A|C Controller \ Console Display driver

[poor_lackey]

I finally had a prototype board running in RX yesterday - figuratively speaking as more running on the seat\console
but A|C controls, Hazards, Dimming all properly plugged in and functioning with standard OEM wiring harness(es).

I still need to update the code to allow for recovery from low power\sleep mode
and need to add menus for setting the clock, testing buttons etc;
So real world hardware\soak testing has to wait a little longer as I only have 1 set of cables.
Also to be honest I forgot to purchase the female end of the power connector (which explains the wires in the lower R/H corner) and need a couple more bits for fine tuning.
A couple of communication RX\TX pairs reversed (the result of a (L)user issue) otherwise I'm very happy with the results.
The sleep mode testing is exceeding expectations with a current draw of 200microAmps +/- (from my budget mult-meter)
so fairly confident that it will never be the cause of 3-4 week flat battery syndrome.

[30405]

Looking good!

[HarSc]

Indeed! That's awesome!

[BarbedWire]

Is this intended to be used with the Metra double DIN kit instead of the supplied electronics?

Sent from my Pixel 4 using Tapatalk

[poor_lackey]

Is this intended to be used with the Metra double DIN kit instead of the supplied electronics?

It can be used with any DIN kit, or none, the controls are self-contained and have no dependency on there being a Stereo.
From experence the cheaper DIN panels EG Ali-Express & E-Bay may require a little filing of the A|C vent panel and trimming of the hazard light switch before they properly.

The board is designed to sit directly on top of the A|C control panel (without a case), so any standard DIN stereo head should fit over the finished assembly nicely. However as a result there is a small compromise, and a little care would be required inserting\removing the head ( if you fit the A|C controls and board first).
- Slide any double DIN head in\out level IE don't drop the back of it below horizontal and onto the PCB.

[350matt]

thats looking very professional
I'm assuming that you intend to sell this as a plug in kit once your development is complete ?
does it need any coding / ardino set-up or will it come as plug and play ?

if its the latter then count me as interested

[poor_lackey]

thats looking very professional

Thanks, I'm pretty please with it myself.

Low volume production is an option, but that wasn't the intention when building it, and whether it's economically viable is a whole other story.
It will be down to cost to produce\deliver x units verses demand, I suspect anything under 100 units would not be cost effective and 100 may be marginal. There are companies that specialize in manufacturing\drop shipping small quantities, but the prices I found when looking for prototype assembly were not as sharp as I'd hoped and in the end I brought the bits and assembled this one myself.

I will be making a couple for a few locals to test in the next few weeks and once there is some real-world feed back (a finalized design), I can look into production costs if there is interest - hopefully there will be a price point where interest and expense balance out.

If there was build run, loading the code would be part of the 'test' procedure so boards would be ready to go out of the box with required cables.

Speaking of which
Does anyone know what sort of connector the cigarette light lamp use?
I can see the bulb holder is a Toshiba unit, but couldn't find matching male\female cable connectors anywhere and if male\female connectors are part of the power wiring loom then there would be no need to cut into the dimmer circuit wiring - and I'm a big fan of not having to cut things.

[350matt]

most of the electrical parts on teh car are made by denso and Sumitomo if thats any help
and I've found this firm in china
https://www.kinkong-connector.com/

make the connectors for Sumitomo so there's a strong chance they'll have what you need - they are happy to send samples for a small fee too

[poor_lackey]

My RX has been sitting in the garage in pieces for weeks (so I could use the A|C cables for testing) and while I don't drive it daily I am a little impatient, so fitted the controller yesterday ( even though the code is a bit buggy with the icon refresh and I still haven't brought a proper power connector).

board_pre_fit.jpg

board_plugged_up.jpg

All of the cables are OEM and require no modification, except:
The hazard light switch cable needs extending, oddly the connector manufacturer does not sell an inline male connector, although they can be seen advertised on ebay and ali-express - I am working on a source of these.

All power/dimmer feeds can be spliced directly into the stereo loom adapter, if the adapter provides pins\wiring – but it seems that many of the 'standard' rx8 stereo adapters don't allow for dimmer (or steering wheel controls) and also looking for suitable option(s).

I have a couple of board changes I am considering;
When I measured the available space I did so with the OEM fascia, and didn't think to check aftermarket items. Something I should have thought about as the double din unit I have is flatter and the board is a firm fit, trimming 2mm off the depth will be possible but will take some juggling.

Also last week I tested by wrapping the board in a chemical body warmer for 2 hours,
zero air flow\circulation @ 55c with the display drawing 20% more power than it can in normal operation.
The board got warm but not overly hot (it was well within acceptable parameters) but I think I can do better, either a different regulator or a heatsink (now I have board fitted I can see there is plenty of space between it and a standard double din stereo).

Next step is to find someone local for a side by side comparison of the dimmer and A|C led brightness, and a trip to the country (no streetlights) as I find the OEM display way to bright in low light areas.

dimmer.jpg

[warpc0il]

Years ago a work college put in many hours developing a carputer system to got into his BMW Mini Cooper, all very fancy and hooked into the CANBUS so he could display anything and even control individual lights and stuff.

It all worked great on the bench, hooked-up to his 12V supply system, and then when fitted to the car and demonstrated.

By the following weekend it was toast

For such a bright guy he'd fallen into the trap of thinking that a car runs at 12V, while most of the time the supply is more like 14.8V, and at start-up can go below 6V. While his very expensive bench supply had been 12V +/- 0.2V between zero and maximum load.

How are you managing the variation in supply voltage?

[poor_lackey]

I feel your colleges pain, and this is one of the reasons I moved away from off the shelf MCU boards like Arduino.

I think (hope) I have covered off most dooms day scenarios,
but nothing is indestructible if you try hard enough, or Lucas prince of darkness is against you.

I literary spent 40+ hours (I wasn't counting and it could have been 80+) looking at 1000's of voltage regulators that are available
and that would work within the available space (and possible heat constraints).

The unit has 2 primary regulators with operational input voltage up to 40 Volts, both are automotive rated and are according to the manufacturers spec's designed to cope with load dump\worst case scenarios.
- 40 volts operational would pretty quickly result in thermal shut down\protection as the board isn't designed to dissipate the resulting heat.
These power inputs also have TVS diodes and reverse input protection.

The regulator that powers the MCU, from the battery 24x7, will operate reliably down to 3.5V in,
- the USB power feed also routes to this regulator, so code can be uploaded without vehicle power.

The main 8V regulator is primary for the display and may eventually be 7.8v (or less) as it will dictate the max display brightness.
This is only powered when vehicle is ACC or running, and should be stable down to 9.3-10V depending on final output voltage.
- I figure if the battery is so flat that the RX won't start, then there is little point in the display being on.
(edit) If this voltage drops below minimum whilst starting, the display will just reset itself once the vehicle is running.

There are 2 smaller low current regulators that provide 5V for the display and 3.3v for the A|C amplifier buffers.
These regulators only power hardware that is needed when the unit is running so they are daisy chained off the 8v regulator output
and as a result inherit all the 8v regulators protection features.


Note:
I am considering replacing the existing 8V regulator with the same model being used as the main 3.3v reg.
There are pro's (space) and con's (cost) and I need to hook one up and bench test the heat generation before deciding.

[warpc0il]

I should write a book (if I wasn't dyslexic) on the months of work it took to engineer a stable 115V 400Hz 3-phase power supply, on a Hawker Hunter T2, which is a 28V DC aircraft, limited by the hardware that existed, and was flight-approved, in the 1970's.

We had to create a platform to flight test the integrated radar and avionics for the yet-to-be-produced Sea Harrier, which was all very different to the existing Harrier GR3.

The highlights involved;
- an encounter between a ground crew headset and a Rolls Royce Avon engine running at full power.
- a rotating mirror chart recorder
- a prototype Tektronix storage oscilloscope
- two 8ft long lead-lined plywood coffins
- a 25lb Howitzer field gun (working)
- lots of sleepless nights
- and a very brave test pilot.

My Boss was originally given the project, which he assumed would be easy but he hadn't considered the interaction between the natural harmonics of the mechanical carbon-pile DC voltage regulators, and the response time of the load regulators and inertia of the rotary inverters used to produce the AC supply.

When his first efforts nearly resulted in the loss of the aircraft, including the RR Technician in the cockpit, while ground testing at Boscombe Down, the project was handed to me.

I was given the original "mule" with the "28V" DC power rails "swinging from +/- 160 volts" as shown on the chart recorder, which was struggling to keep up, and went on to prove that it was nearer +/- 300V using the 'scope; no wonder the engine control systems surged.

Who knew that the NATO part number for the carbon-pile regulator used on the Argosy aircraft has exactly the same digits as a Ordnance QF 25-pounder gun, just in a different order; and it wasn't me that transposed those digits or created a process that ignored the text description in the original requisition...

Very long story short, the combination of changing the regulators, and some tweaks to the inverters produced a result that was self-dampening rather than run-away feedback, but not before a bit of magic smoke had escaped from earlier attempts.

If only we'd had access to the sort of electronics available in the 1980's, let alone today.

[poor_lackey]

I guess they didn't let you keep they let you keep the 25pounder as a souvenir.....?

I know what you mean about electronics and the massive changes since the 80's,
although I think the magic smokes a little too eager to escape in the newer stuff (and I can hardly see most of it, even with a magnifying glass).
Luckily I've little chance of my controller causing me to plummet 20,00 feet, otherwise I might have second thoughts....

And yes write book, I'd read it and I'm sure somewhere there's an ex QM whose still has nightmares about his missing howitzer so he might read it too.

[poor_lackey]

Quick update, just in case anyone's interested.

Test board has been running in the RX for over a month (but it doesn't get used daily).
Currently resizing the physical board (original could have been a bit tight for a flatter stereo bezel).
While I'm at it, working on a switching regulator supply for the LED\display, as the current regulator results in the board getting just a little warmer than I'd like and with it about to loose 3mm width (that's probably 15% + of it's ground plane\heat sink capacity) I think this could impact long term hardware reliability.
The manufacturer of the regulator I picked (based mainly on price and availability) has a design review service and I will be asking for input\feedback from them in the next week or 2 as switching regulator design\layout is primarily black magic (with maybe a touch of voodoo).

Did knock out a trip to the sea side in June of 100+ mile each way (mainly on winding hilly roads) so a good high heat workout.
Small bug with the A|C display occasionally showing incorrect icons, this self corrects and doesn't impact functionality but I will fix it while testing hardware V2.
Otherwise display brightness is a-ok (even in pitch black), dimmer function works fine and the A|C controls are just as responsive as the OEM.
Also managed to identify the manufacturer of the non standard connector required to extend the hazard switch.

So progress, all be it slower than I'd like.

[Phil Bate]

This is excellent work, keep it up

[pfletcher83]

Excellent news i will be first in line for one when they become available

Paul

[poor_lackey]

NZ dropped back into covid lock-down last week, and while I'm 'lucky' enough to be able to work from home; I was quite please to find the last of parts for V2 appeared randomly on the front lawn on the first day .
So apart from disassembling the RX8's console\stereo yet again....
I spent some time at the weekend knocking out a test board, and plan to have a couple more ready for delivery to volunteer testers when we are next allowed out to play.

The switching regulator has made a huge difference temp wise, although yet to test the impact on clock accuracy (the original drifted 3 minutes +/- in a month), the processor also has a registry for offsetting clock drift and I will add a menu option for fine tuning the clock along with an option for personalizing the display 'TEXT'.

The next big problem will be affordable low volume production.....
There are nearly 70 components on a board and it takes me 30-40 minutes just to lay out the parts individually (there's a lot of double checking goes on and it's very very de-hydrating), assembly is a whole other world of pain.

IMG_20210823_192255504.jpg

[poor_lackey]

Slow progress, but any progress is good progress ....
I have a handful of boards ready for real world testing and most of the code bugs sorted.
The (current) global silicon chip shortage has caused a bit of havoc with planning for future development,
so not yet sure if there is a solution other than it being open source\build one for yourself, but will look into low volume production if enough interest.

I have started to knock up a user guide (menu attached) and made a b-grade video of a couple of features.
I have though about adding a default setting for the A|C Mode, temp, and fan speed but not really sure if it's needed?
Feed-back/thoughts appreciated.
Tony

[hachiroku]

Fascinating! Excellent work
Open Source/DIY might be the way to go, I remember having fun with one of the early Megasquirt ECU kits, it'd keep price down and avoid laborious building on your side.