ARP 2600 Voltage Processor--Part Three--Mixer/Inverter Clone--Finished and Working

DiWHYers: If you want to build this eurorack module, you can get schematics, gerbers, a BOM, board/schematic layout, and more, from these guys; their community (here), and this project page. 

 

Hello again!  The mixer inverter ARP2600 clone modules are done--in and working!

ARP DIwhy modules with a friend's vintage 2600.  Got USD $14K-20K to spend? you can get an original...

More info about the project and design can be found in part II (here--original clone attempt...)

You also may want to skim part I (here): original thinking behind the ARP 2600 Voltage Processor project and cloning the "Increase lag time" sub-circuit.

Cloning these 3 sliders and their surrounding circuitry!  Euro format....

I have posted gerbers, BOM etc. for two clone modules. One is the equivalent of sliders 1 and 3 above, with a pot controlling normallized control voltages, the other has pots to attenuate 2 input signals; on both, summed outputs are inverted, same as the original.

To build both you will need 2 boards and 2x front panels:

Front Panels. The one on the left represents ARP sliders one and 3.

In general this is a super easy single op amp build, it took me less than 2 hours to make both versions of the ARP 2600 mixer-inverter voltage processors.

2x inverters with the lag processor you can read about here.

More A-B testing with voltage processors in a vintage 2600....

Easy!!!  

More ARP clones on the drawing board...the Electronic Switch for instance? Until then, have fun and don't breathe the fumes.

ARP 2600 Voltage Processor--Part Two--Mixer/Inverter Clone

Before we begin: Please visit my sponsor's website, PCBWAY; it makes them happy when their analytics tell them I have readers. Click here to visit their homepage and here to find ready-to-fab gerbers for many of the circuits featured in this blog. You'd be helping this blog out a whole lot--visiting their websites from the links above helps underwrite the cost of new projects.

Thanks!

======

Let's keep going with what we started last time: cloning circuits from the legendary ARP 2600 synthesizer--specifically it's voltage processor. 

The subcircuit cloned this time is its Mixer-Inverter:

A clone of ARP2600's "Voltage Processor's" top 3 sliders--a mixer/inverter

Original ARP2600.  We are cloning circuity in the bottom center.

Specifically this time we are cloning the functionality of these three 2600 sliders and 8 jacks!

Looking at the original ARP Schematic (here, page 29) the top three "voltage processor" sliders are tied to an inverting op amp based mixer. Simple enough..... 

So what's the big deal? I think ARP's take on this familiar circuit yields a lot of possibilities in a deceptively small amount of front panel space. Brilliant engineering from ARP, I think.

ARP'S DESIGN

The fun begins with the +10/-10 normals. This means with the sliders all the way to the left you end up with an output bias offset of 0V; the inverted output does just that: it inverts the signals you insert to the immediate left of the outputs.  

However if  you increase the pot's value, and/or put a different control voltage into one or both CV inputs, then mix it with an unattenuated input, you can alter the output's bias offset in extremely useful ways.  

Furthermore if you attenuate the unattenuated inputs (using a passive attenuator, or the ARP 2600's VCA, for example) now you have a 6 channel inverting mixer. Or gang 2 of these together to get more inputs....or use 2 in series to re-invert the inverted signal....or...

EURORACK PORT

The original 2600 design has 3 faders, two with attenunated CV normals, another with a normal Keyboard CV so you can quickly realize a Weather Report inverted keyboard vibe.  But--how to best represent this within our beloved Eurorack footprint?  

In my rack I have to save space so I didn't use sliders. And there was no real way to duplicate the "keyboard" normal in my modular setup. 

Instead, I decided to make 2 variations--one that has the pots vary the +CV/-CV normaled inputs--ARP sliders 1 and 3; the other allows you to attenuate the input signals--a dual version of the 2nd slider from the top. 

To get the full 2600 vibe, you might want to build both?

The photos from this post are the latter variant.  The other has the CV and Mix inputs flipped on the front panel and main board.

On to build photos:

Happiness is new boards from PCBWAY.  Show their analytics that you read this blog by going to their website here.

Each variation requires 2 PCB's, main board and front panel.

The op amp is SOIC--not too tiny....

The pots I use for these Euro projects have a longer shaft than will the knob can, so I cut them to size with a Dremel cutting tool:

The trick to cutting the pots' shafts without shattering a bunch of cutting disks--it's time consuming to replace the cutting disks when they break--was to cut one side, then the other, than back to the first. I did this over and over until maybe 1mm existed between the cuts.  

Then, I used pliers to twist off the remaining metal that is not needed.  

After the knob fitted flush with the front panel.

Ready to test!!!

"Seems working"

Ah, an inverting mixer!  It works, but I have thought of improvements to the main boards' layouts to make fabrication easier.

I am going to redesign 2 of the PCB's then go to PCBWAY with hat in hands to see if they will get me some improved boards.  

Then I will create a "Part III" post and make the gerbers and associated files available to you, my faithful readers.  

So!!  Upcoming--improved ARP mixer/inverter clones, and ARP 2600 sample and hold/2600 audio switch subcircuit work-alikes.  Should be ARParrific!  Until then, don't breathe the fumes.

ARP 2600 Voltage Processors--Part One--Eurorack Lag Processor

DiWHYers: If you want to build this eurorack module, you can get schematics, gerbers, a BOM, board/schematic layout, and more, from these guys; their community (here), and this project page. 

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

Hello again, I went to a geeky synth meetup recently--we went through the ARP 2600 patch book. I wasn't expecting much--I already know all this right? Wrong! The patch book is a wealth of great ideas and covers a lot more than patching a vintage ARP 2600 or 2600 clone. For instance, I didn't know you can create a high pass filter by inverting a lowpass filter output signal and summing it with the original? That's in the book...

We ran out of time...I wanted to keep going through the book, but don't have room for a clone or a vintage ARP 2600 (even if I had USD $10K+ to spend....)

I have 2600-like VCOs, VCFs, VCAs, Reverbs and so on, already in my rack however. 

This time: an ARP2600 lag processor clone

But! I don't have the 2600's voltage processors....these subcircuits are an important part of the 2600's cool identity:

Original ARP 2600 Voltage Processor

With that long ado: in this and coming posts, I will create clones or work-alikes for the 2600's voltage processors. This brings some 2600 coolness to my rack without spending a box of money.

This time: the lag processor.

Dual ARP Lag Processor Clone--done, works! 

The original ARP schematic can be found here, go to page 29; subcircuit is in bottom left corner.  

It's a familiar design, "lag" here means "glide". See my portamento post here for more information about how the design works. 

However, two components ARP used in the 2600's lag processor were unexpected--a .1uF cap and an A1M pot. This means very short portamento times even when the A1M lag pot is "turned to 11". 

But on the clone PCB as well as on an original 2600, the "lag" processor sounds good to my ears, an "Oh Lucky Man" outro vibe.

Lag boards! Shout out to my sponsor, PCBWAY, for helping out.....

Time out--I need to do my usual shameless plug for PCBWAY, this blog's sponsor. They never have dictated what I post and are extemely friendly; I owe them my gratitude. Please visit their site and community page--doing this helps me keep creating these projects and posts. 

Back to it....

ARP 2600 LAG PROCESSOR DESIGN

The module clones the original ARP design, with a few minor changes discussed shortly.

If you build this module, you will need 2 PCB's: a front panel and "main board" in 4HP Euro skiff config. Consider Fabricating the front panel in black with white silkscreen; make sure to check the box "do not use an order number on the PCB".

I used a single TL072, the 2600 uses LM301's; I figure any dual DIP op amp will work. 

My clone also uses a 1K resistor in series with the output jack. This is not on the original design; I put it in there to keep a short to ground (a dangling patch cable that shorts against a metal enclousure) from drawing too much current and possibly damaging an op amp or other components.  

This means the voltage at output might diminish a bit as the lag value increases. This is caused by forming a voltage divider between this 1K resistor and the load resistance of whatever you plug the lag processor into--this may not be what you want. 

If not, put some jumper wire in instead of R3 and R10 and try to be neat when patching or (aguably) use 100ohms.  

Using this protection resistor (or not) is more complex a topic than I initially thought. A forum post about this resistor, to use or not to use, what value should it be, etc., is here.

BUILDING THE MODULE

No surprises here, one mistake is on the front panel, LEV2 is the top pot, not the bottom one (I kept changing this and ended up getting it wrong). This is fixed on the gerbers ready for download at PCBWAY (again, get gerbers here). 

Beyond that the dual lag processor can be built in maybe an hour? Something like that. This is an easy build.

Ready to test....

"seems working...."

MODIFICATIONS AND PATCHING IDEAS:

I have to remind myself to use this lag in unexpected places. The patch book has some ideas. 

For longer lag times increase the value of C3 and/or C5, and/or change the value of the "Lev2" and "Lev1" pots. I left mine at A1M pot and .1uF to match values found in the ARP service manual.

There are two lag processors in this 4HP design, the ARP2600 has only one in its voltage processor section. It would be easy to modify the circuit to normal out one to out 2, putting the two in series, giving you longer portamento times. Easier: use a short patch cable. Viola!

 

OUTTRO

Next time I will clone the Mixer/Inverter section of the voltage processor--an interesting single op amp design that is unique to the 2600 and its clones as far as I know. Until then, don't breathe the fumes....

Restoring a Tek 453 Scope

Before we begin: Please visit my sponsor's site, PCBWAY; it makes them happy when their analytics tell them I have readers. Click here to visit their homepage and here to find ready-to-fab gerbers for circuits featured in this blog. Thanks!

============

The story begins in  maybe 1980, my brother found a discarded 50Mhz Tek 453 scope in his employer's garbage dumpster and took it home. It worked!  He used it for a few years but eventually bought something faster, giving me the Tektronix as a hand-me-down. 

I used it until about 2003.

Restored, working scope.  Beautiful in a 1960's industrial sort of way.

Fast forward many years: I use smaller, faster Siglent scopes; the Tek has been gathering dust for over 2 decades.  

Wasteful--I should make sure it works, then, get it to someone who will use it, right? 

It works here, but at first? nope

POWERING IT UP

Nothing! No fans, no sound, no smoke, no traces visible on the CRT: completely dead. 

Fortunately the schematic and user manual are online (here and here). The 453 is solid state. Also i read that the 453 was designed using no unobtainium ICs or components. 

So--assuming the CRT can be replaced, if it's dead--no matter the issue, this scope can be repaired. 

EASY FIX??

I removed its cover to check the power supply. Sure enough, the power supply was dead (nothing, anywhere); mains fuse was good. Checking mains coming into the power supply: nothing, but it didn't appear to have a short. 

Could the power cable be bad?

**Not** the original mains plug

I opened up the Nema 5 connector and sure enough the positive mains lead was disconnected, so the scope was not getting mains power. Doh!

Easily fixed--I screwed the mains lead back to its screw terminal inside the connector and fastened everything back together.

After that it powered on. Amusingly: after 20+ years as a giant paperweight, the scope's basics worked. The CRT was tired--I had to turn "intensity" to about 70% or more to see traces,  but both channels, one shot triggering, X-Y mode, and the trigger out BNC worked great. 

3D PRINTING SWITCH CAPS

On to cosmetics. I removed 20 years of dust and dirt on the scope and its front panel with a handi-wipe. Easy.

The front panel switches had all but one switch cap missing. I figured modeling the only remaining switch cap with Fusion 360 then 3D printing a dozen or so replacements was in order.  

I hadn't used my 3D printer in a long time and forgot pretty much all of it.  

That's why I have this blog, a trail of crumbs right? 

Original 3D how-to post for my bench is here. 

I rewatched the Fusion 360 for beginners video here.  

The Voxel printer uses its own slicing software--Flashprint MP.  As far as I can tell, other popular slicers like Cura still don't work with the Voxel IIIP, you have to use Voxel's slicer. I had forgotten that too.

I wanted to make the replacement caps to look as much like the original as possible.....I couldn't find a switch cap I liked on thingiverse. Time to roll my own.

The original switch cap was carefully measured and drawn out:

Fusion 360 bench notes:

• Save your design before you start doing anything to it.
• Fusion 360 saves everything in the Autodesk cloud--you knew that right? I didn't.  If you want to save the file locally you have to export it.  
• make your origin planes visible--otherwise you can't tell up from down, really.
• You create 2D drawings in Fusion 360, using a"sketch"  You right click on a plane and then choose "new sketch".
• Hitting "S" in sketch mode brings up a very useful dialog: start typing center and find the center rectangle tool, for instance. Now you can draw centered around the plane's origin. Much easier than trying to find things via the menu and icons--beside, Autodesk keeps changing the Fusion 360 UI, so videos showing you where things are are often out of date/inaccurate and thus frustratingly difficult to follow.     
• Hitting "Q" allows you to extrude, turning 2D into 3D.  
• I could extrude holes with offsets from a flat sketch, but not from a curved one. In the case of this part, I had to cut a 50 mil x 135 mil rectangle so the cap could be pressed on the switch.  I did this from the bottom plane with 100 mil offset, extruding the cut to about 200 mils above the top curved surface. The same cut couldn't be done from the curved top down, but this bottom-up approach worked.
• You can scale the entire part--useful for fine tuning its fit. Lasso the entire part with your mouse, click on "base" in the nav tool to the left, then click "Modify" along the top.  In the dialog that pops up, 1.15 means 115% larger than what you have now. The same can be done in the Flashprint-MP software for my 3D printer, which turns the stl file into a gx file to send to the 3D printer. I found the scaling freatures in Flashprint-MP easier to use vs. Fusion 360's scaling feature. Why did I need to scale the part? The caps needed about a 105% bigger cutout to snugly fit on the switches in spite of careful measurements done previously. I figured the extra 5 percent was needed due to slop in the printing process--not sure, but pretty easily fixed--I expanded the entire STL 1.05 before printing; after that the caps fit perfectly. This may or may not be needed for your printer setup.
• Don't use inches or mils when creating 3D models--ever. Use millimeters instead. Flashprint-MP slicing software only works in millimeters, and sending it stl's crafted in inches drives it nuts.  Fortunately when exporting the stl file from Fusion 360 I could choose millimeter scale, even through 3D model was created in mils; now Fusion-MP was happy with the source file.

Let's print!

I was afraid the base of the part was too small to stand upright while being printed but the Flashprint software was smart enough to create a flat base around the part. It printed perfectly:

After printing I tried different means of removing this flash. 

What ended up working best was to cut the base off with household scissors, trim further with an xacto knife, and finish up with a 300 grit filing stick. If I spent time carefully cutting and filing the switch cap came out almost perfectly--very hard to tell from the original. 

But...I didn't end up spending too much time filing to perfection--no point as I saw it....Instead, I got the switch caps to where they were good enough. Time to move on!

PAINTING 3D PARTS

This was something I had never done. Would the paint stick?  Would it turn the plastic into goo? Not sure. 

But--I used ABS (or was it PLA?) filment and read online that painting either type of plastic was easy. 

I used Rustoleum "2x Ultra Cover" black and grey primer--I found this recommendation on reddit--two quick coats and the switch caps looked great.  

Pressed the caps on the scope--done.  

if anyone else needs the stl or gx switch cap files, get them here.

NEW HOME FOR AN OLD SCOPE?

For the antique scope--will anyone want it?  Not sure, but it was fun and easy getting the old scope ready for use again.  

Is antique electronics restoration is more fun then AudioDiWhy? Perhaps. In another universe.  

Ah! Finished.

Until then/there, Tek out, captain.  

Bench Mixer--Looks Matter

Existential Question: Is AudioDiWHY a dating site? Swipe left, swipe right?  

Use an Arduino to automate the swiping?  

Unlikely.

I have brought DiWHY modules to my geeky audio meetups that sound great (well, I think so) but are --fugly? 

The guys roll their eyes. 

Then I bring in something boring with a neat front panel and layout and they say--dude--cool.

Swipe right?

The bench mixer discussed in today's post exemplifies the latter--doesn't do a whole lot, but looks good, enough to get the coffee date, maybe more, I think.

You can read about the first attempt at the layout and what I was trying to accomplish (not that)--post is here. I went too fast, got cocky, and made lots of mistakes. Sound familiar?  

Fixed it this time, it works, although the L > R callout is backwards on the front panel. Red flag? Deal Breaker? Nope. That's fixed in the gerber I uploaded to the PCBWAY community page for the project. 

Ah yes, PCBWAY.  Shout out to Wendy! Help out this blog and check 'em out. In fact, please!! click this link to PCBWAY--it will register in their analytics that you read this blog; their community home is here.

Let's look at the butt ugly swipe left frac module this date-worthy mixer replaces:

But...but..your profile says 5'-11"!!!

Here's the new one again:

"AudioDIWHY-- No Catfishing"

This is a simple bench mixer--audio at input (2 x 2 with normals to handle mono audio) mixed to 2 screw terminals in the back, which in turn gets connected to a power amp and speakers. 

Two op amps, some hardware, dual concentric pots, and a handful of resistors. Dumb! Easy! That's what she said....

Happiness is new boards from PCBWAY!

On to some useless build photos:

Soldered in the SOIC op amps.  Not too hard to solder, no microscope needed.

Test fit....we'll know more after the 4th date.

 You're a tech right? 

 ....you're gainfully employed?

Not shown in the photos, but it's there: be sure to connect ground on the main board to ground on the jacks board. You can use a 2x 100mil male/female jumper; to get this done fast I soldered 22 gauge single strand hookup wire between the boards.

If you want to take this one home, I posted gerbers, PDFs, Eagle files, BOM, etc. for this module on PCBWAY's Project page--here. 

If you build one you will need 3 PCBs: the jacks board, the board with the op amps, and the front panel. 

For the front panel I'd recommend black solder mask with white silk; check the box at the bottom that says for an extra $1.50 don't put the order # on the PCB. Something like that. 

Let your date pick up the coffee tab for this one, you've been thrown to the mat to many times.

Above all: no ghosting! See ya next time.