Hello from Vancouver Canada

Your chance to introduce yourself and your vehicle
Post Reply
caletron
Posts: 3
Joined: Tue Mar 10, 2015 10:40 pm

Hello from Vancouver Canada

Post by caletron »

My name is Chris, and I'll be starting with a mildly ported Mazda 13B Wankel rotary engine. My "testbed" is a second generation 1986 Mazda RX-7. After getting the first installation sorted out (confident aren't I :) ) I'll be moving on to a more extreme street-ported 13B and Whipple supercharger I have waiting. Finally, if finances allow and I can convince my brother-in-law to sell me his 20G three-rotor then I'll probably have something to satisfy my need for speed for the time being.

I bought the RX-7 new and it now needs a fairly comprehensive restoration. I hope to be finished when the car turns 30 years old this fall. Strangely enough I don't seem to be able to find parts for it as easily as I used to :) . This project will be in parallel with the restoration so it may move in fits and starts.

About me: I began crewing for road-race teams about 1980 and kind of tapered that off ten years ago. Professionally, I was an electronics research technician working at a large university. My nick, caletron is shortened from Caletronics and dates from the mid 80s and there's a story to that.

Around about 1984 I was convinced by copious amounts of alcohol to build an EFI system for a local racer. Andy had been struggling with a turbocharged Hilborn injected 1.6l Mitsubishi. Mix trying to get the right "pill" with 25-30psi boost levels and it was a bit too easy to kill engines along with very expensive forged pistons, rods, etc. What with keeping the rest of the race program going (for instance dry-sump oiling and ground effects at the amateur level was new at the time and so on) he wanted some help.

Since I had been fooling around with the 8MHz 16-bit MC68000 and I had access to an assembler for it (on a roll-around mini-computer!) I had little choice of which CPU to use (at the time the OEM systems were all eight bit and they were struggling to get a good product within that limitation; I wanted to free myself from that at least). A C-compiler (sheer luxury!) was only available to me a few years later. Anyway, a year goes by with things like, for instance, changing to the Vega 2.0l sleeved aluminum block (re-do the optical sensor on the camshaft), trips to Seattle to calibrate a hot-wire anemometer MAF sensor, PCB design tools were mylar, tape, templates and a knife. Updating firmware (I'm not sure that word existed then) was via UV EPROMs.

Remember, no laptop, no feedback except what it sounds like in the shop, at the pits, and driver reports. But we had a very successful project. While it held together. After a couple of seasons and turning up the boost race by race we noticed at the end of one session an oil leak, hmmm, from a crack in the block. Strange, well another block was prepared, fairly promptly blown up and then we knew we had hit the limit. The pressures involved were tearing the cylinder portion of the block from the crankcase :o . We contemplated putting a girdle from valve cover to oil pan but decided done was done.

So, back to Caletronics. I had been driving a '71 Nova and then a '68 Buick Wildcat and was different from my friends in that I drove "American Iron", as opposed to Japanese. That and some stunt driving earned me "Cale" as in NASCAR driver Cale Yarborough. After I started applying electronics to automotive it mutated to Caletronics for my, very small-time, business efforts. Fast forward through things like building a house (only family and friends' hands; no contractors except for gas lines and plaster) ... an aborted attempt at an engine dyno (not enough spare time!) ... started on building a Van's RV-6 aircraft (stopped after a medical condition prevents me from enjoying flying) ... changing career paths from electronics to Linux and systems administration and now I'm sort-of retired.

Other relevant skills: machining, welding, and generally all mechanical skills. Although my electronic skills are somewhat rusty I've tried to keep them up. My electronic design skills are way outdated: I could easily design a op-amp filter or a BJT amplifier, but these days it seems it's more cookie-cutter. Spend more time researching the right chip for the task and then spend very little time actually "designing" the circuit. I'm not saying that's bad but the skill set has shifted. My programming skills are oriented to sysadmin'ing so more in scripting or, say, custom Linux kernel patches but not a whole lot in large C/C++ projects like rusEFI.

I very much admire efforts like this in that I may be good at taking an idea and running with it but could never create or manage a large project such as this (to me this really is more complex than, say, building a house which is just a lot of gruntwork). I hope to be able to write patches for the rotary and contribute to rusEFI that way. At the moment I see a need for bits ranging from rotary trigger shapes and servo control of the Secondary Auxilliary Port, up to staged injectors (stock on my generation of motor) and servo controlled Exhaust Valve (à la BMW motorcycles), water/alcohol injection and supercharger boost control. I also have an idea for a skip-mode idle for bridge- and peripheral-port rotaries to make them clean(-ish) and street-able. For those that aren't familiar with rotaries these port styles are comparable to a piston engine with very large cam overlap with all the benefits and penalties that come with that. Disregarding my racing background I've always wanted to keep my street car a street car: I don't consider drive-ability, economy, and, yes, low emissions as optional. Also, since I am somewhat anti-MS Windows (those who know me use the word fanatical) and pro-Free software I may also contribute in any areas concerned with eliminating any dependencies that may exist (if they do, I haven't found any yet).

I expect I won't be posting too often so I'm making it up with this overly long one :D .

Chris

PS: I've just made my Tindle order. Am I right in thinking the STMF32 Brain Board is copy/pasted from the Frankenso board (or the other way around)? Is the layout on the Frankenso 0.2 board usable?
User avatar
AndreyB
Site Admin
Posts: 14727
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Hello from Vancouver Canada

Post by AndreyB »

caletron wrote:Around about 1984 I was convinced by copious amounts of alcohol to build an EFI system for a local racer... But we had a very successful project. While it held together.
PICTURES?!?!?!

Hey Chris, that's an impressive intro - glad to have you around. I am glad that you do know that at the moment we have only started four-stroke engines, but hopefully there is a chance to support rotary as well. All I know about those is that they are loud as hell on the track and it's all about a "seal" :)

Frankenso 0.2 yes you can solder the chip and some smaller components right onto the board, but the board you'll get in the mail would not have it soldered - we did not get this far with the assembly service we are using so far. Somewhere on the forum there should be pictures/videos of the board which runs stm32 chip "natively". I personally is not capable of soldering this myself.

I actually have a kind-of-dependency ticket for you: https://sourceforge.net/p/rusefi/tickets/158/ :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
kb1gtt
contributor
contributor
Posts: 3777
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Hello from Vancouver Canada

Post by kb1gtt »

Welcome along and great intro. As luck has it my dad has a bunch of crack pot rotary ideas, he's even made some prototypes. He's a crazy old man. I have some knowledge of rotary engines, but my knowledge is limited to some crazy crack pot experimental engines and a couple youtube videos about wankles.

If there is something I can help with I would certainly be willing to help. I'm mostly an hardware fellow, so I can help with schematic type stuff and the more physical parts. I follow software, but have trouble generating it. I correlate it to learning via audio books, I can hear it, but I'm not very good at speaking it.

From what I know, the core software should allow for rotary engines, but will probably need some modifications. Also as you noted you would either need an crank wheel that has already had a decoder written for it, or you would need to get one developed. Do you know VR or Hall? Also the silk screen notes CAM and CRANK, but function wise it's more like "full rotation / primary" and "incremental steps / secondary".

Again welcome along, and feel free to ask questions.
Welcome to the friendlier side of internet crazy :)
User avatar
kb1gtt
contributor
contributor
Posts: 3777
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Hello from Vancouver Canada

Post by kb1gtt »

Oh, also did you find the wiki yet? http://rusefi.com/wiki/index.php?title=Main_Page
Welcome to the friendlier side of internet crazy :)
caletron
Posts: 3
Joined: Tue Mar 10, 2015 10:40 pm

Re: Hello from Vancouver Canada

Post by caletron »

Apologies for the delayed response. As I said I won't be posting (or looking) often.
PICTURES?!?!?!
The 7 isn't much to look at: it's like any other of its' generation. Oh, you probably mean the race car :P ... well I do have slides somewhere (do young people even know what they are?) I'll have to dig through thousands of them (from a young age I envisioned organizing all my pictures as a "later" retirement project so I could put it off at the time). But thinking about it I realized though that I have no pictures of the EFI chassis. I built it. I know what it looks like :) . Not to worry: I know Andy still has it in storage, I'll just have to bring a camera with me next time I go to his shop.
Frankenso 0.2 yes you can solder the chip and some smaller components right onto the board
Good. I just wanted to make sure that if I populate that part of the board it might work even.
I actually have a kind-of-dependency ticket for you: https://sourceforge.net/p/rusefi/tickets/158/ :)
OK. Actually, I haven't looked at any Tuner Studio code at all yet. I figure that's something I'd get to sometime after the first unsuccessful crankover (tongue slides into cheek). To be honest that area of coding is something I have little experience in; so, yes, I may tackle it but it might take me a long time.

caletron
caletron
Posts: 3
Joined: Tue Mar 10, 2015 10:40 pm

Re: Hello from Vancouver Canada

Post by caletron »

kb1gtt wrote:Do you know VR or Hall? Also the silk screen notes CAM and CRANK, but function wise it's more like "full rotation / primary" and "incremental steps / secondary".
The second-generation RX-7 motor (3rd-gen? 13B) replaced the distributor with a dual VR sensor. It looks like a distributor with a blanking plate on top instead of the cap. Inside, the distributor shaft has two, stacked, rotors and pickups. The top has two teeth, the bottom has 24. But, since it's driven at one-half speed of the eccentric shaft (aka crank-shaft) it's essentially one primary and 12 secondary. And since a rotary fires for every TDC event (like a two-stroke) the primary tooth is an indicator for a set of fuel and spark events. The primary tooth is phased between two of the secondary. I'll have to partially re-assemble some of the engine to determine the phasing of the primary tooth relative to TDC of rotor 1. Some mechanical adjustment is available just as in any normal distributor.
Again welcome along, and feel free to ask questions.
Thanks, and I will. Right now, though, I've stripped everything from the firewall forward and there's a bit more rust repair than I anticipated so I'll be burying my head in that for a while.

caletron
User avatar
kb1gtt
contributor
contributor
Posts: 3777
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Hello from Vancouver Canada

Post by kb1gtt »

Great, VR is an input option for the frank series of boards. As well that wheel pattern should be reasonably easy for decoding. I don't think that particular wheel exists, but I think a very similar patterns does exist. If you had an engine where you could log the crank signals, I think it would be quick to get a firmware made with the wheel decoder.

Also single spark per rotation is common and easy to configure. So I don't see any real issues there.

Are there any special algorithms that are needed for Wankel engines? So far it sounds like the standard tuning stuff would work for you, and that the only real lump in the rug is the wheel decoder, which can be made fairly easily.
Welcome to the friendlier side of internet crazy :)
Post Reply