Follow by Email

Tuesday, 31 May 2016

A Pretty Small Hole

I want to drill some small holes in some steel for another project, so I have had a go at some small hole drilling. Actuall a better phrase is small drill bit destruction.

The holes I want are 0.1mm, 0.15mm and 0.2mm. Of these I only had 0.2mm bits so I had a go with them. Several attempts in steel ended up with broken bits after about 2mm.

An attempt in aluminium was successful though, after I had used up all nmy 0.2mm bits and started into my 0.25mm bits.

I used a gcode program that was generated by a Tcl script, this program slowly drills into the material, with the occasional removal for chip clearance.  I think the program is still a bit wrong, as I have broken a lot of bits with it.

Anyway, here's a 0.25mm hole in 2.3mm aluminium:






This is the hole being drilled:


Mini Lathe DRO


I'm looking at building a DRO for the mini lathe and mini mill. The electronics isn't much of a problem, but the mechanical aspect of fixing gauges to the lathe and mill is another matter. I've decided to have a go with the cheap plastic digital calipers that you can get from China for a very cheap price. They have a data output that can be attached to a processor of some kind, which is very useful. The plan at the moment is to have a processor per gauge to handle the data stream and then have a larger master processor that handles the LED displays and does the keypad and clever stuff.

After hacksawing a caliper apart and 3D printing some parts I have the first version of a mounted caliper:






There's a couple of problems with the 3D printed parts, but it all seems to fit and work. It also doesn't seem to interfere with the operation of the compound. I need to put the PCB back on the gauge and wire up to the electronics, but it's looking hopeful. I'm not going to replace the LCD or buttons one the caliper PCB, I'll just use the data output.

From the data output it looks like there may be better resolution than the 0.1mm resolution that the caliper itself displays.


Sunday, 29 May 2016

Sony 7035 Volume Control Replacement
I have been asked to look at a crackly volume control on a Sony 7035 tuner amplifier. This is a nice unit from the 80s or so. The volume control is a ganged potentiometer along with balancecontrols. There's no way to get a replacement after all these years, so the only option is to attempt a repair or replace some parts with something else.

Looking at the circuit diagram, the volume pots are centre tap versions as well, which is used for the LOUDNESS function. Not having managed to find any four gang pots, one option was to use a processor and digital pots. Unfortunately the voltage on the pots is +/- 27V which is too much for the digital pots I coukld find. So, using a dual gabg pot for volume I decided to fix the banalnce to mid position and remove the LOUDNESS function.

I made a PCB on the CNC router


then removed the unwanted copper due to the voltages


Once fitted into the Sony, the volume control works with no crackling.






The original pots had an extended stem which the volume knob attaches to, so I need to machine an extender for the shorter pot I fitted. I hope the extender is solid enough once the knob is attached and that it won't wobble about too much.

The balance is mid position, the LOUDNESS and MUTE don't work, and I fitted a linear pot as the log pot I fitted didn't feel very good. I'll see how this solution works, but a second version may be in order.

I'm wondering if I can put a processor on the PCB and use FETs to effectively create a high voltage set of digital pots. One problem with this is that the processor may cause noise due to switching which the radio may pick up, as it is quite sensitive and will be in the same box.

New Toy Bits

I was asked ot make some missing parts for a 'square peg in round hole' type of toy.


The pentagon and circle pieces were missing, so I made some. The circle was done on the wood lathe, which was a simple choice. Doing the pentagon was harder, but I created another gcode tool and did it on the rotary axis of the CNC machine. It came out quite well. Just needs painting.


Tuesday, 24 May 2016

Freebie Multimeter

When I bought my new Rigol scope, it came with an unexpected extra:





Apparently it was a special offer. It's not an expensive meter, but has a couple of nice features (a backlit display and a 50Hz output) and it's always handy to have another multimeter.

What it didn't come with was a case to put it in.

Fortunately we got a new sewing machine at Christmas, and this was a perfect opportunity to try it out. I ordered some waterproof outdoors type material a few days ago as a trial run for making bags and things, so I cut some of that up and a fw hours later we had a case:


Saturday, 14 May 2016

Pocket Watch Winding Stem

The first stem attempt for the pocket watch was handmade and it wasn't easy to do. It also didn't work very well. So I decided to use the CNC machine to make the next stem. Starting with 2.5mm silver steel rod (2.5mm is the largest dimension of the spindle) the machine gradually turns the profile of the spindle using a 0.2mm D bit.

The setup is able to machine to less than 0.1mm, but the initial attempt failed to give a good profile due to flex of the rod when the tool lowered in for a cut. After adding a wooden block under the rod and running the machine again, a good profile was machined.





The broken stem is on the left, the new one on the right.

The dimensions aren't absolutely correct, but it did work, after the sloping edge was added by hand. The square profile was also added by hand but was filed too small and eventually the corners were rounded off and the stem stopped being able to wind the spring.

I'm going to do a third attempt, with better dimensions and the second slope machined rather than done by hand.

The G code for the machining was done using a custom Tcl script that reads a DXF profile and outputs X-A axis G code to machine the same profile around a cylinder. This can't do the square end, but I'm going to create a new program for square pegs. That will be run as a second stage of machining after the main profile. Creating these scripts as separate tools should hopefully make them more reusable in the future.