CYGN-A

[laserist]

Helipot 1k sine cosine



Bourns version



And I can't get rid of this on my phone...

[Greg]

Those look so WWII, or maybe Apollo era. I wonder what they were expected to be used for.

I agree, swamidog, the KQO can do some really magnificent convolutions. My emotions are still resonating from what I saw when Brian insisted I implement the symmetry exactly as given in the plan. I've wondered for decades if it was my imagination, or did the KQO really have compressed loops around the outer edge, and if so, how on earth was it done. Then suddenly there is a big fat satisfying answer: Yes it does, and it involves some hugely analog magneto-mechanical distortion of a highly clipped signal. Absolutely awesome. Everyone likes spirographics, but the KQO has these unique asymmetries that make more like a character than a mere cycloid.

Interestingly / disturbingly, I have found that the original JFET components called for make no difference in circuit operation when substituted for a similar component I had on hand. BUT the circuit continues to function identically when one of the two JFETs is pulled out of the bread board all together! But the circuit ceases operation if that component is installed with the leads improperly matched. Go figure.

[laserist]

Everyone has seen the old circular CRT radar displays in movies and tv. If you had a sin/cos pot conected to the antenna's azimuth you only have to multiply the x/y signals from the pot with the distance signal to draw the radial line on the CRT and use the return amplitude to control brightness. Everything changes when you go digital, and changes again when you add computers.

[DZ]

Regarding sine/cosine potentiometers. In my opinion, the Elweco Hallpot Resolver is the perfect solution. Sadly, I was made aware a couple months ago that the guy making them seems to not be around anymore, which makes the 2 I have left on the shelf worth their weight in gold. I posted about them on PL a few years ago. The guy who made them said he's run them up to 10K RPM and still had a perfect sine/cosine.

[mixedgas]

Elweco is making a last batch for a client, then looking for a new owner for the product. I spoke to him last week.

Steve

[laserist]

I admit sin/cos pots are at best novelties for laser shows. A joystick had more depth, but if you had a joystick or two - a weighted knob on a sin/cos pot could be spun and the slow decrease in rotation speed was an effect in itself. And they're pretty bullet proof. 25m to 50m rotations was a common spec.

[Greg]

Everyone has seen the old circular CRT radar displays in movies and tv. If you had a sin/cos pot conected to the antenna's azimuth you only have to multiply the x/y signals from the pot with the distance signal to draw the radial line on the CRT and use the return amplitude to control brightness. Everything changes when you go digital, and changes again when you add computers.

I've never encountered such an explanation of radar. From this perspective, it all makes sense. From what I read of the history of analog multipliers, they are a non trivial thing to design. I wonder how the earliest ones used in radar worked.

The photo shows the diamond spirals. The board has the ramp generator from the CYGN-A / B, and also produces an inverted ramp. Ramps are variably attenuated and summed with a variable offset. Sweep will be summed following the SPGN AM function.

If my understanding is correct, the SPGN, due to it using 2QMM units rather than 4QMM units meant it used a monopolar ramp. The current build using the AD633 accepts a bipolar ramp and thus, manually sweeping the ramp offset through zero gives a worthy effect. I'm not sure if the original did this.

[laserist]

The spiral generator was unipolar. Early spiral generators used 533 4QMMs, but the spiral ramp was always unipolar. The reason is the ramp generator could be reset "bam" to zero for the ramp and 10 for the inverse ramp. It was a capability that was used often. The SPGN board layout was pretty full and modifications that damaged prior choreography were worth avoiding.

[laserist]

Your diamond spiral has the same gain for each axis. Usually in the shows I did the x or y 351 master gain would be lower making one axis longer than the other. BTW I liked your spiral popcorn, but the spirals were before where the dog signals were summed into the image path.

[Greg]

I'm open to suggestions regarding adding a reset line to the colormodII ramp circuit.

From what I see, the ramp reset signal enters through the control circuitry for the BB4302. I have this circuitry ready for the chip should one appear, but even if it does, an equivalent should be developed that doesn't involve unobtainium. Steve previously provided some excerpts from an interesting looking tome that may bear on this topic.