Proton Mag Info

[ApBerg]

Nice info http://www.exstrom.com/magnum.html

Regards.

Ap

[Fred]

Hi AP,
Very nice well documented project, thanks for the link!

Hope they will complete it soon.
Regards,
Fred.

[Unregistered]

The micro they are using is obsolete. The AT90S2313 has been replace by the ATtiny2313. If you do build a unit the 90S2313 may be difficult to obtain.

[ApBerg]

Hello Fred, there is now more info, take a look at it.
The Atmel 90S2312 can be replaced bij the ATtiny 2313 ?

It looks like a nice project.. see wat is more to come.

Ap

[ApBerg]

JPG from coils

.....the foto from the coils is not made by me, but from the website... see the first mail.

Ap

[Fred]

Beautifull coil,
great project!.
Really well explained.
thanks,
Fred.

[Unregistered]

Hello Fred, there is now more info, take a look at it.
The Atmel 90S2312 can be replaced bij the ATtiny 2313 ?

It looks like a nice project.. see wat is more to come.

Ap

If you check Atmel's website it describes the '2313 as "mature" and "not recommended for new designs. The tiny2313 is available in a 20 pin DIP and at a quick glance the pinout looks the same.

What is the advantage of that style of coil over a simple solenoid style?

Rick

[Fred]

If you check Atmel's website it describes the '2313 as "mature" and "not recommended for new designs. The tiny2313 is available in a 20 pin DIP and at a quick glance the pinout looks the same.
What is the advantage of that style of coil over a simple solenoid style?
Rick

The 2313 may be "mature" but somebody has made it with it ,and it aparently works.If you think you could improve the design with other processors or coils ,go for it,and share your work as they do .i´m sure you will be welcome.
Regards,
Fred.

[Unregistered]

The 2313 may be "mature" but somebody has made it with it ,and it aparently works.If you think you could improve the design with other processors or coils ,go for it,and share your work as they do .i´m sure you will be welcome.
Regards,
Fred.

It almost sounds as though I am offending you and that is not my intent. I am only suggesting that you might not be able to buy the 90S2313 micro and you will have to use the Tiny2313. That would only mean a bit of their code might need to be changed but nothing significant. It (the Tiny2313) probably would not work any better than the 90S2313. I just had a look at Digikeys website and the 90S2313 has "call" under availability which usually means it is unavailable but an upgrade is listed. I have a working mag that uses the 90S2313 micro. I would like to compare the code they are using to mine when it is posted.

Right now I am working on a Gradiometer that will use a Mega162. The basic measurement principle is the same as the 90S2313. I still have a few bugs to get fixed yet.

I am using a simple solenoid type coil. I am sure there must be some advantage to their style and if there is an advantage to using that style they are I would probably make one.

Rick

[ApBerg]

90s2313 mag

Hello Rick

Nice reply from you ! 'though I am offending you and that is not my intent'.

It is also good that you have a working 90S2313 magetometer .
Can you give some info from you project ? There are some frends that have a great interest in a mag project. There is some info, (Bayout Koeller) very good but not the most easy start.
Can your info be used for a more simple mag ?

Regards.

Ap

[Unregistered]

Not exactly sure what you mean by simple. I built mine totally based on Koehler's design. He provided me with parts of his code that he used to build a gradiometer. The code was incomplete but with a lot of guidance from Jim I got it rewritten to build a simple mag. His schematics are the basis for a gradiometer. The amp circuits are fine the way they are. Only use one relay however for the mag as you only have one coil. The changes involve the 90S2313 connections. Those are minor changes such as an LCD connection. You will need something like an STK500 programmer to write the code to the 90S2313 or whatever micro you choose. You will also need some sort of compiler like Codevision.

The basic principle is exactly as Jim describes. The coil is charged for a specific time period. The time period is set in your code. After the current is turned off to the coil there is a delay of about 80 milliseconds. Now the counting is started. There are two timer/counters involved. The first is counting the amplified precession frequency. The second is counting the number of beats of a 10 mhz clock. The first counter counts to 256 and then overflows and starts again. After it has counted to 1024 the second timer is stopped. That is the count that is used to determine the reading. It is divided by "alpha" to get the displayed reading. Alpha is determined by the number of overflows and the clock. As I said I am using a 10mhz clock and a 1024 count. If you used anything different then alpha will be different.

For testing I downloaded a tone generator from NTH. If I input a frequency of 2048 then it takes exactly half a second for the counter to count to 1024. Half a second on the 10 mhz clock is 5 million. I tested using the raw count and I get counts of 4,999,850 or so. That is very close to the exact number you would expect. If my memory is correct I believe that gave a reading of around 48500. I would have to set everything up to check that number for sure.

I took my mag out in the cold one day and the readings I got were all wrong. I assume it is because when the relay is turned off it mechanically switches to the other contacts and that the cold may have slowed down the switching. That is just a guess. I have tried it in the warmer days and it appears fine.

Jim's new method of switching is much better and I have purchased the parts but have not built it. His new method of measurement is totally different than this method and I don't understand the process.

I am still playing with coils and as I said a gradiometer version using a mega162.

Rick

[Fred]

It almost sounds as though I am offending you and that is not my intent.

Hi Rick,

If i sound offended i am sorry it was not my intent
It´s only that i respect very much others work, and i know how easy it can be to ruin a builders good will when listening to critics that may not seem positive.That´s why i proposed a positive way to do it.
Anyway thanks for posting your info, as Ap i am interested too by the mag and i am following the thread .But i understand nothing of programming language.Just electronics.
Good luck and best wishes,
Fred.

[Unregistered]

Hi Fred

Having an electronics background puts you ahead of me. I had no programming or electronics experience. My electronics was six weeks in high school and about 3 hours per week at that. Anything we did used vacuum tubes. For instance in Jim's writeup he describes amplifiers and how they may need to be modified for different frequencies. You would know what he is talking about. I always have to find someone to explain it to me.

The programming is something I am slowly learning. I understand now what Jim has done but I could never have done it on my own. Whether you built something patterned after what I have done or the one that started this discussion, you will have to program a micro. The program I use is Codevision as a compiler and I use the free version. The NTH program I mentioned is also free as trial version. The expense lies in buying the STK500 or similar programmer. There are plans on the net to build your own. I paid about $80 when I bought mine a few years ago. There is also a newer programmer out called a Butterfly but I am not familiar with it but I believe it is even cheaper.

Along the way I learned something called In System Programming or ISP. It is a better way to program the micros as you do not have to remove it from the mag and install it in the STK500. However I don't believe you can do that with the 90S2313 as the pins used for ISP are the same as the ones used for the LCD. I would go to something like a mega8515 or the one I am using, the mega162. The code will work on them as well. You would have to change a line at the start to read "include mega8515.h".

If you do decide to build one I will ask Jim if I can share the code with you. I doubt he will have any problem with it.

Rick

[ApBerg]

Progy

Hello Rich.

Some years ago I made a little pcb for programming the 90S2313, it is some sort ISP programming. The info for it I found on internet.( can not find it now..) Connect it to the com port from a pc and use ponyprog as software. As power a 9 volt batt. is used. There can be better ways.... but it works.

If you have some files for programming .. ?( sourse code ?) I like to tray it !
I am not a expert in programming but like to experiment with it in the time to come.

Regards

[Unregistered]

I checked on this computer and I do not have the source code for the 90S2313 on it. I stopped working on it a couple of years ago and moved to the mega162. I will have it on a cd or on my old computer. I will find it and send it, hopefully tomorrow.

It is done in Codevision and I have no idea how it will work with any other program. There are 3 versions of Codevision. I use the Evaluation version and it is free so you may be better off to download it to your computer. Your programmer sounds like it is setup similar to STK500. I have a power supply that my son built instead of a battery.

Give me a day or two to find the code and I will try to post it as an attachment. I don't know if you can do that or not but I will try. You will be able to open it in Notepad or similar program and then just copy and paste it to Codevision. You will have to start a new project in Codevision first. It has good help files so it should not be difficult. Even I managed.

Rick