Page 38 of 66

serial bluetooth pipe

Posted: Tue Nov 04, 2008 1:32 pm
by GregsGarage
GregsGarage wrote:I was thinking a serial to bluetooth adapter is what we need, don't know if such a thing exists. :shock: .


A quick bit of googling and I found this;
http://www.sparkfun.com/commerce/product_info.php?products_id=582
It's a serial to bluetooth pipe. Bit pricier than a cable but should do the job.

Just had a look at the Maplin site and the cheapest SatNav I could find was £70. So that was a good buy at £50. I do agree that it is a possible cheap option as far as the hardware cost goes, but it may be more work to hack. Just when you have got it hacked, that unit may no longer be availble so you start again, figuring out how to hack the latest "cheap" sat nav. Although the N810 is more expensive I think any software developed for it should be usable on more devices since Nokia is using a version of the Debian linux system. Any code written for the N810 will also run on the N800, provided it is running the latest version of the operating system. It should also be possible to use the code on Linux versions of cheap laptops such as the Eeepc.

Posted: Tue Nov 04, 2008 1:48 pm
by retepsnikrep
qdos

Link please, make serial no etc. Specs? Have yout tried getting other software to run on it. It must have an autorun setting presumably, and runs the fairly basic looking interface as soon as the OS is runing, autoexec.bat, config.sys? OK OK I'm showing my age and the limit of my skills here.

Although I'm happy to liase with/assist whoever attempts to make the nice interface software. It won't be me as I know next to nothing about windows programming. We/I can supply simple 9600bps serial data over the usb lead, it's up to whoever then to do the rest.

Greg

That bluetooth thing looks interesting and I can probably do the Master end and supply it data, but again someone else will have to do the other end and all that software.

Posted: Tue Nov 04, 2008 2:24 pm
by GregsGarage
My thinking was that it could be fed to the CANTOP program. Might try and download the source code for it and see if it will compile and run on my Ubuntu Linux machine. If that works then it should hopefully be able to compile it for the Nokia. But first I need to test the new software you wrote for the slave boards. :shock:

Posted: Tue Nov 04, 2008 3:11 pm
by qdos
Peter the Sat Nav I mentioned is by an obscure company marketed as AyToBe it runs Windoes CE NET 4.2 Core

ay to be ATB083
3.5" LCD touch screen
GPS built in
SD Slot
Mini USB interface
Ir-DA with remote

There's a fair few websites out there about reprogramming Sat Navs and I'll try to find some for you but it's not uncommon that you can do this with a lot of Sat Navs

Yes you can get to the autoexec and reprogram it to enable you to run different software other than the default boot into the Sat Nav software. I've not done much other than literally play about with it though others have done and I'd be happy to see if I could do something but I doubt I have the skills to I had spoken to Charles in Leeds quite some time ago and he might be the man to do it but the point is that it is possible I'm quite sure of it and cheaply too :D Another bod that might have the skills or know a man who can is Rob Sharpe as he works with Motorolla designing mobile telephone interfaces. I think they use Java though

Posted: Tue Nov 04, 2008 6:55 pm
by retepsnikrep
This quite interesting for the maplin A2B thingy.

http://forums.digitalworldz.co.uk/13912 ... rsion.html

Seems you can access various bits.

Posted: Tue Nov 04, 2008 9:09 pm
by qdos
Yep indeedy and I believe you can do this with quite a lot of Sat Navs just the AyToBe was a bargain basement price so I got one to play with.

I was intending to customise it with a different Flash Screen on start up and to put in special routes (Sunday Drives sort of thing for recreational country jaunts) but in delving deeper it struck me that these Sat Navs could be modified to do a whole range of customised tasks and would be ideal for your display unit I think. I've not had the time to look much more into it but I was thinking of using it as a unit for some EVs I'm playing with. It would be really good to have someone on board to look into this. They are readily available and many people buy them already so they become fairly cheap effectively as they are kinda paid for already as a Sat Nav.

I'll dig out some other bookmarks on them later on if you like but I think you've sussed what I'm getting at :)

Posted: Wed Nov 05, 2008 9:58 am
by GregsGarage
retepsnikrep wrote:Does your charger have an adjustable output voltage pot? Could you hack it like I'm doing?


Peter, I had a brief look when I got the charger but couldn't find it. It should be there somewhere though. I will have to have another look, it will have a voltage sense circuit to determine the charging voltage, just have to find it.


I was thinking about how we are calculating the cell voltage. The new code is just removing some of the calculation steps that the pic does, and there by improving accuracy. But are we calculating the right value?

How about if instead of calculating the cell voltage and comparing it to are various values, we instead calculate what the various values are in relation to the adc value that the pic gets when it reads voltage. We then would just need to do a one time calculation on all the min and max voltage values, before the main program loop, store the results and then all the pic has to do is to compare the adc value against these calculated values. This should be the most accurate way possible of doing the calculations and also speed up the program, possibly allowing lower clock speeds. 8)

I don't think that this approach would work very well for the digital versions, but should work well for the analouge slaves. I may try and build a slave board tommorow, and unless someone beats me to it, work on an example code for my idea.

Posted: Wed Nov 05, 2008 11:05 am
by retepsnikrep
GregsGarage wrote:I was thinking about how we are calculating the cell voltage. The new code is just removing some of the calculation steps that the pic does, and there by improving accuracy. But are we calculating the right value?

How about if instead of calculating the cell voltage and comparing it to are various values, we instead calculate what the various values are in relation to the adc value that the pic gets when it reads voltage. We then would just need to do a one time calculation on all the min and max voltage values, before the main program loop, store the results and then all the pic has to do is to compare the adc value against these calculated values. This should be the most accurate way possible of doing the calculations and also speed up the program, possibly allowing lower clock speeds. 8)


I'm a bit confused. :?

The new code does improve accuracy slightly (by 1 bit in fact) not because it simplifies the calculation, but my allowing a much bigger starting number (63xxx) which is almost as big as the 16 bit maths can use(65335). So we are using the limited 16 bit maths resolution to it's max.

I don't think you can do what you suggest because of the way we calculate the cell voltage.

Remember the result from the adc changes depending on the pic supply voltage, not the voltage being measured which is always 1.235v.

We calculate the actual Picaxe supply voltage from the adc result using the rather strange formula.

Have a go by all means. :shock:

Posted: Wed Nov 05, 2008 12:32 pm
by GregsGarage
O.K. Peter, let me explain myself a bit better. At the moment we read the value RefADC and then the pic has to calculate the cell voltage every time it goes through the loop. This makes it easy for us humans when we want to change values for LoadOn, MaxVOptoOn, Etc. but means the pic is having to carry out calculations every cycle. What I am proposing is that instead of comparing CellV to LoadOn, MaxVOptoOn, etc, we compare
RefADC to RefVADC/LoadOn * 2 ,RefADC to RefVADC/MaxVOptoOn * 2, etc. (where RefVADC=63,200)
Of course RefVADC/LoadOn * 2 only needs to be calculated once because it never changes!
8)

We don't even have to have the pic do the calculations we can just have a spread sheet calculate the values or a table in the remarks with values such as ;
    356 = 3.55 volts
    351 = 3.60 volts
    346 = 3.65 volts


The accuracy is now determined by the hardware and not the formula that we use, and the pic is doing less work so might be able to clock it slower. 8)

Posted: Wed Nov 05, 2008 2:04 pm
by retepsnikrep
Have a go at some new code Greg by all means. I'm not convinced your proposal is better than what we have with the new formula, but you may be able to persuade me :wink:

I'm working on the Master Proportional Charger control at the moment. Got my asus eec pc thingy very neat, loaded picaxe editor etc. Do you think wife will notice if I take it on holiday with us soon? :shock: