Page 1 of 1

Fw. improvement - PANIC BUTTON

Posted: Tue Feb 23, 2021 7:50 pm
by JRD McLAREN
I have (another one stupid) idea to improve rusEFI FW ...

Some other standalone ECUs have something like "Panic Button" function.
It is "user triggered" button, for any reasons ...
User push the button when they feel / or hear when something is wrong ...

On HW side it is "grounding" one of the analog input.

On SW side ...
The "stupid ones" just highlight some parts on log file, and "we" can find any area on log file much easier.
- I'm not sure, if TS or MegalogViewer can serve this ...

The "smart one" has added value like:
- Boost Cut
- Ignition retard
- Fuel add
- Cut off engine
- etc.

What are you thinking about this ..??

Re: Fw. improvement - PANIC BUTTON

Posted: Tue Feb 23, 2021 7:55 pm
by AndreyB
Please start with an explanation of who and why would need that new button. What is the big picture here?

Re: Fw. improvement - PANIC BUTTON

Posted: Tue Feb 23, 2021 8:23 pm
by puff
This could be a handy tool to make timestamps of interest in the logfiles... This would probably make the further analysis of logs a bit easier.

Re: Fw. improvement - PANIC BUTTON

Posted: Tue Feb 23, 2021 8:41 pm
by AndreyB
puff wrote:
Tue Feb 23, 2021 8:23 pm
This could be a handy tool to make timestamps of interest in the logfiles... This would probably make the further analysis of logs a bit easier.
:( So far this is "let's have a magic button which reads user's mind and does exactly what users wants! more power! less power! log timestamps!"

Re: Fw. improvement - PANIC BUTTON

Posted: Tue Feb 23, 2021 9:09 pm
by JRD McLAREN
AndreyB please calm down. I'm not your enemy ..

PUFF explain a basic function of this button.
(when you hear misfire, hear knocking, feel lazy engine rev up - due to lean/rich mixture, etc.)

I can see or hear the opinion of other people.

For brainstorming:
every ecu with one free analog input, has this "basic function" ...
but it call something different ...

Re: Fw. improvement - PANIC BUTTON

Posted: Tue Feb 23, 2021 9:17 pm
by blundar
The way I read this, there are two somewhat independent requests.

1. Hit button - get log event. This is a pretty common feature in a lot of software I use. The idea is a very sound one. Particularly in long datalogs, it's often hard to pick out the moment(s) of greatest interest. If you have a button activated by the user that creates an "entry of interest" on the log, then that can make all the difference in identifying juicy moments. This conceivably could be implemented by tagging a analog channel that is logged as "Interesting Analog" and wiring it with a pull {up,down} to change its state. You can use the changed state of "Interesting Analog" to find juicy log bits.

2. Hit button - Panic/shutdown. This is not a terribly common feature on street cars but is commonly accessible on race cars in the form of some button that shuts off power to _____. I think this feature could be implemented in ECU logic but I question the value keeping the ECU active while shutting down boost/fuel/ign/etc. would have. I think that some form of "valet mode" where power was limited by a function of DBW/rev limit/boost limit/etc. could almost be adapted to a "panic button" because the reduced "valet" parameters would probably also suit "panic" mode.

Re: Fw. improvement - PANIC BUTTON

Posted: Tue Feb 23, 2021 9:58 pm
by AndreyB
blundar wrote:
Tue Feb 23, 2021 9:17 pm
The way I read this, there are two somewhat independent requests.
Thank you! It's fundamentally easier to understand once things are not piled together.

https://en.wikipedia.org/wiki/Divide-and-conquer_algorithm is a fundamental concept.

Re: Fw. improvement - PANIC BUTTON

Posted: Wed Feb 24, 2021 12:12 am
by kb1gtt
blundar wrote:
Tue Feb 23, 2021 9:17 pm
2. Hit button - Panic/shutdown.
Something like this is required on monster trucks that are preforming a show. The kill switch is controlled by any one of three people. If the driver gets knocked out any one of the staff members can press the button and kill the engine before the truck makes it to the crowd. The kill switch requires a constant radio signal to allow the engine to work.

Re: Fw. improvement - PANIC BUTTON

Posted: Wed Feb 24, 2021 12:14 am
by mck1117
it sounds like there are actually two asks here that are getting confused:
  1. kill switch: "please stop injecting fuel now"
  2. limp home switch: "please limit power output and do something safe so I can pull over"

Re: Fw. improvement - PANIC BUTTON

Posted: Wed Feb 24, 2021 12:53 am
by kb1gtt
I think there is a third inquiry / request. A push button to highlight stuff in the log.

Re: Fw. improvement - PANIC BUTTON

Posted: Wed Feb 24, 2021 8:24 am
by JRD McLAREN
1. - Timestamp in LOG file, or any other mark in LOG file, and no other action is needed ...
2. - Safety function, ign. retard, boost cut .. etc ...
3. - Kill switch .. stop engine immediately

???

We have first one ...
blundar wrote:1. Hit button - get log event. This is a pretty common feature in a lot of software I use. The idea is a very sound one. Particularly in long datalogs, it's often hard to pick out the moment(s) of greatest interest. If you have a button activated by the user that creates an "entry of interest" on the log, then that can make all the difference in identifying juicy moments. This conceivably could be implemented by tagging a analog channel that is logged as "Interesting Analog" and wiring it with a pull {up,down} to change its state. You can use the changed state of "Interesting Analog" to find juicy log bits.
First and second can be the same button ...
Third "KILL SWITCH" separate .. ??

Re: Fw. improvement - PANIC BUTTON

Posted: Wed Feb 24, 2021 5:10 pm
by deaner
I reckon mental resources would be better suited to building better failsafes via corrections or limp mode. The ECU will find fault and shut things down far faster than a person can compute that theres a problem and manually hit a button. I used to like gauges and warning lights in my cars. Nowadays, I let the ecu figure all that out after setting failsafes lol. Ive seen the ecu catch stretched timing chains, bad tanks of fuel, failing timing belt tensioner and failing injectors and wastegates...things a person would never be able to catch and react to in time. Just my .02.

Re: Fw. improvement - PANIC BUTTON

Posted: Wed Feb 24, 2021 8:19 pm
by JRD McLAREN
Computers doing thing faster in every conditions... that's true ...
but not better in every condition or right way ...

(they need any kind of sensors .. but what they do, when sensors fail ..?? )
(or wires, pipes, hoses etc.)

The first requirement for "good automated safety function" are good maps / and all other data.
You must create maps and collect this (right) data.

This type of button may help you "find this data" ..

Re: Fw. improvement - PANIC BUTTON

Posted: Fri Mar 05, 2021 12:24 pm
by Simon@FutureProof
I think within the plan for "smart features" the concept of a panic trigger could be included.
A single flag based on a pin state that manually triggers "limp mode" (what ever that may turn out to be or the user chooses it to be).

It might actually be worth including that as part of a test mode for limp, same as we have for other feature test buttons. My understanding there would be that it only needs to set the same flag as the limp triggers when a pinstate is high/low to work.

I will try to incorporate this into the logic for the smart features as I can see the utility, there have been times I have seen something odd but not been able to mark the log in any way to indicate when exactly it was.

Re: Fw. improvement - PANIC BUTTON

Posted: Sat Mar 06, 2021 5:44 am
by dexter_lab
it will be great if the logging is made on "sliding window" for example 10sec and when you press the button to record those 10sec

Re: Fw. improvement - PANIC BUTTON

Posted: Sat Mar 06, 2021 4:09 pm
by Simon@FutureProof
I don't think that is possible, that would be tunerstudio side and probably not something we could implement.