Our products to date have been EMX based and used only a 16x2 LCD character display, and until a few days ago I hadn’t even looked at GLIDE :-[
However a potential new client is very keen on a touch screen and WiFi, so I brought out the EMX Dev system for a play, with the idea of putting together a new G120 based board with support for Redpine/ESP8266 WiFi and a manageable display of around 3.5" - 4.5"
Using a nominal 3000mAh 18650 (Pink LG) cell and LDO supply, our EMX based board presently draws around 200mA and has a 10 hour run time. I really don’t want to have to change the power supply setup, or lose too much run time.
I’m thinking of using a serial display (SPI ?), mainly because I find all those close spaced wires a headache when hand-assembling runs of maybe 5 at a time, and also I have a Voltera on order and I’d like to be able to prototype with it! So I’m hoping to find a low-power display which is proven to work with the G120 and GLIDE, with minimal interface wiring required. The display will be fairly static, although some real-time graphing may be called for later. Choice or capacitive or resistive touch would be nice. Can anyone recommend something?
Speaking of GLIDE, on the forums I saw there was a designer for it originally called Motif and then renamed Visual Design Studio, but I can’t seem to find it anywhere, does it still exist?
Not sure about a display but there are a few options with SPI but I would recommend the full RGB interface for speed. I’ve just received a Newhaven 4.3" sunlight readable display (830cd/m3) with cap touch and it’s absolutely stunning visually. Low resolution compared to the 800x480 I am used to working with but for some applications that need to be outside, this is going to be a nice solution. What is your power budget for the display?
As for the Glide editor, have a look on here for Glide Studio by Blackdogspark. It’s not 100% and can crash (entering a width or height 0 normally does this) and it does slow down after a while of use but as long as you save often, it really does make creating displays for Glide a breeze. I have contacted the author and once he sorts out some home issues, he promised to make it open source so we can tinker and sort out the issues.
I was really hoping for something less than 50mA, the power supply could handle an additional 200mA or so for WiFi and display combined, but the battery endurance drop off is non-linear with current and around 200mA is the sweet-spot for the LG Pink cells, the best I have been able to find. We can’t use inductors due to the magnetic fields, and the LDO is ok down to 3.6V but 3.3V regulation starts dropping out at 3.5V with a 200mA draw.
LDO’s or buck regulators are not ideal for using with battery supplies if you want to get down to the lower voltage end of the battery. A buck/boost type is the ideal way to do this. It will provide a constant output with higher or lower voltage input.
Have you considered one of the few e-ink displays that come with touch? If your display isn’t update often it might be more suitable? They are slow though for refresh so not ideal if you are updating something on the screen in real time.
@ Dave McLaughlin -
We used buck/boost regulators on our early boards, unfortunately there were some areas where they would fail due to the magnetic field saturating the inductors’ cores, despite magnetic shielding. Being within a few inches of a 320,000 Amp DC line will do that!
Using an LDO and running from a fully charged 4.2V down to 3.5V gives us the bulk of the energy in the cell, and a full-power run time of around 13 hours, using the LG Pink, which has a different discharge curve to most 18650 Lithiums. A standard TrustFire black or similar may only give 1 hour in the same test!
There are a few inductorless boosters around, but not with sufficient current for our needs.
It was only recently I thought of using a super-cap to bridge the time the device is in the peak field areas, something that could be worth giving a try.
Anyway at this stage I’d like to be able to pitch a low power touch display as an update to the existing design with minimal changes (= minimal risk). But as an alternative I’m thinking to leverage the display/wireless/power efficiency/cost of a mobile phone (I’ve seen new Android Kitkat phones for $23) and reduce our board to a low-powered slave doing the analog inputs, buttons and LEDs, connected to the phone by BlueTooth or USB.
Thanks for the pointer to Glide Studio. I’ve had a few "Program has Stopped Working"s with it, have to learn what to avoid! But if we can get hold of the source and fix those it could be very useful.
I did look at some of those. Problem is I don’t know exactly what the client wants to use the display for. Maybe he’ll throw in a moving colored graph at some stage. I’m planning just to quote on providing a graphics display which can mimic our 16x2 LCD for now (he wants a price by Friday) and leave the details for later.
Did you ever try toroids (sounds like a new candy)? They radiate/pickup less (perhaps 20dB) than typically wound inductors.
@ Hoyt - No, we didn’t, and I’m sure that by playing around with different coil configurations and shielding methods we could have found something that worked, but the logistics of doing on-site testing wasn’t good, and the LDO was the best option at the time. The magnetic field is due to a large DC current, and has the effect of saturating the inductor core, it isn’t an AC radiated effect. It also kills auto ignitions, wipes credit card magnetic strips, stops pacemakers, etc.