Automotive +5V

[GigaJoe]

Hello,

I have some device that has to be installed in a car , and powered from +5V from the car source , consumption approx 100-150 mA ( no measure , assumption).  Where the +5 come from i have no clue and possible load on the bus as well.

When I open the box it PIC18F2480, + 3 MCP4912

Device basically do input as a MAF voltage,  and a output skewed data ,   overclocking the engine ....

An electronic design itself are terrible , like basic \ simplifying \ cutting corners design ;  but multi-layer PCB .... 
for DAC:   Vref == Vsuppl == +5 from the car with noise up to 6.5V spikes ...
There is no any input  / output even basic protection for dac\adc,  MCU , dac out  or ADC input  not buffered
I found just one ceramic cap for a Vsupp.  voltage;  2 place where suppose electrolytic caps,  empty.   

I need some advice to make +5V better,  as I cant make an active filter , i need some passive one,  I even has no clue how much capacitance I can load on this +5 bus ....

I probably can cut Vref pins for MCU and DAC  to supply from some better source like 100ohm 1000uf    RC filter -  I dont know ...   I need some advice

Device:


+5V: (not my measurement)

[sokoloff]

Trace where the two [presumed] capacitor circles on the right side of your picture go.

If they go across Vcc and GND, consider putting 16V electrolytic caps that will fit into the case on both those footprints.

[Andreas]

consumption approx 100-150 mA ( no measure , assumption).

never: this is usually the current that 5-10 sensors have together from the 5V sensor supply.

Where the +5 come from i have no clue and possible load on the bus as well.

A ECU has usually 1-4 sensor supplies that deliver the 5V for sensors like accelerator pedal some pressure sensors ...
Each can deliver 50-150 mA typically.

When I open the box it PIC18F2480, + 3 MCP4912

so in sum about 10 mA.

I need some advice to make +5V better,  as I cant make an active filter , i need some passive one,  I even has no clue how much capacitance I can load on this +5 bus ....

shure that the measurement is done with the right ground connection?
(you have to measure directly between sensor ground and sensor supply).

Usually the inrush current has to be limited on the sensor supply.

by the way: using soccets for the processor is a no no in automotive applications (vibrations).

And I fear the cirquit does not withstand a short cirquit of the 5V to the 12V battery voltage.

with best regards

Andreas

[james_s]

Perhaps look at some older stock ECUs? I repaired the ignition control module in a 80s VW years ago and it had a bog standard 7805 in it, but the regulator was what had failed.

The ECUs in my 80s and '90 cars have a slightly fancier 3 terminal linear regulator with a couple of electrolytic capacitors, nothing exotic.

[GigaJoe]

hook a scope to +5V

in AC mode random spikes ...



Seem better , need a filtering Mhz spikes ....

sooo ... trying to explain in seq, we feeding that +5V   it need filtered:   
10ohm   resistor - ceramic cap 0.1uf - some ferrite  coil  -  ceramic caps 0.1uf -  20uf el. cap -  5.6V zener (crowbar)  ---  device.
  Vref:  from croqbar zener  - 200 Ohm - cer.cap 0.1uf - el. cap 20uf.  (dont know  ... it looks strange )

the main  uncertainty:   will it work , due to RC has some reaction time ,   what happens when engine start ... i dont know ....   as MAF voltage will skewed due to Vref RC till it fully charged ...

can it be that caps was removed by very that reason ....  it basically shorted +5 bus at beginning   ?

do separated psu from 12V  + Vref ....  an ideal situation ,   but ...  it a separate box .... power dissipation, heat ,  ...  will it reliable  ...   
 

[SparkyFX]

I have some device that has to be installed in a car , and powered from +5V from the car source , consumption approx 100-150 mA ( no measure , assumption).  Where the +5 come from i have no clue and possible load on the bus as well.

As Andreas stated, probably from the ECU´s regulated 5V supply, which the MAF is connected to.

Device basically do input as a MAF voltage,  and a output skewed data ,   overclocking the engine ....

Most likely will modify the Intake Air Temperature sensor, that is included in most newer MAF-Sensor assemblies.
It is kind of snake oil for cars, as it will fake a higher air density this way, lead to more fuel injection, but will mess with the O2 sensor values and then get it downregulated again - and btw. increase emissions, which would make it practically illegal for road use over here.

[Rerouter]

to make the 5V better, just supply your own regulated rail

Use these regulators in all manner of automotive circuits without issue.
https://au.mouser.com/ProductDetail/RECOM-Power/R-7850-05?qs=YWgezujkI1KPMF6NhJFaJw%3D%3D

[JPortici]

You are going to use the dedicated sensor power supply, which is very clean and very well regulated and protected on over/undervoltages. Anything you see on your scope is being picked up by your probe

Just follow the instruction manual. (They wouldn't be selling THAT particular device at THAT particular price if it didn't work as intended. They are expert in the art of muntzing)

[DBecker]

Just follow the instruction manual. (They wouldn't be selling THAT particular device at THAT particular price if it didn't work as intended. They are expert in the art of muntzing)

Errrm, this apparently the kind of device that sits between the MAF and ECU and tries to trick the ECU into providing a richer mixture.  Look them up.  They usually claim "up to a billion more HP, or more".  They never work as described.  The MAF is the input sensor in a feedback system.  Eventually the ECU changes the long term fuel trim to negate the richer mixture, or it raises a lambda control fault.

Most similar devices are a just a resistor, perhaps a diode and two resistors.  Since this has a microcontroller, it has the potential to be better at tricking the ECU for longer.  But that seems unlikely.  It's probably just better-looking snake oil.  It's actually anti-Muntzing adding components to look more sophisticated.

As for providing a better regulated power supply, that probably won't create a better result.  The MAF might be providing a proportional output, and the ECU using that specific 5V supply as the reference voltage during the ADC conversion.

[JPortici]

I happen to know that device very well (let's say it's from a competitor) and they have very specific reasons to use as few components as possible and socketed MCU.
Also, it's not snake oil as there is a measurable effect

More components are required as different cars are "tricked" in different ways. Sometime you want to add/subtract in specific conditions which is a nightmare to do in the analog world. Some times you have to adjust the supply of the sensor. Some times you have to output completely different values and finally some times you have digital sensors in a network so MCU is a must. Chiptuning is much more interesting and complex than what you think

[GigaJoe]

Right,   Burger motorsport device  JB4  ....

There are  2 basic components ... software code , what I have no clue and probably some art of spoofing data ...
And a hardware part ...   that in front of my eyes ...  and:

I literally see missing components, no blocking capacitors
PCB design upside down ,  square pin 1  on PCB not as on chip
mess with positive and negative label
traces going nowhere
6 channels DACs in total -  for 2 MAF output
can bus receiver with differential input, don't using a twisted pair
why using DB-25 , and only a few pins , do not parallel it for reliability .
socked MCU ?
plastic case not much water resistance,  at least need conformal coating ....
Then ... there is no even basic protection for input\output signals ports , crowbar zener at least ...  DAC no buffered ...   Vref == Vsuppl without even attempt to clean it ...

Why it cost so much and designed literally as a prototype ...  Like a basic proof of concept not a further ...

I use to fix some radio units for my friend service mechanic,    2-3 cases , I do remember a total shielding , and a a serious filtering where radio, admitting it analog ...   Automotive power full of noises ...

Forgot to say ... to hook bluetooth   they use a screw needle to attach on +5v wire ..  ;  as device used , it already traces of oxidation on some pin's connector, ( white thingi)  ...

[JPortici]

Since i'm obviously biased i'm not going to comment on the hardware design, so.. your words
I can just say that despite everything it works well enough that they have a strong following from people driving certain brands in certain areas, which explain the pricetag (and my "just follow the instructions, you'll be fine)

I'll also point out the obvious, that device (as ours) can do many different cars with many different sensors. Let it be one or two or ten. If numbers don't scale up it's actually cheaper to just have one single assembly. They use a socketed MCU so they swap different processors (with/without CAN, with different firmwares for different functions) and if you check their forum they were able to engage in the users, which will swap configuration and micros between each other, hats off to them

[HwAoRrDk]

why using DB-25 , and only a few pins , do not parallel it for reliability .

It appears they have a variety of models all sharing a common form-factor, some with a lot of external I/O, some with few, so I guess they chose a DB-25 as a 'one-size-fits-all'. Although their use of DB connectors inside an enclosure is a bit weird.

socked MCU ?

They sell 'upgrade chips' for the end user to swap by themselves, so a socket is understandable, if again a bit weird and outmoded, given the USB and bluetooth functionality on offer in some products. Why no firmware upgrades using those mediums? Maybe these upgrade chips are add-on parts; I notice some of their products have dual PIC microcontrollers inside, so perhaps one can buy the second chip to upgrade the capabilities of a more basic unit?

plastic case not much water resistance,  at least need conformal coating ....

So they can up-sell you a wanky machined billet aluminium enclosure for an additional $40. At least you get something with proper weather protection (O-ring seal) for your money...

[GigaJoe]

how about rs232 TTL cable with ftdi chip for US$30 as add on ..
or check  fuel wires , 2 wire + needle injector - $30
and Bluetooth android application another $30
and finally  ... Bluetooth module $140 ..