ADA STD-1 cloning help

[dj_death]

I decided cloning the monster, ADA STD-1, but i will need help by everyone who can. I have redrawn the schematic in Eagle, because i want to make the pcb by the nets created by the schematic. This is only the signal circuit without LFO and clock circuits. I think i will build them as a separate pcb. I tried to remove some sections to make it smaller but it will be huge for sure. It's the instrument version with FET input. I removed the remote switching circuit that's used for effect bypass. Also one spdt switch can be used for Regenaration ON/OFF instead of the transistor-4007 circuit. In this schematic switches and pots are not shown, but are spotted as pins. I will try using an lm317-lm337 instead of the mc1468 regulator which seems to be a very rare part. Also opamp supply is not shown yet.I used the factory schematic but i can't identify some values. Please check the schematic to see if i made any mistake in resistor and capacitor values. I'm not sure if some resistors are 5.1K or 51K etc. Please suggest other sections that can be removed whithout losing the character of the effect. Maybe some filters could be removed but it will lose some of it's quality. I think it should remain complicated. Any suggestion and schematic correction is well appreciated.
Here's my job so far. http://i12.photobucket.com/albums/a248/dj_death/std1_schem_Signal.jpg

DJ_Death

[dj_death]

Now i have completed the signal schematic adding lm317-lm337 to provide power supply instead of mc1468. http://i12.photobucket.com/albums/a248/dj_death/std1_schem_signal.jpg Also 80% of CV-LFO-Clock schematic completed. The values that i'm not sure have question marks.http://i12.photobucket.com/albums/a248/dj_death/ada_std1_cv.jpg

[Mark Hammer]

Boy you ARE a sucker for punishment, eh?

Seriously, though, that's quite an ambitious undertaking, and one that will demand VERY meticulous bookkeeping and lots of time to spend.

A buddy and I were pondering producing a "reduced" version of the STD-1 since he was, at the time, able to score enough MN3011's to make a limited run of a commercial product.  It's a year and a half later now, and whether the chips are still available is unknown to me.

We were trying to do something that would fit in a 1590BB and accomplish many of the more desirable functions using 4 or 5 pots and a handful of toggles.  The original allowed for a ridiculous amount of flexibility, and I suspect that many of the things it was able to do in analog at the time, have since been surpassed in the digital domain.  So, perhaps the thing to do is to focus on what it can still do better in analog than can be done in digital.

"Better" is a word with many connotations.  One is certainly a nicer tone, but it can also mean easy to change on the fly.  I think one of the things that the STD-1 did "better" was in being able to produce a truly stereo chorus and flanger sound.  Did you need to be able to assign EVERY tap to ANY output?  I think not.  Did you need to be able to take a regen feed from any tap?  I think not.  So, conceivably, a reduced version simply consists of being able to fade "groups" of taps in or out of the output, and master control of the delay range via the clock.

If you have plans to build it into a rackmount chassis then use what you want.  If you have plans to build it into anything smaller, like a 1590BB or DD, then I recommend SIP-style op-amps and 1/8w resistors as a way of conserving space and making layout more efficient.

[dj_death]

Mark thanks for all your advice. I have completed the CV-LFO-CLOCK schematic. Although i would like some verification for the part values that have question marks because i'm not sure if the values are correct. Also does anybody know why cd4041's outputs are connected together. That looks very strange because every input has two complementary outputs one inverting and one non inverting. These outputs are connected . What is the reason of doing this? Someone would say that it could cause problem to the chip.

[Mark Hammer]

I'm not sure about the 4041 in particular, but it gets used as a clock buffer to achieve faster clocking rates.  Paralleling multiple invertor sections provides a clock pulse with more current, enabling the clock pulse to overcome the substantial input capacitance on the clock input pins of the BBD. 

Why does that matter sonically?  Because production of a "seamless" continuous sampled signal relies on instantaneous handoff of sample N to the next stage.  When there is too much input capacitance, relative to the drive current capabilities of the clock generator then the clock pulse gets altered from crisp square wave to rounded peaks.  The practical impact of that is that the clock generator changes from going "Now SWITCH...SWITCH...SWITCH!" to "aaaaaaaaaaannnnnnnd......SWITCH.......aaaaaaaaaaaaannnnnd......SWITCH....aaaaaaaaaaaaannnnd.....SWITCH!", leaving gaps where the instant handoff should be.

The MN30xx series of BBDs had roughly 700pf of input capacitance on the clock pins for every 1024 stages.  Datasheets for the MN3011 show input capacitance on the clock pins of approximately 2300pf.  Consider that the MN3005 was 2800pf but hardly ever clocked fast enough to produce chorus or double tracking, let along flanging.  The STD-1 assumed that the MN3011, in all its 3328-stage splendour, would be clocked fast enough to generate chorus and flanging (though obviously not of the A/DA Flanger variety).  Consequently, to bump the MN3011 up into that range, buffering of the clock lines, via a 4041, is needed.

[puretube]

 

can`t see your clock-schemo: got all pins correctly?
the + and - supply, too?

[Mark Hammer]

 

[dj_death]

Thanks Mark for the detailed explanation for buffering. Actually i knew that it was used for buffering the clock outputs, i don't understand why  the buffer outputs are connected with the inverter outputs, because when for example the output of a buffer is low, the output of the inverter will be high. That means that a high output is connected with a low output. Why they didn't use an inverter or buffer chip such as 4049 or 4050 like Ultra Flanger. Is this configuration used because cd4041 and the other "cd" chips are powered with VDD at gnd and VSS at -15V. Here's the link with the final cv schematic http://i12.photobucket.com/albums/a248/dj_death/ada_std1_cv.jpg.

[Mark Hammer]

See page 6 here and tell me if it doesn't look familiar: http://hammer.ampage.org/files/Anderflange1.PDF

[StephenGiles]

Dave Tarnowski was reason for that 4041 configuration.

[Mark Hammer]

Well that's a lucky guess.  Dave's the reason for about 80% of everything that can't be attributed to Anderton, Penfold, Beigel, Bode, or Moog fer cryin' out loud!

[puretube]

again: check out the 4041 pin configuration.

read the datasheet for that chip carefully...

watch the tiny dots on each "inverter"/"buffer" in the datasheet and the schemo.

don`t take a "41" for a "49" or "69"...

[dj_death]

You are right Mark in this flanger 4041 is used very cleverly. But in ada all the inverter outpouts are connected toggether and all the buffer outputs too. In Anderton's flanger 4041's pins 1,4,8,11 are connected as buffer output and 2,5,9,12 are connected as inverter output.
In std-1 1,4,9,12 as CP1 and 2,5,8,11 as CP2. Am i missing something?

[Mark Hammer]

You are right Mark in this flanger 4041 is used very cleverly. But in ada all the inverter outpouts are connected toggether and all the buffer outputs too. In Anderton's flanger 4041's pins 1,4,8,11 are connected as buffer output and 2,5,9,12 are connected as inverter output.
In std-1 1,4,9,12 as CP1 and 2,5,8,11 as CP2. Am i missing something?

Yeah.  Probably an accurate original schematic from what I can tell.

As Ton keeps reminding us: don't trust schematics blindly.  Especially if they are "posted around" and especially if they are circuits of any complexity.  A number of original Maestro service manuals that Rick Lawrence was kind enough to send me some years ago had errors in the schematic, with things omitted and other things connected where they didn't have a hope in hell of working.

I doubt that manufacturer-issued schematics are deliberately mis-drawn to mislead the public, since they are produced for authorized repair centres to transform a nonfunctionng product into something the owner can love again and recommend to their friends for purchase.  Rather, there just isn't the quality control in their production.  I suspect this is because service manuals are an afterthought (you don't design stuff to break) and less corporate resources and time are allocated to producing them and checking them over than are allocated to checking over stuff coming off the production line.  If it was cars, where someone might die or be injured and launch a company-destroying lawsuit, then of course there will be substantial quality-control of the product and of repair/technical advice to those in charge of repair and servicing.  If it's a $100 pedal that no one's life depends on, then considerably less effort is put into checking over the technical info sent out.

[dj_death]

Ok Mark i think i'm gonna change the schematic, there is no chance for the circuit to work like this ,in my opinion.

[puretube]

OK: never completely trust a schemo on the web...

the pin-numbers of U24 on the ADA schemo are correct!

BUT: the inverting/non-inverting depictions are "not so fine"

see the 4041 datasheet...

AND: big thanx for pointing me the way to a more powerful drivechip than the 4049/-69... 

[dj_death]

Moreover even if the 4041 outputs connected correctly the inverters separately from the buffers, their outputs will not be complementary they we'll always have the same phase and so the MN3011 will not work. The 4047 osc out pin 13 should connect to all input pins 3,6,10,13 of 4041. The same as Anderton's flanger clock.

[puretube]

sorry - I don`t care what you wanna do,
but I know I`m right about those pin-numbers in the ADA to be correct...

the C.A. schemo is correct, too...

BUT: the "C.A." has to create 2 opposite phases from a single clock (4046, pin4).
        the "ADA" has to maintain the 2 opposite phases it receives (4047, pin10&11).

the pin 13 of a 4047 will put out a not-neccessarily symmetrical (concerning duty-cycle)
clock of twice the frequency of pins10&11, the latter being the wanted frequency in the "ADA".

and: yes, I`ve watched the signals on various pins of the 4046 & 4047
carefully for days on end...

and studied the truth-table of the 4041, and drew up a detailed partly schemo
of the ADA drive circuit with distinct pin-out
before posting reply#15...



[EDIT]: a minor drawback is the price of the 4041 compared to hex-inverters,
and its not-so-widely-spread availability... 

[dj_death]

OHHHHHHHHHHHH  You 're right powertube ada schematic is 100% correct in the clock section. When for example 4047's pins 10,13 have values low , high respectively, pins 1,4 of 4041 will be low and 9,12 wich are inverter outputs will be, low. At the same time pins 2, 5 of cd4041 will be high (inverter outputs) and 8, 11 (buffer outputs), too. Thanks everybody.

[Mark Hammer]

My apologies to you both for the false alarm, and to A/DA as well for besmirchng their reputation.

On the other hand, we DID learn something from Ton's excellent detective work.