Pulled a blinder and then curtains.

We have a roller blind behind plantation shutters in our bedroom to block out the light. To get a better light seal I mounted it back to front, close to the frame of the shutters. That means the ball chain chord that raises and lowers it is a bit awkward to use as it partly behind the blind.

Some blinds are geared, so that the pulling force required is less, but ours isn't. It is six foot long and I also added some steel balls to the tube that runs along the bottom to make it hang straighter. Therefore it needs quite a strong pull to raise it, so I always had a plan in the back of my mind to automate it and in August I finally started working on a design.

I have a random collection of gearmotors, so I pulled out one that seemed about the right speed and torque and printed a pulley to mesh with the ball chain and fit the shaft.

Short Heatfit inserts that I bought from Stefan at CNC Kitchen are great for printed gears and pulleys.

A quick test with a battery showed it easily had enough torque to raise the blind, taking only about 200mA at 12V, but the ball chain likes to jump out of the pulley. It would need an impractical amount of tension on the chain to keep it in, pretty obvious really given the shape of the pulley teeth.

My solution was to trap the chain between two such pulleys and use the second to drive a threaded shaft with a nut that slides along it and triggers limit switches. The up limit switch is adjustable on a slide mount to fine tune the opened stopping position. The down switch is fixed but I can adjust the idler pulley on the shaft to adjust that limit first.

My original plan was to use 6mm studding in 6mm bearings but I found that it was a very loose fit, so I use 8mm studding and turned shoulders on each end to fit the bearings with my CNC lathe. I used a collet in the spindle nose to avoid marring the thread.

I tested this with a two way / two pole toggle switch to set the motor direction and the limit switches wired in series with diodes to bypass them, so that the motor can only back away from either switch once triggered. I mounted it to the wall with a printed bracket and was able to test the mechanism worked mechanically and seemed reliable.

For the electronics I decided to use an ESP8266 module running Tasmota firmware and control it from my Home Assistant server that runs on a Raspberry Pi MK4. In Home Assistant the blind just looks like two switches, open and close. I set up automations to open it at 8 AM and close it at sunset. Tasmota is configured with two interlocked relay outputs with timeouts and two buttons for manual operation. 

To drive the motor I had an L6203 full bridge driver chip lying around from RepRap days but I can't even remember where it came from. It's rated to 3A, so overkill in this application but I haven't had any use for it in about 14 years, so I thought I might as well use it up. The only problem was its package is a Multiwatt11 that doesn't have its pins on a 0.1" grid. I found I could fit it on a perfboard at 45° but it was looking like the electronics were going to be embarrassingly large compared to the mechanism. 

Around the time I was thinking about this a member of the marketing department of PCBWay coincidentally left a message on my blog offering free PCBs in return for a review, so I decided to take them up on their offer and use a PCB to make the electronics much smaller than I could make it on perfboard.

It is about 10 years since I last designed a PCB. For one off projects I normally just scribble a schematic on paper and then use perfboard, or for mains projects I use Veroboard and remove all the unused tracks to get enough creepage clearance. It is a lot quicker than designing a PCB with CAD and getting it manufactured and shipped but most of that is waiting time. So PCBs are worth while for more complex projects or ones like this where I might need more than one.

I downloaded the latest version of KiCad, which was more than 1GB! It must have changed a lot since I last used it but seemed fairly easy to use.

It is powered by a separate 12V power supply, so the cable to the window is just figure of 8 flex and it actually fitted through a gap under the shutter's frame. A tiny buck regulator from AliExpress provides 3.3V for the ESP-12F module. 

The limit switches cut the inputs to the H-bridge, so it doesn't rely on the firmware to not crash the blind. In fact, the Tasmota firmware just turns the outputs on for a fixed time that is just a bit longer than it takes to open the blind, about 30 seconds.

I thought it would be a good idea for the state of the blind to be discoverable remotely, so I used the normally open contact of the limit switches to pull two spare ports low. This was a mistake however because the GPIO2 port is connected to the blue LED on the ESP12-F module and is defined as an output, at least during start up. As these connections actually allow the firmware to crash the blind I decided to not connect them and two pin connectors could be used.

The buttons allow manual control of the blind. Pressing once starts the blind moving in the specified direction and pressing again stops it, so you can manually position the blind anywhere on its travel.

The programming connector is only needed once to load the Tasmota firmware because it can be updated over the air after that. So I don't solder it to the board and in fact there isn't room for it in the case.

The PCB design is an odd mix of through hole and surface mount because those are the parts that I had to hand. 

Because I wasn't paying for it I used four layers to simplify the routing. The inner layers are ground and 12V.

I also went for a routed outline because space was so limited that I needed a cut-out in the corner to miss the limit switch. I also only had space for one screw hole, so I printed rails in the case to hold the left side and the front right corner, so rounded corners make it easier to slide into place.

This is how tight the packaging got:

I modelled the board in OpenSCAD because the 3D model from KiCad didn't have the switch positions at the correct height from the board. I could only import it as an STL file whereas with an OpenSCAD model I can query the hole and component positions.

I extracted the component placement data from the footprint positions file that KiCad produced using a python script. This needs a lot more work to make it general though because through hole PCB footprints have the origin on pin one, whereas NopSCADlib has the origin in the centre.

I did all this modelling and the case design while the boards were being manufactured by PCBWay. 

I uploaded the design to PCBWay on the 12th of September and impressively they were dispatched on the 15th. The quoted build time was  4-5 days. I really liked the way the website shows you which stage they are at in the process as they progress. Four layers involve extra steps. 

Clicking on the "View Details" links shows videos explaining each process step.

I also ordered a solder paste stencil for the surface mount parts and this was made as the last step. I expected it to be made in parallel but it didn't add much to the time as it seems to a be fast process.

The package was picked up in Shenzhen by FedEx on the 16th and delivered in the UK on the 20th, so only 8 days from order to delivery. The minimum order was five PCBs but I actually received six. I think probably they make an extra one in case one fails the quality checks but if they all pass they send you it. I vaguely remember that used to happen sometimes when we ordered prototype PCBs at work.

The PCBs are great quality.  All the features are well aligned and the outline was very accurate as it fitted my 3D printed case perfectly. 

The solder stencil worked well. I forgot that the apertures might need to be smaller than the pads, depending on the stencil thickness, to get the correct amount of solder paste. I just used the default file exported from KiCad and it seemed to give the correct amount of solder.

Here it is populated:

I chose a white solder mask to match the white PLA I used to print the case in case any of it was visible but that wasn't really necessary, green would have been fine.

I used a T-962 reflow oven to solder the surface mount components. It was the first time I had used it since buying it from Elektor a year ago, so I had to apply the upgrades detailed here:

  

This was replacing the masking tape with Kapton tape, adding a cold junction sensor for the thermocouples and replacing the firmware with an open source version.

I used leaded Chip Quik solder paste on the default oven profile for leaded solder and it worked great without any calibration. I mainly use leaded solder at home because it gives better results. In this case it has the advantage of not reflowing the internal parts of the ESP-12F module, which I assume uses lead free solder.

Here is the finished unit installed. I used white PLA, so hopefully it wont melt in the summer sun.

Flushed with success of this project, my wife wanted me to automate some curtains to keep the house warmer in winter. It was a race against time to get it working before our last holiday.

The gearmotor I used (Nidec GMAG 404 327) was nice because the motor has a transorb mounted on the armature to stop the back EMF from the coils causing sparks, which generate wide band RF interference. I looked it up and found out it now costs £162.36 from RS components, so I looked for a cheaper alternative for the curtains.

I found this DF-ROBOT FIT0492-A from PiHut for £11.40:

It was a challenge to make a gear_motor class in NopSCADlib that can draw both but I manged to do it, with a lot of parameters.

So it is possible to make the blind controller with the cheaper motor, although making the design handle both motors was also tricky because the clash avoiding constraints are different for each motor because their shapes are so different.

I haven't tested this version but it should work. I published the design as a NopSCADlib project here: github.com/nophead/Blinder.

I designed a curtain puller around this new motor.

It clamps around the curtain pole, which has a hole drilled through it for the shaft. A length of picture cord goes from one bobbin, around the idler pulley clamped to the other end of the pole and back again to the other bobbin. The curtain rings at the open ends of the curtains are each attached to the forward or return cord so that they move in opposite directions.

The motor shaft has a printed spur gear to drive the threaded rod which activates the limit switches. It also acts as a coupler for an extension shaft to drive the bobbins.

It nearly worked but I ran out of time to finish it before going on holiday. I couldn't get enough grip with the 3mm grub screws because the heads would strip first. I need to replace them with M4 to get more grip. Hopefully it will work then, with perhaps some guides to stop the cords tangling.

I built 12V power supplies using some potted modules cased in a short piece of aluminium box section for fire safety.

To turn the module into a PSU you need to add a mains filter, a transorb and fuse to the AC end and a couple of filter capacitors to the DC end.

In summary, I populated three of the PCBs and they all worked first time. The PCBs were only \$25.97 for 5 pieces in 4-5 days, if I had been paying for them. The solder mask was \$10 in 1-2 days. FedEx shipping \$22.89.

Now that I have got my reflow oven set up I will make more use of PCBs and SMT parts. PCBWay make it very quick and easy to get good quality PCBs for a good price. They also offer PCB assembly, CNC machining, sheet metal fabrication, 3D printing and injection moulding but I haven't explored those. Hopefully I can do my own 3D printing!

Mooshimeter Mod

Last year I bought a Mooshimeter wireless 2-channel multimeter. It is a multimeter front end that links to your mobile phone with Bluetooth and displays the results in an app.

It is handy because you can read it remotely, it can measure voltage and current simultaneously and display power, it can log to an SD card and graph the results. It can also speak the results.

Despite all these good features my "go to" multimeter is still my EEVBlog branded Brymen BM235. So my Mooshimeter sits in a drawer for most of its life. When I get it out it usually wants to do a firmware update, which needs fresh batteries. Because it has no off switch the best you can do is put it into shipping mode. It still flashes its LED occasionally, so the batteries run down over a period of several months and are then not up to doing a firmware update.

Devices with no proper off switch are a pain if you only use them rarely because the batteries are always flat when you come to use them. This is particularly a problem because modern Duracell batteries seem to leak and corrode as soon as they are flat. This didn't used to be the case. I found some very old ones that I had abandoned in outdoor devices that I expected to be corroded to hell but in fact they were not corroded at all, despite being well past their use by date. In contrast I have had many corrode recently that were flat, but still well within their use by date. I have stopped buying Duracell and now use Costco's own Kirkland branded ones. It is too early to say if they corrode or not.

So I normally remove the batteries from devices I use rarely but with the Mooshimeter this involves removing the casing that is held together by two screws. I decided to add a switch to it but it has a cat III safety rating and cutting a hole in the case would void that.

I had two ideas to get around this: the first was to put a normally closed reed switch in series with the batteries and 3D print a cradle with a magnet in it to turn it off. My second idea was to use a mercury tilt switch to turn it off when placed upside down. I ordered both but as the mercury switches arrived first I implemented that and it works well.

I decided the easiest point to break the battery circuit was the link between the two cells. For some odd reason that is a copper fill rather than just a track. It is on the top side of the PCB so I had to desolder one of the contacts to get at it.  Fortunately the battery contacts have thermal relief connections so I just cut two of those to isolate the pad.

I then made an insulating washer out of Kapton film. I used that because it is very thin and can handle soldering temperatures. I stamped it out with a hole punch but I found it very difficult to get the hole concentric. This is my third attempt that was just good enough:

And here it is in place:

I reinserted the clip over the top and resoldered it. I then soldered the tilt switch between the two now isolated battery terminals.

When upright the contacts are bridged by the mercury, which is very low resistance. When I turn it upside down the mercury flows to the top of the bulb and isolates the batteries.

So all I have to do is remember to place it upside down in its case. If you want to attach the meter to something moving then the reed switch idea is the one to go for. I think it can probably be mounted in the same place if you use a neodymium magnet. You can get it nearer the back of the case by mounting it on the other side of the PCB but I don't know if that affects clearance distances for class III.

Telequipment D32 trace rotation fix

Another ancient piece of equipment I have is this Telequipment D32 mains / battery operated portable oscilloscope.

Telequipment is an old brand name used by Tektronix in Europe. I think it was made in about 1972 for servicing TVs. I bought it in a broken state from a friend of a friend in the mid 1980s. The six NiCad D cells had gone short circuit and the mains transformer was burnt out. I replaced both of those and as far as I can remember that was all I had to do to get it working.

I did get a lot of use from it as for many years it was the only working scope I had. I haven't used it much in recent years though because I now have several much better scopes. It is the only battery powered one I have, although I do have two USB digital scopes which are battery powered if I run them from a laptop.

I powered it up fully expecting the NiCads to have died again, but amazingly they still work to some extent. The only thing wrong with it was the trace was slanted and the trace rotation control didn't have enough range to correct it fully. It seemed to max out in one direction and then have a large dead band.

I did a bit of research and managed to find the schematic. I must admit that somehow the fact that a beam rotation circuit is needed because of the earth's magnetic field had completely passed me by, or else I once knew and have since forgotten. This is despite building my own scope with valves (vacuum tubes) when I was 13 and using CRT scopes for many years after.

The rotation circuit is below. After looking at it and making some measurements I was surprised to find it to be a design fault rather than a faulty component.

The base of TR301 can be varied between ±7.5V with R303, so one would expect the emitter follower to drive the coil over that range with a V_BE offset. The problem is the coil has a resistance of about 1kΩ, so the pull down resistor R314 can't drag it below -3.5V. This means the control has no effect for a significant part of its travel and the range of the correction is asymmetrical.

To fix the problem I removed R314 and replaced it with a PNP emitter follower. That gives push-pull symmetrical drive between about -6.8V and 6.8V. The only dead band is a small one around 0 due to the V_BE drops (classic crossover distortion). It also has the advantage of not wasting the current through R314, all the current now flows through the coil.

This is a view of the board before the mod. It is ancient technology, all discrete transistors and hand routed PCBs, including vias made with what looks like soldered in brass rivets.

I don't know why all the transistors are socketed. The only other time I have seen that was in a Russian transistor radio, but they were germanium PNP transistors, so may have been unreliable. These are silicon planar epitaxial so should be reliable and never need replacing. Perhaps they did it because they thought soldering would damage them. Or perhaps they were just old school designers used to valves.

Anyway the scope is a joy to work on because all the PCBs hinge out to allow access to both sides with the scope still operational. See www.youtube.com/watch?v=BOeEbyIllcM for a look inside a D32.

Here is a view of my modification: -

I replaced R314 with a 2N4403 (any PNP transistor should work) and linked the base to a handy via. Don't worry about the proximity to the ceramic capacitor's lead, it is the same net.

So a simple mod but I don't understand why I need it now. Perhaps the earth's magnetic field has changed since I last used it. It was about a decade ago!

Update

The scope has always had a delay between switching on and starting up. The power LED comes on about a second after it is switched on.  This got longer and longer and then it stopped starting reliably at all. I had always assumed the delay was caused by a capacitor charging in a bias circuit for the inverter but on examination of the schematic I fount that wasn't possible because all the capacitors have a low impedance supply.

It turned out the be the on off switch. I measured about 3V across it in the on state. The contact resistance measured 0.3Ω but that should only drop 0.3V at 1A, so it seems to be a strange non-linear resistance and it must get lower over time as it heats up.

The switch is part of the brightness potentiometer and it is operated by pulling the knob out and pushing it in. Normally potentiometers with switches operate it at the start of the rotation but this allows the brightness setting to be retained. It is also a small form factor so I didn't think I would be able to find a replacement. As it is riveted together I didn't think I would be able to repair it or even get any switch cleaner into it.

It is actually a two pole switch, the second pole is used to change the charge current of the battery to include the scope current when on. I figured 0.3Ω would not affect that circuit as it has 150Ω in series, so I swapped the connections over. It now starts instantly although it does take time for the tube to warm up of course.

Mendel90 GitHub catch up

I finally found time to update GitHub with some Mendel90 changes that I have had in the works for a long time. The problem with releasing them sooner was that they were all not quite finished and / or would make unintended knock on changes to the kits I was producing. In particular the changes I did to make a Huxley90 in a hurry for the TCT show and the E3D mods kindly contributed by Philippe LUC that conflicted greatly with it, so needed a lot of work to merge.

OpenScad

I also updated to the latest version of OpenScad. The upside was that hull and some of the 2D operations are much faster. I was also able to replace all the calls to minkowski with offset as I was only using it for 2D offsetting. The net result is it is now four or five times quicker to generate the preview and the STL files. The downside is that the 2D sub-system now uses fixed point coordinates but the rest of OpenScad doesn't. This makes it difficult to get 2D and 3D geometry to match up. For example, an extruded circle now has slightly different vertices to a cylinder of the same size. This created a few degenerate triangles requiring that I changed the way I constructed some objects in order to get nice clean STL files.


The solution in the case above was to make the cylinder slightly bigger than the circle used to make the pointer.

On the up side it seems OpenScad has got better at handling unioning exactly coincident faces since I first wrote Mendel90, so I could remove some of my small offset bodges to avoid z-fighting.

Another benefit is that the X end brackets now slice correctly in Slic3r, as the bug that caused internal faces to point the wrong way has now been fixed. Skeinforge doesn't care about face orientation, it just counts edges to work out what is inside and what is outside. Other slicers got confused and filled in the nut cavity.

Along the way I discovered that, although OpenScad now has trig functions that are accurate for multiples of 90 degrees, etc., it doesn't use them in rotate, or vertex creation for circles and cylinders. It converts to radians and uses the library trig functions. Degrees can never be represented accurately as radians in floating point because Pi is irrational, not to mention transcendental. To get round this I now override the built in rotate with a user space version that uses the accurate sin and cos degree functions.

module rotate(a)
{
 cx = cos(a[0]);
 cy = cos(a[1]);
 cz = cos(a[2]);
 sx = sin(a[0]);
 sy = sin(a[1]);
 sz = sin(a[2]);
 multmatrix([
  [ cy * cz, cz * sx * sy - cx * sz, cx * cz * sy + sx * sz, 0],
  [ cy * sz, cx * cz + sx * sy * sz,-cz * sx + cx * sy * sz, 0],
  [-sy,      cy * sx,                cx * cy,                0],
  [ 0,       0,                      0,                      1]
 ]) children();
} 

Not surprisingly every STL and DXF file generated is now slightly different numerically but hopefully not dimensionally. I made a stable branch to record the state before these global changes, just in case. GitHub has some excellent image and STL comparison views but unfortunately it gives up if more than a handful of files have changed and there are hundreds in the Mendel90 tree.

Wade's Block

After a few people started to report broken or cracked Wade's blocks I strengthened it a bit around the bearing block. I also made the bearing sockets a bit bigger so there is less stress created pressing them in. Kits from around March 2015 have shipped with this version.

X Carriage

When Philippe LUC created the E3D branch he fixed a few bugs. One of these was that the X carriage top was only 2mm thick, when the design intent was 3mm. This was due to the fan duct using the same variable name. Whoops! I have updated it now in the main branch. I also made the nut traps for the fan bracket screws deeper to allow for longer screws and to allow them to be withdrawn further without loosing the nuts. This makes it easier to remove and replace the duct. Simply removing the washers is an alternative.

E3D Hotend

I temporarily parked Philipp's mods in an E3D branch until I could merge them. I have now updated the master branch to support E3D V5 and V6 hot ends with this one line change to the config file. The generated files for V6 that are different from standard build are in new folders dibond_E3D and sturdy_E3D and I have deleted the temporary E3D branch.

There is no room for the right hand wing nut because it clashes with the hot end's fan. Fortunately the carriage has always had nut traps to allow the screws to be inserted from below. A plain nut above can then be used to secure the extruder.

Primarily the things that change are the Wade's block, the fan duct and the fan bracket. The Wade's block has no extension to avoid losing more Z build height than necessary and a plain screw hole on the right end instead of the hex socket.

The fan duct has to slope downwards to avoid the E3D heatsink. That creates a sloping bridge that is also skewed horizontally. I haven't found a slicer that handles this properly yet, having tried Skeinforge, Slic3r, Cura, Kisslicer and even paid for Simplify3D! I have blogged about their failings in another post here: hydraraptor.blogspot.co.uk/2016/01/a-bridge-too-far. Any other slicers I should try?

Another bug Philippe noticed is that there was almost no clearance between the fan and the belt. Fortunately the belt is twisted so it actually does clear the fan. I have added more clearance as
Philippe did. It makes the fan bracket and fan duct 2mm longer. If you print either from the new files be sure to print both or the duct will be misaligned.

I also improved the internal shape of the duct a bit. From this: -

To this: -

It probably doesn't make a lot of difference but a comparative test of various fans and ducts will be the subject of a later post.

Even with the shortened Wade's block the E3D V6 hot end is 4mm lower and the V5 is a bit longer still. If you retro fit it to an old machine you will lose 4mm Z travel. If you are building a new machine then there are alternative files which add 4mm to the height of the frame and lengthen the Z smooth rods and threaded rods on the bom. That also has a knock on effect on the shape of the spool holders and the dust filter. If you use the larger sheets be sure to get the correct size rods and use the correct spool holder parts to match the frame.

New Lighting Options

I redesigned the lighting system I described here to work with some commonly available LED light strips. These consist of an aluminium PCB strip that slides into an aluminium extrusion with plastic end caps, which I discard. Instead of printing a bar to hang the lights and camera from I now add printed end caps to the light strip and uses those to hinge it from the frame edge clips. I then hang the camera from the strip with its own hinge.

The strips come in 500mm lengths but they can be cut at discrete points between every third LED. They are described as "50CM 5050 SMD 36 LED Warm White Aluminium Rigid Strip Bar Light Lamp" and I bought them from bgood2010 on eBay.

I got some from another seller and although the eBay picture looked the same the extrusions where actually not as deep. The STL files on GitHub are for 8.6mm deep extrusions and are generated by light_strip = RIGID5050_290. Setting it to Rigid5050_290 generates the clips for 7mm deep extrusion. Other sizes can easily be accommodated as long as they are rectangular. The definitions are here.

Rather than waste the off-cut I mount it above with a second pair of end caps that clip onto the main light strip. These are set back just far enough to avoid the build volume in the unlikely event you print something tall at the back edge of the bed. This is calculated by the model with lots of trig and Pythagoras maths. Set show_rays = true to see this view showing that the camera and lights are pointing at the centre of the bed and the build volume is clear.

Another light strip that can be selected is this one: FSRP3W, discovered by Alzibiff.

Again the end caps are removed and replaced with printed ones that clip into the screw channels in the extrusion. There is no room for the plug so I just solder the wires on.

It looks neater and gives a more diffuse light but is not as bright as the double strip of 5050 LEDs and is more expensive. I bought it from www.ledlightingandlights.com.

The only problem with these light strips compared to my original Sanken ones is that they are unregulated, so they flicker when the bed switches on and off. I described how I fixed that here. I also need to update the mounting for the Raspberry Pi to accommodate the plethora of new Pis that have appeared since my original design.

Huxley90

The Huxley version is scaled down in the same way as the Sells Mendel was scaled to make the Huxley. It has a build volume of 150mm cubed and uses NEMA14 motors, 6mm smooth rods and M3 fasteners for the frame. There is a good photo of it alongside the full sized machine on Ivor O'Shea's blog post.

The NEMA14 motors have about half the torque of the NEMA17s when driven with the same current. The Y carriage and bed have about half the area hence half the mass, so that is about right. Also a NEMA14 has half the mass of a NEMA17, so the X carriage also has about half the mass.

I believe the flex in the middle of the rods is proportional to the length cubed times the weight divided by the bar radius to the power of four. The length of the X rods is almost exactly 75% of the Dibond version and the diameter is obviously 75% as well. The relative flex then boils down to 0.5 / 0.75 = 0.67. So going down to 6mm rods is justifiable as well. Everything scales very nicely physics wise.

As the design is fully parametric shrinking it should have been easy, but because vitamins don't scale perfectly lots of snags arose where things clashed. A typical example was the x_motor_bracket. The NEMA14 motors are smaller but the raised boss around the shaft is the same size. This makes the bracket a different shape and it then needs a support to print it.

Half a truncated teardrop with a crutch!

The heated bed was made with veroboard and coincidentally has the same resistance as a full sized Prusa PCB, so the machine takes the same amount of power but heats up about twice as fast. There is no room on the frame for an ATX PSU, so I used an external XBOX 200W PSU. I couldn't find a spec for the 5V standby rail but it seems to supply enough current to power a Raspberry Pi.

Direct Drive Extruder

The extruder is where the scaling fell down a bit. The original Huxley used a Bowden drive to make the carriage small. I didn't fancy that but I didn't want to have the carriage as big as a geared extruder would need, so I went for direct drive with a NEMA14 and 1.75mm filament. 

The filament needs about one third of the force to feed and a Wade's has roughly 1:3 gearing, so a direct drive NEMA17 is about equivalent. The NEMA14 has half of the torque, so it is a bit under powered. I used the smallest drive pulley I had which was a mini hyena from Laszlo Krekacs' Indigogo campaign. Unfortunately I don't think the small diameter version is available now. I could probably make one from a hobbed bolt if I needed to or hob one from scratch.

It feeds PLA fine at 200C but isn't able to pull it off a spool. I will try a spool holder with a central bearing rather than the rim bearings to see if that is low enough friction. If that doesn't work I might try a powered filament supplier like the one on the first Up printer preserved here. Or maybe even try Bowden drive.

The design is parametric so there is a NEMA17 version suitable for Mendel90. I just need to adapt it for a commonly available drive pulley. It should just be a matter of adding a description here.

It can also use the E3D hot end but that doesn't fit between the bars on a Huxley90.

So that is Github up to date and hopefully correct although I haven't tested a lot of these changes.

I noticed that Blogger is now a lot worse than it used to be. Headings and pictures are now a nightmare.