Activity – Musing on Stuff

Home-made Internet-controlled Irrigation Valves

July 14, 2020July 14, 2020 firmspear_adminLeave a comment

“Timer-based Irrigation Valves with Internet Control.”

This is the latest variant of a project I started last year which used a Raspberry Pi and Homekit bridge software to control greenhouse irrigation. The solution worked well, but the software was a clever adaptation of a Java-based HomeKit bridge developed by a colleague. In essence – way above my technical pay-grade.

This year I fancied adding another bank of valves, so looked around for a means to do this myself!

I happened upon the Sonoff 4CH Pro R2. This nifty device basically has 4 independently controllable latching relays, Wifi connectivity, phone app control and can be AC Mains or 5-12V DC powered.

Of particularly note about the Sonoff range is that they are based on the ESP2866 chip and that there is a strong community behind an alternative firmware called Tasmota and the Sonoff devices either directly support the ‘DIY’ flashing of the firmware or they are pretty hackable – depending on the device.

Tasmota provides extensive support for controlling the devices over various protocols including MQTT and offers great integration capabilities to Home Automation solutions such as OpenHAB which has been ‘on the list’ of things to do…

Step 1 though, as described in this post, was to get the valves up and running using the Sonoff app.

The Bill of Materials

This is not a particular cheap build – but something with the same capability available commercially such as the Gardena Watering Valve Box V3, with three solenoid valves would set you back about GBP 200-250 – without timer or connectivity.

A commercial alternative: Gardena enclosure and valves

This is what I used:

Sonoff 4Ch Pro R2 as above- the Pro is needed as it has the 12DC input (as well as 433 radio control as an added extra..)

Some Cat 5 cable for connecting the Sonoff enclosure to Valve enclosure

2 enclosures – one for the Sonoff and one for the Valve enclosure. Note: these boxes are IP55, which is ‘low pressure jets from all angles’. If you want something more robust weather-wise, go for a IPx6. (The enclosure for the Sonoff is *just* too small internally because of the screw mounts. It means trimming the corners of the Sonoff – it does have mounting lugs so space to do this but if you don’t want to do that, source a bigger enclosure)

3 12V Solenoid Valves – these seem great value at under GBP 14 each

4 Garden Hose Tap adapters – best to get as good quality as you can. Mine were pretty budget.

Plastic piping: 1 x BLK PP Plug Thread 1/2 (SKU 350390), 3 x BLK PP Tee Thread 1/2 (SKU 350268), 3 x BLK PP Nipple Thread 1/2 (SKU 350299), 4 x BLK PP Reducing Socket M x F Thread 3/4″ x 1/2 (SKU 350341) – all from here. (Pipestock’s postage charge is quite chunky but the service and speed of delivery is very good).

An old 12VDC 2A wall wart I had kicking around (mine had a 90 degree connector – this may be important for clearance in the enclosure if you go for the same enclosure I did)

A rough total cost including shipping, where applicable, would be about GBP 110-115.

The Build

The ‘plumbing’ part is pretty straight-forward.

The approach I found to work well was to create three sub-assemblies of the Solenoids pairing each with a T-connector and reducing connector:

Your valves are likely to be ‘one-way’ – you may see a flow arrow on the underside of the solenoid or check the spec – the site I got mine from had the spec in PDF form. Make sure everything is (water) tight and use teflon tape if so inclined.

When the valves are joined you will not find it easy to tighten the sub-assemblies further.

Now join the sub-assemblies with the parts curiously called the threaded nipples, together with the threaded plug one end and another reducing part the other:

Now would be a good time to do a leak test:

Worth running at as much pressure as you can to test

Drilling the enclosure requires some skilful measurement – it’s tempting to use the cut-outs as a guide but they are not an exact spacing in relation to the valve assembly – or at least in my combination that is what I found. Use one cut-out as the starting point and measure the other two. Measure for the inlet and drill. (I used a flat wood bit because that’s what I had. You have the choice of using a 22mm in which case you may need to sand out the cutouts by a half mill or so or use a 25mm and accept a you will have a ‘gap’ around the valve threads on exit. I did the ‘hard’ or rather time-consuming way – using a 22mm drill bit).

You should have something like this:

You now need to ‘wrangle’ the solenoid valve assembly inside the enclosure (having first removed the hose tap connectors you may have used for the leak test). Fortunately, there is a lot of flex in the enclosure (at least the one I used). With the hose connectors attached the valve assembly is ‘secured’ within the enclosure. You could provide additional support to the assembly if you wish. It seems fine without….

Now for the wiring. I spade-terminated a length of cable for the +ve supply of each solenoid and ran that to a connector strip. The -ve connection I just daisy-chained to each solenoid and also back into the connector strip.

Then whatever length of cat5 you want (from one enclosure to the other). I doubled the twisted pairs so three pairs are supply to respective solenoids and the forth pair is common.

I ran the cat5 out through a drilled hole for expediency. Ultimately some panel mount and wired IP68 4 pin connectors like these would be a nice touch but I didn’t have the time to order them….

Finally, the Sonoff unit is installed in the second enclosure. Annoyingly, the internal measurements of the enclosure weren’t available and it meant trimming the corners of the Sonoff with a multi-tool. Not a problem for me but you may want to source a bigger enclosure if it bothers you doing this.

The relays on the Sonoff do not supply power, so you basically need to have power going to the ‘com’ tab of each relay and then each solenoid wired to a respective normally open tab of a relay. Thus when the relay is closed, 12V flows to the solenoid.

There is probably a much more elegant way to wire this but essentially I cut the connector off the wall wart 12V supply and put both ends (+ve and -ve) into a connector block. I connected +ve to each of the ‘common’ tabs of three of the relays. So essentially 12V power goes to each terminal and also powers the Sonoff through the 12VDC input.

Then I wired the NO (normally open) terminal of three of the relays to the three pairs of cat 5 that correspond to the +ve supply of each of the solenoids.

I’m not sure this dodgy ‘circuit diagram’ helps but you’ll figure it out…:

OK – it’s not pretty or clever but I’m powering the sonoff and the three common tabs of three relays with 12V

The App

The Sonoff ‘comes’ with iPhone and Android app support – the ‘ewelink’ app. It’s OK … in that it’s not rubbish and it works reliably. You can set schedules and see status of the relays:

The Result

Works as expected! Reliable. No leakage.

Liked it so much I built 2.

The Future

There is more to come hopefully.

First, flash the firmware with Tasmota. This offers some connectivity benefits to OpenHAB which I plan to implement, and would give HomeKit integration and decent rules-based behaviour and data capture and monitoring.

Add some sensors. With Tasmota, the Sonoff’s I2C pins can be exploited for sensors. I’ve got some DS18B20 2 wire temperature sensors. I plan to get some I2C moisture sensors too. Maybe also some rain sensors.

That way, the solution can provide feedback on actual moisture levels and I can automate the schedules based on this, and with the outside unit, compensate for wet days.

Probably also plan to build one or two additional units for the ‘top garden’!

My version of the Arduino-based Turntable Tacho

April 18, 2020April 22, 2020 firmspear_adminLeave a comment

My version of the Arduino-based Turntable Tacho

“This is based on a diyAudio project, with some little tweaks of my own.”

There is a thread over on diyAudio that’s a few years old but still very popular. Some folks figured out how to use an Arduino Uno with a sensor to accurately measure the rotational speed of a Turntable.

Maybe if you’ve got a modern TT or a model renown for rock-solid speed and low wow & flutter then this is of no interest. But some of us have, well, older machines.

It’s really very simple – the hard work (the software or ‘sketch’) has been done for you, and it’s basically about connecting power to the Arduino, wiring up the sensor and positioning it.

You may not be surprised that I’ve built a few iterations of these for a couple of reasons!

First, my Linn Sondek LP12 is over 30 years old and is belt-driven, with an AC motor and a (reasonably) simple power supply driving the motor. (known as the ‘Valhalla’ power supply – essentially an AC regenerator). The rotational speed is adjustable – by mechanically orientating the motor spindle off the vertical by lighting or loosening a couple of retaining screws (!). Second, I’m part-way through (!) building another diyAudio project – a multi-phase sine-wave generator that has much more flexibility than the Valhalla (including voltage supply and phase offset), but which doesn’t have a speed feedback loop, so manually knowing the RPM is useful. Finally, I have a couple of other TTs – 2 Linn Axis in fact that I plan to eBay but which need servicing – These have potentiometer speed control but are also not rock-steady all the time.

The prototypes..

So, I build the original project on an Uno per the thread and just used a breadboard to connect sensor to board, power etc. Checked the speed – all worked great. I used the IR obstacle avoidance sensor which is a pretty ubiquitous Arduino sensor board but the Hall effect sensor, that measures when a magnet passes past, is another option. To make the sensor work, I used some 3M reflective tape on the side of the turntable outer platter.

However, perhaps because the Axis platters are a bit shinier, the IR sensor wouldn’t work for them. And it wasn’t very portable, given I hadn’t boxed it in an enclosure.

That got me thinking. I could i) use a Nano so I could box it in a smaller form-factor, ii) change the display for a little OLED for the same reason. Finally, iii) make it so I could use one of two sensors, depending on TT – so have a Hall effect sensor as well as the IR sensor just for good measure. (a tiny magnet, blu-tacked to the platter switches the sensor in this case).

I made a couple of changes to the original sketch to accomodate the different PINs on the Nano and the OLED display. I powered the Nano with a 9V PP3 (which you might know as a 1604x) since the Nano has a Vin and on-board regulator that can handle up to 12Vs. Given the device would only be used for a minute or so at a time, I wasn’t too concerned about power consumption.

It worked out pretty OK – in that it worked.

I housed it in a little metal project box. Frankly, a too small project box really, it was a bit of a messy cram. I had a small issue cutting the screen opening (originally I did it for a 128×64 OLED, but actually used a 128×32 one) and stupidly glued the screen and PET/perspex screen ‘cover’ I used from recycling some packaging by using – Superglue.

End result was the glue reacted with the PET or Perspex and it ended up cloudy.

Usable – but ugly.

The best bit was the off-board sensors, boxed in little cases, connected by a cable with 2.5mm plugs.

The ‘final’ iteration

It struck me that I could house the ‘works’ in a slightly bigger enclosure, include the different types of sensor, which could be switched between them and still have it ‘handheld’.

So, the bill of materials was as follows:

Enclosure, Arduino Nano, SPST on/off mini rocker switch, mini rocker (on-off-on), IR sensor, Hall effect sensor, OLED screen, NeoDymium Magnet

The on-off-on will switch the sensor signal to the Arduino. Rock left, and the left sensor is activated (The reflective sensor in my case). Rock right and the right sensor is activated. They are obviously never both active at the same time. (If you chose to do this, just remember the switch pins will likely be opposed – use a continuity check if in doubt!).

(I initially shared the signal wire and switched the 5V supply. The Hall effect sensor was fine with this but the IR sensor didn’t like it at all. Not sure why, possibly to do with the fact the signal cable is pulled high and the program measures HIGH to LOW transitions – but essentially above my technical pay grade). Wiring it the way I have done has the disadvantage that you are powering both sensors all of the time but my ‘on’ time is always less than a couple of minutes, so no real issue)

They will share the same signal (and ground) cable.

A rough layout of the components helps work out connector lengths etc.

The software firmware/sketch includes a provision to calibrate the actual value seen from the output of the crystal rather than assuming the ‘spec value’. I’d not done this on previous iterations – basically because I didn’t have a frequency counter. But now I do, so…. worth a quick check…

….. and at 15,630.60 Hz it was adrift from the theoretical 15,625 by about … 0.04% – so not too significant but worth dong if you can.

(Said another way, instead of reading 33.333 as you see in the picture, the uncalibrated sketch would have calculated 33.321)

While I modified the sketch for the OLED display from the original PCF8574 LCD library and otherwise hacked about with it by commenting out pieces I didn’t need – If you are minded to make one and are looking for the sketch, I later found somebody else had done something very similar but with tidier code so you’re better off going here for it.

Marked out on the enclosure the display and switches and then used the old, time consuming, ‘drill many holes and file’ approach to making good…

Calibrating the SW from the actual crystal frequency (Optional step!)

Pretty much got as good as I thought was possible by hand… fairly pleased….

Connected the connecting wires.. and that’s basically it.

The result

Per the initial picture, it works excellently…

The turntable now runs at 33 and 1/3 with 180g Vinyl and needle down (it runs slightly faster needle up no record on – 33.35).

The output is consistent irrespective of which sensor – IR reflective or Hall effect, is used. (To use the Hall effect, I blu-tacked a really small Neo-Dymium magnet to the platter side. For the IR reflective (obstacle) sensor, I taped a small 3M self adhesive reflective strip to the platter side.

Re-conditioning an original 1950’s Tannoy 12 inch Silver Monitor Canterbury

April 12, 2020April 13, 2020 firmspear_adminLeave a comment

Re-conditioning an original 1950’s Tannoy 12 inch Silver Monitor Canterbury

Tannoy 12″ Silver monitor in Canterbury Enclosure

“This was intended as a quick ‘spruce-up’ of an original but tired Tannoy speaker – it took a bit longer than expected and isn’t perfect but I was pleased with the results”.

If you’re interested in vintage Hi Fi then you’ve likely heard of Tannoy’s original range of concentric speakers. If you’d like to know more – have a look here.

The concentric design incorporates the tweeter cone inside the centre of the base unit and is still widely used today for studio monitors.

This range of original speakers – even ‘naked’ without cabinets, now command eye-watering prices (well in excess of £2K for example) …. and …. my father-in-law had one stashed away in his garage!

He’d bought it new, but was lucky enough to get it for a knock-down price when Cole Brothers of Sheffield were shutting down their audio department some time in the late 50’s or early 60’s.

For anyone that’s interested, it’s marked LSU/HF/12/L which is the 12 inch silver monitor, regarded by many as their best (well the 15 inch one is at least). The driver serial number is 023941 and it’s cabinet 549.

Since it was very definitely languishing away, I offered to ‘fettle’ it so we could eBay it and he could have a useful pot of money for a holiday etc.

My objective was to do a light, sympathetic improvement – I didn’t want to overdo it as some purists might consider that to reduce the value (that’s an excuse for otherwise saying I’m not a furniture restorer by any stretch of the imagination!)

Half the promise is now fulfilled. (I still have it in my living room since he was keen that someone in the family got some use out of it. When it’s fully broken in (!) I’ll get it sold – for now I’m enjoying it!

The starting point

The cabinet’s lacquer was distressed, there were a few dark ring marks from mugs being left on top, a few chips to the veneer and the grills were pretty filthy.

The ‘restore’

I rubbed the unit down with wire wool, artificial wire wool (3M) pads and various grades of wet n dry.

In some places I used spirit and a razor blade. The lacquer was caked on!

The task was made much harder by the fact the grill (a plasticised type of cloth weave) was not removeable – at least I couldn’t see how without totally taking the unit apart. Each quartered side has grilled panels separated by a small wooden detail – quite tricky.

I suppose all told I put about 30-35 hours into it over a period of a couple of weeks!

I found some veneer samples on the ‘bay. In some places such as the end corner, the match is excellent – in others it’s still pretty noticeable (in some part due to my very basic veneering skills no doubt!) but at the very least I have protected the surface.

I had a decent amount of success with the dark marks and rings using propriety ring remover (which I suspect is just Oxalic Acid).

I then looked for an oil that had more oil than varnish – a Danish Oil and slapped that on. I followed it up with Renaissance Micro Crystalline Wax – this is great stuff and I’d used it previously to give my Linn LP12 a really nice finish and I’ve used it on metal too – a proper conservator’s tool!

The grill cleaned up somewhat but as you can see not perfectly.

(The paper-y thing is actually the original paper Tannoy label, attached by screw to the rear!)

The Result – Part 1

The next problem was how to hook it up!

Being a single speaker (and back in the day they were rarely sold in pairs as this pretty much pre-dated Stereo), I needed a mono amp. So…. I went on the ‘bay and bought a Cyrus Smartpower Plus Mono/Stereo amp. ‘Course, this is a power amp so I also needed a pre-amp. I thought this would be a good time to experiment with a passive pre.

Basically, there are a band of Hi Fi aficionados who swear blind about passives.

In essence, (and being no expert) a ‘pre-amplifer’ is pretty much a misnomer anyway – for the most part the volume knob on a pre-amp actually attenuates, rather than adds gain. So a passive pre-amp usually has some input selector and and some form of potentiometer to actuate the line-level signal. No power required.

An active pre-amp by contrast can add gain, provide fancy (remotely) relay-switched inputs and can provide a preamp buffer which (as I understand it) can be useful to match impedances when the source and the output might otherwise be incompatible.

So, I bought a Tisbury Mini Passive Pre-amp!

As an experiment, I tried it in my main Hi Fi rig – replacing the active Roksan Caspian M1 Pre-amplifier. I’d recently bought this (guess where!) but back in the day (2010s etc.) this retailed for over £1,500…. Not cheap.

I don’t want to start an active/passive war but I *think* I preferred the passive!! However, I’d bought the passive for a purpose, and for me, part of the Hi Fi set-up I have is pure aesthetics – I love the Roksan Caspian’s rack look (strangely many don’t!). So the Caspian is a keeper!

So, the way the Tannoy is wired up is: FLAC (lossless) -encoded music is Roon-served to a Chromecast Audio, into the passive, and into the Cyrus, set on mono mode, powering the single Tannoy.

It sounded really good in fact. Really ‘warm’ but still distinct. Great range from high to low frequencies so I was very happy!

The Crossover

Of course, over the next few weeks, it didn’t stop me wondering about the speaker’s crossover..

Crossovers have capacitors and capacitors don’t last for ever.

So I thought I’d have a closer look.

It’s pretty simple. What’s interesting is the custom two cap configuration in a metal enclosure.

I also found the circuit diagram online:

So I decided to build a replica. I ordered a bunch of parts – including some Hi Fi-spec’ed (and priced) jumbo caps and induction coil of the correct values.

It looks pretty good assembled!

So I unmounted the original and replaced it with mine – using all original screw holes etc. to preserve as much originality as possible:

I decided to check the caps with my cheapo ESR meter (and sound like I know what I’m doing at the same time). One was good, the other not.

The Result – Part 2

In a word – amazing! Clearer, more defined. A really great speaker – just wish I had two, and that it was actually mine!

It’s currently in the corner of my living room, and I can stream all my music via Roon.

The clarity and richness of the sound has made me want to find a way to stream my Vinyl from my Linn Turntable to it too – but that might be another ‘project’!

Battery powering the Muffsy Phono Stage

April 12, 2020April 16, 2020 firmspear_adminLeave a comment

Battery powering the Muffsy Phono Stage

The Muffsy Phono Stage – Before

“There is nothing clever about this, but it’s effective and is even quieter than the already quiet adapter & on-board AC/DC power supply powered version.”

If you haven’t heard of the Muffsy Phono Stage you should check it out here . In essence, it’s a kit option of an open source phono stage design – the CNC Phono Stage that has a pretty wide following in the diy audio community such as on Audiokarma.org for example. The CNC itself is based on some application notes from National Semiconductor for the LME49720 op amp.

Anyway, the nice things about the Muffsy version are – i) It’s a full kit minus AC Adapter & Case, ii) specifically designed to fit in a nice, small, cheap case readily available online (the 0905), and iii) has great build instructions and support, if you need it, from from the enthusiast behind ‘Muffsy’.

It a bit of fairly simple soldering and you end up with a quality phono stage. Anyway I built one (and then a couple more as it happens).

The phono stage is powered by a separate AC-DC low ripple power supply, which fits in the same box.

I had a few issues with hum initially (which is how I know the support is so good), but it sounded great.

I also have a Rothwell Rialto phono stage which retails for about £400 and which HiFi Choice magazine said a few years ago: “This unit sounds utterly sublime.”

To my ears, both are frankly great, but if the Rothwell has the edge, it sounds ever so slightly quieter. ( less hiss, needle up, same photo stage gain, same pre-amp settings)

So – why not see if powering the Muffsy by battery makes any difference – why not indeed!

There is going to be two parts to this – modifying the phono stage and building the battery box.

The Phono Stage

The phono stage has an AC socket and a single way switch, and a separate board for the PSU. Out comes the AC socket, switch and power supply.

Because the phono stage requires a +v/0v/-v dual rail supply, we’re going to feed the phono stage with a shielded three core cable and use a DPST switch on entry (to ensure both + & – voltages can be cut). Then we need to fab a three core cable for inside the enclosure. I decided to braid 22awg wire, and cover in a copper shield, tied to 0v, just for good measure.

One of the cool things about the Muffsy Phono Stage is the backplate, designed for the 0905 enclosure. I wanted to reuse this so I needed my new DPST switch and my dual rail DC supply to fit in the pre-cut holes. No problem with the SPST round rocker – there is a DSPT version of the same thing. For the 12mm power socket – I found GX12 ‘aviation plugs’. Genius, and they look good too.

Apologies for the lighting but hopefully you get the gist. The GX12 in the backplate bottom left presents 3 wires – +12v/0v/-12V. Common (blue) goes straight to the choco block at the rear of the case, red and black go via the DPST switch to same. The cable is routed as far away from any signal paths as possible, and is shielded.

(Yes, I l know it’s a different version from the main image – but I said I have a few of these!!)

I said I don’t do this for a living! Sometimes the patio furniture becomes your workstation on a sunny day during ‘lockdown’.

3 single core braided 22AWG, adhesive copper tape covered for a shield (and then later covered in insulation tape)

The Battery Box

Not that tricky.

Bought a very inexpensive pine trinket box on the ‘bay. (By the time I’d stained it, varnished and waxed it, it would probably have been cheaper to buy a pre-made diamond encrusted box but that’s not really the point).

Two 12v SLA batteries . And a charger .

My cunning plan was the box would either charge the batteries (in parallel) or power the phono stage (with +12v/0v/-12v). But given the capacity of the SLAs, it didn’t need to do both at the same time.

So even I managed to work out that I’d need a 4P3T switch, some sockets – my GX12s and power-in from the charger, and some cable.

The charger had a couple of croc clips to charge the battery and that wasn’t going to work. So I broke into the charger (literally – somewhat destructively as it was so cheap it was all glued together) and soldered a replacement cord with DC barrel pin from my collection of old wall warts that matched the barrel socket I took out of the phono stage…

It’s all a bit messy and there is some permanent damage but since it was going to sit on the floor anyway – I’m happy enough.

So here is what’s needed circuit wise:

Essentially, the batteries connect to the four terminals of the ‘OFF’ setting of the switch. Then, thrown ‘right’ (as you look at the diagram), the switch connects ‘-‘ and ‘+’ to create the common 0v, and the left battery’s ‘+’ terminal forms +12v and the right battery’s ‘-‘ terminal forms -12v.

When charging, the switch has a couple of jumpers on the ‘left throw’ (again as you view it from the diagram) that place the batteries in parallel.

In this way, there is no danger of the circuits ‘crossing’.

(You might read some stuff about not charging batteries in parallel. They are probably, right, but it possibly won’t end badly. YMMV. I guess the advice is – Don’t just follow blindly what some guy on the Internet says, especially when he admits to not being an expert at this…)

Assembling the Box (and supply cable)

Basically, you need a couple of holes for the in/out, and one for the toggle switch. (And if you like, add a couple of status LEDs across the in & out (with appropriate resistors) to give some visualisation of the box’s state. (i.e. charging or suppling).

When you wire up the connecting cable it’s obviously ~~helpful if~~ essential you maintain the pin positions!

I used screened 4 core cable (because I couldn’t find 3 core).

The female connectors on the cable screw satisfyingly onto their male counterparts on both boxes.

In hindsight, the silicone 18AWG connecting cables were tricky to solder to the little switch and I probably should have used single core 22AWG throughout (I was concerned I couldn’t crimp the 22AWG for the battery connectors but since I ended up using 22AWG for the GX12 as I couldn’t get stranded 18AWG to solder to the small legs, with a couple of connecting blocks to connect to the already soldered switch, I would have saved myself time and effort swearing by using the thiner hook-up wire.

The results

A/B testing phono stage with (temporarily sited) Hi Fi

It’s hard to A/B something when you’ve hacked A to pieces, but I’m pretty sure it’s sounds less hissy now at high volume, needle raised. (Even though it was really quiet before).

[EDIT: I’ve A/B’ed it against the Rialto – it’s way quieter – needle up, same phono stage gain, same pre-amp settings. Very happy!]

[Btw – I suspect my original issue with the Muffsy was the AC/AC adapter; I bought a cheap one rather than the one ‘Muffsy’ recommended – go figure!]

I’ve not used it for long enough to comment on battery life (not that that matters with the always-available charger. Nor whether, if the batteries have slightly different characteristics (which they will), there is any issue with either charging or with the phono stage because of voltage variations on the separate rails. (although the PSU it replaced had slight variations anyway presumably due to the Resistor and Cap components not being 100% matched).

What I did think, after the fact, is that since my Rothwell Rialto uses 24v DC from a wall-wart, I could have incorporated this as an option by placing the batteries in series through another switch option. I might have a think about that as a ‘future upgrade’, and how I could ensure ‘off’ before ‘on’ as I cycle between options. If I can’t do it passively, I could always use an Arduino and Relay Board….

Restoring an Anglepoise 90

March 20, 2020July 15, 2020 firmspear_adminLeave a comment

Restoring an Anglepoise 90

“When I was a kid, my great aunts had an original black Anglepoise. Now the Anglepoise has become iconic and synonymous with Pixar “

But back then, it was just solid utilitarian British engineering!

Being older, I now realise it was a model 1227. They still make these!

Unfortunately, I have no idea where the lamp ended up. But, I was clearing out my parent’s house and found it’s younger sibling – the Anglepoise 90. It’s not got the same seminal square pyramidic base, and it was painted in ’70’s light green.

And it was paint scarred.

So I thought I’d ‘restore’ it. My idea was to sand it down and polish it to a mirror finish since it’s (mostly) Aluminium.

Unfortunately I didn’t really do any ‘before’ pictures.

Also, I didn’t succeed in the mirror-finish. Transpires I didn’t have either the equipment or patience or both to complete that!

It does look pretty cool though, in a polished-ish, burnished final state.

And I did complement it with a red silk braided cord that is quite striking.

The ‘restore’

I disassembled the lamp. I did take a couple of pictures for reference (not enough as it transpires but more on that later).

Then I used paint stripper and rubbed it down progressively using finer grit wet n dry – I probably used about 8 grades, from 240 including 1500, 2000 and finishing on 3000.

I then bought some polishing mops for both my Bosch GRO 12V (basically a ‘Dremel’ like multitool) and my 18V Bosch drill, along with various jewellers polishing compounds.

Long story short, and possibly it was just lack of patience, but I figured I was never going to get a mirror finish using the polishing mops and doing it hand-held.

So I changed tack and convinced myself I’d be happy with the a more rustically burnished finish.

(In truth, I really like it now).

I even decided to leave some traces of the old paint on a the swivel mount….

Reassembly

I bought a length of red silk covered braided 3 core flex.

My Anglepoise 90 was double insulated, but you don’t seem to be able to buy two core braided, so for good measure I decided to earth the lamp too. I basically connected the earth core to a screw on the lower articulated element and checked continuity on the other metal components too. This then meant ‘only’ feeding two twisted cores through the inner supports.

One very tricky part of the reassembly – the lower inner support that holds the cable on its run up to the lamp holder contains (in my version) a plastic insert designed to securely hold the flat, double insulated original flex. I couldn’t for the life of me get this out, and I couldn’t get the two strands of the braid up it either.

The only way I managed was to ‘burn it out’ using an old 8W pencil thin soldering iron. It cleaned up afterwards, but I’d rather not have done it that way.

I’d left a considerable time between disassembly and reassembly – probably three months, as other stuff got in the way. So I’ll admit to having been a bit puzzled as to some of the pieces on reassembly! Fortunately, I was able to look on eBay where there are loads for sale and some good people had taken sufficient pictures to help me work through what I should have done it the first place. Many thanks!

The resulting lamp looks good so another pleasing ‘project’ completed!