
E 11.71 5 0 3 varies
6 5.51 4 0
7 6.06 5 varies
8 0.755 6 varies
7 varies
8 11.61

12v, well regulated supply, no bypass LEDs active. Maximum current draw: 33mA

madbeanpedals
Build Pic ManOWarDX

madbeanpedals
Schematic ManOWarDX

madbeanpedals
Error, error, error! ManOWarDX
Problem: When the modulation switch is set to the center position to turn modulation off, the rate LED continues to blink
for a while before fading out. This also means the LED/LDR combo is active until the LED goes dark.

Cause: The IC pin attached to the center lug of the On/Off/On switch was left floating in the off position. This means any
current sitting on that switch continues to drive the LEDs attached to it. Because there is a path to ground through the
resistors that current will drain off but can takes several seconds.

Solution: Put a 2M2 resistor from IC7 pin3 to ground. When the switch is in the off position, this has the effect of nearly
instantaneously draining that pin and making the LEDs go dark equally as fast.

Why the mistake happened: Most likely because when I originally prototyped the MoW I used an AQB MOD PCB for the
modulation and no switch (rather I used a resistor in line with the LDR so the modulation essentially turned off when the
depth was turned all the way down). Even though I went further and built both the MoW and MoWDX production boards
in addition to my prototype I did not catch the design mistake. Most likely because those pesky LEDs stay lit in sync with
the peaks of modulation. IOW, if the LED is fully lit and you put the switch to center, you have the staying lit problem. If the
switch is put to off when the LED is in the fully dark part of its swing, it stays dark.

So, my apologies for the mistake. It was pretty noobish on my part and once someone reported the problem I pretty much
knew right away why it was happening. Finding the easiest solution actually came from forum member Cybercow. Thank
you so much for figuring this out, Mark!

Two images below:

Put a 2M2 between the pin in the orange box and one of the two ground points in the purple boxes. The lower one is a
ground via and even though it is very small it is possible to solder to it if you are careful.
Second image is the fix on my build.

madbeanpedals