durwood

Ok, I have been trying to find a solution for integrating my stock buttons.

Here is the issue, they are held high @ ~2V. When the button is pressed it goes to ground. I have tapped into the buttons as shown in this thread, but I need to make sure no voltages feedback into the stock circuit board. Can the Fusion brain "monitor" the button without feeding voltage back into it?

Is it something I should even worry about?

__________________

2k1Toaster

You would have to use an analogue input. Which I prefer anyways.

But then connect the "interesting" button wire (2v->GND) to the analogue input. No voltage will be fed out, and it should register as long as you ground the Brain to the chassis as all car installs will be.

If you use a digital input, it wont register because it needs to be 5v to trigger.

__________________

durwood

I know it could register on a digital input if the theshold is high enough to tell if it's high not low, but the problem is digital input are typically held high via a pull-up resistor just like how a USB gamepad is setup. That's what I was kind of wondering.

The analog inputs are just an A/D conversion correct? I guess that should work, but the brain only has 10 analog inputs huh? Hmmm, I think that would be enough inputs. And if I already have a rotary volume control (Rotary encoder) tapped into, can I send the 5V/0V oout to the Brain? Would it be as simple as connecting the two wires to two digital inputs to control volume? I've already done the volume by tapping into a mousewheel and using autohotkey to change volume in roadrunner, but if this would do everythign in one shot I'd be extremely happy.

Oh, and amp turn on delay with instant off before computer finishes shutting down.

Sorry for all the questions. I was interested in this way back before you guys had a working product and I lost track of the whole thing.

__________________

2k1Toaster

Actually the digital ins have 1 pin go to the chip and 1 pin is +5v. The pin input takes voltage. It doesnt go to ground. It is at ground already. So you can inject up to 5v. It wont register unless very high though, so 2v wont cut it.


Yup 10 analogue inputs. And also you can combine them as OEM systems do. It seems like you know a bit of electronics so with some resistor magic, you can get a single wire that outputs different voltages at different presses.


Technically yes, but as for the sampling of the Brain it is "slow" in electronics terms. It is maybe 33Hz. It can go up to 180Hz for reading which is much better but CPU cycles suffer. So depending how fast you click, it may not register.

We are working on a digital encoder to analogue voltage converter with a faster sampling rate. So for now it may work. In the future it will work.


Turn on delay on boot is 100%. Turn off I am still trying to find the acpi call to know if the computer is shutting down. But if the program is shut down, the outputs will turn off (unless otherwise specified).


Glad that you came back!

__________________

durwood

Well I couldn't do that because one button might feedback into another, but I think I could make do with 10 button control. SO basically, I could tell your software once the analog level falls below level "X" perform this keystroke?


Ok that is probably the problem I have with my current setup. I don't know what the clock rate is for a mouse, but between the rotary encoder, autohotkey and RR, the delay is quite large, at least larger than is desired.


Cool. I was looking at picking up a powermate and itnerfacing with that but that might be another option.


That's prefect then. As long as it shuts down before the soundcard powers off, then it should serve it's purpose. I assume turn on delay would happen after it is already booted into windows-correct?

__________________

2k1Toaster

They shouldnt feed into eachother. Just build a simple resistor DAC around them and voila!

Like this:



By delay do you mean how fast from spinning the wheel to doing something? If you spin slower, does it work faster?

From the time the brain reads the value, to when the software spits out the command is usually under 3mS. The reading though is only done every 30mS on average. It can be sped up, but CPU cycles suffer.

The problem comes from the value. The encoder sends out short pulses usually, and then it must count those. Well if the brain doesnt count when that pulse is high, it will miss it, so it is sort of like a lucky guess situation for something like that. That is why what we will have will be an encoder to a analogue output based on absolute position. As long as the brain samples 2 times for every time it spins all the way around it knows what direction you are going.


Yes the outputs are all off until the brain communicates with the PC and then if the initial state is set to on, it will power on the output (amp). Then if the brain loses communication with the program, the outputs start their individual timers. When their timers are up, the respective output shuts off. The timer can be from 0 seconds (instant) to 30hours. This has a delay of about 1-2 seconds though initially to make sure it really has lost communication and your outputs dont just flicker each time it gets a hiccup! I will have the program sense shutdown calls through software and send out a call right before it goes into a suspend/shutdown mode instead of relying all on the hardware to figure out it has gone blind.

I will have the program watching for shutdown calls, and start the timer then instead of

__________________

durwood

Ah-well I'm a little lazy so I'd just sacrifce loosing two buttons for now. 10 should be enough for most of everything I need to control.

As far as my current volume control situation, If I turn the knob fast, the volume doesn't increase that fast but it will ramp up after I'm done turning. I guess that is just becuase autohotkey is simulating keystrokes and it takes it little bit to "catch up". It can really catch you off guard if you aren't careful.

I need to get some monies together so can order one. Time to go sell some misc car audio gear.

__________________

turbocad6

you can use 2 brains at the same time too, now 12 is no problem, you've got 20

__________________
MY NEWEST INSTALL:modded infiniti fx with big screen - "If my typing sucks it's probably because I'm driving....................."

first windows carpc install........my liquid cooled LVDS screen :D

durwood

I have a hard enough time working with 1 brain- with two they might fight against each other.

I'm trying to scrape together enough $$$ for one so two is no go. Maybe later I'll pick up a second one.

__________________

2k1Toaster

You can actually use 127 of them at the same time now.

__________________

durwood

__________________

Blue ZX3

I was doing this with a mousewheel and girder, and yes, there was a lag time, and not a instant event the way it should be.


I personally would not recomend one of those. I started out with one and had problems once or twice a week coming out of standby and having no vol control(had to unplug and replug to get vol ctrl back)...so I ditched the idea. Now, it's just a expensive paperweight!


My final solution to this whole problem was to use girder and have it directly control the audio lines, which act just like a reg head unit, 85% instant change!

__________________
RoadRunner...The #1 FE, PERIOD.

RR Media >>HERE <<

Updated RR Config (Current Ver 1.0.1.0)
Old RR Versions
Complete CFX PSD Set
RRSkinEditor v.71
RR String/Variable Cmd Ext Plugin
DFX Btn Editor

greatwhite

Might this thread be of some use?

http://www.mp3car.com/vbulletin/inpu...-controls.html

I don't think it's quite what you're looking for, but it may give you some ideas.

__________________