Looks amazing.
1 Like
Thanks for the nice comments. It’s not without suffering and questioning why I got myself into the project 
I’m seeing doubles 
:dizzy:
I remember when you got the double box delivery, jealous.
What a Great Wall setup though! I love the lighting and the wood panel looks.
Gives me “house on the rock” vibes.
For those who don’t know, house on the rock, is an eccentric dude’s home in Spring Green Wisconsin USA which turned into a museum of oddities over the years. It has the worlds largest carousel but not a single horse on it for example.
Point being it feels like a warm yet mysterious place, perfect for music making imo.
2 Likes
That’s some wallpaper right there 
I’ve also toyed with the idea of two 2600’s (and two A4’s!) but space will only allow for one atm. But I can see after spending hours on a killer 2600 patch that you’d want to make a complimentary one on a second machine without tearing down the first.
4 Likes
Does your guy know why the VCA level may drop depending on the chosen gate source ?
Bit cheeky tagging Scot like this, possibly not an original 2600 thing either, but then again, it has its own quirks and often there’s a good reason …
here’s my mind bender
i have a very very simple vanilla patch on the blue marvin, just about as simple as it gets …
one osc into the filter
quickish ADSR envelope opening the closed filter
longer AR env on VCA
i was revisiting CV triggering of that voice which is where the ARP has one of its interesting quirks wrt gating
i noticed something odd when using my Korg SQ64 and it manifested as a quieter signal by a noticeable margin, probably less than 3db but no need to measure, it’s plainly lower
well that’s weird, the only thing i vary across comparisons is the gate - nothing else is changed !!
so i compare the following
Manual trigger (via button)
S/H ‘auto’ gating (via switch)
vco 2 square (mono-polar 0v > 10v)
S/H manual gating (via internal clock out) (mono-polar 0v > 10v)
also A4 CV out set to gate 0v > 10v
all of teh above behave the same, they all fire the envelopes when i send a feed of the gate multed or via Y splitter - and they all sound the same level wise
the outlier was the original device selected which inexplicably sounded quieter than the manual trigger method - this was the SQ set to V-Trig 10v !
I also fired up the square LFO from the Brother32, presumably bi-polar circa 5V +/- (i didn’t measure) and it delivers the same lowered output - again, i am just changing the gate source … the envelopes are clearly firing, but for some reason they outcome is a lower overall volume … wut ?? … i can’t get my head around this
wondering if this is something that affects the og or if anyone has any leads (no pun intended)
what i started to imagine, albeit that to me it made no sense, was that maybe the gate was a bit lightweight from the Korg, far from it, it’s pumping out an 11v gate on the 10v setting and it is actually higher than other gates - the max on A4 is ten so i can’t test 11 on that
does it make any sense that a gate input (sole source split/mult … triggering the gate/trig simultaneously) in excess of 10 could manifest in some biasing of the VCA or some other thing which drops the volume
i may see if there’s a calibration on the Korg or a way to adjust that in due course
but for now i’m flummoxed - ftr, just using a regular voltmeter, nothing fancy, but the stead state part of the gates is easy to verify and is as mentioned
i need to review the Behringer documentation and maybe google the quirks given that the circuit should be broadly cloned
Otherwise looking for any confirmation of the B2600 observation and any credible ideas about what could trigger this - i can’t imagine there’s anything being overlooked, but i’d be happy to have my understanding straightened out
The one thing i didn’t test was teh SQ powered by DC, but the 5v usb device charging it is good for plenty more amps than required and it’s kicking out 11v anyway
@anybody what gives … ?
1 Like
maybe this helps, I’m not sure… and to be honest I often forget the details of this on the 2600, especially when I’m changing the way I’ve got it setup (what I’m sequencing it with). and maybe you’ve addressed this above since you mentioned multing the signal to gate and trig… the trig input will fire the ADSR only. so if you’re using it with the switch in the AR section up, you’ll need to mult that signal. I think it has to be 10V as well (which can be accomplished in the voltage processor if your sequencer can’t do it; but sounds like you’re set there). I instead typically use it in switch down position, with a 10V gate sent to the “S/H Gate” input. this fires both AR and ADSR.
3 Likes
If I am understanding your post correctly, this is indeed the way the original behaves. Keep in mind that the AR is linear, the ADSR exponential (you can change this with patching). As you likely know, the latter is probably going sound a bit quieter than the latter.
Both of the 2600s I have owned required pretty high signals to provide anything like full volume. In fact, I almost always need to run my sequencers through a voltage processor modules to give me enough oomph.
Theoretically it should trigger the envelope if it exceeds a positive voltage of 10v. However, I suspect that the culprit is in the shape of the pulses being output from the SQ-1. And indeed, according to discussions on Gearspace, it seems that the 2600 might be a little too old and pokey to grab onto the rising edge of the SQ-1’s gate signal. The trick here is, apparently (I haven’t tried it) to run the SQ gate through the 2600’s lag generator, leaving the slider set fully to the left, and then use the output of the lag generator to the S/H GATE input in the AR/ADSR section, with the switch in the down position (natch). I imagine that you might have to nudge the lag slider a micro-smidgen.
Give it a shot and let me know if it changes anything. I’m interested.
6 Likes
Not at all. Even after all these years I am still learning about the 2600 - it’s a very deep machine with nearly unlimited possibilities. But it has plenty of odd quirks (looking at you, S+H clock!) and loads of less-than-obvious features (Ring mod as VCA, etc). I am only too happy to share what little knowledge I have gleaned over the years. One of the coolest things about all of these clones coming out is that there will be more users accruing more knowledge to go around.
2 Likes
Really nice.
1 Like
Thanks! I’m really proud of my work quality on this one for sure!
I’m rarely proud of my work openly, even though I should be more often (because I do make some cool stuff).
4 Likes
Looking forward to some experiments,will report … fascinating thoughts on this, not gonna lie I like the quirks, it makes you change course, a bit like retuning a guitar
1 Like
A quirk that drove me mad, but that I grew to like: the External Clock Input, you’d think it would override the internal clock, affecting the rate of the Electronic Switch and S/H GATE input. Nope. It only affects the rate of the Sample & Hold.
So if you want to, say, trigger the envelopes and keep the S+H in sync, you need to send your clock to the Multiples and then to the External Clock Input, along with the S+H/GATE jack.
This made me crazy for years and I assumed my External Clock Input jack was broken. Nope. It’s just a 2600 design quirk. Now, I have come to enjoy setting the internal clock rate (and thereby that of the Electronic Switch) against an incoming clock for interesting off-kilter rhythms.
Speaking of the Electronic Switch - it is bi-directional, something not entirely obvious at first glance.
Also not obvious - the Ring Mod input in the VCA section inverts any signal plugged into it. ARP really liked their inverters… 
4 Likes
Such a cool shade of yellow. Not an obvious choice but it really works.
1 Like
fwiw, I could get the SQ-1 to trigger my 2600, using either of the two gate/trigger options I described above. the SQ-1 gate outputs 10V. I’ve had to use the voltage processor trick with a Doepfer Dark Time though. until I discovered there’s a jumper inside to select between 5V and 10V gates.
now my 2600 has had extensive modding done to its guts, so I don’t know that all 2600’s behave this way with the SQ-1. sounds like no, from what you’re saying you’ve read on GS.
2 Likes
From what I can see, I think it sometimes works with some SQ-1s and some 2600s. At various times, I’ve seen 2600 manuals that claimed 10v, 12v, and even 15v were needed. I can’t imagine wanting to shove 15v into my precious 2600! 
I never tried an SQ-1 with either of the 2600s I’ve owned, so I’ve no personal experience. I have had volume discrepancies, but the Synthesizers.com Signal Processor module always gets me out of trouble.
4 Likes
Most typically used op amps in synths past a certain point (and certainly most transistors) are happy with 15V. (Assuming 15-18V supply.) Seems odd though that it would take 15V to trigger any function. Tolerate 15V certainly, but requiring that level seems strange. Things were different way back though, and some interesting power choices were made.
3 Likes
Yeah, I was always suspicious about that one - but just 'cause they said it, doesn’t mean it ever actually required it. I don’t know offhand of any change to that circuit through various versions, but I’m no expert. It’s only ever taken 10v on either of them I have owned - a Tonus and then a 2603 off the final lot (so no potted circuits! Yay!), so up to ten years between 'em? That always made me doubt the stated 12v and 15v.
I just mentioned it because it seemed even ARP was vague about the required voltage, leaving the rest of us guessing for 50 odd years.
2 Likes
Yeah, I’m not an expert on things of that era myself, or of the design choices of the time. I’m sure there were a lot based on which components could do what. Things like non-symmetrical power rails, etc. baffle me 
2 Likes
Indeed. Of all the synths I’ve opened, ARP’s are the most unpredictable. 2600s are almost always potted (except the final iteration) so who knows what lurked inside that epoxy? However, the Odysseys were never potted and of the four (supposed) Mk2s I’ve owned, not a one had the same guts. Heck, I had a Mk2 that was entirely a Mk1 in a Mk2 case, and I’ve had two Mk2s with close serial numbers that had completely different oscillator circuits. Now, any of these might have been built, “restored”, or repaired with whatever parts were on hand, and that only adds to the head-scratching.
And then there are the potential mods! My current 2600 is a crow’s nest of fragile wires and dangling daughter boards, giving me no desire to get in there and fix its few issues (I repaired a broken VCA, noise generator, and replace the power supply, but it was scary business).
Mind you you, I have little idea how any of it works, so I wouldn’t attempt to understand the actual circuitry. Luckily, there are smarter people than myself around. Speaking of which, what are you doing Saturday… 
4 Likes
I’ll throw the SQ1 into the mix when I deep dive into this, but FTR I’ve been referencing my SQ64’s gates thus far given that it’s my elected 2600 buddy for now.
I may have forgotten to mention that I tried the Auto gate (sh clock switch down) and fed a solitary gate signal to that, that fires both too as Scot reminded me but there was something about this method which didn’t hang together iirc as it’s otherwise the simplest way to use in a ‘normal’ both-envs-together way
All this nuanced discussion on gate profiles has me wondering if there’s an economic way to log cv signals in the time domain ?
Okay - I have cracked this, and to be honest, it only makes me love the charm of the ARP 2600 design even more - I like these little things which work differently amongst synths, it introduces unexpectedness if you make too many assumptions … thus the reason to explore and begin to understand/utilise/master etc
So firstly, I’m happy to report all is well with the Blue B … this is a 2600 thing and I think it merits an exploration on your part if you’re an owner (especially recently) of any variant
The important takeaway, for anyone not wishing to delve a bit deeper (useful illustrative patch shown below), is that it’s important to discard the idea that gates are a binary thing … on and off … start and stop … as this is not the full picture with the 2600
This explains my confusion when I was perceiving reduced VCA output even though I was successfully ‘triggering’ both envelopes (reduced, wrt comparing internal ‘gating’ in contrast with certain external gear, pretty much everything except the A4 … kudos due to Elektron here for delivering the goods)
The other thing I’m annoyed I forgot about, but had never confronted in practice (best way to learn/remember), is that the 2600 AR (or maybe best though of as an AHD) is akin to a slew limiter (or a second onboard lag processor, but with independent slew rates for onset{A}/end{R}) so in this case the H(hold/‘sustain’) level is in fact the gate voltage level … so the ‘gate’ in this case is not merely acting as a binary/timed switch but it acts as the terminal voltage level at the end of the attack slew
What I was witnessing in my previous tests was that the relatively ‘reduced’ volume was determined by the lower gate strength of the devices used (with the sole exception of the A4)
All the goodies you can throw at the ‘gate’ within the 2600 deliver a nice healthy 10v as does the A4, but the advertised ‘10v’ gate from the Korg SQ-64 (although seemingly pumping out 11v as per my meter) was clearly not on spec - same for SQ1, but that is apparently set at 8v, so its effect on the perceived volume shouldn’t have been a surprise had I not thought of a gate as being a binary switch that works when it reaches a threshold voltage (Microfreak/Crave/Doepfer Dark Energy don’t deliver any better than the SQ1 or SQ64, but it may only be the SQ64 that indeed promises 10v, I haven’t checked the respective specs)
Incidentally, when using the A4 (which permits setting a gate level of your preference up to 10v, I noted that there was some ‘triggering’ when the output was as low as 3v (iirc, that’s what’s needed to activate the AR/slewer) and if my memory serves me the gate (but not yet trig) voltage for the ADSR is North of that (5-6v) but well short of 10v … and the test patch below beautifully illustrates the interconnectedness and what is happening with the 2600’s envelopes depending on gate levels
In this examination it hasn’t seemingly been necessary to examine gate ‘profile/shape’ but it would be unwise to discount that this could play a factor given the other nuances of the design
In contemplating how best to present this, I decided the best way to learn is NOT to read, but to TRY ! so I came up with the simplest method that should apply to all variants including the og, so with that in mind B2600 owners need to put the two envs in the mid(1x) time factor state. Repeat is off, voice mode mono and trig mode was left on single for simplicity)
The rest of the patch is a simple square from VCO1 sent through an open filter, the filter feeds the VCA which is opened by the AR, VCA to Mixer - see image - everything is in place, note that I haven’t ringed the VCO1 freq (forgot), that’s an important level, it’s difficult to quantify with the log value, I find it much more useful to vary pitch than volume when auditioning modulators. Keytracking is on and there’s no Midi but despite the initial osc freq being in the 5k region, the actual frequency is offset internally, just be sure that what you hear is about 440Hz (with no(i.e. zeroed FM) ADSR modulation), that is to allow headroom to hear the two stages of pitch modulation the patch illustrates … it’ll make sense when trying this out (it’s just much easier to perceive volume changes in modest frequencies and freq changes too)
For comparison between internal and ‘external’ (i.e. voltage from an alternative source) it is simplest to use the [S/H Clock] input to trigger the envelopes (although we’ll be inserting here and breaking the normal from the 10v clock)
A recap, when the S/H switch is down a 10v signal will activate both the AR/ADSR identically to pressing the Manual button, by using that input we have a metric for gate strength now, so we can vary the incoming gate level triggering (or not) the envelopes … also note, the Trig in although adjacent to the AR actually relates to the ADSR only, the Trig in is nothing to do with the AR, and as you’ll see, there’s two ways to activate the ADSR depending on gate voltage, so by varying the gate strength we actually have three envelope permutations !! corresponding to let’s say circa 3v 6v 9v
3v will trigger the AR slew and as we’re using this for the VCA it’s relatively quiet
6v will trigger the gate of the ADSR (but not the Attack stage yet) … so there’s a secondary slew type envelope using the DSR, where D and R are akin to A and R of an AHD, and in this case, H is modulating the freshly opened gate which targets VCO1 pitch, so the pitch is raised now (start with zeroed D & R and lift them a little to see the slewing of pitch)
9v will properly trigger the ADSR, i.e. it triggers the Trig in now, so it actually utilises the Attack phase of the ADSR and with the attack phase jumping to 10v, you get an even higher pitch trace
(PS it’s not exactly 9v or 6v btw, just making it read simple with 3/6/9 … it’s probably 3/5/7 on the B2600 … og ARPs may have different sensitivity thresholds)
So to explore the effect of Gate level (and keeping it all in house for the 2600) we’ll use the voltage processor to offer up an attenuate-able 10v signal that we’ll ‘gate’ on off with the Electronic switch - set the Sample and Hold Rate around the 10% or circa 1second)
That’s what we feed to the Gate input, and we can attenuate the inverted -10v (i.e. 10v) by varying the slider (red arrow) left to right … left is attenuated, right is ‘open’
So by varying the attenuator on the net 10v (a tad less coming out of the B2600) signal you can get 4 envelope states
-
Both Off
-
AR On (low volume)
-
AR On (mid volume) + DSR (i.e. an untrigged gated ADSR acting like our AR/slew as the Decay stage acts like an attack stage by going from 0v to sustain level)
-
AR On (High Volume) + ADSR (i.e. a trigged & gated ADSR, so the Decay stage is Max to Sustain level (for a long enough gate as per our controlled experiment) )
Hopefully experimenting with the patch below will help consolidate your knowledge of how the 2600 envelopes are activated - that should make using the independent gate inputs and the ADSR Trig input a bit more intuitive going forward
If you can get a gate of e.g. 8v say, then it will ‘fully’ trigger the two envelopes, but you just won’t get the same oomph from the VCA as the AR envelope is slewing a ‘lower’ voltage of 8v as opposed to the 10v you get internally from the S/H clock (or Manual, or LFOs(untracked osc) etc etc
Thus my original perceived (i.e. non) issue … so lower gate voltages can work, but they don’t necessarily equate to what’s possible with the 10v internally, depending on chosen application - it’s probably not too significant with the AR as VCA … I was just puzzled as to the why it was quieter
Anyway, hope this little patch helps, it helped me to simplify it like this … the 2600 is an amazing utility device as well as incredible sounding
Start with this config (especially ADSR levels) and have fun varying the VP slider to see the 4 phases or activation
PS the original manual is incredibly helpful, but I’m not sure it elucidated the initial way the ADSR can be ‘activated’ (sans A i.e. sans Trig)
Sorry for loooong post !
11 Likes