After having used the low profile vise for a little over a year, I can say that it’s great for general purpose work but slow for precision work. For woodworking it’s fine to place it with a speed square. For high precision work though it has two disadvantages: (a) it needs to be dialed in with an indicator each time and (b) due to the M10 screw backlash it cannot provide a repeatable Y origin; a new Y probing is required with each operation.
Time to make a secondary vise. Requirements:
- lower profile
- stronger
- one fixed side
- a set-and-forget origin
The initial idea was that by using the table slots the components do not need to be connected by rods. After sketching a couple of variants I ran across a design by Rob Lockwood.
https://www.instagram.com/p/BYM994IAXtb/?taken-by=rob.awesome.lockwood
That was it. Using low profile edge clamps would offer a minimal and very sturdy solution.
Here’s the old one helping make the new one.
Btw, this is a recent discovery, a 1/4” shaft fly cutter. I was surprised to find one that small. Pretty sure they call it a fly cutter cause the chips it cuts fly everywhere. Unless of course you have an UMCS.
Behold the UMCS. One plastic folder’s cover cut in two, attached with t-nuts and screws to the Z extrusions, the two pieces held together with split pins, pleats at the bottom for jumping over obstacles.
Pockets.
Profiles.
Set and forget. (I like how the reflection from this angle makes it look like transparent)
Cutting the step on both fixed and moving sides.
Now this is what this is all about. Cutting the step in place grants two things: it’s perfectly square and at a known location. It’s taking its own WCS with a persistent setting (G10 L20 P6 X0 Y0). In the rare event that something happens that cancels this setting (e.g. homing switch replacement), it’s easy to reset after cutting a fresh sliver.
Hello G59.
Edge clamps in place.
720 lbs of clamping force on some test stock.
4 Likes
I missed how its tightened. But am still drooling. 
Excellent post! 
You place the stock on the step of the fixed part, then slide the moving part to hold the stock from the other side, then fix the moving part to the table. At this point both parts are holding the stock but loosely.
Then tighten the little clamps with a torque key. With tightening, the clamps shift sideways. Some clamps achieve this with eccentric screw heads. These particular ones achieve this with a conical slot.
Since the UNC countersunk screws have an 82 deg angle, by moving the screw head down by d the clamp moves sideways by d*tan(41) .
3 Likes
Are the plans for it electronic anywhere? Purchased?
1 Like
All dimensions are custom to my setup. It will be faster and less confusing to draw it from scratch. Here’s all you need:
2 Likes
Nice idea, did you not like the red clamps shown above?
Where did you source those clamp pieces with the conical slots?
Oh, I like my red clamps and my step clamps and my fixture clamps and my self-centering vise. I find workholding the most fascinating phase of the whole from-idea-to-part chain due to the endless challenges and the problem solving and the thinking ahead. Some are stronger some are less, some allow facing some don’t, some allow profiling some don’t, some provide a fixed origin some don’t, some work best with flat stock, some work better with block stock. There’s not a single solution for everything. You can’t ever have too many workholding tools 
The new custom vise can be used for facing, deep profiling and is my fastest option for fixed origin.
The link was posted a couple of messages higher, but let’s include it here as well for reference along with some other options:
Carr Lane Tiny Vise (low profile)
Carr Lane Tiny Vise (regular)
Mitee Bite Pitbull Clamps
Mitee Bite Fixture Clamps
1 Like
Getting ready to tame my spaghetti of electronics.
Test fit.
6th side will be the power supply.
4 Likes
Top. Outlet fan, relay switches, emergency stop. I looked for a 120mm computer fan, there’s a ton of options in this type. This particular one comes in four flavors. I picked the one with the highest flow:noise ratio. Also added a rubber frame underneath to eliminate any chance of vibration.
Right. Inlet air filter.
Front. Motor power supply, probe port, probe parking (more on that later).
Back. Main switch, relay AC sockets, USB port, motor ports, homing switch ports, VFD port (0-10V & enable).
Standoffs on the controller board for clearance for the solder joints and also for passing some cables underneath.
Thermal compound on the relay heat sinks.
Bottom. Secondary power supply, relays, controller. This particular PS is a triple output: 24V for the controller, 12V for the fan, 5V for the homing switches. The controller can take anything between 12-24V but I found that the PWM to 0-10V conversion gives better results at >12V (better as <0.5% deviation between commanded and real voltage throughout the whole 0-10V range).
2 Likes
The controller is called Phoenix CNC Controller. It’s a driverless ATmega328P board. Here’s its Tindie link.
Instead of writing why this and how I feel about it, I’ll just copy paste my review.
I was planning to build a new GRBL-based controller for a recent CNC upgrade. I would normally need an Arduino Uno, a screw shield and a noise filtering circuit. To be able to add a VFD in the future I would also need a PWM to 0-10V circuit and a spindle enable relay. The Phoenix controller contains all five components in one sleek small factor package.
I remember seeing the Phoenix in the past and after revisiting recently I was pleasantly surprised to see that the new SMD, smaller footprint version was released. The timing was perfect. In my case shipping was intra-EU, so super fast and without customs.
The PCB, assembling and soldering is of the highest quality. This is a low production item and it shows. You can also tell from his youtube channel that Hayri is not only knowledgeable, he is down to earth and a genuinely nice person. He was always there to support when I needed to ask or confirm something.
This has been my second Tindie order and I don’t know if I got lucky, but both cases have been gems. People say Tindie is the Etsy of electronics and this is true in the sense that in high volume marketplaces there is typically a seller-to-product relationship. Here it seems more like a father-to-child relationship. People devote time and care and take back pride. That’s exactly the case with Hayri and the Phoenix controller.
6 Likes
To settle on final cable lengths the four faces (front-bottom-back-top) are interconnected but flattened. Every now and then, when new cables cross two faces, a fold check takes place to make sure they bend ok. Ideally the box should have been wider to allow for better cable routing and management. I will have to live with that though even if it feels a bit crammed, cause there’s simply no bigger space to fit the controller.
Bonus tip. The motor power supply uses a Molex Mini-Fit Jr. connector. I had made one a while ago, but also found a cheap PC cable (EPS12 extension) that uses the same connector and thought to give it a try. Factory crimping should be better that mine.
I only need 3 out of the 8 cables, but don’t wanna buy an extractor just for one time. I used a couple of staples instead.
Bent in L shape, pushed on the sides of the unwanted pins to unhook the pin tabs from the connector body. Then the pin can be released by lightly pulling the cable.
1 Like
Assembled. Front-left.
Back-right.
Back. One thing I forgot to mention. Using a non-conductive material for the case, I needed another way to ground the connectors (and therefore the cable shielding). I had plenty of adhesive copper tape left from grounding the dust deputy a while ago. It can be seen close to the left edge. The USB pigtail is also grounded.
Probe parking. The cable on the other side connects to the ground of the speed control circuit for the VFD. When the probe crocodile (i.e. probe ground) is parked the speed control circuit is closed. When not (i.e. clipped on the bit) it’s open. I can see myself hitting resume after probing without removing the crocodile first. This is no problem with the Makita, as there is a physical switch, but imagine what would happen with a VFD when spindle speed is controlled by the gcode. This is essentially to protect myself from my idiot self.
Connected.
Stickers.
Disco.
7 Likes
everything you do is so well thought out.
Respect man!
@anon68752607 Thanks. I’m still learning.
@AngusMcleod Main reason is aesthetics. I just like how I can see the internals. It’s a little more sci-fi 
There are some other minor, but practical reasons.
It was very fast to cut. With aluminium it would take more operations to cut. I’d probably have to do a lot of chamfering to prevent cuts and curses.
Then a transparent material makes aligning parts to holes a little bit easier.
Transparency also allows for visual debugging. There are leds on the relays and on most outputs of the controller.
1 Like
This domain seems expired. If you read further down from that point there are some more references. And you can of course google for more.
Yes, it’s a C31/40–D–DBS–P–ER20 .
Why Teknomotor?
I did not want a random no-name spindle. I wanted something accessible but also reliable. These are precision made in northern Italy, carry high quality bearings, are tested for run-out and come with a balanced collet nut. In case something goes wrong there are spare parts available. In case there’s a warranty issue I know who to talk with. I had a long discussion with Kevin Damen of damencnc.com about options. He also happens to regularly visit Teknomotor’s factory and knows these spindles inside out.
Why 0.73kW (1HP)?
This particular spindle weighs 4.3kg (9.5lb). The machine runs on 15mm rails, it’s not a VMC. 0.73kw translates to almost 100cm^3/min MRR in aluminum and more than 500cm^3/min in wood. So far I have cut 7cm^3/min and 50cm^3/min respectively. That’s more than 10x room to grow. I’m happy with that.
But wasn’t the Makita the same power?
In paper, yes. In reality, no. An AC trim router vs a 3 phase spindle is not a fair comparison. The spindle watts are output, its real capacity at the bit. The Makita is less than half at the bit. Maybe a bit more at peak. The spindle watts are sustained.
Here’s a hint about the efficiency difference:
That’s an ER20 (i.e. max 13mm or 1/2") on the left. Another two substantial benefits are the lower RPM and of course speed control via g-code.
1 Like
Indeed, 1HP is pleeenty for the kind of work we do!
Nice piece of kit, the square body is also very convenient 
I know you dont buy stuff without doing the research, thanks for the write-up.
What is left to upgrade, season 5?
Hiwins - check
Servos - check
Spindle - check
ATC?
