Great potential, but disappointing out of the box

FESCH.TV INFORMIERT:

TL;DR: this is my main, but only after quite a bit of work, and I am very disappointed in GAN for releasing it in the state that they did.

Out of the box, this puzzle is unstable and pops very easily, even on the tightest tensions allowed by the numerical GES system. The tip magnets are also extremely strong, to the point that they would have an impact on my solves. Finally, my first copy was missing a magnet in one of the centers, but thankfully I was able to get a replacement. In my opinion, all of these issues are inexcusable, especially at this price point and by a company as highly esteemed as GAN.

Fortunately, I was able to find a solution to the majority of the performance issues that I had. The first thing you need to do is remove the tips, which is not trivial on this puzzle. The best way I found is to insert a plastic razor blade between the tip and the center about in the middle of the piece. Pull out slightly and push up to remove the tip. This requires more force than most people would expect, so don’t worry too much about breaking it. The first modification I did was adding a washer on the center stalk under the nuts. The washers I used were from an MF3RS2, but anything of a similar size will be fine (most post-2015 3x3s made by MoYu and their sub-brands will work, along with various others). This mod effectively tightens the tensions by reducing the amount of travel in the center piece before it hits the top of the nut. It greatly improves the stability and significantly reduces popping, although it does not eliminate pops completely. The other modification I did was removing a magnet from each tip. This proved to be extremely difficult and there is a significant risk of damaging the puzzle if you are not careful enough. The first thing I recommend trying is inserting a thumbtack into the gap in front of the magnet. Try to press it under the magnet and pry it up. It’s important to be careful not to deform the plastic in front of the magnet too much, as doing so can prevent the tip from going back on. If you are unable to get the magnet out in this way, insert the thumbtack in one of the smaller holes around the magnet and gently press it into the plastic with a hammer or something similar. This will expand the hole and loosen the magnet slightly. Go all the way around the magnet and repeat this process in every hole except the larger one in the front, but be careful not to overdo it, because it is possible to go all the way through the plastic piece, which will damage the center under it. After you have gone all the way around, try the first removal method again in the front hole. If you are still unable to get it, try tapping the thumbtack a few times with the hammer to shove it under the magnet. However, the plastic is extremely thin in this area and you will go through it very quickly, so be careful. I recommend the first mod to most people, but only attempt the second one if you are confident in your ability. If I did not have multiple puzzles, allowing me to swap pieces, I would have at least one damaged center, which would make the tips slow and scratchy.

After the mod(s) are complete, there is a very simple yet effective setup for this puzzle. I added Weight 5 on one of the torpedoes, the main surface of one of the centers, and on each tip. If it is too slow, add a bit of DNM-37 on top of the Weight 5. I am currently using green nuts on 0.6, but experiment to find what you prefer.

Doing all of this will result in a fast, smooth Pyraminx with excellent corner cutting. Pops are possible, but they should not be much of an issue unless you have extremely inaccurate turning. While I am not fully used to this puzzle yet, I am confident that it is currently the best option for people who are willing to put in the effort. However, if you would rather have something that requires less work, check out the QiYi MS. While the performance isn’t quite on the same level as the GAN, it is a solid puzzle and doesn’t have any major issues out of the box.

==>