__________________________________________________________________________________________

www.analogwizard.com

The DDRM Expander Story

Analog Wizard provides professional Services for Audio/Studio/Synth repairs and development.

There is not enough space left in my synth-studio for an original Yamaha CS80 (and the decision got even
easier regarding the prices to pay for such a service intensive weight of 120 kg :-) But getting a deckards
dream (DDRM) was a must and waiting for the announced expander was almost a pain. Once I got it, I
straight connected it to my DDRM and played around. Cool at the beginning... but after some jamming, I
found several weak points and started to realize, that the unit, in my opinion, was far from perfect.

▪ Not even one single LED showing “ON and ready status” (extremely minimalistic)
▪ Some faders behavior where far from optimal and started to annoy me...
=> ADSR Release adjustment
=> RM-Speed adjustment
=> DL-Tone adjustment
▪ Rather dull sound compared to the direct signal from DDRM

I decided to get it on my lab table, open it up and have a close look to see what was going on inside.

During this, I found even more weak points:

▪ Erroneous SW in the Delay FX. If you put the fader fully up there is a fix delay of about 160msec
with absolute strange behavior of feedback/tone => according to Black Corp it is a feature!!!
Not a feature to me… (adjusted the fader range to start at shortest delay)
▪ Unfortunately, the delay time cannot be adjusted without disturbing zipper noise (could maybe
be fixed with proper delay software)
▪ Master Volume sets headphone and FX output in parallel but headphone is extremely loud,
making it impossible to use both outputs at the same in the best possible way with high output
level on the line out to optimize SNR and reasonable levels on the Headphone (I split headphone
and line out with line out having now fix high level)
▪ Mono Output has a 6db higher output level than L/R and can easily run into distortion (I tuned
the output driver for 6db less output level)
▪ The input circuit monoes the two inputs and limits the high frequency range too strong for my
opinion. Trough-out the whole unit there is a substantial loss of high frequency on the direct
signal path!
▪ Level matching of each FX stage is far from optimal, there is at least a 12db waste in S/N ratio (I
fixed the level matching internally) You get either more noise or distortion... hence where the
overload LED came from
▪ Too much of a level drop when switching to Tremolo mode (not fixed yet, will need an additional
analog switch)

I then decided to fix as much as I could on above issues. Below you will find my mods.

Missing LED’s

During my investigation I found, the expander would gain much by adding some LEDs and showing some
internal “behavior” to the user. Here is what I added:

__________________________________________________________________________________________________________

2021/03 © www.analogwizard.com [email protected]

________________________________________________________________________________________________
www.analogwizard.com

▪ Power ON LED to get a first sign of life if the unit is switched on

▪ An additional LED, showing ADSR ENV, even if it is not heard (@ zero modulation amount)
▪ Adding an FX clipping level LED to allow for best settings regarding SNR/Audio performance.

Adding the Power-ON LED

The easy way is to add a LED to one of the power rails. But I also wanted a fault detection on the +/-12V
power rails and a LED that shows if BOTH DC Levels are ok. With my solution, if one of the DC-rails drops
below 6Vdc (or the total of both below 18Vdc), then the LED will go dark.

And here is the mod inside the expander unit with the red LED next to the power switch. A small piece of
pcb with the Z-Diode, the resistor and a molex plug is all that is needed.

Adding an audio signal clipping LED

My testing showed, that it is very tricky to get the right input level setting for best SNR but without
distortion in one of the three FX-stages. Inside the unit, three FV-1 chips are running the three FX sections
(chorus/tremolo, delay, reverb). These chips have an integrated overload LED output. Oring all three
overload outputs together would make a perfect single overload LED to signal any overload within the FX
chain. That was exactly what I did.

__________________________________________________________________________________________________________

2022/01 © www.analogwizard.com [email protected]

________________________________________________________________________________________________
www.analogwizard.com

The schematic shows the oring of the 3 overload outputs via D1,D2 and D3. In the picture you can see 2
of the FV-1 chips connected to the small PCB with the electronics to drive the single LED.

Adding an ADSR Envelope LED

Once I was into signaling internal states, I thought it might be cool to get an optical feedback of the
envelope settings. Did it… (added PCB behind front)

For me this was a very useful optical feedback and was showing the ADSR even without listening to it
(ring modulation intensity set to 0).

Fader behavior optimization

I hate fader/pot action ranges, that are not musically spread across the fully available range like abrupt
changes within a small area or hardly any action in a certain range and therefore “compressing” the rest
into a smaller range making it hard to set values rather precisely. Fader and pots can easily be set as
appropriate by setting the start point, the end point and the range curve (like linear, logarithmic, antilog
or special curves like midrange stretch or midrange shift).
I am not going into more Details for this, but I fixed the Faders to my “musical” demands just by adding a
few resistors around the faders and testing different values for best performance optimization.

I did this for:

- ADSR Release adjustment
- RM-Speed adjustment
- Delay-Tone adjustment
- Delay FX range setting for shortest Delay without the zero position “anomaly”

Treble loss and dull sound

The input stage monoes the 1/2 inputs with a first 0db input amp. In the feedback, there is 100k (R155)
together with 47pF (C154) that dampens the frequency at 20Khz by 1.3db. I changed tis input amp to
150k with 15pF to give it a bit more gain and less HF loss. The same situation was with the output amp
__________________________________________________________________________________________________________

2022/01 © www.analogwizard.com [email protected]

________________________________________________________________________________________________
www.analogwizard.com

(R154/C153) adding another 1.3dB HF loss. I changed the ouput to -6db gain with 47k feedback. This
moves the gain bandwith up an octave without changing C153. Now the overall sound was ok for my
ears. It seems, that some DIY Kits have wrong feedback capacitors delivered. You need to check!

Level Matching of each FX stage

During my analysis I found that there could be a better level matching between each FX-stage. With this,
the distortion point of each FV-1 chip will be reached more or less at the same input level setting. This
helps to get the best possible SNR with the unit. In case of reaching the clipping point, it will be shown
with the clip LED mod anyway and independently of the stage that actually clips the signal.

I changed R74 from 10k to 11k leaving the RM output with slight less gain. This showed better matching
of signals with or without ring modulation.

I reduced the input level of the delay FX by 6db (R41/R51) by changing the two resistors to 12k. I then
compensated this on the FX output amp by changing R38/R46 to 12k as well boosting the signal by 6db.
The overall gain remains, but the clipping point of the delay FX matches much better with the rest of the
unit.

I then adjusted the REV FX balance mixing by changing R111/R123/R122/R132 to 20k. I also adjusted the
overall gain of the reverb wet/dry mixing amp by reducing R108/R109 to 68K. As a nice side effect, this
extended the HF bandwidth and was again an improvement on the dull sound.

Last I changed the headphone gain with R142/R148 to 30k giving 6db lower gain and again a better
bandwidth for the final signal.

Firmware Bug
My unit was delivered with a PIC 16F1508 for Midi Processing. Unfortunately, the Version 1.2 has
problems with triggering the RM every time a new note is played. The ADSR sometimes hangs and stays
in sustain mode! An update should help, Black Corp. says. Easy if you own a PIC programmer.

And I want more!

With the mods described above, I’m now totally happy with the expander module. Overall noise figure is
now quite good. The ring modulator is what I wanted and it’s great. All the fader settings are CV based,
thus leaving room for more easy mods... I will definitely go for a CV-Input on RM-Speed and RM-
Modulation Amount.

My wish list doesn’t end here. But the rest would be to much as a mod, but imagine…
▪ a hardware that would be laid out in full stereo with switchable mono input mode.
▪ A bypass switch that gives the pure Input that I still want to use (without juggling around jacks...)
▪ adding midi remote functionality (configurable) that adds:
o frequency offsets to the RM speed based on highest/lowest note played
o frequency offsets to the RM by pitch wheel, modulation wheel or any other CC
o RM modulation amount by pitch wheel, modulation wheel or any other CC

I need to stop here and go playing 😊

__________________________________________________________________________________________________________

2022/01 © www.analogwizard.com [email protected]