|
|
|
#31
|
|||
|
|||
if all the sensors are wired why not bring the hard wired signals through the firewall and have the bluetooth located in the cabin?
__________________
"NEW" 1984 300D TD needs some love |
#32
|
||||
|
||||
Quote:
__________________
Current Stable
|
#33
|
||||
|
||||
I'm really struggling with the software end of things. I've been able to reliably get data across the Bluetooth from the Arduino to the Android and I've been able to get the gauge program running locally, but when I try to bring the two code pieces together it has thus far failed. So I decided to take a bit of a break and focus on other pieces of the system.
The way I'm going to source the oil, trans snf intake temps is using a very slick DS18B20 temp module. I got this breadboarded with 4 sensors, so that part is ready to go. A common way to get oil and trans temps is to T into the line with the sensor. I have no interest in adding 6 additional points of potential leaks, so I'm going to use an external pickup. I'll clamp an aluminum block with the sensor epoxied inside to the line going to the respective cooler. Here's the trans temp pickup block ready to install the sensor. The oil temp pickup is identical, though with a larger hole.
__________________
Current Stable
|
#34
|
|||
|
|||
if they are steel lines may want to wrap the lines with something to insulate the line from the block so you don't get electrolysis with the dissimilar metals. if they are aluminium and you can bolt it right on maybe try using thermal paste like you do on processors
__________________
"NEW" 1984 300D TD needs some love |
#35
|
|||
|
|||
Quote:
Un parquet? Un parquet? You have a parquet? Sans blague?
__________________
Strelnik Invest in America: Buy a Congressman! 1950 170SD 1951 Citroen 11BN 1953 Citroen 11BNF limo 1953 220a project 1959 180D 1960 190D 1960 Borgward Isabella TS 2dr 1983 240D daily driver 1983 380SL 1990 350SDL daily driver alt 3 x Citroen DS21M, down from 5 3 x Citroen 2CV, down from 6 |
#36
|
||||
|
||||
Major breakthrough on the software front today. Got the app on the Android capturing simulated data from the Arduino through a wireless Bluetooth connection and presenting it on analog gauges along with digital readouts. I've had the pieces working, but this is the first time the structure of the system has been working end-to-end. To convert from simulated data to real data, is trivial... it was actually quite a bit harder to code simulated data that would exercise the gauges across their entire range, than it will be to drop in code for the actual sensors.
Yesterday I got the TIT (turbine inlet temp) port installed on my extra test pipe, so I'm ready to order the probe and amp for the thermocouple. Also last time I was at the junk yard I found a GM 3-bar MAP sensor on a Saab turbo. This will give me my boost signal. It was $2. I'm already contemplating the first addition to the v.2 system - a compass display. Something like aircraft panels use but with a car bug.
__________________
Current Stable
|
#37
|
||||
|
||||
While waiting for the TIT (EGT) components to arrive, I've been making progress on the circuit board, moving from breadboard to prototype. The components are the power supply, the Arduino, the Bluetooth module and the USB programming module. The thermocouple amplifier will go in the open spot in the center. There's still a fair amount of wiring yet to go on the board, not to mention the 10 wires that will go to the plug.
I also got the 1Bar MAP (vacuum sensor) and 3 bar MAP (boost sensor) tested and calibrated.
__________________
Current Stable
|
#38
|
||||
|
||||
I like it
I'd definitely use thermal paste (or at least antisieze) between the clamp-on sensors and the lines. Insulation isn't a bad idea either, that block of aluminum looks like a nice heatsink.
__________________
$60 OM617 Blank Exhaust Flanges $110 OM606 Blank Exhaust Flanges No merc at the moment |
#39
|
||||
|
||||
Progress continues...but I have one question - which side of the transmission provides the supply to the cooler in the radiator? It seems like it's the drivers side, but I can't find a definitive answer.
Got the board wired up as far as I can go without the thermocouple amplifier module. Also got the relay box mounted in the car and mocked up ready to start doing the chassis-side wiring. The boost sensor and the vacuum sensor are also mounted and ready to go. I also tackled the piece of the project to include a vertical card compass and got the compass working with simulated data. The compass module will be here in about a week. Unfortunately I don't think there will be room inside the box... of course it probably doesn't make any sense to have the sensor module in the box anyway due to interference.
__________________
Current Stable
|
#40
|
||||
|
||||
Not using the stock 3bar boost sensor?
Why not use the phones compass? Lookin good!
__________________
http://superturbodiesel.com/images/sig.04.10.jpg 1995 E420 Schwarz 1995 E300 Weiss #1987 300D Sturmmachine #1991 300D Nearly Perfect #1994 E320 Cabriolet #1995 E320 Touring #1985 300D Sedan OBK #42 |
#41
|
||||
|
||||
Almost there...
I've been running the system successfully in the car with one temp gauge as well as boost and vacuum gauges. Awaiting the other temp sensors and the TIT thermocouple. The thermocouple amp came in today and that's wired up and appears to be working - reads 0 as I would expect without a thermocouple attached. Also added a temp sensor to the board itself to monitor the temps inside the relay box. The Arduino's normal operating range is -40 to 85C but the absolute max rating is for -55 to 125C. With a sensor on the board, I can monitor where it normally runs and make adjustments as needed...maybe adding a temp switch to cut power if it gets too hot. The gauges are a bit hard to read sometimes, so I added a plain old digital read out to the selection of screens available, which now include 6 white gauges, 6 dark gauges (night use), 2 large gauges (boost/TIT), debugging (raw data) and now a digital read out. Also changed the needle colors of the darker gauge sets to yellow and made them a couple pixels wider. It's quite interesting to monitor the vacuum. It drops quite quickly under braking, and then quickly builds back up. It was also interesting to note that when driving in the mountains at 4000 ft the boost gauge read -1 with the engine off - makes sense since it's reading absolute pressure not relative pressure. Tomorrow it's off to the junk yard to research fittings that can be used/modified to install the intake temp sensor in the filter housing and some connectors for the temp sensor wires to allow disconnecting if needed.
__________________
Current Stable
|
#42
|
||||
|
||||
I got the DS18B20 temperature sensors in over the weekend. They are really nice. Started to install the intake sensor and blew it - drilled a hole way too big. Back to the drawing board.
I finally had time to mess around with the baud rate on the Bluetooth, and as Winmutt suggested, that really helped stabilize things. I still padded the front and back of the packet with bytes that I can use to check for a complete packet before I try to parse the data. But alas, there was apparently no free lunch. I found that with the higher baud rate I was actually getting slower packet transfer. Turned out after much head scratching and troubleshooting, I realized that I made two changes at the same time (bad idea). I added the live code for the temp sensors at the same time I changed the baud rate. The temp sensors are really slow and brought things down to about 2 samples per second - not bad but makes for jumpy needles on the boost and vacuum gauge. So realizing the problem, I changed the code to only sample the temp sensors once every 10 seconds while boost and vac get sampled 5 times a second. There should be no noticeable temp change in 10 seconds so this is just fine. As soon as I did this, the data rate jumped back up where it should be. I also redid the graphics on the second set of gauges to lighten them up and increase the font size on both the gauge face and the digital read out. This will make things a ton more readable on the small phone screen. I was a little concerned about the engine bay temps for the Arduino, but I realized that if the temps were really going to get above 100C, that would mean that the washer fluid would boil... and it obviously doesn't. The Arduino is mounted way to the front and in the opposite corner of the engine bay from the turbo/exhaust, while the washer fluid tank is right adjacent to the turbo. But I've placed a temp sensor directly on the board, so I can monitor it in real time. Also added a virtual "led" indicator to show "connected" and "receiving data" conditions. Here's my straight digital readout.
__________________
Current Stable
|
#43
|
|||
|
|||
You might change the scales on all the gauges so the "normal" value is at the top. Then when everything is normal the needles are all vertical.
__________________
. 1st MBz: 1982 300SD 2nd MBz: 1987 300SDL 3rd MBz: 1995 S420 4th MBz: 1987 190DT 5th MBz: 1984 300SD w/1983 300DT engine 6th MBz: 1999 C230k I'm 3rd owner, got it w/57,235 miles. and manages Mom's 2007 R320 CDI |
#44
|
||||
|
||||
Quote:
Here is the latest addition to the gauge face library. The Oil, Trans, Arduino and Intake are all using the same sensor and use the same scale, so I decided to present those temps color coded on the same gauge.
__________________
Current Stable
|
#45
|
|||
|
|||
Very nice, but the labels seem to be in backwards order compared to the temp readout's "normal".
To me, it looks like it should be: Intake, Arduino, Trans, Oil. Then the label is closer to, or nearly over the normal temp for each of the indicators.
__________________
. 1st MBz: 1982 300SD 2nd MBz: 1987 300SDL 3rd MBz: 1995 S420 4th MBz: 1987 190DT 5th MBz: 1984 300SD w/1983 300DT engine 6th MBz: 1999 C230k I'm 3rd owner, got it w/57,235 miles. and manages Mom's 2007 R320 CDI |
Bookmarks |
Thread Tools | |
Display Modes | |
|
|