Synth Repair

::: EML Synkey = Sexy :::

These pics are 'borrowed' from an Ebay auction.  I have an EML 200 silver face that is absolutely beautiful.  I consider those rare but the Synkey is Much more so. I love the idea of a Punch Card memory system!  Not bad for a little company in Connecticut!






Brochure borrowed from here.

::: IF :::

::: Simmons SDS V : Terms of Refurbishment :::

I had a Simmons SDSV in storage in the underground garage of our apartment building.  It's exactly like the other one I have written about here before.  It had the Kick, Snare, Tom, Tom, Tom cards and an MFB sequencer that i installed in the last two slots.  One day I went to the basement garage to retrieve it to loan to a friend for a recording project and got a surprise.  Much to my dismay, the cement ceiling of the garage had a leak and it happened to run along that ceiling to directly above the box i had stored the Simmons in.  I didn't even worry at first after the initial shock because i had wrapped the boxed gear in plastic for storage. This is where things got ugly...

Apparently the water had saturated the top of the box and began to run into the large wrapped bag protecting the gear.  I really couldn't believe it, the entire bag was FULL OF WATER.

Not just crappy rain water either.  I was mineral rich leeched water from the cement Which had solidified again and formed a crusty hard calcium like layer all over and inside of the Simmons.  After the shock, anger, depression, and acceptance of this catastrophe i decided with resolve that i had to save this instrument.  I felt like it was my duty to do so.  It took a long time to accomplish and a subsequently longer time to be able to get my thoughts together into a post on the subject.

So it begins.  Part One: The Damage Done


::: IF :::

::: Garfield Electronics Mini Doc :::

While we are on the subject of Garfield Electronics gear... I have been trying to get this one going for a little while and i haven't quite figured it out.  I got it as 'completely working' from Ebay and of course its not.  Luckily i have another one that works great to compare it against.  As with most topics i discuss on this blog, it's impossible to find a manual let alone a service document for these.

Upon receiving this unit i noticed that the triplet triggers weren't firing but the straight divisions were.  When i opened it up the arpeggio trigger wiring looked completely different from the other unit i have.  So i reconfigured the wires the way i assume they are supposed to be and it didn't fix the problem.  I also replaced some caps that were leaking and obviously bunk.

At this point i decided maybe one of the chips had failed.  They are all cheap and easy to find so i bought one of each and plan to start at one end and hope to fix it before reaching the other!  These multi pin chips are a pain in the butt to pull so i am socketing the replacements...

More to come...


::: if :::

::: Garfield Electronics Master Beat :::

I picked up this Garfield Electronics Master Beat on a whim.  I was hoping it was capable of generating tempo internally so i could sync some gear with it and use it as the master tempo source. It looks like it will do that which is great!  I also wasn't expecting it to show up with the manual but it did.  Total bonus as there's not much info about these availible on the web.  I think i will scan the manual in to add it to 'the google'.

Sadly, the shipper did a terrible packing job and all of the rack ears are bent.  So i figured i would do a little straightening before i cleaned it up.


::: IF :::

::: Roland TR 808 :::

The true magic of the Roland TR-808 Is in its lively sound.  It really sounds stunning in a way no samples ever seem to capture.  It isn't just the analog circuits performing in real time.  It's the fact that your patterns have not only notes, tempo, and pattern but also accent! The accent is programmed just like any other instrument on the grid and has a master amount adjustment by the kick drum controls.  Which ever steps you choose to accent really pop out and add a huge amount of dynamic energy to the pattern.  Its very impressive.  Super dynamic and sounds great!




Cool pattern sheets for famous songs: Rob Ricketts TR-808 Programming Posters

Manual:  Roland TR-808 Manual

::: IF :::

::: Roland Juno 106 Repair Part 2 :::

So the Juno 106 from here came back to the bench over the holiday.  But not because it had a failure again! The band noticed that sometimes the patch edit lights would come on even if they hadn't adjusted any sliders.  When it did the LFO would turn up to maximum and make the VCO all out of tune.  Of course when they came to drop it off we couldn't get it to happen, so i decided to try again later.  When i did so i noticed that if i picked up the front left corner of the synth the problem would suddenly appear like clockwork.

I opened it up and reseated all the cable harnesses.  I also decided that the problem wasn't really an issue with the electronics but the chassis.  The end cap screw mounts were all in need of some love.  I had replaced the screws but if this was to be a reliable instrument i needed to modify the way the chassis fits together.  I decided to drill out the original screw holes on the bottom row of the end cap mounts.  This would allow me to install bigger screws that fastened with a star washer nut instead of relying on the thin steel bracket that was stripped out.

I also decided to fix the right back end cap hole as it was partially shattered.  My fix doesn't look pretty but it is on the inside so no one will see it anyway.  I used small cuts of tooth pick and epoxy to strengthen the cracked mounting hole and rebuild the missing chunk to support the new screw properly.  I was unable to use the new screw and nut fasteners on the front right end cap hole because that nut would be under the keyboard and impossible to get back off.

Oh yeah, i almost forgot!  I tracked down some shallow IEC connectors and replaced the silly 2 prong Roland one with a properly grounded 3 pronger!


::: IF :::

::: Crumar Bit 99 : Part One :::

What an odd beast.  I remember seeing one back in the 90's at a local music store.  Remember those!?  Although the price was right I didn't buy that one because it had lost it's memory and was full of nonsense patches.  This type of problem makes many of these old instruments seem catastrophically scrambled. But upon inspection the scramble has a surmountable solution.  You can replace the battery and either sysex dump or tape dump the patches back in and often you are good to go.

The real issue with the more rare instruments like this one is availability of service info.  For instance it's easy to find ( almost ) any Roland service manual to trouble shoot a synth but Crumar and many rarer instruments can be mysterious about their ailments.

What are the power supply voltages supposed to be?

What was the final OS version?

How do i calibrate the voices so that they match and perform properly?

These are the mysteries that lie beneath the hood of this Crumar Bit 99.  It's built like a tank. The metal chassis is rugged and fully hinged like old Roland synths.  The voices are all firing ( nonsense at the moment ) and the display is working.  There are some messed up keys that i am still sorting out.  Some were broken but i used epoxy and toothpick pieces to reattach the hooks and re enforce them.  The other key problems have to do with triggering properly and triggering with velocity.  That's right!  this synth has VELOCITY on the keybed.

That my friends is why this is an awesome instrument.

It's much like the old Ensoniq Mirage keybeds which i have fixed many times in the past.  There's a spring that is stretched through a window at each key.  The spring moves from one contact to the contact on the other side when a key is pressed.  The CPU determines what the velocity of the key press was from the way the spring travels across the divide.  Pretty raw and clunky but it does work as long as the contacts are clean and the springs are not mangled.  Several of the springs are really mangled and barely working on this keybed so i am going to go scavenge an old parts Mirage keybed for springs.

But i have been unable to get the tape dump to take.  It fails half way through.  There are error codes that come up but without the service manual i don't know what they are!  The manual doesn't list any user issues or troubleshooting info.  It's otherwise a fairly well laid out user manual though!  I do have a sysex dump of the patch info and i will probably resort to that.  I read somewhere on the net that people have issues with the tape inputs on these so maybe this one suffers from that too.  I imagine it's lumpy power or old capacitor related interference.

There will be updates!


::: IF :::

::: Factory Racked Simmons Clap Trap :::

This appears to be a Simmons Clap Trap that was Racked at the factory back in the day.  It was an Ebay auction so these pics aren't mine.  It looks neat and has the Simmons sticker on the back.  The knobs are cool too!  

::: Simmons SDS 6 Schematics :::

::: Mastering App for Iphone From Howie Weinberg!!!~ :::

OMG, there is low depth to the lows in software design!  Now you can master a record wile on the subway, or in jail, or while taking a jog! Mastering App for Iphone!


::: IF :::

::: Oberheim OB-8 Auxiliary Info :::

There's a lot of mystery beneath the hood of the Oberheim OB-8. From the hidden page two parameters to the OS B5 DSX / Midi parameters and Reassignable Mod / Pitch bender orientation.  I figured I'd put a post together to cover a few of these and repost some arcane knowledge about Oberheim's almost most advanced synthesizer.  Consider this my first Synth-wiki-dump:

Article one:




The OB-8 has many functions that a user would never guess existed from

looking at the control panel labeling.  This is a cheat-sheet to avoid

constant reading of the manual.  Most of these functions are hidden by

virtue of requiring some combination of simultaneously held buttons,

or some other such thing that doesn't seem obvious.

*  Change split point:

Pressing the SPLIT button normally takes you in and out of keyboard

SPLIT mode, but you can also hold down SPLIT while pressing a key to

change the split point.  Split points are non-volatile, and will

survive patch changes and shutdowns.  However, they are also stored

with split patches, so selecting a new stored split will recall its

stored split point as well, overriding whatever is currently set.

*  Transpose keys on LOWER keyboard:

The keyboard may be transposed by any interval, but the "lower" part

in a split/layer can only transpose upward.  (Nevermind that in a

layer, both patches cover the whole keyboard -- there's always one

patch that is assigned through the UPPER button and one that's

assigned with the LOWER button.)  The reason the "lower" patch can

only go upwards is because the interval is selected on the keyboard

relative to C0, the lowest C.  Press and hold down SPLIT or DOUBLE

(as appropriate) and LOWER at the same time, and select the upward

interval above C0 on the keyboard you'd like to transpose to.

*  Transpose keys on UPPER keyboard:

This is the same, except transposition can go up or down because the

reference point is now Middle C (C3).  It's true even if you're

transposing a layer ("double"), and both patches are covering the

keyboard.  Hold down SPLIT or DOUBLE (as appropriate) and UPPER at

the same time, and select the interval above or below Middle C.

Transpositions of both the UPPER and LOWER components of a split or

layer _are_ stored with a split or layer patch!  Recalling a split

or layer that was saved with transpositions active will also recall

the transpositions!

*  Detuning one side of a split or layer:

The "lower" side (the patch assigned with the LOWER button) in a split

or layer can be detuned relative to the upper one by holding down the

LOWER button and turning the OSC2 DETUNE knob at the same time.  This

control is only active for LOWER.  (You can't hold down UPPER.)  Like

the other special hidden settings of splits and layers, this one too

is actually stored with the patch.  Recalling a split or layer that

had a relative UPPER/LOWER detuning in effect will also recall that

detuning interval.

*  Reset all split/layer settings to defaults:

To reset the split point to Middle C, cancel all traspositions,

relative detuning, etc.), hold down whichever button applies

currently (SPLIT or DOUBLE), and press the MANUAL button.

*  Recalling/writing splits and layers:

Contrary to the way patch selection normally works on Oberheims when

not in split or layer mode, in this case, the GROUP buttons (which

are similar to Roland's concept of "bank" buttons, to select a bank

of 8 patches) now behave just like regular patch buttons.  That is,

they don't select a bank of patches, they just select a single patch,

and have equal weight to any of the eight regular patch buttons.

This means that in split or layer mode, rather than having as many

possible patch storage locations as there are GROUP button combinations

_multiplied_ times the regular patch buttons (which on Rolands would

typically be 64), here you only have the 4 GROUP buttons _plus_ the

8 regular patch buttons, for a total of 12 patches.  In split or layer

mode, a GROUP button is just another patch button, and it selects a

split or layer all by itself, not in combination with a button from

the regular buttons.  However, your 12 splits and 12 layers do not

share slots, and you always have 12 patch locations of each (12 stored

splits and 12 stored layers).

To recall a stored split or layer hold down SPLIT or DOUBLE as

appropriate, and press one of the GROUP or regular patch buttons,

keeping in mind that any of them is just a patch button as far as

splits and layers are concerned.  (There are no "groups" or "banks".)

To write one, do the same thing, but hold down the WRITE button until

its light comes on first, and then do it.  All transpositions, detunes,

split points, etc. will be stored with the split or layer.

*  Modulation lever peculiarity:

The modulation lever on an Oberheim has a spring-loaded center

position like a pitch wheel, but does nothing if you push it away

from you.  That's just normal Oberheim behavior.

*  Pitch bend lever peculiarity:

Contrary to the norm adopted by every other synthesizer company, on

Oberheims, pushing the pitch bend lever forward makes the pitch drop,

and pulling back on it makes the pitch rise.  This is completely

backwards, but it can't be changed without rewiring the potentiometer.

That's also just normal Oberheim behavior.

*  Modulation lever's LFO waveform selection:

This is perhaps the finest example of hidden control panel functions

the OB-8 has to offer!  First off, the RATE and DEPTH knobs near the

pitch/modulation section of the keyboard _pull up_.  Yes, that's right,

you can pull up on the knobs like the headlight switches on older cars,

which turn but also pull out to turn the headlights on.  (Bet you would

have never even thought of trying that...)  Naturally it's not labeled.

And here's now you select LFO waveforms for modulation:

Triangle Wave Push down the RATE knob while doing

doing nothing with the pitch bend.

If it was already down, pull it up

first so you can then do it.

Square Wave Pull up on the RATE knob while doing

nothing with pitch bend.  If it was

already pulled up, push it down first

so you can, and then do it.

Rising Sawtooth Wave Pull up on the RATE knob while pulling

back on the pitch bend lever.  See

above for if the knob was already up.

Falling Sawtooth Wave Pull up on RATE knob while pushing the

pitch bend lever forward.

White Noise Push the RATE knob down while pulling

back on the pitch bend lever.

Sample & Hold Push the RATE knob down while pushing

forward on the pitch bend lever.

This is really how you select waveforms for modulation lever LFO's.

Okay.  You can stop laughing now.

*  Setting non-zero amounts of modulation all the time:

Since Oberheims have a spring-loaded lever for modulation rather than

a non-spring-loaded wheel like everybody else, it's not possible to

just park the modulation setting at a non-zero position and leave it

there while playing on the keyboard like it would be anywhere else.

However, there's a DEPTH knob, which if turned up, will let you set

how much modulation you'd like to have running all the time, even when

you're not pushing the lever.  The trick here is, it doesn't do

anything at all when it's not pulled up like a headlight switch.

(That's why I'm filing it under non-obvious.)  The advantage to this

is supposed to be that you can find a setting you like, and jump to it

immediately by just pulling up.

It may be worth noting (though it's not as unconventional -- lots of

synths have this feature), that there is a third source of modulation

control possible, beyond the lever and the pullable depth knob, if you

have a variable pedal connected to the MODULATION jack on the back.

In that case, you modulation amount would be a sum of the current

lever position, the DEPTH knob position (if it's currently pulled up),

and the pedal.

*  Setting pitch-bend range:

The vaguely labeled "AMOUNT" button near the pitch bend lever is a

toggle that can run in one of two modes, depending whether it is lit

or not.  When it is off, the pitch bend range is fixed, and is always

a whole tone up or down.  If you turn AMOUNT on, the pitch bend range

is programmable.  Doing that is, of course, another hidden function.

Hold down AMOUNT, and at the same time play any key in the lowest

octave of the keyboard up to C1.  This will let you set the pitch

bend range anywhere from a semitone up to a full octave.  The setting

becomes non-volatile, and will take effect whenever the AMOUNT light

is on.  Pressing the AMOUNT button again takes it back to its fixed

nature, which is always one whole tone up or down, and you can toggle

between the fixed setting and the user-defined one that way any time.

*  Accessing the arpeggiator controls:

The buttons and knobs that control arpeggiator funtions are mostly the

same buttons and knobs that control pitch bend and modulation lever

functions.  The MODE button switches the meanings of those controls

between the two roles; when the light is on, you're looking at the

arpeggiator settings.

*  Keyboard HOLD function with the arpeggiator:

You might not think to try this, but the KBD and HOLD buttons can be

made to come on at the same time.  This way, the arpeggiator will

arpeggiate notes held down by the HOLD function _AND_ notes being

held down manually on the keyboard, both.  You must hold KBD and HOLD

down together, because normally pressing one makes the other go off.

*  Arpeggiation types:

The UP and DOWN buttons do individually what you would expect -- UP

makes the arpeggiator play upward, and DOWN makes it play downward.

Less obvious is that you can have them both on to make it play up

then down, and having them both turned off puts it in random mode.

Also, random mode tries to give slight preference to the first note

played, on the presumption that it may be your tonic.

*  Using arpeggiation ranges:

Another wonderfully obscured function, the OB-8 arpeggiator can

memorize and recall any of five programmable transposition settings,

which then become used only in the arpeggiator.  (They have nothing

to do with regular keyboard usage.)  Also, there is a sixth mode for

not using any arpeggiator tranposition, which is the default.

It's hard to explain what this does.  Obviously, when you're in the

default mode, the arpeggiator plays what you played.  This is selected

(if you're not already in that mode) by holding down MODE and pressing

the LEFTMOST of the six buttons below the pitch/modulation levers.

(That's the one labeled HOLD, the first button in that row.)

Ignore the labeling on the buttons; these functions are totally hidden

and have nothing to do with any of those six buttons labeled functions.

But if you hold down MODE and this time press the SECOND button in that

row (labeled KBD), you enable the first tranposition.  Since the OB-8

comes with all the transpositions we're talking about here set to

octaves, that means the arpeggio will spread out over an octave beyond

the spacing at which you actually played it.  However, that interval

can be user-defined (see below) to whatever you want.

If you hold down MODE and press the THIRD button in that row (the one

labeled OSC2 ONLY, but disregard that label for this), you've enabled

TWO arpeggiator transpositions -- the one described in the last

paragraph, plus this one, and the two intervals combine.  So now if

all your arpeggiator transpositions are still set to octaves as they

are at factory settings, your arpeggios are spreading out over two

octaves beyond what was actually played.  If they're something user-

defined though, then it's your first user-defined interval plus this

second one.

Hold down MODE and press the FOURTH button (labeled, irrelevantly,

"AMOUNT"), and you enable the first THREE transpositions for the

arpeggiator -- the two described above, plus this one.  At factory

settings, that makes your arpeggios span three octaves further than

actually played.  You get the idea.  It goes this way for all six

buttons below the pitch bend and modulation levers, without regard

to their labeled functions, so long as you hold down MODE while

you're selecting them, and so long as you remember the leftmost

one is the one that turns these transpositions off (its programmed

interval is NO INTERVAL), which means pressing that combination

makes the arpeggiator play your arpeggio unmodified.

Almost makes you wonder why they bothered labeling the control panel.

Half the functions on the OB-8 have nothing to do with anything it

says, and are accessed by holding down this and this and pressing

that while you play something else.

*  Programming arpeggiation ranges:

As mentioned above, although OB-8's came pre-programmed with all five

of the user-definable arpeggiator transposition intervals set to one

octave, giving up to a five octave spread to a played arpeggio pattern

(remember the first of the six buttons turns the feature off, and is

not an interval), those intervals can be changed individually to

something other than octaves.  To do that, hold down the MODE and

ARPEGGIATOR buttons at the same time, and then play five notes on the

keyboard, one at a time, not as a chord, while holding down the two

buttons.  Those five notes will become the new transposition intervals

for the arpeggiator, and will become non-volatile, and independent of

any currently selected patch (i.e., a global setting of the keyboard).

*  Pedal sustain time:

It is labeled on the control panel (at least on OB-8's that have their

PAGE2 funtions labeled), but it's still worth noting that unlike every

other keyboard's sustain pedal policy, on the OB-8, the time that a

note will sustain while the pedal is down is a programmable parameter

of a patch.  On any other keyboard, holding the pedal down is simply

equivalent to holding the note down indefinitely.

*  Arpeggiator sync:

The arpeggiator sync jack on the back of the keyboard is completely

undocumented in the manual.  I have heard from one source that it

expects a +5V trigger, and I have also heard that sending a loud

note from a drum machine into that port will suffice.  Although the

OB-8 can support MIDI, there is no function to sync the arpeggiator

to MIDI clock, and that continues to be true even if the factory

MIDI is replaced with the Encore retrofit.  (The Encore retrofit for

the Roland Jupiter-8 _does_ sync the arpeggiator to MIDI clock though.)

*  Waveform selection:

When setting the waveforms for oscillators 1 and 2, selecting sawtooth

or pulse is done as expected with the buttons so labeled.  However, it

should be noted that turning both SAW and PULSE off selects a triangle

wave, and turning them both on creates a composite wave from both the

sawtooth and the pulse wave.

*  Square waves:

Square waves can be obtained by selecting pulse wave, and turning the

PULSE WIDTH knob fully to the left.

*  Individual oscillator control of pulse width:

More unlabeled joy!  You can change the pulse width of OSC1 or OSC2

individually by holding down the PULSE button of the oscillator you'd

like to change, and turning the PULSE WIDTH knob while it's held down.

Normally, the knob changes pulse width for both oscillators at once

(provided they both are set to a pulse wave as far as waveform

selection for each oscillator goes).  Once the two oscillators have

been given differing settings this way, they can be made them same

again by turning the PULSE WIDTH knob all the way to the right or

left while _not_ holding down anything.

*  Switching oscillators on and off:

The buttons to turn oscillators 1 and 2 on and off (allowing you to

create simpler one-oscillator-per-voice patches if you'd like) is in

a rather confounding place -- in the FILTER section.  Don't ask me

why they're not in the OSCILLATOR section.  Also, OSC2 has the option

of being on at either HALF or FULL volume.  (Don't worry, there are

labeled buttons for that.  You don't have to hold anything down.)

Also, there's a white noise generator, and it's turned on in the

FILTER section too, right where you'd expect a noise generator to be.

*  ATTACK settings:

This isn't related to the control panel, but is an important fact about

the OB-8 nonetheless: The factory documentation distributed by Oberheim

for calibrating the envelope generators is _WRONG_, and if followed,

will result in the OB-8 being incapable of generating fast punchy

attacks.  The OB-8 has actually gotten a reputation for having slower

attack than its predecessor keyboard, the OB-Xa, because of the number

of OB-8's that have been incorrectly calibrated _following Oberheim's

own instructions_.  In the service manual, where it says to calibrate

the EG's to as close to 0.000 volts as possible, that should actually

read "as close to -0.256 volts as possible."  That is because there is

a 0.256 volt difference between the reading your voltmeter will get

while the system is running in calibration mode and the setting it

will have after the calibration is finished and it is returned to

normal operating mode.  That will manifest itself as sluggish attacks.

This only affects the calibration of the EG's (i.e., the ADSR).  All

the other calibration instructions should be followed literally.

*  Envelope reset:

In a bizarre but nifty feature, the ADSR cycle of any sound can be cut

short as it fades out (the RELEASE stage) by hitting the WRITE button,

the button normally used for writing patches.  If pressed quickly and

not held down, that button serves as a cutoff for still-dying sounds.

*  Sampled Vibrato LFO

The LFO's aren't too badly labeled, but on thing that's not too obvious

is that all three of the waveform selectors can be on at once, and in

that case, it will do sample & hold from the separate LFO generated by

the modulation lever (rather than from noise as normal S&H would do).

For regular noise-derived S&H, turn the bottom one on as labeled on

the panel.

*  Portamento bend:

In PAGE2 mode (which is itself an alternate control panel mapping,

before we even start talking about hidden stuff), there is a button

labeled PORTAMENTO BEND, which makes all notes gliss from a

programmable interval above or below the actual note played to the

actual note.  Naturally, programming that interval requires holding

the PORTAMENTO BEND button down (while in PAGE2 mode) and playing

the desired interval on the keyboard, using C2 (the third C on the

keyboard) as a reference point.

*  Reset all PAGE2 parameters:

If your PAGE2 settings have become simply mad, you can reset all PAGE2

parameters to defaults and retreat to the relatively simple world of

the OB-Xa (which had no PAGE2 mode) by pressing the PAGE2 button twice,

holding it down the second time, and while it is held, pressing the

button (for this purpose, irrelevantly) labeled F-ENV, which in this

context functions as a PAGE2 parameters reset.

*  Loading cassette dumps made on other OB-8's

Some of the calibration settings on an OB-8 are handled by the CPU, and

are thus actually non-volatile software settings rather than physical

positions on trimmers on the circuit board.  Because of this, those

parts of the calibration that are done that way are lost when the

battery gets replaced.

Oberheim wanted to save keyboard techs from having to go through a

calibration just to replace the battery, so they made it so that every

cassette dump also includes the non-volatile calibration parameters in

the dump.  Since every tech already knows to save the customer's

patches before replacing the battery (if it's not gone already), it

makes getting your CPU-managed calibration settings back afterward

happen more or less automagically.

The bad part is that anyone wanting to use the WAV files on the

Internet which contain recordings of the binary tape leader tone and

data from an actual cassette dump of the OB-8 factory patches to get

the original factory patches on their OB-8 will also be loading the

calibration settings of whoever that OB-8 belonged to.  That probably

will not sound very good.  It also strongly discourages trading of

patch banks between users using cassette dump as the medium.

If you ever do load someone else's cassette dump, you will want to

review all parts of the calibration procedure that are not made with

a physical trimmer adjustment (i.e., all software-driven settings)

after you get the sounds loaded.

It is probably a much better idea to trade sounds with sysex dumps,


a)  Not all OB-8's had MIDI at all, from any source.

b)  The original Oberheim MIDI could only sysex dump single patches,

and had no function for sending a whole bank in one dump.

c)  The Encore retrofit can send whole banks, but it was undocumented

in earlier versions the printed manual, and you might only know

you can even do it if you've looked at the PDF file manual now

available on Encore's web site.

So cassette dump is the only way of saving patch banks that all OB-8

owners can do, but you still probably don't want to do it unless you

don't mind checking your calibration afterward.

Article 2:



Status Data Bytes  Description

1000  xxxx 0kkk kkkk  Note off. (See notes no. 1-2.)

0vvv vvvv   0vvv vvvv = note off velocity: always 40H.

1001  xxxx 0kkk kkkk  Note on. (see notes no. 1-2.)

0vvv vvvv   0vvv vvvv = 40H

1011 xxxx 0ccc cccc   Control Change. (if enabled).

0vvv v000   0ccc cccc = Control number (01=mod lever).

0vvv v000 = control value.(range 0-78H. Lowest

3 bits are ignored).

1011 xxxx 0ccc cccc   Control Change. (if enabled).

0vvv vvvv   0ccc cccc = Control number (40H = Sustain


0vvv vvvv = control value.(0 = off. 7FH = on.)

1100 xxxx  0nnn nnnn  Program select. (if enabled)..

0nnn nnnn = 0 through 77H.

1110 xxxx 0vvv vvvv  Pitch 0end change LSB (see note 3).

0vvv vvvv  Pitch Bend change MST TRANSMITTED DATA -- SYSTEM MESSAGES

1111 0000 0H     System  Exclusive Oberheim I.D. no.

0ddd dddd   Device number. OB-8 = 01H

01H   Command Byte 1 : Program data dump follows.

0ccc cccc  Command Byte 2  : Program number.

data    Program data.     (see note 4)

F7H    End of System    Exclusive Status Byte. OB-8 MIDI Implementation Version B-5 August 15,1984   Page 2


Status Data Bytes  Description

1000 xxxx  0kkk(kkkk  Note off. (See notes no. 1-2.)

0vvv vvvv  0vvv vvvv = note off velocity: ignored

1001 xxxx  0kkk kkkk  Note on. (see notes no. 1-2.)

0vvv vvvv  0vvv vvvv = 0: Note Off.

0vvv vvvv not = 0, velocity ignored.

1011 xxxx  0ccc cccc  Control Change. (if enabled).

0vvv v000  0ccc cccc = Control number (01=mod lever).

0vvv v000 = control value.(0-78H. Lower 3 bits

are ignored.)

1100 xxxx  0nnn nnnn  Program select. (if enabled).

0nnn nnnn = 0 through 77H

1110 xxxx  0vvv vvvv   Pitch Bend change LSB (see note 3).

0vvv vvvv  Pitch Bend change MSB


1011 xxxx  0111 1011  cccc cccc = 123 (7BH) : All notes off.

0000 0000  vvvv vvvv = 0. The OB-8 turns off all notes that

were turned on by MID!.

1011 xxxx  0111 1100  cccc cccc = 124 (7CH) : OMNI mode off.

0000 0000  vvvv vvvv = 0. The OB-8 turns OMNI mode off and

turns off all notes that were turned on by MIDI.

1011 xxxx  0111 1101  cccc cccc = 125 (7DH) : OMNI mode on.

0000 0000  vvvv vvvv = 0. The OB-8 turns 0MNI mode on and

turns off all notes that were turned on by MIDI.

1011 xxxx  0111 1110  cccc cccc = 126 (7EH)   MONO mode on.

0000 0000  vvvv vvvv = 0. The OB-8 has no M0NO mode. When

this command is received the OB-8 switches to

OMNI on  / POLY mode and turns off all notes that

were turned on by MIDI.

1011 xxxx 0111 1111  cccc cccc = 127 (7FH)   P0LY mode on.

0000 0000  vvvv vvvv = 0. The OB-8 is always in P0LY so no

mode change occurs. All notes are turned off that

were turned on by MIDI.

OB-8 MIDI Implementation Version B-5  August 15,1984  Page 3


1111 0000  10H    System Exclusive   Oberheim I.D. no.

0ddd dddd   Device number   OB-8 = 01H

01H    Command Byte 1 : Program data dump follows.

0ccc cccc   Command Byte 2 Program Number

data    data (see note 4 for data format)

F7H    End of System Exclusive Status Byte.

1111 0000  10H    System Exclusive   Oberheim I.D. no.

0ddd dddd   Device number   OB-8 = 01H

00H    Command Byte 1 Program data dump Request.

0ccc cccc   Command Byte 2 Program Number

F7H    End of System Exclusive Status Byte.

1111 0110 -   System Common Message : Tune Request

OB-8 MIDI Implementation Version 8-5  August 15,1984   Page 4


I. xxxx : Basic Channel number minus I. i.e. 0000 is CH.1. and 0001 is CH.2.

range : CH.1-8.

2. kkk kkkk = note number. Range 24H-60H

3. Sensitivity of the pitch bender is selected in the receiver. Center

position (no pitch change) is 2000H, which is transmitted ExH-00H-40H.

Maximum transmitted value is 7F40H. (The 6 lsb's are not looked at by the



Sent as 4 bit nibbles, right justified, LS nibble sent first.

:  BIT 7  :  BIT 6  :  BIT 5  :  BIT 4  :  BIT 3  :  BIT 2  :  BIT 1  :  BIT 0  :


BYTE 0 :  VCF REL (6 BITS)            :    LFO WAVE   :

:                :   2               1  :

-------------------------------------------------------------------------            :----------

BYTE 1 :  VCA REL (6 BITS)      :            :UNISON:

:        :   0        :          :

---------------------------------------------------------------------------------------------                                                           BYTE 2 :  VCF DCY (6BITS)    :  FILTER: OSC 2:

:        :     FM   :   FM   :



:        :    I        :      0   :



:        :    I        :       0   :


BYTE 5 :   VCA ATK (6 BITS)   :  OSC 2 : OSC  1 :

:        :   PWM  :  PWM   :


BYTE 6 :  VCF SUS (6 BITS)    :  NOISE : 4 POLE :

:        :       :  :


BYTE 7 :  VCA SUS (6 BITS)   :  OSC 2  :  OSC 2 :

:        :    ON     :  HALF  :


BYTE 8  :  VCF M0D (6  BITS)   :  OSC 1  :  TRACK :

:        :    ON     :             :


BYTE 9 :  VCF RES (6 BITS)    :    PW1   :   VC01   :

:        :   180 '    :   180 '   :


BYTE 10 :  VCO 1 PW  (6 BITS)   :   VCA    :  F-ENV  :

:        :   MOD   :              :


BYTE 11 :  LFO FREQ (6 BITS)   :  SYNC   :   OSC 1  :

:        :              :   FM      :


OB-8 MIDI Implementation Version B-5 August 15,1984  Page 5


BYTE l2 :  FM AMNT   (6 BITS)  :              :              :

:        :     5        :       4     :

-------------------------------------------------------------------------        VOLUME      --

BYTE 13 :  PWM AMNT (6 BITS)  :              :              :

:        :     3       :        2     :

-------------------------------------------------------------------------                            --

BYTE 14 :  PORT AMT  (6 BITS)  :              :              :

:        :      1       :       0     :


BYTE 15  :  VCO2 DETUNE  (6 BITS)  :      VCO 2 PW        :

:        :      5       :        4     :

--------------------------------------------------------------------------             :              --

BYTE l6  :  VCF FREQ  (6  BITS)  :               :              :

:        :      3       :        2     :

--------------------------------------------------------------------------             :              --

BYTE l7 :  VCO2 FREQ  6 BITS)  :               :              :

:        :       1       :       0     :


BYTE 18 :  VC01 FREQ (6 BITS)  :  SPARE   : LEGATO :

:        :               :     PORT. :



:        :       2       :      1       :

--------------------------------------------------------------------------              :--------------

BYTE 20 :  PEDAL SUSTAIN (6 BITS)  :                :    PORT   :

:        :       0       :    BEND   :


BYTE 21 :  FM VIB RAISE (6 BITS)  :     LFO    :  FM DLY  :

:        :  TRACK   :   INVERT  :


BYTE 22 :  PWM VIB RAISE (6 BITS)  :    PORT   :  PORT      :

:        :   QUANT  : MATCH    :


BYTE 23 :  FM  VIB DELAY (6 BITS)  :    180 `     :      90 '    :

:        :                :                :


BYTE 24 :  PWM VIB DELAY (6 BITS)  : PWM DLY :    PWM    :

:        :   INVERT   :  QUANT  :


BYTE 25 :  VOICE DETUNE (6 BITS)  :    EXPO    :   CONST  :

:        :    PORT     :   PORT    :


BYTE 26 :  BEND AMOUNT (6 BITS)  : LFO RATE :     FM     :

:        :    DELAY   :  QUANT  :

---------------------------------------------------------------------------------------------------- OB-8 MIDI Implementation Version B-5  August 15,1984  Page 6 MODES

The OB-8 defaults to 0MNI ON upon power up. If the OB-8 is a receiver, it will receive on all channels. If the OB-8 is the transmitter, it will transmit on one channel. (selectable)

The OB-8 may also be operated in OMNI OFF mode. If the OB-8 is a receiver, it will now receive ONLY on the selected Basic Channel. If the OB-8 is used as transmitter, it will now transmit the upper half of the keyboard on the Basic Channel, and the lower half will be transmitted on the Basic Channel + 1. Pitch bend, progam select, etc. will be transmitted on both channels. The Channel Split Point is the same as the regular Split Point. (default is middle C.) THIS MODE IS INDEPENDENT OF SPLIT MODE.


NOTE: Functions must be enabled on source AND destination machines to work.

Switch Function

A  Enable/Disable program change and program dump.

Power-On default: disabled.

B  Enable/Disable Pitch bend and modulation controls.

Default: disabled.

C  OMNI ON/OFF. Toggle OMNI status. Power-On default is OMNI ON

(led is lit.) (see MODES)

D  Channel display/select. Press and hold down D button to display

or select the Basic Channel.

WRITE Dump current STORED program to MIDI. NOTE: SWITCH "A", "PROGRAM


TRACK Sequencer Re-Enable / Turn off MIDI Notes.

IMPORTANT: The OB-8 cannot RECEIVE MIDI info and be run by the

DSX sequencer simultaneously (due to hardware design.) So, to

prevent MIDI data errors, the sequencer is DISABLED upon

receiving any data from MIDI IN. This condition is displayed by

the TRACK led on page 2. When you no longer wish to use the OB-8

as a receiver, and you want to use the DSX, disconnect MIDI IN

and press the TRACK button. The led will go out, the sequencer

will work normally, and any notes turned on by MIDI will be turned off.

Power-On default: TRACK light off, Sequencer Enabled.

Article 3:

OB-8 Revision B5 Software

Operation Guide

8/15/84 This revision of OB-8 software (version B5) has been made to improve the operation of the OB-8 with the DSX, as well as to add some new MIDI features.  For a complete explanation of the OB-8's existing features, please refer to the OB-8 Owner's Manual and the OB-8 Revision B3 Software Operation Guide.  To verify the software version number of an OB-8, press the PAGE 2 button twice and hold it down the second time it is pressed (the PAGE 2 led should now be lit).  Wile holding down the PAGE 2 button, press and hold the SYNC button.  While holding both switches down, the PROGRAMMER leds will display the OB-8's software version number.  If the B led in the GROUP section and the 5 led in the PROGRAM section are now lit, the software version is B5. This version has all of the features of version B3, as well as the following NEW FEATURES: 1. The Sustain Footswitch on the OB-8 has now been added to the MIDI interface.  This means that when two OB-8s are connected together through MIDI, the Sustain Footswitch on th MASTER OB-8 (the one using MlDI OUT) will also control the Sustain on the SLAVE OB-8 (the one using MIDI IN).  This feature can also be used by other synthesizers that transmit and recieve Sustain Footswitch information on MIDI.

2. When an OB-8 with MIDI is connected to a DSX Digital Sequencer, the MIDI output of the OB-8 will send out the notes played by the DSX through the MIDI output.  This allows additional synthesizers with MIDI to play the same thing that the OB-8 is playing from the DSX.  Now with version B5 (and a DSX with version 3.00 or above), it is possible to have the DSX control the OB-8 AND a synthesizer connected to the OB-8's MIDI OUT INDEPENDENTLY. By assigning the DSX's CV outputs to the OB-8 MIDI (see "DSX REVISION 3.00 INSTRUCTIONS"), the DSX will send whatever the DSX's CVs are playing to the OB-8's MIDI OUT without the OB-8 playing it.  The notes that the OB-8 are playing will NOT be sent to the MIDI OUT, so that the two synthesizers can play completely independently.  This also means that when this feature is used, and the DSX is not playing anything on the CVs, no notes will be sent to the OB-8's MIDI OUT regardless of what is being played on the OB-8.

3. The overall speed of operation on the OB-8 has been increased to allow the DSX to communicate with the OB-8 faster, resulting in increased accuracy of the timing of notes played by the DSX.

4. The range of vibrato lever has been changed for better compatability.

Article 4:

Oberheim OB 8 Improved sound mod!!!!

This is from the archives of Haible Juergen

As I posted before, the main difference in sound between the

OB-8 and the OB-Xa is a bass loss of the OB-8 due to a 22Hz

highpass in the signal path.

While 22Hz don't look that dramatic, remind You that the phase

is fucked up at much higher frequencies. You can see the saw

wave distorted to an exponential slope instead of a linear one

for the entire low octave. The pulse has extreme overshots

(differentiated), and the triangle is hard to describe, but not a

triangle at all anymore.

In my earlier posting I suggested increasing coupling capacitors,

but this affects the autotune routine which is optimized for speed

and will produce errors with the enlarged time constants.

Last night I tried an alternative method: Compensating the phase

shift of the voices in the output stage (after the autotune loop).

The "bad guy" actually is the coupling between the VCF and VCA:

it's a 75kOhm resistor in series with a 100nF capacitor.

To compensate for this, we have to put a similar network into the

feedback loop of an opamp.

To avoid additional stages, I decided to change the frequency

response of the opamp that converts the final VCA's output current

to a voltage.

Now here's the way to go:

(1) Replace the resistors R505 and R506 (100k) with 1M resistors

on both voice boards (4 resistors in total). If You stand in front

of Your open OB-8, these resistors are located near the right

edge of the board, next to a TL082 opamp. Cutting them out

is easier than desoldering them on the 2-sided pcb.

(2) Now build a small network of a 120k resistor and a 56nF cap

*in series* and connect the whole network *in parallel* to the

new 1M resistors (solder it right on top).

That's all. If You want to do an A-B crosscheck, modify one voice

board first and cycle thru the voices. You will hear the difference,

and You will also see it at a scope.

If You find the original OB-8's sound more pleasant, just short the

four 56n capacitors. Or solder 4p1t switch across the caps to have

immediate access to both sounds.

For my part, I have the change hardwired, without a switch. If I want

less bass, I can do this with the mixing console ...


Though this works fine on my OB-8 (and I don't know any reason why

it shouldn't ob Yours as well), I guarantee for nothing.

Article 5:



::: IF :::

::: Finally, a book all about Drum Machines! :::

In keeping with the Movement CMS Drum Computer post i did, here's a full blown book about all the special drum machines that have been! Beat Box

Also posted here

food for addiction!

::: IF :::

::: Movement MCS Drum Computer :::

I swiped this info from an Ebay sale a few years ago.  I hate that Ebay isn't searchable beyond a few months.  It's just as useful as Vintagesynth or any other web site that talks about vintage instruments. Movement MCS Drum Computer


"This is a very rare British made drum machine from the early 1980's. According to wikipedia approximately 30 were made! Famous users included The Eurythmics (can be heard on Sweet Dreams, 1984), Phil Collins and The Thompson Twins.

It is capable of both analogue synthesised sounds (Simmons style) and sampled drum sounds and has 14 voices in total (2 per card). Each voice can be switched individually between analogue and 8bit sample playback - quite an unusual feature. Patterns are created using the internal computer and chained together into a sequence much like any other drum machine.

Overall physical condition is excellent, the casework is free from cracks and there are no missing knobs / keys. Of course, there are some signs of use which is to be expected. Most obvious are some scratches on the orange casework (can be seen in the photos) and a few missing knob caps.

Functionally, there are a few issues which I shall list below. None I believe are serious. Voice 1 - doesn't switch between analogue/digital playback - digital always seems to be selected.

Voice 5 - very very quiet.

The CRT display is faulty and does not display an image. However, as you can see from the photos the internal computer is alive and well. I've demonstrated this by connecting an external monitor to the video out on the back panel.

There are some crackly pots.

I was unable to get the midi input to respond.

Pattern 0 seems to crash the computer when fully erased - the others are fine. Other than the above it seems to be in good working order."

The pictures were so beautiful i had to keep them.  I'm glad i kept the text as well!



It's so Beautiful!  I'd love to make some beats on this!

::: IF :::

::: The Oberheim that never was :::

I swear i saw one in a dream.  The Elusive OB-Xpander.  Not the Oberheim Xpander.  A ghost of the Polyphonic Past... :::

OB Xpander Vintage Synth book


I got this pic from the Vintage Synthesizer book.

::: IF :::

::: Sequential Circuits Pro One MIDI Upgrade :::

So here we go, it's part 2 of the Sequential Circuits Pro One CPU/MIDI upgrade!  I had ordered some parts and was waiting for them to arrive when I left off in the previous tech post here.  I got the cables I ordered and they worked out just as I had hoped.  I broke this tech project into 2 parts because I had some intense and somewhat stress inducing decisions to make about the second half of this affair.  The CPU upgrade was pop and drop and non destructive but the MIDI part of the project required drilling! I believe the saying goes something like "measure twice, cut once".  I tend to  measure by eye 3 times and then cut later after measuring one last time.  It seems to work for me this way.  Here's a breakdown of what it entailed:

1, Holes were drilled in the bottom plate of the synth to add the MIDI daughter board.

2, Holes were drilled in the back ( this is the scary part! ) of the synth to add MIDI in out and through jacks.

3, I made a wiring harness out of RC servo cables to attach the MIDI board to the new CPU and to the MIDI in out and through jacks.

4, The wiring harness was done so that there's connectors between all the boards which makes future work easier to accomplish.


I was most concerned with making the MIDI jacks I was adding look good on the back of the synth.  I also wanted to mount them on the 45 degree angle so that they would underline the Pro One logo.  I used a small drill bit to make pilot holes for the 3 jack locations and a larger stepped drill bit to enlarge them to the final size.  The MIDI board was placed on the opposite side of the instrument to keep it away from the audio path and closer to the power section.  I flipped the orientation of the RC cable harness wires so they wont get confused if the synth is disassembled in the future.

PS: In case you are wondering why I only attached 3 of the 5 connections for the MIDI jacks, The only connections actually used on a MIDI cable are 2/4/5. 1/3 don't do anything!  I believe that under the original MIDI specifications there was 5 cables allotted for future implementations.  There was probably supposed to be a MIDI II convention that never got written or adopted.  Those last 2 wires could have allowed faster parallel data transfer or maybe a separate sysex / note transmission bus implementation.  Many manufacturers like Motu made MIDI interfaces for computers that would stream MIDI much faster between the software and the interface than the interface would stream to the MIDI devices.  This allowed individual MIDI in/outs to actually be more in time with each other as the MIDI interface multiplexed the data stream from the computer to the individual hardware outs.







::: IF :::

::: Sequential Circuits Pro One CPU Upgrade :::

This is part one of a two part series on the Sequential Circuits Pro One.  The Pro One is a classy sounding and nicely laid out mono synth from the same lineage as the SCI Prophet 5.  It was released during the era of Big Expensive poly synths and their more compact Mono brethren of the late 70's and early 80's.  Its sound is versatile and can cover any ground between Lead and Bass and all the way to great drums and percussion. :::

There's a limited CPU on board that can record basic, i mean BASIC, sequences and a decent Up/Up Down arpeggiator that is fun to use.  I happened upon a site for MTG, Music Technologies Group, that offered a replacement CPU that adds MIDI note in out plus other controller input to this classic synth.  This is the first part of the process.  I plan to complete the MIDI in outs and other CPU routings later but i found that i wanted to get some extra parts first.  The MTG kit includes the CPU, MIDI in out jacks, a MIDI daughter board, and some connecting wires.  I felt that i wanted to have connectors to and from the jacks and MIDI board so that the Pro One would remain easy to service and modular in layout.  I decided to get some RC servo extensions as they are 12" long and have 3 conductors in a nice small package with male and female receptacles.  I will complete this project as soon as I have all the parts from Amazon!


While i had the machine apart i found that there was a broken pot that had been carefully repaired.  I re-repaired the shaft of this pot and cleaned the others to help improve the function and performance of the face plate controls.  It looked like someone had used superglue the first time.  I used epoxy this time so it will last longer!  I also noted that this unit has the power transformer mounted to the chassis and not the board.  In early versions the board mounted transformer tended to snap off and destroy the inside of the synth!


Many people lament the construction of the Pro One and accuse it of being flimsy and shoddily made.  I think it could be better as well, but i have seen worse too!  The pots are all plastic shaft and are prone to breaking off if you abuse the instrument.  Especially because they are not secured to the faceplate.  I think that the faceplate is a cool design, but it is plastic and tends to flex more than it should considering it is the support for the board that is underneath it.  I have considered attaching a bar of some kind under the front lip of the top plate to strengthen it for the future but i don;t have a clear idea as of yet on how to make that happen.  Maybe for a future post!



The CPU just pops in and it's ready to go.   Here's a list of features from the site: Features:

  • Plug-n-Play! No soldering required!
  • Supports the original functionality including the sequencer and arpeggiator.
  • Each sequence can be up to 256 steps (compared to total of 40 steps on the original).
  • Sequences are retained in memory even after power-down. The CPU module does not use a battery.
  • Power-on settable parameters for clock start mode, arpeggio up/down end notes, arpeggio down mode and arpeggio gate time.
  • Sequence Tie mode allows for a variety of note lengths.
  • Sequence End mode lets you chain the two sequences.
  • Optional: If you want to make your own MIDI interface, instructions are in the manual. If you want MIDI hardware included, see CPU+MIDI.

::: IF :::

::: Oberheim DMX auxiliary parts :::

Many days I find that it's the little things.  I've been meaning to do this for a while and as i was at the hardware store today i happened to remember this plan so i finally got it enacted. The Oberheim DMX and DX have hinged lids for adjusting certain sound parameters,  In the case of the DX you can lift the lid to change eproms to achieve different drum kits and percussion.  In the case of the DMX you can change the pitch of each voice under the hood and also interchange voice cards as seen here to achieve different sounds with different sized eprom samples.  Every time i am under the hood on either of these i find myself wishing that the thumb screws used to secure the lids had gaskets to keep them from damaging the paint on the lid.

So i got some rubber washers at the hardware store to put in between the screws and lids to protect them better.

It's the simple fixes!


::: IF :::