What's new
What's new

Can the Precision Matthews PM-728VT CNC kit able to do micro machining?

Hi Donkey Hotey:
You're right...I don't.
(Barring the fact that I don't actually care what gender the OP turns out to be), if that's the case I've wasted my time...
But I'm going to assume it's a "he" and that "he" is serious and really wants to know how to go about tackling "his" goal, and can "he" do it credibly on that low end machine "he's" contemplating.


OP, note the pause after each pass in the video, allowing the tool and workpiece to cool down under the mist. Also note that the saw is conventional and not climb cutting. Nice to have an example so you don't have to discover these methods the hard way. (Check out the shattered saw in one of the pictures you posted!)
Last edited:
I was recently reading some papers on heatsink design (I wanted a passive cooler for a project). There is quite a lot of research out there on this stuff and the gaps/heights are not that obvious. There are a few calculators to design optimised cooling for various designs and some of those are free. Some of the factors are totally unrelated to the fins themselves as well. Interesting stuff. Just google a bit for the various papers (lots are referencing air transfer, but I think you can extrapolate to liquid cooling as well as it's a parameter)

Marc (the German guy) does FEM simulations on the computer to make efficient professional water block designs (I wonder if he is using a RTX 3xxx/4xxx GPU to aid in that and if its water cooled using his own waterblock designs):

You found calculators for air fins or for water-cooling, care to share the "free" ones?

I also found Marc's website where he sells his water cooled GPU water blocks, seems like top-notch stuff better than the popular stuff the new kids buys:

He is discussing his most recent deisgns and implementation on this forum:

Really interesting stuff going down there if anyone into advanced GPU/CPU water block designs.

Seems like I'll need to wait until I can do these cool things, since I changed my plans on buying a bench mill and would rather buy a used bridgeport knee mill, refurbish it, then use the BP knee mill to help me make my own precision CNC, as we all know some parts needs to be made on a precision mill/CNC to make another precision CNC.
Last edited:
Back to making heat sinks on a CNC mill
Guess If doing that on a light duty mill I would set 4 vises on the table and program the machining toward long travel. that would likely be the most ridged method.
The weight of the vises would help and I might even add a steel plate or something to add more weight to the table. likely have to make a gang saw arbor. Might even put the out end of the gang saw arbor in an out-board bearing.
The vises might be set @ 90* so the gang arbor would be vertical.

Possible a used CNC Bridgeport or the like would do that work well.
An OK used cnc mill can be bought for about that same price
Last edited:
Hi All:
I understand the desire to help the OP to the best outcome we can, and part of that is to offer him design alternatives to the decisions he's made.
But what he's hoping for is advice on making the design he's already chosen, not to find the best possible design.

Best possible design? The best design is micro fins, if you saw my original post, it's about micro fins that I want to make.
However you are not, nor I'm I real professionals in micro fin design. However an IT guy like me knows one thing for sure, you guys knows how it's made. So I took my notes that it's really simple and use a sitting saw.

I have no intensions to create new cutting edge (no pun-intended) micro fin designs, that is going out from my expertise and getting into a boring process of advanced thermodynamic engineering which requires good mechanical + material science engineering knowledge, sure I can dig into this but it's not my cup of tea of trill and enjoyment.
I'll only get maybe 5 degree Fahrenheit performance in better micro fin designs, honestly not worth it, I'll just stick whats proven to be best now.

So this micro-fin tech has reached its best limits of performance, right now engineers needs to get creative in how to use this tech more effectively and creatively. This is where IT guys like Marc and myself have some interesting ideas we believe that would be more effective, this is why odd people like me are hanging around here, to take some notes and info.

A corollary to that is that we are invoking processes he couldn't possibly get near for the capital he hopes to spend.

So if we get back to what the OP intends:
We've decided a slitting saw is going to work better than an endmill for what he's trying to make.
I've invoked the possibility of picking an alloy that's easiest to machine.
We've decided how this thing should be fixtured...standing up on a plate.
The OP has found someone who's shown it is possible on a cheapo crap machine, so that's put to bed too.

No, spending the $1 million capital with no clients and just to make prototypes makes no sense what so ever, even investors from Goldman Sachs will say that. I assume you have no idea how to business coming up with that logic.

Bill gates, Steve Jobs and Jeff Bezos all started low, cheaply, low capital and slowly. They eventually got clients and then big time investors.
Please my friend, learn how to do business before making terrible false conjectures on capital requirement for new entrepreneurs trying to introduce new products requiring high precision machines.

Some of the old and wise on this forum would say to buy used old ass precision machines instead of the latest and greatest new machines where you can play Tetris on the display monitor, to start getting into the game.

How much or little money he's going to make, how ideal the design is or not, how elaborate his fixturing has to be to make a million a month...it's all kinda irrelevant for this conversation IMO.

BTW I do business and have business license as any business professional, during COVID I was scalping on GPUs making more than double the profit. I bought them for $400 and sold them for $1,000 (without tax and shipping). (As again this is just the GPU and not an entire PC computer, please keep that in mind).

You might ask me how in the world did I got a CONSITENT supply of multiple Nvidia RTX 3060 TI GPUs during the prime of COVID while folks like you couldn't get any from BestBuy but had no choice to buy it from scalpers like me 🤣.
Hint: Be a professional in the business... know other professionals in the industry (I got them directly from Nvidia's authorized distributors where they only do business with licensed businesses and not directly with consumers).

Talk about profit, you didn't even paid attention to my thread but rather criticize me.
So let me make things clear again:

I'm not planning to sell a $300 GPU water block alone (I might if customer asks), this is where little money is to be made.
There are too much competition and big players making thousands of water blocks, customers are overwhelmed by the choices they have.
As you smartly assume there is indeed little money to be made here, however never assume anything, Marc is doing this over a decade, he is probably making a pretty penny, but watching him on YouTube still rocking the cheapo handy BF30 says otherwise... you may be right on this because the return of revenue by now means he should be rocking some quarter million dollar swiss made machine aiding in making GPU water blocks. I assume he does this as a hobby and not professionally.

BTW I have seen a YouTuber which goes by "BitTech" who has IT background and Precision machining background. Makes custom water cooled PCs. He started using a cheap CNC less than $7K (German made) and got shit ton of rich ass customers, made excellent revenue and now playing with the big boy CNC machines which you guys brag about.

I plan on making big money by selling products for at least $3,000+ in making entire custom made gaming/server/research/database/enterprise grade computer units, professionally water cooled where the water blocks are 100% engineered and made inhouse.
Not many people/companies are doing this because it's complicated and need to know the market of each field, it's a specialized business which I know will work due to my experience in computer science and computer engineering business.

Please tell me how many people/companies you know in America (USA) that makes professional computers (yes, the entire PC) where the CPU + GPU + Motherboard + PSU + Case are all 100% water cooled using 100% custom made water blocks for the consumer level?
Trust me no body, I guess I'll be the first one, thats why you got a computer guy here dicking around in precision machining.
Space/defense tech companies does this, but not for the consumer level, nor do they provide services for consumers to make 100% custom made water cooled PCs, they don't have time for that business.

Ahh... yes... if you're catching along you might be getting the "big picture" where the money of all of this is coming from, right?...
You have that filthy rich ass motherfuckers who would shit on you with pile of cash for that sick ass one-of-kind water cooled master piece.
If you can't find the clients or come up with innovative ideas which customers will buy, continue to work for your boss 9 to 5 at the machine shop until something comes up in the mind.

I can look at a simple garbage can and come up with new ideas what will be hella profitable selling better garbage cans (I have plans to sell smart garbage cans which can throw garbage bags away and then re-bag it self, I actually spent few weeks engineering one in CAD).

This sounds like voodoo to you because how many precision machinists also have an IT/engineering + business background?
Answer is none, because IT/engineers gets paid way more than precision machinist and precision machining is much easier than doing calculus or advanced computer science programming.

So there is a big ass gap in this special business field I am first creating/getting into because you need to have three backgrounds:
1) Business (Sell products that are innovative or solving a problem which no one provides and make profit from it).
2) Computer engineering (Understand how computers are created and made, therefore you know your clients and what they like to play with).
3) Computer science (Yes, I'll be making my own advanced software to cater my products professionally, this could be a big factor why customers will buy your product).

What happens when an IT/engineer starts learning precision machining?... lots of trippy cool products comes into this market.
Don't believe me? You think I'm talking out of my ass?

Real life example:
"Optimus" GPU water blocks, if I remember correctly the CEO is an engineer and made products for the aerospace industry, left that industry and got into the IT industry, starts selling CNCed water blocks at the consumer level for PC gamers at $50 a piece. Rest is history...

If you're still not catching along, I plan on selling products (complete working computers) for no less than $3K... I hope it makes more sense why this is such a big deal to "me".
I could practically get 5 customers for each month, that will be $3K x 5 customers = $15,000 revenue... Man I'm going to make so little money LMFAO 🤣🤣🤣🤣

So it's IT guys likes me who accidently learned about precision machining, I never thought in my life there are dudes at the bottom of the world with dirty hands holding "1-tenths" all day on a machine called a "Bridgeport" which people also calls a "knee mill" made since the 70's till this day. At first it didn't make sense to me how precise the machinist slang term "1-tenth" was or what it even meant until I found out it's close to 25 microns. I'm immensely hungry for new information (I first googled how IT semiconductor machines are made in photo-lithography, to my surprise I jumped into the rabbit hole to learn more, then I started to learn about precision motion control A.K.A "CNC", then started to figure out that CNCs are irrelevant, the real important things are precision granite surface plate, precision squares, precision calipers and all precision metrology tools are the most crucial things needed to help make precision CNCs/machines but thats not all, the skilled precision machinists needs to be your best friend to help you know how to use these metrology tools.) and found precision machining a work of art, great importance of advanced development because of it's foundation to make IT stuff (Intel/AMD/Nvidia..etc).
Last edited:
Nice to have an example so you don't have to discover these methods the hard way.

I'm extremely ASHAMED that Marc was doing this way back 10 years ago... because I always wanted to make these but didn't realize you could make em with CNCs.

Yes, Marc figured it all out. But I got many great hints on this forum well before finding his videos, everyone said to use sitting saws right at the get go.
Last edited:
Possible a used CNC Bridgeport or the like would do that work well.
An OK used cnc mill can be bought for about that same price

But now that I got my eyes on a Moore Jig Grinder Model #3, can't I use this machine hooked with an arbor and a micro sitting saw to do this operation? This machine can hold tolerances down to 2 microns or so... and will take very slight light passes.

Best of all the flat surface on the water block where the main GPU die chip will directly rest/sit upon on needs to be precision surface ground for the metal thermal paste to be effective.

GPU water block.png

GPU die chip:

RTX 3080 die chip.png

So am I not better off with the Moore jig grinder model #3 for this task?
Basically it can do many tasks all in one machine, I think of it like an all-in-one laser jet printer if it makes sense.
I can do the sitting saw cutting operation + the surface grinding to make these small and intricate details.
Last edited: