Jump to content
LegacyGT.com

Flinkly's OBXT


Recommended Posts

This part is especially interesting. Might be nice to see some data from your setup; if you plan on sharing, of course! :)

 

oh, i'll share.

 

i wish i could get someone to provide a set of cast manifolds, and an uppipe to convert my TS header to a SS turbo inlet. Would love to datalog a tuned VF52 on OEM headers, then ELH headers, then the VF52 TS on ELH headers. get data from SS OEM, SS ELH, and TS for comparison. but that's two tunes more than i need, 1 custom welded part, and a set of OEM manifolds w/bungs installed. and on an old platform that most don't really care about. :)

 

As for the EGT sensors, haven't figured out how i'll datalog it fast enough yet. Can get a $600 logger, i guess, but have better things to spend $$$ on. I'd want at least two datapoints per rotation, giving me 8 datapoints per period, and moar would be cool for resolution. So 5 inputs (4 egt and either time or another car sensor, to link romraider logging to the separate egt logs...), lets say 7k rpm max, so ~600 hertz or more. Alot for arduino logging to sd thru spi. Will need alot of buffers, a fast arduino, and to be a better programmer.

 

in this case, installing sensors is the easy, and even cheap, part. :)

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

  • 2 months later...
  • Replies 146
  • Created
  • Last Reply
  • 2 months later...
Custom Cubby 4x Guage Pod - Version 2.0!

 

So, with efficiency in mind i switched to 4x little OLED screens, re-purposed from the old-school phone industry i believe. Since Adafruit didn't have a large enough OLED to really fill the cubby up nicely, i decided on 4x smaller screens. One of the HUGE benifits of the OLEDs (or LCD etc.) is that i can display the NAME of the sensor above the sensor value. before i was really scratching my head trying to figure out how i'd get little letters on the faceplate of the 7-segment version above. NO MORE!

 

Anyways, as time permits, i'll stop by again and list out what I'm going to get it to do, and hopefully will have one installed in my car in 2 weeks (depends on the shop who's making the new cubby really, and the holiday comin up).

 

 

 

http://legacygt.com/forums/attachment.php?attachmentid=219475&stc=1&d=1447801275

 

Any updates on this? Did you get it all working?

Link to comment
Share on other sites

i did have it working and installed in the cubby...

 

then i ripped it all out to (mostly):

 

- switch the 1/8" plexi cover for ~1mm thick glass (looked like crap.)

- add red film to change from white OLED to red (match car)

- add more program functionality...

 

 

I've actually got the CAD for the new faceplate ~90% complete. changed to more of an individual gauge look for the setup, designed the bezels to be aluminum, then a few plates of something (3d print, aluminum, whatev), custom cut glass (re-purposed microscope slides ;) , cut by hand - low tol.), a sheet of red transparent plastic used for picture/video lighting, and also figured out that datalogging it all super fast was useless (spent ages trying to do buffers, faster writes, etc - then figured out sensors aren't that quick...).

 

all in all, like all my projects, it never ends. get it working only to tear it apart to make it better. will get some new red tinted pictures up soon. since you asked, i'll finish the CAD and get it sent to the shop for manufacture.

 

thanks for the nudge

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

i did have a terrible picture of it at night lit up in the cubby. i guess i didn't post it. will try to find it and post. probably didn't cause it looked aweful as it must have been on my old phone camera - like 5 or 8 mp and terrible.

 

EDIT: guess i deleted the in-car photo. it was terrible, and was hard to see what was up. here was a picture of it together at my desk. Not sure why it looks so blue... i guess it is a high kelvin white.

IMG_20151209_174108487.thumb.jpg.98d57b589cdc159c682e252fbf93e26b.jpg

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

- PLX AFR gen 4 wideband sensor, and their sensor module

- a pressure sensor of the appropriate range from Dwyer for Fuel

- same for Manifold pressure/boost (broke my PLX boost sensor module...)

- an AEM brass sensor for oil pressure (before i sourced the others from Dwyer)

- and some type K thermocouple probes for EGT. 1/16th diameter, ~3" long, through a custom board with an AD8495 thermocouple amplifier

- and i was going to break into the stock coolant sensor to be able to log that as well.

 

probably won't watch the fuel, MAP, or coolant often, but i'd like to watch them at the moment for different reasons. Same with EGT. probably won't always have them ready to display, but working towards the option.

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

more detailed info, mostly for record keeping:

 

EGT Probes:

- 1/16" inch diameter Type K thermocouple probes

- Response time of ~.55 seconds (~4 combustion events @ idle, ~37 @ 8krpm)

- ungrounded (for circuitry/readings), 3" long probe

- planning to put compression fittings in header at exhaust ports

- routed into AD8495 thermocouple amplifier - to output 0-5V signal

- probe can handle temps up to 1300C

- $55 each for thermocouple

- $26 each for SS swagelock compression fitting

- $6.6 each for M/F mini high temp connectors

- $6.85 each for AD8495 chips, Z rating (increased accuracy)

 

Fuel/MAP sensors:

- Dwyer 626 316L sensors with Packard electrical connection

- Response time of ~.3 seconds, .25% FS accuracy, 0-5V output

- MAP is PSIA 0-50 psi (0-14.7 vacuum, 14.7-35.3 boost)

- Fuel is PSIG 0-100 psi (~43 psi nominal, raising to ~63 @ 20 psi boost)

- Made a custom SS/Al part for fuel sensor, not sure about MAP locating yet

- $135-$160 depending on sensor

- $28 for 9' packard plug (2 for these, got one more for oil pressure)

 

Oil pressure sensor:

- AEM 30-2131-150 150 PSIG Brass Sensor, also packard connection

- plan to install at front oil galley with Kinugawa Sensor Adapter M18x1.5 -1/8 NPT

- $65 for sensor and un-wired packard plug

 

PLX SM AFR gen 4

- uses Bosch LSU 4.9, so more robust than my earlier 4.2 (closed reference cell on 4.9)

- $144 for sensor and sensor module

 

Adafruit 1.3" diagonal 128x64 OLED screens:

- SPI interface for shared datalines and fast transmission (usually half main chip clock)

- ssd1306 display chip can communicate at 10 MHz (limiting speed)

- 12C, the alternative protocol can only handle 3.4MHz max (maybe enough for gauge screen refresh...)

- $19.95 each

 

Adafruit Feather M0:

- 48MHz Arduino board to run at least most of the gauge/datalogging

- $19.95, but am using more breakout boards too (RTC, microSD, etc.)

 

 

Figured i'd normally have AFR, MAP, coolant temp and oil pressure on the four screens, and then have a button to throw an arduino interrupt to switch all 4 screens over to EGT (and back with another button press). we'll see....

 

And maybe a second button to datalog. figured i could set the Arduino RTC to the laptops clock (when uploading code to arduino) and record the time at the beginning of each datalog, and then just record ms from start at each subsequent datalog. Romraider records the laptop time when it starts a datalog at as well, so can sync up Arduino datalogging to Laptop datalogging. in theory...

 

 

Oh yeah, 0-5V and 0-3.3V considerations. The newer, faster arduino chips run at 3.3V (electrical industry trends to lower power devices, 0-1.8V is the next step), vs. more generic "older" chips that run slower and at 0-5V. Being that the Cortex M0 chip is a new arm chip, it runs at 0-3.3V, which means that it can only read analog inputs up to 3.3V. This is a problem for 99.9% of the sensors I've found, which output 0-5V signals generally.

 

Two options:

- Hope to only use the 0-3.3V section of a 0-5V output (not sure how kosher this is, as i expect too much over 3.3V would be a bad thing for the chip/input)

- "Convert" the 0-5V signal into the 0-3.3V range - with a voltage divider (?). This means you get lower "resolution" but full sensor range. depending on the accuracy needs or actual accuracy of the sensor, this probably doesn't matter.

 

I've got parts to do the voltage dividing, but we'll see what happens. nothing is ever simple... :)

 

also, as an example, for the EGT probes i need 0-5V to read 0-1000C. 900C should be the highest i see, and the sensors can only handle 1300C. 0-3.3V only provides 0-650C, where from what i've read most readings will be at 800C. i could also use the AD8495 chip to just shift my readings up, to get something like 0-3.3V giving me 340-1000C, which i imagine would be perfectly fine. will have to look into shifting the AD8495...

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

CAD picture of next version that's at the shop for MFG. Smaller parts will be black anodized aluminum bezels, and the plate is 16 gauge aluminum plate for mockup, but is the width and height of cubby area inside door. The point being that i can still close the cubby door for hiding when needed (parking in sketchy places).

Untitled.thumb.png.b34f4b27d4db0a6d264a0cba23cc29fe.png

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

i did have a terrible picture of it at night lit up in the cubby. i guess i didn't post it. will try to find it and post. probably didn't cause it looked aweful as it must have been on my old phone camera - like 5 or 8 mp and terrible.

 

EDIT: guess i deleted the in-car photo. it was terrible, and was hard to see what was up. here was a picture of it together at my desk. Not sure why it looks so blue... i guess it is a high kelvin white.

 

This set up is awesome and very much like something I want to do with mine, but I'm rather attached to the cubby hole as I actually use it to hold things. I've worked with Arduinos and such, but just don't have the time to make something like this, but this build has made me want to try and do something with my own build.

 

When I first saw the pictures of your display, I thought they were vacuum florescent and got really excited, but OLED works too. :)

Link to comment
Share on other sites

i mostly got pushed to the OLEDs cause i wanted smaller readouts and the ability to change/display the "name" of the measurement was nice. They're also low power and SPI.

 

First attempt at this was using a bunch of single 7 segment LED displays, but after much work and two attempts, i realized i'm not an Electrical Engineer as both prototypes sizzled out of (working) existence. I hadn't figured out how to display the names of the readings either, or at least in a way i was happy with.

 

then looked through alot of display tech, but didn't like alot of it (LCDs are ugly, but thought about a positive LCD with red backlight...), and most just weren't sized/pixelated to give me what i was looking for (4 displays in the cubby).

 

then i found these OLEDs. super easy to work with through arduino, low power, several sizes (going to replace the clock with a .96" 128x32 too)

 

 

 

The only real downside is the lack of color (not really an issue to me, could use a color OLED, but less options and more code), and the lack of brightness control. this could be solved with a darker red tint layer, or maybe a smoked layer and a red layer together. there are ways around it though, but in a perfect world you'd be able to PWM/control the brightness to match the dimming of the cluster. alas, not this time.

 

Also, looking forward to trialing an inverted display mode to "flash" the readings on the screen if something is amis. say, EGT over 900C, PSI below 15 (or something), etc.

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

  • 4 months later...

Nothing to report, except the growing list of maintenance and things that need fixing (don't worry, not posting it). Me and the OBXT have been living it up for 6 years. crazy how time flies. Another kiddo on the way, due December, so should get anything "important" done before then. had that idea last time though, and all i got done was the wheel refinish and install with new tires.

 

here's to another 6 years, hopefully more productive (as far as keeping up on maintenance and upgrades).

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

Clutch dampener delete yet?

 

negatory.

 

been sick for 2 weeks and wife has been killing me with stress cause she's been home from work (among other things, she pulled up the carpet in our master bath. and no, she isn't putting a new floor in. we haven't even chosen a kind of floor... :spin:). she goes back next week.

 

 

the electrical system has become my new top priority. Need to check the baseline battery drain before and after removing what i think is the power feed to the janky window modules. if that's the source of my battery issue (and probably even if it's not), will be reverting back to stock for my windows.

 

Co-worker mentioned that i might want to see where all the hacking was done, and maybe i can just get a new door harness rather than trying to put things back together properly.

 

After that, i think i need to find my cold start screech. i think it's the starter, since it's only there when starting.

 

oh, and reconnect that wire i accidentally cut under the dash (purple with silver stripe?). turns out it's something to do with the PWM/dimmer circuit, and not the headlight selection switch. headlights still work. :)

 

That's it for the stuff that's currently broken (that i know about).

 

 

Then i think i just need to take my car in to HMS and pay them to do a pre-tune checkup (compression, leakdown, intake leak, etc.) as well as modifying my TS DP for length. if i get it done there, i don't think my wife will bat an eye at the cost (not that it will be much). will fall into the Auto maintenance/repair budget, which i never use.

 

THEN, i just need to roll the car over to the build parts i have already and get with Cryo for a tune (TS VF52, deletes and ID1000's essentially).

 

 

all well before December...

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

  • 1 month later...

Mostly for my own records, but i stumbled upon this over on ecomodder. Have always been planning to make a rear belly pan (axle back), and always figured i'd just angle it up towards the rear bumper, which will probably be 5 degrees or so. but apparently that's not good for drag in the wagon configuration. no angle is, need it flat. so flat it will be...

 

http://ecomodder.com/forum/member-california98civic-albums-black++green-picture5282-effect-rear-diffuser-design-angles-drag-according-body.jpg

* Build Thread * 26.53 MPG - 12 month Average *
Link to comment
Share on other sites

Hey Flinkly, I came across your build thread via your posts on the Romraider forum. The custom arduino gauges look awesome. Like you, I'm not particularly happy with the PLX gauges and would like to build my own.

 

Can you share your arduino sketch?

 

Sorry if this seems like a lame first post!

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.




×
×
  • Create New...

Important Information

Terms of Use