External battery pack to extend the battery life on wireless tags

A game of tag anyone ?

I have installed a network of low cost wireless sensor tags to monitor our homes here in the UK and also in France. These are really useful little devices that measure temperature, humidity and motion such as a door opening. 

They store their readings and can be interrogated via the Internet to see a profile history.   They also have limit alarm trip options that push messages to email to let you know if anything untoward is happening.  The product website is here.

Tag device
Image of a tag device

As examples of their use, here in the UK we have one fitted inside our freezer to check that the temperature is within limits.   Should the freezer fail or the door be left slightly ajar we get a message. 

In France we have one installed inside the spa pump cabinet to check the temperature over winter to check that the pipework does not see freezing temperatures.  This works in conjunction with a greenhouse tubular heater that is also in the spa cabinet wired to a frost stat set to 5C degrees.  If the temperature drops the pipes are protected by the heater coming on.   If the AC mains fail the heater will not come on and the sensor tag warns us to ask the neighbours to visit and check the electric distribution panel as we do experience frequent power outages.

Under normal circumstances the tag internal CR2032 battery will give a reasonable battery life approaching 12 months.  These are not normal circumstances however and due to COVID we have been unable to visit for over 12 months.  As a result all the tags at our house in France have dead batteries.  Their use inside freezers and outside in the cold degrades the battery life achieved. This got me thinking about fitting a larger battery pack to the tags in France to help longer battery life monitoring while we are absent.

My solution is not elegant but functional.   I have wired an external battery pack containing two AA batteries to the tags.   The battery box is a commercial item and is available in packs of five on EBay.  They come fitted with an ON/OFF switch and more than enough cable to connect to the battery contacts inside the tag. 

The tag case can be sprung open with a fine screwdriver or wood chisel and the PCB removed.   The dead battery can be slid out of the PCB battery holder. There are a string of five module probing lands at the foot of the PCB and after some checking I discovered that two of these are connected to the battery holder contacts.   The external battery pack wires can be fed into the case via a 3mm hole in the top edge.   The wires then pass through the old battery holder down to the bottom of the board and are soldered in place.   The tag is re-assembled, batteries fitted and switched on.  The tag will bleep if all is well.

How to wire the external battery pack
Internal view of the tag and how to wire the external battery connections

This electrical modification works but it is physically a bit gangly and scruffy.   A few minutes on Fusion 360 produced a simple 3D printed holder to contain the battery pack and the tag.   This has a slot so the temperature sensor is not obscured and a sprinkling of holes to allow physical mounting as appropriate.

Fusion 360 pictorial view of the tag and battery holder
Fusion 360 pictorial view of the tag and battery holder
Final assembly of the battery pack and the tag
Final assembly of the battery pack and the tag. The dot on the battery pack is a retaining screw.

In theory this modification should dramatically improve the operational life of the tags between battery changes but time will tell whether it is borne out in practice.

Similar or related subjects : –

Tormach PCNC440 Z axis oil manifold access

Turning a Dribble into a Flow

The right hand side oiler for the Z axis on my 440 has always had a poor oil feed to the slideway.   I had tried to clear all the pipework that is easily accessible but with no improvement.   I was conscious that deeply embedded in the spindle assembly was an oil manifold that fed the left and right slideways and the ball screw.   Could this be the source of the problem and how was I going to get at it without a major strip down of the Z axis spindle assembly ?

Looking down from the top of the spindle assembly I could just see the manifold in a hole in the casting.   If I could get to this I might just be able to work on it without a major strip down.

First of all an apology that I do not have any photographs of what follows as I was so engrossed and so grubby that I left my phone secure and out of the way.  The process will become obvious as you progress and is not overtly difficult.

The stripping process was to remove the two screws holding the door interlock switch, remove the four screws holding the wrap round cover, remove the holding pin on the power drawbar and remove the two bolts holding the spindle motor.   With all these fixings and parts freed off it is possible to lift the outer cover up and over out of the way and to lay the spindle motor inside it.   It is tight to shimmy the cover around the power drawbar body flange but it is possible without removing the drawbar piston assembly.

With all this removed it fully reveals and gives access to the cavity containing the manifold.   It is a short term joy because it is pretty much impossible to undo the right hand side oil feed as this is from the end of the manifold.

Light bulb moment – if I drilled two 6mm holes in the assembly casting opposite the two hex head screws that are holding the manifold in place, I could get a hex driver T bar to remove these screws and free the manifold from the casting wall.   This would allow the manifold to be moved to work on it.

Here is a drawing to help locate the two holes giving the distance from the back and bottom of the casting. (The lower black and blue assembly is my fogbuster mounting).

Once the manifold is free to be moved around it is possible to remove the right hand feed pipe and remove the associated length of pipework to clean it.   I found the manifold end of the pipe had not been cleanly cut and was restricting oil flow.

A word of warning – the oil feed to the ball screw is a semi rigid pipe and this terminates in the side of the ball screw.   Do not overstress this length of pipework or there will be tears.

Assembly is the reverse process.   Apologies once again for the lack of supporting pictures.

I can report my right hand side Z axis slideway oil flow is much improved.

Similar or related subjects : –

Having it and not needing it or Needing it and not having it ?

Changing the Drive Belt on a Cowells ME90 Lathe

One of the posters that Jimmy Diresta sells says “I’d rather have it and not need it than need it and not have it”. The saying is apt and often strikes home.  This is not just in terms of larger workshop assets but also in the small scheme of things like workshop tooling.  You know the time you spent making a jig for a job and thought ‘all that extra time and effort to just make that and what do I do with the tooling now ?’

I think it is a saying that is close to the heart of many hobbyist no matter what the medium you are working in.   It does explain why our workshops are full of ‘stuff’ that we accumulate on the ‘just in case’ basis.  How many screwdrivers do we really need ?  The answer of course is ‘one more’.

I believe there should be a sub clause to Jimmy’s poster – “Needing it and Having it yet not being able to Use It”.

I have a Cowells ME90 mini lathe which is a beautiful piece of engineering and I seem to remember it was my first real mechanical engineering purchase.   For 364 days of the year it sits looking forlorn at the back of the bench asking to be valued, loved and used.   When it is called into use it is indispensable.   Usually.  On a recent once in a blue moon 365th day when it and only it could perform a task for me I found the drive belt to the headstock had perished.  You could almost see the grin on the ME90s face.   Gottcha mate, serves you right for not looking after me.

Thankfully the drive belts are standard sewing machine belts (#MB410) and are readily available both direct from Cowells or numerous sources on the Internet including Amazon.   A replacement was ordered and it arrived quite quickly.

Now to the nub of the problem – how to fit the belt ?   Looking at the headstock it suggested that maybe the whole assembly had to be lifted off and split but the cap head screws for this which went down into the baseplate did not want to budge.   I looked at the spindle and it seemed to have differing diameters that at first glance would not allow it to be removed out of the bearing mounts.

Cowells ME90 headstock assembly
Cowells ME90 headstock assembly for reference while following the belt replacement instructions working left to right

Rather than risk a regretful step I emailed Cowells and very quickly got a support reply from Colin.   For all future intrepid belt changers here are his instructions : –

The only way to fit the belt between the 3 step pulleys is to dismantle the headstock assembly.

Its quite simple really:-

Start at the left hand side of the headstock.

1, Unscrew the knurled gear retaining nut.

2, Pull off the 20 tooth gear ( be careful not to lose the tiny Woodruffe key beneath it).

3, Unscrew the round adjuster nut that butts against the large (64t) gear. -You can use a pair of pliers/grips if you put some emery cloth in their jaws.

4, Slacken the M5 grub screw ( or take it out) in the 64t gear.

5, Pull this gear off. (If it is reluctant to budge then, its probably due to a burr underneath- see below for advice).

6, Slacken the M4 grub screw( or remove) in the little collar that abuts the headstock pulley inside the headstock channel.

7, Slacken (or remove) the grub screw in the central vee of the headstock pulley.

8, Slacken the tension on the two bearing adjuster journals- these are the large cap head screws you see on the top face of the headstock body.

9, It should now be possible for the headstock spindle to eject toward the tailstock.

Clean all parts thoroughly and re-assemble in reverse.


If you have trouble removing the 64t gear then, make sure all grub screws are removed as above. Screw back on the knurled gear retaining nut and with a hide mallet, gently tap the headstock spindle toward the tailstock.

As I said in my thank you reply to Colin, I felt like a hybrid version of ‘stupid boy Pike’ and ‘Rodney you plonker’.   (UK sitcom specific joke).

Enough said ?

Similar or related subjects : –

Helping get cable down a duct

A minor post but it might help someone.

I had a need to run an armoured cable down a duct to an outbuilding.   The duct had thoughtfully been installed a long time ago with the potential for use in the future.

The future arrived and it was a daunting cable run of over 25m.   Armoured cable is pretty rigid which helps on the straight runs but when it comes to bends in the duct it had a mind of its own.

After struggling for some time I had the thought that some form of leader was needed to navigate the bends.   After searching the workshop I decided a cable tie (zip tie) might be rigid enough but flexible enough.   I snipped the fastener ratchet block from the tie and taped the residual length to the end of the cable as below.

A liberal coating of DC4 silicon grease and the cable shot down the duct and round the bends like a rat up a drainpipe.

Similar or related subjects : –

Supply, IC2 and USB connections over CAT5 connection

CAT5 Breakout Board

One of our group of ‘silver experimenters’ is building an Arduino based celestial camera tracker.   This will be deployed in the garden and he needed all control to be routed back inside the house.   The garden installation consists of a USB webcam mounted on a servo controlled platform all powered by 12V DC.

We pondered long on how we might remotely connect to the garden.  The crucial thought was that the Arduino servo board was a two wire interface using the I2C format data exchange.   Given that the USB needed four wires and the DC supply two wires we had a need for an eight core cable connection.  It seemed like a length of CAT5 cable would do the job and we could elegantly use standard CAT5 sockets.

The PCB was designed in Design Spark and milled on the Tormach PCNC440 using FlatCAM.

There is a problem with running USB over more than 5m but I did some tests at 10m and all seemed fine which should be adequate for the application.   

The breakout boards had a male and female USB connector fitted and the connections had to ‘cross over’ on one of the breakout boards to maintain continuity.   We also paired the Data + and Data – connections with the +5 and Ground twisted pairs in the CAT5 so the Data + and Data – were not twinned together.

Nothing technically magic but a simple solution to a project need.

CAT5 breakout board for USB, I2C and DC supply
CAT5 breakout boards for USB, I2C and DC supply

  Similar or related subjects : –