hifi28 - good memory, with of course, a back-story for fun.
The crossover modification is at SN4901, May 1987, two years in. I don’t remember the exact change - I have only the revised layout from which I have derived the schematic. But I do remember the back-story.
Let’s start with the end: the modification was very slight, making a change in the roll-out of the upper midrange. Generally Jim revisited designs to tweak crossovers to accommodate driver tendencies and burn-in factors that become apparent as time passes. Or if a driver gets a production change (such as adhesive, etc.), slight engineering updates may correct it. The CS2 was uniquely a completely stable product, having no such driver changes and requiring no real tweaks. Jim was pleased that he had gotten it right the first time out.
Enter Larry Archibald, the fairly new publisher of Stereophile Magazine. Larry was supportive of our products, but always had the last word. He had written critically about the CS2’s upper midrange ’edge’ in various ways. It turned out, learned behind the curtain, that all of his listening to the CS2 had been without their (required) grilles because "he doesn’t listen with grilles". On a second round of review by one of his staff, Larry heard the CS2 with grilles and declared that "the problem had been fixed". Now enter Kathy Gornik, our co-founding marketing director whose job included press and public relations. Being an essentially political animal, she deemed fit to give Larry an ’out’ - by Jim ’fixing’ the crossover that Larry could point to as the remedy. Jim’s desire to have a product that needed no changes through its life took back seat to Kathy giving Larry permission to praise the upgraded performance without losing any face.
I don’t know whether it would be possible to reconstruct who did what when, the exact change and so forth. I do not remember changing our archived CS2 reference samples.
Bottom line: a small crossover change was made at serial number 4901 as well as adding the viscous rubber strips to mechanically unify the grille with the baffle. If anyone has access to an early CS2 and wants to send an XO photo noting component values, I’ll be glad to document the change. If anyone updates their CS2 crossovers, I’ll use the revised values. But the big deal is to use the grilles which include a tweeter wave guide and full baffle diffraction control which considerably upgrade performance.