Accessing a Vigor 130 modem

As intimated earlier, my Vigor 130 modem is not accessible as it sits behind a router. The 130’s address range is 192.168.2.1/24, while my LAN sits on a 192.168.1.1/24 range.

Try 1

Although Draytek have an article about using tagged & untagged VLANs to access these modems behind routers, the setup didn’t work for me (I suspect their use of the ‘0’ VLAN tag as shorthand for an untagged VLAN didn’t work for the particular router I have).

Try 2

Instead, I want to set up a route policy and any requests for the address range of the modem should be passed back through the WAN2 interface to be resolved there. At least I think that is how it is done. Using Routing / Load-Balancing Route Policy I setup a policy so that any requests on the 192.168.2.1/24 range go via WAN2, and should go to the modem. However it doesn’t work.

Try 3

I think I have the wrong concept about the route policy and have to instead use a virtual WLAN such as WLAN5? An excellent article explained it better, however it was for another type of modem. Dantheperson’s explanation was that by adding a secondary static route and pointing that at the WAN interface, “Now when your PC routes the request for 192.168.0.1 to the router, the router knows it can access that address via the WAN interface, instead of forwarding it to the default “internet” interface.” This is what I want to do, however I think that the Vigor router uses the permanent virtual circuit (PVC) idea to do the same thing. So my current setup is Route Policy –> virtual WAN5 which sets up a VLAN on the WAN2 interface. It still doesn’t work.

Argggggh!

Advertisements

Strange latency spikes

I’ve purchased a new Vigor 2862Vac to be the main wireless and router at my house. I used to use a Fritz!Box 7390 and loved the ease of use combined with the technical power of the combined modem, router, firewall, VoIP and DECT base station. But things are getting older and with new VDSL/2 capabilities I have swapped from a cable connection to a fibre-to-the-cabinet through another provider, Andrews and Arnold. They are stupendously great, and techies all the way. Don’t expect any hand-holding.

But I’m starting to see some external wierdness after installing this new router. I have enabled ICMP from the WAN (ie. , being able to be pinged from the Internet) as I run an external monitor to test my broadband access from outside the home. This works using a Firebrick device that can sustain many thousands of ping requests per second and attempts to access my public internet address once per minute, then graphs the results.

The graph is showing some strange latency spikes:

Probes from outside

Although I haven’t noticed any particular instances of slowdowns locally, this is a little worrying as my graph was nice and clean before, and those yellow spikes are a little regular to me … in fact, too regular at 30 minute intervals every couple of hours.

So a little digging – I run a Vigor 130 modem on the front-end that does the VDSL dialling and syncing. This is great, as I can then take the signal via Ethernet connection into whatever I like: a switch, my own router, pfSense running on a PC, or a commercial AP/firewall such as the Vigor 2862Vac. What does cause an issue is that the Vigor 130 runs on its own IP address range, and once it passes the connection to the router it is no longer accessible. I need to access it to see if the spikes are coming from it or the router.

Of sealing wax, and cabbages

I’ve a new phone, and that means a new phone charging stand.

Although the EU standardised on a common external power supply for small electronic consumer devices in 2009, followed by many international and national standards bodies, my new phone uses the USB-C plug. This isn’t a problem as I use a USB charger that has five slots from which I can run either micro-USB or USB-C cables, depending on the need.

What is far more interesting is the charging stand.

For every new phone that I purchased either for work or leisure, I had to replace the charging cable and get a new one for my work backpack, buy a new phone cover, get a new desk charging stand plus a new car holder while driving. This added considerably to the cost of a new unit but was often overlooked. Since retirement I’ve done away with a lot of these distractions and instead:

  1. Use a universal car holder without active power. The reason I can do this is that new phones:
    • hold their charges for much longer, up to 3-5 days for some Huawei phones, and
    • I don’t drive as much, and journeys are shorter.
  2. I still get a new, custom fit phone cover.
  3. Buy a generic desk stand which I adapt using Sugru. These stands are much cheaper than the branded ones and with some plastic magic easily adaptable to the shape of the new phone.

This is how it is done.

First you need the phone to be in its cover, as the dimensions will fit exactly to the outside of the holder. Place the Sugru into appropriate areas of the holder – for me it was the rear kickstand which supports the phone at a correct angle to engage with the USB connector, and the bed of the USB connector. Depending on what generic stand you get this may be different.

Using cling flim, Saran wrap or whatever it is called in your country then cover the back of the phone and press it into the bed of the stand, engaging the USB connector and allowing it to rest gently on the back of the stand. I then leave them for a short time, perhaps several minutes, and gently remove the phone and cover allowing the cling film to stay for some while longer.

Once the Sugru has gone ‘green’ (easily able to be cut but not so hard that it is difficult to work with) then I trim off unneeded parts and neaten. The photos show a couple of examples of a micro-USB and USB-C stand for specific phones. The overall fit is tailored and exact which reduces the strain on the power connector and prolongs the life of the phone.

Egg robot Number 1

The first egg robot was hatched when Bruce Shapiro in 1990 dreamed up the Eggbot™. Bruce published it as an open-source design for both software and hardware however the name is trademarked. While it is available for purchase as a whole machine, I wanted to see if I could 3D print one myself since the initial cost was large.

I found another design by Probot that seemed to do a reasonable job, and although many of the other ones such as the SphereBot are larger this one seemed to produce reasonable quality and I had most of the parts:

  • 308zz bearings
  • 3D printer to print parts
  • Arduino
  • silicon egg cups moulds
  • stepper motors
  • stepper motor controllers

… and that was about it. Most of those I had, the rest I could order easily (for example, I had Nema-17 steppers and this only required 28byj-48 ones, either 5V or 12V). So, in a fit of decision I launched into building it and surprised myself with how quickly it came together.

The base and cover are easy to print with the sides of the cover bending into a living hinge. The ends are bolted shut and there is a sliding swing arm and end stanchion. The living hinge is an interesting design choice which leads me to believe that the designers were very familiar with 3D printing as it avoids overhangs in the upright sections of the print. It does make a very neat and attractive package when finished.

The challenge comes because of the interaction between different things which were not designed together: hardware and software. The original hardware has an actual PCB created to control the movement, whereas the design I am using replaces that with an Arduino running software called EggDuino. That software has to emulated the responses and obey the commands expected from the original PCB so that it can print things. All this is controlled by an extension to the open-source software called InkScape that sends commands to the egg-robot! Lots of moving parts in this design and 3 different things created by different people does give lots of room for error. So the process goes like this:

Inkscape → Eggbot extension → Arduino running EggDuino → egg

The ‘tail stock’ – if this were a lathe – is based on a printed part with a 308zz bearing inside. The 308zz are a very common type of bearing, sometimes called ‘skateboard bearings’ as they are ubiquitously used for skateboard wheels.

Once completed the whole machine is connected to a computer by USB cord and you use Inkscape to create and print the eggs.

Blutack and masking tape

I’ve started masking the base PCB of my WiFi doorbell and I am using both masking tape and Blutack. I discovered the Blutack while going through my stationery drawer looking for something similar to the latex masking fluid which is used in real electronic waterproofing. I had thought of using an eraser which would be soft and pressed onto the micro-USB port would mostly seal against the conformal spray. The Blutack is now my best friend for masking, even if the spray dissolves it some and leaves it sticky – but the easiest way to remove Blutack is using more Blutack, which worked very well on the PCB board.

I decided to leave the camera alone as it was already in a unit, and previous experience has taught me that getting near lenses with any type of fluid or compound invariably means that you leave smears, no matter how careful. As it is a single unit I’ll seal around its seating and hope that that helps prevent major water ingress into the rest of the mechanism. If the video feed fails, I’ll know better next time!

The speaker at the bottom is also hard to waterproof: cover it in conformal fluid or seal it in somehow would mean that it likely could not be heard, and similarly for the microphone which needs to vibrate to detect speaking. I’ll simply seal around the edges of those components and hope that their internals are okay with a little moisture.

Miniature speaker at bottom of unit

The door-press button is difficult, as it needs to move to trigger the micro-switch, but isn’t sealed around its edge – I assume that rain against the front of the unit seeps through capillary action around the edges and into the major internal space, and thence via condensation onto other surfaces. That’s bad, as it possibly means I cannot seal one of the areas of major ingress.

Press button is … difficult

Experimenting is fun, so I am going to try a thin film of plastic wrap (sometimes called ‘cling film’ here – other names elsewhere!) glued to the rear of the face plate to see if that works. The plastic film should bend enough to allow depressing the micro-switch, but the real challenge will be getting it sealed against the back of the face plate. I will try a rim of silicone sealant to which I will press a square of plastic film.

Overall the task took me the whole day. While that seems a little crazy (and you may prefer watching box sets on television), I saw it as a learning exercise and bit of a challenge! I’ve thought through how to waterproof a very difficult piece of equipment and although the proof that it’s successful still remains, I am overall pleased with the result and would attempt something similar with greater skill.

The doorbell does work now and I have tested it both via pressing the push button and the mobile app. Bring on the rain!

Waterproofing electronics

I’ve been interested in getting a WiFi doorbell for my house. Having looked at the issues surrounding the two market-leading brands, decided to go the cheaper route and purchased a no-name brand from China.

The doorbell and separate chime are excellent, but are not waterproof.

I experienced this after having had the doorbell working for a week or so – England is anything but dry and before Christmas a downpour lashed our front elevation for around 18 hours, thoroughly wetting the outside of the unit. While the camera still worked and I could connect to it from my phone app, I did not suspect that water had entered the compartment. When it did occur to me was after one door press by the post delivery worker that left a lingering ‘ding dong’ in the air after it should have stopped.

The next day the lingering ‘ding dong’ did not stop at all and I had to remove power from the chime unit. I opened the unit and saw moisture inside and dried it out, but realised several days later that I could not connect to the unit from the mobile app and so have taken it off the wall to fully waterproof it. Thankfully the ringing had stopped but I worried that it could re-occur in the middle of the night or while we were away.

Components of the video doorbell

I will use electronic conformal liquid in a spray form to coat the PCB and internals, and will try to either glue non-moving parts in place with silicone sealant as well as putting a bead around the external rim of the back-plate. My challenge is that there is several holes where ingress is needed for the microphone and the speaker! How to waterproof those? And the press button will still have to move inwards to depress the micro-switch and if glued solid it could not do this.

The reason the High Street died

In the UK the major shopping district is colloquially called ‘the high street’. Other countries may call it the CBD (central business district – but more generally this is the office and skyscraper part of a city), downtown, or simply ‘the shops’. My local town here in the UK has one named The High Street and this isn’t unusual for England as it is the most common street name. With the advent of online shopping in the early 2000’s and the rise of Amazon there was much consternation at the plight of bricks-and-mortar shops on the high street. TV programmes that aimed to remedy the flight away (notably ones featuring Mary Portas – ‘queen of shops!’) were common, there were government grants and industry worked with academia to study the issues – I know, because in my professional life I took part in one such study.

But the shoppers continued to flock to the online retailers and everyone was puzzled. Discounts, robo-shoppers (research online, shop in store without browsing), price matching and all manner of techniques were put forward. Card networks offered to analyse patterns and events were held in target towns. Yet, from my personal and anecdotal experience I’d have to point at one major reason why people are turning away from shopping in person especially at the larger out-of-town warehouses.

Rudeness.

I’ve just come back from picking up some items from one of the large chain store brands here in the UK and my experience was infuriating. I struggled to the checkout with a large TV and no-one offered to help, whilst the store was mainly empty with assistants wandering the aisles. We had a gaming console plus assorted other items and our purchases were a sizeable part of £1000 – so not insignificant. At the checkout the assistant immediately started up-selling product insurance and when I declined insisted on asking why? He also wanted to know my street address and email, and when I also refused those insisted that the purchase would not go through without them. Actually, the store already had them as I have an account in their online shopping system and had reserved the console earlier. But what shook me was the general abusive approach taken to the sale; countering my every decline, challenging my refusal to take insurance, insisting on more personal detail and so on. As we struggled to lift the items from the counter he stepped back and turned away – not even offering to hold them while I took some to the car. Perhaps it was just me, or a latent racism on their behalf – but this retailer as a whole has a ‘bad’ rating on Trustpilot with five-figure negative ratings. I felt unwanted and used.

Recently another large out-of-town shopping warehouse has gone bankrupt and when I read forums on Toys-R-Us and their decline many people mentioned the shambolic store layouts and disinterested assistants. If shopping is all about experience, as many retail consultants believe, then it is important that shoppers feel great about what happens to them, not abused. I’ve spent long enough in the cut and thrust world of business and even time in retail as a IT consultant, so I understand the pressures and analytics which drive certain behaviours. Scanning time for checkout operators, forced meet-and-greet, up-sell and cross-sell all sound good on PowerPoint but in reality shoppers are looking to feel good and have an enjoyable experience, not just acquire goods out of necessity.

I’d have to identify rudeness as a major reason why the in-store experience puts off many shoppers in England in the larger chain stores. Contrast this to a recent bookstore – surely an endangered breed if ever there was! – and the utterly other experience of an assistant looking people in the eye, asking how they were and did they enjoy that type of genre, wishing them well with their purchases and offering to go the extra mile and lookup stock if needed. You felt warm and appreciated and even though I seldom purchase physical books now, I have repeatedly gone to that store simply to browse. I felt good, I felt human, I felt wanted.

Adventures in waterproofing 3D prints

I’ve started printing 3D vases which look really nice in metal filaments.

3D prints are prepared by making the shape in software and then ‘slicing’ it into layers which can be understood by the printer. Usually this means that several layers are placed on the bottom, the side walls are built up by multiple passes of the print head, the top is a repeat of the bottom and the middle bits which are not seen are filled by ‘infill’ in special patterns. All this makes for strong objects.

There is a special mode of slicing which ensures that the print head never lifts off the work being printed, and never ‘retracts’ the filament nor moves backwards. This is sometimes called ‘vase’ mode and makes the print into a long spiral all the way to the top. There are few objects for which this works properly since it makes the sides only one layer thick and there is no infill. On vases or drinking glasses, cups this is ideal however and with the proper sized nozzle and good filament the results are great.

One issue is that usually these are not waterproof. Filament and deposition manufacturing by their nature leave microscopic holes in the print and while tuning the slicing parameters and the printer can have some effect, generally it is just luck when things hold water.  I’ve decided to look into lining the inside of objects to see if their water-tightness can be improved and so far I’ve read or seen of a few ideas:

  1. A wax or waxy substance rubbed into the surface then allowed to dry,
  2. liquid latex, of the type used for face painting,
  3. PVA wood glues of different types,
  4. general purpose epoxy resin, sometimes used in the fibreglass manufacturing industries. One I have found is called ‘Epolam,
  5. acrylic conformal coating used in the electronic industry for coating PCBs,
  6. … or general spray paint coatings.

Of these, the conformal coating sounds good but is very expensive. I’m going to try some of the other ones first to see if they are effective and attractive, starting with the cheapest which seems to be the wood glue and liquid latex. I wouldn’t want to mar the appearance of a beautiful vase by having a thick visible coating, but I also don’t want that vase to leak.

Redemption

The roots of the word ‘mortgage’ mean something which is paid until complete. Some macabre definitions float around the internet including those which take the old French or Latin and read them in English, which is always a dodgy way to interpret languages. I saw ‘death-pledge’ being touted as one due to ‘mort’=death and ‘gage’=pledge, which sounds too simplistic.

Different countries also have other expectations about how to finance buying a house – I was surprised to learn in Turkey that colleagues of mine there would take a loan from the property developer, and repay in a matter of years rather than decades. I’ve always lived in countries where the loans are taken through financial organisations and paid over decades with low interest rates that compound because of the extreme length of time involved.

The process of removing this debt is called ‘redemption’ and is removal of the pledge-until-payment and release of the deeds which show title in ownership. While that is nice, it feels a little medieval and reminds me of the long history of land title and ownership. Real land reform is quite rare and has mixed outcomes, not always helping either the rural poor nor the lower classes. I do admire the land reforms in Scotland which include returning tracts of land to common or community ownership.

It is nice as I redeem my own house to learn of the different ways that governments and rulers – some quite surprising – have redistributed land and tried to encourage greater wealth and ownership.