DIYEFI.org Forum
http://forum.diyefi.org/

Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS
http://forum.diyefi.org/viewtopic.php?f=56&t=2842
Page 1 of 2

Author:  volvoguy [ Sun Feb 18, 2018 11:45 am ]
Post subject:  Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Clockwise rotation of the Hall effect 4/2 DSM/Miata/MX-5 pattern with a cam degree wheel.

The standard rotation direction is anticlockwise, the change in direction has reordered the highlighted event numbers inner<->outer.

  • 70/110 for outers
  • 140 for long inner
  • 72 for short inner slot
  • 54 degrees from leading of long inner to leading of outer contained within long inner
  • trailing inner edges 360 apart
  • sensor offset not known

A diagram to match the one in the other CAS thread:

Image

Those events occur at the following crank degrees relative to each other:

Image

Fred, shall I try to populate the rest of the tables from the other CAS thread too?

Author:  Fred [ Sun Feb 18, 2018 1:42 pm ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Wait! Why do you want to do this? Does using the Yoshifab adaptor spin it the opposite way? EDIT: Yes, of course it does. Mounted on the back of the head in stock form, and Volvo block dizzys do go clockwise.

If so,you can't even just flip the disk as the two sensors are offset from each other. Better to get a 24+1 disk for it and just run that decoder?

When my Yoshifab adaptors arrive I'm going to have a FreeEMS-specific disk made up and write a decoder for it, it'll work better in every respect than any of the stock ones or current naive aftermarket ones.

Author:  volvoguy [ Tue Mar 13, 2018 11:09 am ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Fred wrote:
When my Yoshifab adaptors arrive I'm going to have a FreeEMS-specific disk made up and write a decoder for it

If I'd bought a Yoshifab adapter instead of bonding a Volvo distributor to a Hall effect CAS I'd probably do the same. I've asked around about whether or not the optical CAS and Hall effect CAS internal assemblies are interchangeable within the cast housings. If they were I could swap my Hall effect assembly for an optical one and use the new disc. Perhaps unsurprisingly, because the units are so interchangeable no one seems to have ever tried swapping the sensors between housings.

So I think my options are
  • To buy an optical CAS, get lucky and have the sensors fit in the Hall effect housing, on ebay looks like £80.
  • To buy Yoshifab's adapter and an optical CAS which will come in hopefully a little under £300, making it the most expensive thing on the car.
  • To use the Hall effect CAS, and have to get code written that no one else will probably ever need.

Author:  Fred [ Tue Mar 13, 2018 11:44 am ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Honestly, writing the decoder wouldn't be too difficult, probably. If you're up for it, I can do code review until it looks right.

Fair enough re yoshifab adaptor cost - EU/UK import duties are fucking horrible. Any details on the "bonding" process you did?

If the Mazda dizzies are anything to go by the castings are different and you can't chuck an optical into the hall one (IIRC).

If you go semi sequential/wasted spark you can just use the crank. If you don't want to tolerate low RPM limitations until the code gets some rewrite, you can modify the pattern with an angle grinder to reduce the count to 20 - 1 or something like that quite easily. Though I spun the B234F in my 240 sedan to nearly 6k on the 60-2 setup tonight. Dog fooding it so that if it's an issue I will be highly motivated to sort it out for everyone :-)

80 quid sounds like too much for something you should be able to get for near free from a junk yard or from a forum or something. I need to go looking for Mitsu CAS units for when my yoshi adaptors arrive. I'm in no rush though, and I haven't shipped Josh back his ECU, either, so eye for an eye, it's all fair :-)

Author:  volvoguy [ Tue Mar 13, 2018 5:22 pm ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Fred wrote:
Any details on the "bonding" process you did?

Yep, wrote a thread here that photobucket has dinked a little bit. I could rewrite with new links.

Fred wrote:
80 quid sounds like too much for something you should be able to get for near free from a junk yard or from a forum or something.

Yeah. A lot of scrapyards here seemed to disappear ~10 years ago though. The price of scrap steel went up. Brownfield land was getting snapped up for crappy townhouses and tenements. Then around 2009 I think 'cash for clunkers' shipped thousands of cars abroad that would have trickled into scrapyards over the following few years and kept them going.

I could probably scrounge one up cheaper on a forum though, yeah.

Fred wrote:
Honestly, writing the decoder wouldn't be too difficult, probably. If you're up for it, I can do code review until it looks right.

I'd love to give it a bash, what's involved?

Author:  Fred [ Tue Mar 13, 2018 8:01 pm ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Forgot about that! Thanks. FWIW there is a firefox plugin that fixes photosuckit - I must have installed it in this instance because it worked fine.

Not much, copy the existing one into a new file called similar and insert the word "Backward" into the name somehow that makes sense, then change the sync logic around to suit your data above (that I didn't review, yet). It's probably half an hour's work for me, but maybe a bit more figuring stuff out for you. Give it a shot and link your attempt and we'll go from there.

Author:  volvoguy [ Sun Apr 15, 2018 4:55 pm ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Fred wrote:
copy the existing one into a new file called similar and insert the word "Backward" into the name somehow that makes sense

Could you point me in the direction of the originals? I looked everywhere I could think of and only found hex files and terror.

I've been working on a rather compact exhaust manifold to hang my TD05HR-16G6 up next to the B230, will soon have photos :)

Author:  Fred [ Mon Apr 16, 2018 10:49 am ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Keen to see, especially if it's reproducible for my 240 setup because I have the exact same turbo or at least, the slightly shittier one that doesn't end with 6, not sure.

As for the source, you didn't look very hard! :-D

src/main/decoders/*4and*
src/main/decoders/inc/*4and*

Enjoy! PS, you may need to modify the sync logic significantly depending on how things line up. I didn't check the pattern for reverse sync. I guess it's probably very similar, though, just with angles swapped for clockwise vs anti.

Author:  volvoguy [ Fri Jun 08, 2018 11:06 am ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Fred wrote:
As for the source, you didn't look very hard! :-D

It's always in the last place you look, right? I had actually never even opened the vanilla folder before :oops: I was looking through repos and hosted files.

The .h file I had no problem editing, or at least largely understanding. I've marked up every line I've edited, and uploaded them to my fork.

The .c file reminded me that a computer science bachelors does not mean that you can dip into code and have a clue what's going on. I edited the #include and @comment statements... Is there much in that file that needs editing/understanding?

Author:  Fred [ Sat Jun 09, 2018 1:51 pm ]
Post subject:  Re: Clockwise Hall effect 4/2 DSM/Miata/MX-5 pattern CAS

Dear Volvo guy, I took my 16v 740 for a burn tonight :-)

I see your diff includes some angle and ordering changes which seem reasonable, and your source file is almost identical to whence it came.

Question, you say "sensor offset not known" but how did you come up with the *relative* timings of the inner/outer sections without that or without spinning it and scoping it? This may mean the work I did tonight for you is wasted, or not. Depending on how you answer.

I've done most of the changes to your C file, but not finished, so not pushed, yet. Need a clearer head to complete the task, but if you want to duplicate work for your own edification, look for the lines around 'correctEvent' and modify those if statements and contents thereof. We can compare branches afterward. Or if you just want the result, I should have it for you tomorrow. :-)

Page 1 of 2 All times are UTC [ DST ]
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/