Importantly it should give a much smoother surface and locally work harden the brass.
Should be able to see the difference between the standard and the hammered one, the top one is standard
Well the balls turned up yesterday and today I have been swapping them over, half way through it and a few pics of my work so far.
This thread has been a great read and thought provoking for the tweakers looking for general improvements or low cost upgrades.
Could we get some summaries of best results and reviews on SQ improvements?
Mana works well. Don’t know why
@Dozey, I think you posted this on the wrong thread; you want the ‘Equipment Platforms’ thread:
This thread here is specifically about vibration control using ceramic and metal supports, particularly DIY jobbies.
Mana uses metal spikes to support glass. Seems in keeping with the thread title to me.
I made my own cups out of 1 inch diameter CARBIDE inserts that I use at work (machinist) . I bought the balls at a local bearing shop, I believe they were less than a dollar each. Carbide is harder than steel, then the glass sits on the steel just like Naim do it. I have not done any A-B testing, but I am confident they work. The ball is only contacting my homemade cups on three points, just like Naim do it. image|666x500
I had to use a wire EDM machine to mill out the triangle shape in the middle of the carbide insert.
I take it the carbide is tungsten carbide.
Quite impressive being able to ‘cut’ tungsten carbide; doing it mechanically rather than using spark plasma erosion would be ‘challenging’ to say the least! Incidentally, I was surprised at it’s electrical conductivity, I’d have expected it to be almost purely covalently bonded.
The one trouble with that design is that it doesn’t have quite such a difference of sound velocities across the interfaces, instead it relies on the incredibly small contact patches between the carbide and the steel ball to limit energy transmission (I assume the balls are a tough/hard bearing steel). This is a different approach to my design.
The reason for my design specifying brass and silicon nitride (or silicon carbide) is to both reduce the contact patch AND maximise the gradient of sound velocity at the interface, so increasing the reflection of sound and so reducing the transmission.
Yes it is tungsten carbide, we use these in a facemill to cut steel. The wire EDM machine can cut through almost anything, and is very accurate. Originally there was a round hole in the center, for a screw to hold it in place on the facemill. The 3 points of contact is what Naim use on their cups, so I cut the Center out the same way. I could of cut any shape I wanted really, once I had the program made, I just stuck the next piece on the machine and hit cycle start. I made 21 pieces in about 5 hours of machine time.
I took I photo of what the carbide insert looks like before I start, it is a round hole with counterbore for a flat head screw. I have a few spares in a drawer.
Well they’re certainly not going to get worn out in a hurry!
Very cool, nice to have access to such equipment.
I have wondered how Naim mass produce the ‘rounded triangle’ hole in theirs on an economic basis? I assume that Naim use stainless steel for their ‘polos’ :0)
Well if Naim’s cups are made of steel, they can be machined on a CNC machine very quickly. They would use carbide cutters, and have some kind of fixture to hold them in place. Carbide cutters will not cut the pieces I made, since they are carbide themselves.
Indeed, I have seen EDM in action at the local Rolls Royce aero engines plant. They were very blase about it!
Laser cut?
Back in the day, and if we needed to make a lot of em, I think that our workshop would have used metal working broaches
Just had a quick look at my Naim ‘Polos’ the triangular ish holes look like they have been cut using a CNC milling machine.
I did notice that the edges on which the balls impinge are starting to indent, that is the knife edge is not what it was. Tungsten carbide isn’t going to suffer that problem, although it might be more prone to chipping.