SianaG

I know it's an old question, but you asked someone to reply if they stumble along, so here goes. You need to ground the SENSE pin of MicroUSB socket on OTG processors such as the one on Raspberry Pi, else the processor's USB controller is configured as peripheral endpoint port, not as a host controller.Also make sure to not confuse the fundamentally somewhat misnamed D+ and D- pins.Additionally USB HID devices, wireless dongles or directly wired, are LowSpeed devices with relaxed signal integrity requirements. Bluetooth dongle should be a FullSpeed device with somewhat more stringent requirements. It is recommended to not use loose wire but compliant cable and untangle only the last cm or so, as little as you need. If using loose wires, it's a good idea to twist them together. Length ma...

I know it's an old question, but you asked someone to reply if they stumble along, so here goes. You need to ground the SENSE pin of MicroUSB socket on OTG processors such as the one on Raspberry Pi, else the processor's USB controller is configured as peripheral endpoint port, not as a host controller.Also make sure to not confuse the fundamentally somewhat misnamed D+ and D- pins.Additionally USB HID devices, wireless dongles or directly wired, are LowSpeed devices with relaxed signal integrity requirements. Bluetooth dongle should be a FullSpeed device with somewhat more stringent requirements. It is recommended to not use loose wire but compliant cable and untangle only the last cm or so, as little as you need. If using loose wires, it's a good idea to twist them together. Length matching is a good idea, but within a mm is plenty. But usually this is all reliability wankery and for the most part, even remarkably crude connections usually work or would work at least some of the time.

Nothing speaks against getting lenses from a fitted frame and designing a new frame that has the exact same glass positions though, but it can be adjusted in the temple and nose area for more comfort though, right?Then again lots of things would likely work for me personally, because my lenses are essentially really strong horizontal cylinders, no spherical strength, nearly zero angles.

You are quite correct. The highest capacity 18650 size cells are approx. 3500-3600 mAh, made by Panasonic (ex Sanyo) and LG Chem. Which is honestly quite impressive already, i mean anything beyond approx. 1600-2000 is no longer terribly trivial, which is about how much you can expect from cells made by Chinese chemical companies, no matter how much they exaggerate.I'm not sure i'd take those cells of unknown origin even for free, i mean what it they grow a lithium dendrite inside and short out? There's still plenty of energy in there to make the consequences uncomfortable.

Perhaps instead of parallel, you accidentally connected them antiparallel, i.e. plus to minus and minus to plus. This is actually a short circuited series connection, causing the maximum possible current to flow, as much as the cells will emit. However, between the common terminals, the voltage is close to 0V, so there was no electrical damage to charge controller.That the wires got red hot means that the cells took thermal damage too. Toss them away, you can no longer use them.Another possibilty is much less likely, that one of the cells gained high ESR and another low, due to repeated series charging without full balancing, causing a difference in apparent voltage when they were connected, which caused compensation currents to flow. But it's difficult, nearly impossible for that to o...

Perhaps instead of parallel, you accidentally connected them antiparallel, i.e. plus to minus and minus to plus. This is actually a short circuited series connection, causing the maximum possible current to flow, as much as the cells will emit. However, between the common terminals, the voltage is close to 0V, so there was no electrical damage to charge controller.That the wires got red hot means that the cells took thermal damage too. Toss them away, you can no longer use them.Another possibilty is much less likely, that one of the cells gained high ESR and another low, due to repeated series charging without full balancing, causing a difference in apparent voltage when they were connected, which caused compensation currents to flow. But it's difficult, nearly impossible for that to occur at such currents as to heat up the connections red hot, because then ESR is current-limiting. What normally happens is that it either levels off or the battery depletes itself and becomes useless, but no extreme temperatures.

Have you had luck with your project? I am not familiar, but it appears the display in such vehicles is usually not an LCD of any kind, but a vacuum fluorescent discharge tube display, VFD. It's a collection of thin shaped glass vacuum tubes that are coated with a Luminophore on the inside that determines their colour; correspondingly, while they can be manufactured in various colour, their colour can not be changed afterwards. I am not aware of any possibility to get a custom one-off VFD made for you.It's possible to reverse engineer the signals driving the display, and then create a drop in replacement, based on a microcontroller interpreting the signal and driving a graphic display of some kind, an LCD or OLED. With OLED, there is the question of suitability to vehicle use. The displa...

Have you had luck with your project? I am not familiar, but it appears the display in such vehicles is usually not an LCD of any kind, but a vacuum fluorescent discharge tube display, VFD. It's a collection of thin shaped glass vacuum tubes that are coated with a Luminophore on the inside that determines their colour; correspondingly, while they can be manufactured in various colour, their colour can not be changed afterwards. I am not aware of any possibility to get a custom one-off VFD made for you.It's possible to reverse engineer the signals driving the display, and then create a drop in replacement, based on a microcontroller interpreting the signal and driving a graphic display of some kind, an LCD or OLED. With OLED, there is the question of suitability to vehicle use. The displays aren't intended to run for hours on end, they degrade with use. Also their brightness is typically 10 times weaker than VFD. Also very few colours are available, you may be stuck with a colour that is more similar to that of VFD blue than your chosen LED blue. With LCD, there is an aesthetic issue with lowered contrast and potentially bad viewing angles. Monochromatic LCDs usually allow replacement of LEDs lighting them, and also LCDs with adjustable RGB backlight are available. An advantage of using a graphic display on a microcontroller is that you are not stuck with static symbols, you can adorn them with effect graphics, such as the frame of PRNDL gliding across the display or magnifying the currently active mode, beautiful, easily legible font for your odometer, etc.I think the most promising course of action is getting an SMD LED display custom fabricated. You can likely drive them with simple electronics just to reduce the current and voltage from the original VFD signalling. While you can get numeric 7-segment LED indicators, you may need again signal re-encoding with programmable logic ICs or such, and there's a possibility of colour mismatch. What i'm thinking of being most promising, is laser cutting a light barrier from black Delrin or acrylic. That can then do everything at once, both your PRNDL and your numbers. You can fill it with diffuser epoxy and set it on your custom designed LED board. You'll need a manufacturing technique like that anyway for your custom elements, and if it's automated, it won't matter whether you're doing a tiny element or a whole screen at once, and it won't matter whether you're cutting a handful of shapes into it or a hundred, and you get a uniform result.Do you have a photograph of your display at hand so i know whether i'm thinking in the right direction? I am thinking of a display which has PRND321, a set of 7-segments for odometer, a couple of extra symbols around, but no text capability. If you have text dot matrix display which has distinct single-pixel spaces between the characters, it's likely an HD44780 signal-compatible VFD, and you can get HD44780 LCD replacements. They come with width of 8, 16, 20 and 40 characters and height of 1, 2 and 4 lines.

Curious. I have not considered the possibility that they would use a colour gel to tint the display. Or that one could do that afterwards.Message center is a monochrome graphics LCD. You shall probably find some SMD LEDs on the board beneath it, that can be replaced. I am not familiar with specific display model employed, nor are there any data sheets - it has apparently been manufactured by Optrex Corporation, Japan, specifically for use by a division of General Motors, and has not been available generally or used anywhere else. Most companies wouldn't be caught dead ordering or a 45x28 LCD. For reference, the display model designation is DMF-50796H. If it's anything like i imagine it to be, there should be a metal shield retained by twisted tabs on a PCB, and merely removing the shiel...

Curious. I have not considered the possibility that they would use a colour gel to tint the display. Or that one could do that afterwards.Message center is a monochrome graphics LCD. You shall probably find some SMD LEDs on the board beneath it, that can be replaced. I am not familiar with specific display model employed, nor are there any data sheets - it has apparently been manufactured by Optrex Corporation, Japan, specifically for use by a division of General Motors, and has not been available generally or used anywhere else. Most companies wouldn't be caught dead ordering or a 45x28 LCD. For reference, the display model designation is DMF-50796H. If it's anything like i imagine it to be, there should be a metal shield retained by twisted tabs on a PCB, and merely removing the shield should give you access to the LEDs.An interesting hack, with possibility of grand destruction, is inverting the LCD. On top of the glass is usually a polarization film that can be peeled off and replaced with one mounted 90° off-angle. Tough luck if the filter is instead sandwiched between glass layers.Apropos generating an NTSC composite static picture. There are a few possibilities that i have in mind.The simplest is probably taking the logic board from an old digital photo frame. This seems like the ideal use case for these devices, as they are really intended to cycle between images in their intermal or external storage. Ones manufactured before 2006-ish often have a composite display inside of them, and correspondingly also have an AV-out socket with the same signal. Sometimes, same trick can be applied to portable DVD players from that early era, if they have a USB or SD card socket, and you can remove the drive. Newer devices always drive the display digitally, but if there's an AV-out socket, you can still use them, but they can be extremely rare, it seems they died out 5 years ago. Challenge will be finding a device with AV-out, and one that doesn't have an annoying splash screen and goes straight to picture viewing.The cheapest possible hardware is Raspberry Pi Zero, although the availability is TERRIBLE. It has composite AV out on an unsoldered header pin. Challenge will be developing a custom operating system which boots in a fraction of a second and then just sits there displaying an image, by stripping away everything possible and rewriting the boot system. This kind of software job is actually right up my alley. Maybe they're losing money on it, i don't know, something like $10-$15 would seem like a more reasonable price than the $5 they officially offer it for, but even a full Pi isn't all too expensive. As to Pi alternatives, there's C.H.I.P. which is "$9" but so far vapourware, and Orange Pi which is like $25 and actually available, but as these are based on basically undocumented Chinese ICs, it might be difficult to cut the software down. It might seem cruel to chop down a fully featured computer as a single picture generator, but can't argue with the prices. All kinds of nominally simpler devices that i considered reprogramming for the purpose would end up being similarly or more expensive anyway, and much more effort.Then there's a bunch of possibilities with custom hardware, but honestly they don't seem all that attractive. Luckily the PAL/NTSC colour encoder ICs no longer cost $20, so at least it's not insane. Or maybe i.MX233.What is it that you're imagining? If a black screen or simple black and white low-resolution text will do, i think that's something i can probably do on a $1 board :)

Do i read you correctly, you're attaching a 12V battery (DC) instead of dynamo (AC)? Then yes, you don't need the rectifier portion. You also don't necessarily need the capacitor C1, because the battery has a low enough ESR, but if you already have it, you can leave it in.In order to charge smartphones and tablets via USB, the port that you connect them to must identify as a USB host or as a charger.To implement simple charger identification according to USB 2.0 SDP (limit: 500mA or 100mA, device dependent), connect D+ to GND via 15 kOhm resistor and D- to GND via 15 kOhm resistor.To implement simple charger identification according to Battery Charging Specification, connect D+ to D- straight or through 200 Ohm resistor (depending on version), current limit: 1.5A.To trick Apple devices ...

Do i read you correctly, you're attaching a 12V battery (DC) instead of dynamo (AC)? Then yes, you don't need the rectifier portion. You also don't necessarily need the capacitor C1, because the battery has a low enough ESR, but if you already have it, you can leave it in.In order to charge smartphones and tablets via USB, the port that you connect them to must identify as a USB host or as a charger.To implement simple charger identification according to USB 2.0 SDP (limit: 500mA or 100mA, device dependent), connect D+ to GND via 15 kOhm resistor and D- to GND via 15 kOhm resistor.To implement simple charger identification according to Battery Charging Specification, connect D+ to D- straight or through 200 Ohm resistor (depending on version), current limit: 1.5A.To trick Apple devices into charging at 500mA, place 2.0V on both D+ and D- via suitable resistor voltage divider networks. Aim for total resistance between GND and +5V no less than a few thousand Ohm.To trick Apple devices into charging at 1A, place 2.8V on D- and 2V on D+.To trick Apple devices into charging at 2A, place 2V on D- and 2.8V on D+. Note that you shouldn't do that if you use 7805 as your current controller, because it can only supply 1.5A. The voltage will drop, and the Apple device will detect a charging fault, and will then refuse until it's connected to another charger.Most Android and Windows devices will be OK with an Apple charger as well.To confuse yourself even more, please refer to MAXIM Application Note 5801.

I think "600mA" is a filty lie and you shouldn't trust it. I think it'll do closer to 100mA on average, or less.