CYGN-A

Yesterday I ran across a pdf of the IEEE Journal of Solid State Circuits from April of 1971 that contained this diagram:

Attachment 57377

It looks pretty simple, and it is, but I started looking for that circuit for a long time ago. Those triangles are inverting op-amps. The grounds on the plus inputs aren't shown. That's basically Laser Images' CYGN-A. There's a couple of Zener diodes added in line with R6 that was copied from this example from the IC Op-Amp Cookbook:

Attachment 57378

R1 and R2 were broken into a two fixed resistors and a dual reverse log potentiometer that controlled coarse freq. R3 was broken into a fixed resistor and a 50K linear pot that controlled fine freq. and introduced a bit of a gain difference between the sine and cosine signals, but nobody cared enough to spend the extra money to use a dual potentiometer in line with the reverse log pots.

I've wanted to post this for a long time, but wanted to do it this way...

Terrific information. I've been avoiding mentioning the CYGN-A so as not to appear too greedy in my appetite for Laser Images' secrets. I do love those "Ahh... THAT'S how it was done!" moments. Sorry for the delay in posting results from the complete CYGN-B, my hands are a bit full until mid next week. And I expect to be home and back to my parts inventory in late January. Also, thank you Steve, for the obscure information you sent that will increase the probability of successfully rebuilding the SPGN (Spiral Generator Card).

The attached image provides an annotation to what Brian just posted that makes it a bit clearer what we are discussing here.

Any words of wisdom regarding operational parameters of the 2QMM I could test to discern if it is working, if I endeavor to build some?

Quote:

---

Originally Posted by Greg

Any words of wisdom regarding operational parameters of the 2QMM I could test to discern if it is working, if I endeavor to build some?

---

I'd just use AD633 - 4 quadrant multipliers in place of the two quadrant modules that were scattered throughout the Mark 6. The output is the same if you're driving them with 0-10 volts... They're what I was using in my build when I broke my hand and Jon offered me a Mark 6 back in 2013.

Wondering if I'm on the right track regarding substitution of the 2QMM for AD633. (see attached)

Quote:

---

Originally Posted by Greg

Wondering if I'm on the right track regarding substitution of the 2QMM for AD633. (see attached)

---

That's close, the 2QMM was a current multiplier. "I" in & out.
Attachment 57501
The resistors are 22k 1%
note: the inverting input to correct for the inverted signal due to the op-amp.
And you don't necessarily need to do the whole current thing if you're not trying for plug and play into an existing Laser Images design.

Quote:

---

Originally Posted by laserist

That's close, the 2QMM was a current multiplier. "I" in & out.
Attachment 57501
The resistors are 22k 1%
note: the inverting input to correct for the inverted signal due to the op-amp.
And you don't necessarily need to do the whole current thing if you're not trying for plug and play into an existing Laser Images design.

---

Sorry, my drawing is wrong. The resistor on the input to the AD633 shouldn't be there. Basically the input to the 2QMM was a summing junction and the output was always coupled to an inverting op-amp with a 22k feedback resistor.

I was going to ask how to deal with the unavailability of the IH5012CPE Virtual Ground Analog Switch (excerpt from datasheet attached), which is used in the SPGN, but once again, they are not unavailable. Ordered a batch.

Regarding the CYGN-A, am I correct that the component placement chart and the PCB in the attached photo are not the same circuit?

Quote:

---

Originally Posted by Greg

I was going to ask how to deal with the unavailability of the IH5012CPE Virtual Ground Analog Switch (excerpt from datasheet attached), which is used in the SPGN, but once again, they are not unavailable. Ordered a batch.

Regarding the CYGN-A, am I correct that the component placement chart and the PCB in the attached photo are not the same circuit?

---

The board pictured is the cycloid rotator.

Quote:

---

Originally Posted by laserist

The board pictured is the cycloid rotator.

---

That's great news! The CYGN-B needs the Z axis to fully be what it was. Ok, I'll be patient and keep putting together the parts order for the cards you and Ron have given me info on, and post when I get home in a couple of more weeks.

Quote:

---

Originally Posted by Greg

I was going to ask how to deal with the unavailability of the IH5012CPE Virtual Ground Analog Switch (excerpt from datasheet attached), which is used in the SPGN, but once again, they are not unavailable. Ordered a batch.

Regarding the CYGN-A, am I correct that the component placement chart and the PCB in the attached photo are not the same circuit?

---

Greg,

Here are actual scans of the card that I just dug out of the garage.

Ron

Instrument Grade PhotoMos relays are what I just bough to deal with a similar problem. They do not have the elegant correction scheme to deal with distortion that some of the newer chips post 1990 have, but they work for DC-audio just fine. . I use them at work.
I grabbed 20 of these lately when they were on sale at 99 cents as they retail around 7$ each in the through hole package.

Panasonic/NAIS AQV258 Solid State SMD Relay



G23378 - Panasonic/NAIS AQV258 Solid State SMD Relay (goldmine-elec-products.com)

I'll know how well they work on scaled down ILDA signals this weekend.

These are also rated for audio:

LH1518AAB

Steve

notes on preparing for the CYGN-A build:

substitute original UA741CV with UA741CN
substitute original RC4558DN with RC4558N

Ambiguities regarding capacitor values (attached graphic): Probably fine to go with 0.01uF and 0.033uF, but in the documentation, there is ambiguity about what these values are.

More interestingly, I wonder why Oscillator 3 has a different value of capacitors than 1 and 2.

Oscillators 1&2 were typically high max frequency oscillators and 3 had a lower max frequency. Since the CYGN-A was designed for the Mark 4 the caps were changed with the advent of the Mark 6's G-124 scanners.

Combining parts for the SPGN with the CYGN-A in the same ordering frenzy. See attached graphic for considerations regarding part NE536 specified in documentation vs. LF356BN seen in Ron's photo.

Dogloid + SPiral GeNerator (attached image)

Quote:

---

Originally Posted by Greg

Dogloid + SPiral GeNerator (attached image)

---

CYGN-B?, swept spirals, fixed rotation, variable rotation, AOM chopper, and stars too...

I'd like to get this problem out of the way. Steve, perhaps you could help with this: I need to scale the 0 to 2.5V dogloids signal output to +-10V. The values line up exactly with the case given in the document (excerpt incl. schematic attached) but something isn't right (see attached annotated scope captures).

Brian, The pinout format page of the SPGN documentation suggests the module exposes an interesting variety of input and output signals, which I may request your help deciphering in a couple of months when all the components arrive and the build begins. Do you have any sharable recollections for this thread regarding Carl Hannigan? Other than your earlier comments and seeing his name as the engineer of the SPGN, I don't know how and when he fit into the history of Laser Images Inc., but I guess he was an impressive person.

Unless I'm missing something, the 11K resistor is pulling down the 1.25 volts from VREF and offsetting your signal. Lose it. To get 10V p2p you want a 66K feedback resistor.

Sorry I meant to get +/-10V...

Request for wisdom on a couple of points:

1) Attached graphic shows the nodes involved in wiring the frequency control pots to a CYGN-A oscillator. Any advice regarding where the grounds and interconnects go?

2) Is the rotary ganged switch labeled Spiral Select related to the SPGN?

Attachment 57657

The freq pots are not grounded. Oh and my drawing has a error. I'm pretty sure the 1meg pots were dual reverse audio taper.

The spiral select switch was the image source for the spiral card. Originally on the Mark IV the spiral source was a fixed quadrature osc on the spiral card with the Mark VI and maybe Laserock the existing switch was retasked as a spiral source select. Originally three positions and eventually five. CYGN-A, CYGN-B, KQO, and two others I can't quite recall were the new sources.

1. CYGN-A
2. CYGN-B
3. KQO
4. DANUBE 8's
5. DANUBE 0's

And thus, a slice of history is revealed in a log book entry. The photo shows this circuit built using a KIA324P quad op amp. The output is a beautifully tunable quadrature. But the output amplitude is only slightly over +-1V. Is this correct? I notice the 10K in the summing amps suggests some gain there.

As the SPGN schem receives my closer attention, I realize that I don't have a clear idea of what this circuit actually does. I see the inputs for an external cycloid. It is like a VCA in which a shaped ramp modulates gain? Or is the ramp for the "sweep"? Is the sweep a varying offset in one axis? What is PDM out? Spiral in rate control? How was the reset used? Your comments, as always, and as your time permits, are very much appreciated.

The spiral card had 3, 0 to 10 volt inputs from the 351 data, and two digital inputs. Out rate, in rate and sweep were the analog signals, spiral reset and diamond spirals were the digital inputs. So you could control how fast the ramp generator ramped up and down, reset the ramp, and the sweep was summed to the image input. The ramp generator produced a ramp and an inverted ramp. Two pairs of 2QMMs modulated the input/sweep as spirals and inverted spirals. Analog switches could apply the other ramp to one input of the multiplier pairs for diamond spirals. I've never looked at the ramp gain. I assumed it's 0-10 V, but the key is the resistance before the 2QMMs. The PDM output was a square wave at the spiral frequency that was used as an offset on one axis of the fixed cycloids. Originally used in "Summer Madness" in Laserium I.

An error causing the low amplitude of the CYGN-A oscillator was found. Photo shows the three oscillators.

Other photo shows the collection of the less common semiconductors which should be able to become the SPGN, except for the Burr Brown, with which this build will have to do without. If I may ask, what is the part of the circuit that contains the dual transistors doing?

The dual transistor in the spiral ramp circuit is similar to the 2QMM circuit. I'm guessing, but I believe it was to get a log frequency response from the 0-10 volt up down ramp control voltages. You see it in audio synthesizer circuits where they're after one octave per volt VCOs.

I'll send the BurrBrown chip with my next shipment of Laserium artifacts. It just occured to me that it might be a cool effect if you used the ramp shaper on just one axis. It might be similar to the KQO symmetry effect.

Quote:

---

Originally Posted by laserist

I'll send the BurrBrown chip with my next shipment of Laserium artifacts. It just occured to me that it might be a cool effect if you used the ramp shaper on just one axis. It might be similar to the KQO symmetry effect.

---

Definitely I'll try that if / when we get that far. And thanks for offering the last known remaining Burr Brown chip!

The info in many instances is not possible to read with certainty from the SPGN schematic. Thank you very much Ron, for the great scans of your board, without which this would be much more difficult. After much deductive reasoning using the scans, parts placement charts, component lists, and the schematic, the following points of interest turn up:

The documentation contains an earlier and later version of the SPGN. Ron's board is the later version. Some parts in this version have different values, and there is a resistor (R52 470 ohm) that does not appear on the schematic. There is also a diode and some jumpers on the back of the board, but I haven't enumerated the diodes and caps yet.

R30 = 330K on the schematic I assume must be a mistake. It's listed as 82K in the parts list.

A comment is sought regarding the Y shaped schematic symbol appearing on the analog switches. I'm not sure how to interpret that. (see attached)

Quote:

---

Originally Posted by Greg

Definitely I'll try that if / when we get that far. And thanks for offering the last known remaining Burr Brown chip!

The info in many instances is not possible to read with certainty from the SPGN schematic. Thank you very much Ron, for the great scans of your board, without which this would be much more difficult. After much deductive reasoning using the scans, parts placement charts, component lists, and the schematic, the following points of interest turn up:

The documentation contains an earlier and later version of the SPGN. Ron's board is the later version. Some parts in this version have different values, and there is a resistor (R52 470 ohm) that does not appear on the schematic. There is also a diode and some jumpers on the back of the board, but I haven't enumerated the diodes and caps yet.

R30 = 330K on the schematic I assume must be a mistake. It's listed as 82K in the parts list.

A comment is sought regarding the Y shaped schematic symbol appearing on the analog switches. I'm not sure how to interpret that. (see attached)

---

Look on the right side of the drawing under the conspicuously unused revision block, and those are the inputs to the "Y"'s. The gates generate an inverted version of the spiral reset, sweep direction, and diamond spiral signals to turn the 5012 switches into SPDT switches. R-30 is on the "delta frequency" input. Now there were laserists and there were Laserists. Laserists tended to expand the capabilities of their machine. Sometimes by fixing things that were just wrong, and sometimes changing things to make it better. An 82K resistor instead of a 330K resistor would give a much greater range in the "fine frequency" adjustment of the spiral rate. There was a diode or two added in 1979 to prevent a really unlikely lockup of the spiral ramp circuit. I remember this because it was put in backward on the machine that was my first installation. Spirals didn't work - and "Laserock is almost all spirals", said the tech on the installation - and BTW the tech who put the diode(s) in backward - but in the end he was a hero - we discovered that if you shorted the input to a comparator with +15V there was a ramp. No control, but "you can't always get what you want..." He sat at my ankles for the shows that night with two test probes stuck into his hand. As CR4 & CR5 aren't on the parts list bat are on the layout, I'm guessing they're to keep the up and down ramp voltages from going negative. Perhaps never a problem before changing to an 82K resistor?... Not really sure about the 470 ohm resistor.

With regret, I never saw a performance of Summer Madness. The rhythm of that bass is also in something very well known that appeared a couple of years later. (Something about diamonds I think.)

Might that be a test point on the output of A4?

Woah, there's a rail indicator with no voltage and polarity.

Yes, that's a test point, and I believe the voltage indicator should be +15, but it's 3am... Adjust the trim pot so the voltage at the test point is zero when the sweep data byte is 10000000 or maybe 01111111 (5v). At 5 Volts in there's no sweep. At zero it sweeps max. At 10 it sweeps max the other direction. (Not the other axis, that's what sweep direction does.)

Summer madness was nice. The lumia effect used an interesting standoff made of a thin perforated metal. It was made up of about 1/8 inch holes in a hex layout, and was maybe 2x2 inches square. They would dip a piece of the metal in a lacquer, and rotate it slowly with the metal part vertical until the lacquer setup. The result was a hex array of 1/8 double concave lenses made of the lacquer. They weren't good lenses - which actually added to the effect. As I recall the standoff was about 3/4 of an inch above a rotating disk, bu I don't remember which one was used originally. As Richard alluded, we tended to mix and match as time went by.

There was this whole proprietary thing going on way back when. It was understandable. A fair amount of effort went into developing the stuff that made Laserium stand out. The day of the premier at my first installation the director of the planetarium walked up with about 100 sheets of paper and asked if I'd look at this laser documentation and tell him if it was any good. So I looked through it and said yes it's good - it's ours. Somebody, who will remain nameless, (In all honesty, because I never bothered to remember his name.) had redrawn a number of Laser Images drawings with their xompany's name in the header block. Stupidly they drew them with exactly the same layout and call outs. So not only were A1, C1, R1, & U1 through N, all in the same relative places, they all had the same call outs. Embarrassing for somebody.

Someone did a review saying Laserium could become a standout art form of the future, or something like that. Ivan loved it, but I don't believe something becomes a stand out art form in secret. Now that Ivan's gone, I'm willing to roll back the secrecy for whoever is interested.

Greg,

I can't attach the "6B alignment procedure" PDF because of the 5MB attachment limit (it's 14.2MB), but you can get it here:

https://media-archive.exploratorium.edu:5001/sharing/1RFDPfuPx

There's a section on the spiral generator in there. It's a rough document... Hope this helps.

Ron

Quote:

---

Originally Posted by Greg

Woah, there's a rail indicator with no voltage and polarity.

---

That's absolutely great! Thanks Ron. That has a lot of important info I haven't seen before. It even mentions the KQO, which so far has been shrouded in mystery. The hand drawn waveforms are charming.

Photo shows the current state of the SPGN build. I'm sorry I can't post the wiring diagrams I draw up for these module rebuilds, but let me know if you want them for the archive, and I can email you them.

Photo shows the SPGN wiring complete ready for testing. There is an empty socket to receive the BB4302 if one appears, but for now I'll use an externally generated ramp.

The alignment procedures document posted previously by Ron makes it clear which documentation packages are my most sought at this time: FXCY, AUMD, and IMRT

It appears that the AD633JNZ 4 quadrant multipliers I bought are counterfeit garbage. I should be able to get the real ones in a couple of weeks or so. Sorry for the delay.

Brian and Ron have come through again and provided some more previously unseen documentation, which includes the circuit called the KQO. This circuit is supposed to produce a spiral shaped cycloid. In the following hasty build, the P Channel JFET component had to be substituted for what I had on hand. This substitution may or may not be contributing to the observed problem(s) with the output.

The photo shows the clock signal, which I believe is doing what it is supposed to: a +15V DC duty cycle of around 80msec that briefly drops to ground.

I have a potentiometer between the L and M nodes (KQO "Q" control) but presence or absence of this resistance doesn't affect the output at this time.

The XY mode output shown is the 0 degree and 90 degree outputs, from nodes N and P. One can see there is some evidence of quadrature, and something that looks like a fly back, but something isn't right. Investigation continues.

Wiring error discovered. Two of 3 sections of the circuit appear to be working.

That is a beautiful thing...

Steve

Hey greg, You'll need this too:


Attachment 57733


The symmetry effect is nice & subtle. The Q Pot was a 10K, ten turn linear, and the symmetry was (I think) a dual 10k linear.

Yes, I will need that. Thank you Brian. I agree Steve, that is a beautiful thing... but where's the rest of it? The KQO I remember had a flyback that looked similar to the one seen in the spiral circuit I built in the early 1990s from 2N2222 BJTs and the >1 Henry ferrite cup core from Bell Labs, first mentioned here:

https://www.photonlexicon.com/forums...3-CYGN-B/page8

The FXCY KQO in the photo looks like something perhaps is being over driven, and the flyback curve is hiding in a dot, waiting to unfold from somewhere. Investigation continues.

The retrace you remember was a product of the G-124s scan-set and PDM scan amps. A signal that comes to a screeching halt on an oscilloscope behaves much differently on the dome. Wait until you see the signals after they get past the clippers...

That's good to know, as it means that this circuit is performing to spec and further investigation is unnecessary. The previously posted photo has the signal passed through the clippers, though nothing is being clipped due to pads attenuating the gain from the 0 and 90 degree outputs. The clippers appear to be amplifiers that simply clip at a level close to the supply rails.

I have a memory of it not really showing up on an oscilloscope, but when the KQO signals are run through the dual pot the out of balance signals into the clippers produced a very cool symmetry effect within the body of the KQO image without effecting the outer edge. As the gain of one signal increased and the other decreased as the dial was turned off of 5 - one axis became elliptical within the body of the image. Perhaps it was a tiny phase shift writ large by the G-124/PDM scan amps...

Quote:

---

Originally Posted by laserist

I have a memory of it not really showing up on an oscilloscope, but when the KQO signals are run through the dual pot the out of balance signals into the clippers produced a very cool symmetry effect within the body of the KQO image without effecting the outer edge. As the gain of one signal increased and the other decreased as the dial was turned off of 5 - one axis became elliptical within the body of the image. Perhaps it was a tiny phase shift writ large by the G-124/PDM scan amps...

---

A fortunate gem of a recollection, that. The signal does not look like the image. THAT is the secret of the KQO. I wouldn't have guessed it, but there it is, and a beautiful thing too.

I think this was used in Mars in Starship.

It seems to me that Keyed Resonance Oscillator (since Q is resonance) is more descriptive than Keyed Quadrature Oscillator, as the 6b is full of quadrature oscillators. It even looks like a big letter Q.

This creates a problem in capturing a KQO. Perhaps someone who knows more about GS devices than I do would comment on the feasibility of obtaining a position signal from the unused wires, since these are running in open loop.

I meant GS100PD, not GS120PD.

See attached for a very sniped CX-660 position input schematic and one from a Tropel.

I prefer the one with inductors.

Care and Feeding part one page 10 has how to wire galvos without AGC feedback circuit, tie pin 4 to pin 12 in that case.

Steve

There are two ways to approach this. One is to record the feedback signals from your 100PDs. The other is to build a circuit that makes modern scanners behave like non feedback scanners when you want. You could even play the amount of "ringing" on your dogloids and whatnot.

Quote:

---

Originally Posted by laserist

There are two ways to approach this. One is to record the feedback signals from your 100PDs. The other is to build a circuit that makes modern scanners behave like non feedback scanners when you want. You could even play the amount of "ringing" on your dogloids and whatnot.

---

I hadn't thought of that. Following that idea, a circuit that imposes the desired overshoot, ringing, phase shift, distortion, etc. under user set depth control (like effects pedals), and without moving parts, and passes the signal through as +-10V which can go to existing summing amps. Either of these approaches however, is outside my range of ability to design.

Two op-amps hooked up as differentiators would be a good start, but I'd likely just do it with a micro-controller sampling the inputs and modifying to output. It wouldn't be KQO but you could get 98% there by ramp shaping spirals non symmetrically.