The Development of a Gauge System

I've put this here as it's relative to Glenn K's Supra and the Syvec's ECU. When you buy something off the shelf and plug it all in and enjoy it's benifits, he is a small insight into the development cycle of someone doing it as an aside. As the full colour screens are in a very early prototype state, I'll start here, the old Mono or "Fail" gauges have already been well documented in another thread.

After the disappointment of the MONO LCD system certain obvious gaping holes in the concept were apparent. Sunlight blotted them out and there were hard to read when you could see them.

There was nothing else for it, OLED full colour, high contrast, replacement was the only real option, it's current and it ticks all the boxes. These units have a co processor that does all the drawing for you and has a 4gb SD card on board for logging and graphics storage. They can communicate via RS232, I2C or Dallas 1 Wire protocols with a base micro controller.

The base Micro controller talks to the Syvecs as well as capturing 8 channels of independent analogue (TTL 0v - 5v) data @ 10 bit resolution and talks to the Syvecs inputs to control boost, calibration and traction control selection. The serial packets are decoded, formatted and transmitted on the bus for the screens to decode and interpret accordingly (Like my own can bus system).

While the OLED screen modules would appear to be the utopian answer they have some serious flaws. The first is lack of RAM (small screens have 510 bytes) and screen memory is direct accessed and it's not possible to double buffer the drawing and swap them. As a result anything you draw and move about flickers like a bastard. They suggest you play movies selecting frames to simulate a gauge (so say 360 frames to simulate a needle moving around), I may have too,however, with all the permutations of gauges that would eat my 2gb graphics space quite quickly. I need to be able to render a software needle over the gauge graphic and come up with a cunning horizontal raster drawing scheme to eradicate the flicker. This I have done, I should have the working gauge running by Friday evening having done the donkey work last week. A video will be taken and posted up.

One of the usual big head aches is low level access to the hardware. The manufactures will no doubt protecti their IP by only allowing access to the functionality though their libs. I reckon I could treble the speed by being able to hit the hardware direct. We will have to wait and see.

Here is a picture of the my current test gauge:

http://www.nod-land.co.uk/assets/images/Dsc00112.jpg

This was downloaded from the net from a picture (it's an AutoMeter) and messed about with in Paint Shop. I've replaced the numbers with hand drawn ones now as these one anti alias really bad and the needle is suffering raster shear.

Mark at life racing suggested CAN is the way forward when talking to the Life/Syvecs products. Looking at the can message packet structure of the F88 and the info therein (There is even turbine speed for Wez ), I have to agree. Can 2.0 to SPI adaptor is 21 euro's so there is no longer a reason to avoid it

Am very impressed by this

I'm currently in talks with 4D Systems (The people who make the OLED module in Oz) who for the record are very approachable and react fast to their forum helping their customers. I'm hoping they will grant me access under the bonnet and then there will be some pixel magic happening. They are sending me a pre production of there newest development, they haven't told me what it is or what it does so I'll have to wait and see, I'm rather excited.

There was a small issue with one of the drawing commands with regard to the SPI bus being closed after execution. I was expecting it to remain open for subsequent drawing on a continuation basis, but unfortunately it resets the address back to the start which means having the set the address manually each line and that causes a massive slowdown. Hence no movie and why it's not finished. 4D are going to alter the function so it works to my spec as they agree it would be better that way (I never expected this to honest, full marks to 4D systems for not letting Ego get in the way of progression).

I now have Power FC comms information and a cheap source of adaptors for use with any Power FC software $170 rather than $350 for an FC Data logit.

I thought it worth mentioning that these products are being developed in conjunction with Ryan @ 2 Bar Tuning and will be advertised and sold through his Trader Account. My posts are purely for your information. I don't want other traders thinking I'm taking the urine by circumventing the trader policy.

I thought it worth mentioning that these products are being developed in conjunction with Ryan @ 2 Bar Tuning and will be advertised and sold through his Trader Account. My posts are purely for your information. I don't want other traders thinking I'm taking the urine by circumventing the trader policy.

This really is great work, Lyndon. As soon as I get some spare time, I'll be in touch to finalise a deal with you. Which reminds me, I still need to speak to Ryan. ARRGGHHH!!

Great work Lyndon

Have you seen the can gauge offerings from ProEFI?

http://www.proefi.com/Pro_Can_Display_Ver.html

http://www.proefi.com/Can_Gage_-_52mm.html

http://www.proefi.com/ProEFI_Can_Display_Ver.html

I've been keeping an eye on ProEFI over the past couple of years, seems to be some nice developments coming out of that camp.

Lyndon - do you think your gauge system could be adapted to read info from OEM diagnostic ports?

Have you got that guage hooked up to a laptop? whatever it is, it is boosting at 5psi!

Have you got that guage hooked up to a laptop? whatever it is, it is boosting at 5psi!

OEM OBDII should be possible eventually. Concentrating on Syvec, AEM, Link and PFC to start with.

The gauge just does a 360 sweep for render testing, it just happended to be there when I took the picture.

OEM OBDII should be possible eventually.

That's something I would be interested in

Small Update

After a bout of ill health and one thing an another, I've regained some focus on this project. After lots of sucking and seeing, the current plan is to build these into a Supra Dash replacing the Speedo, Fuel and Water temp gauges with 3 screens, 1 x 3.2" and 2 x 1.5" colour TFT screens leaving the warning lights and Tacho dial as is. Metal Monkey will be making the fascia plates and I'll try for a white EL back lit RPM gauge so the light is evenly distributed what even colours are chosen (Stainless, White, Blue, and Black fascia colours are now available).

I hope to have Glenn K up and running by the end of Feb on the 3.2" TFT Speedo Dial with the Fuel and Temp to follow. Where injector duty cycle info is available Range and MPG functions will be available. This requires CAN BUS on the Syvecs which is also in hand.

There will be a stand alone touch screen 3.2" Syvecs Can commander Interface available later this year.

There will be a stand alone touch screen 3.2" Syvecs Can commander Interface available later this year.

Do these first bud as i have a que of people after them!!

Really looking forward to the dash clock replacement gauges

Lyndon,

It's in my diary!

Keep up the great work.

Cheers,

Glenn

Do these first bud as i have a que of people after them!!

I'll need the loan of another S6 to do the can bus data capture on

Default Can packet configuration has a lot of stuff in it that is not needed but is missing a few vital ingredients. I seem to remember you saying the can packet information structure can be user defined, is that so?

I'll need the loan of another S6 to do the can bus data capture on

 

Default Can packet configuration has a lot of stuff in it that is not needed but is missing a few vital ingredients. I seem to remember you saying the can packet information structure can be user defined, is that so?

Will get one sorted in feb mate as we get like 50 units in.

The can can be user defined, you are correct

is there any further development on these yet?