Alternator Control box for standalone MS install for NB
 

The following information is courtest of AbeFM. I credit all of the following information to him and his website.


Ok, picked one up - had a little trouble with it for a bit, eventually proved to myselve stuff was working by putting at 20 turn pot between power and the signal in line to the altenator, and finding out that ~0.980 volts got it charging to around 14.3V.

Anyway, armed with said knowledge, I found out the correct wiring for this. Thought I'd note it here for posperity:
To wire up NAPA part number:
New Voltage Reg Napa online MPEVR38SB $15.29 "VR38SB"
Connector for VR above Napa online ECHVRC38 $12.64 (8.99?) "VRC38"
on a '00 miata:

Ground the case, I found it convenient to mount it near the radiator overflow bottle.
Hook the BLUE wire to switched(?) " 12v" battery voltage
Hook the GREEN wire to the Grey/Red wire on the alternator, pin 1-O on the factory ECU.

I let line 1-T hang, I'm not really sure what it is, it seems to be measuring the voltage of one coil of the generator, before rectification. Not sure what that's good for, maybe you could get engine RPM out of it, or total altenator capacity.

Anyway, hooked it all up and everything works well enough. It seems to only charge to like 13.5-13.7, I think this is due to the Mopar part only putting out 580 mV, where I got better charing feeding ~980 mV. Much more than that and you're rapidly feeding 15, as much as 18V into the system.

http://abefm.smugmug.com/photos/238575862_wgHfe-M.jpg

http://abefm.smugmug.com/photos/238575575_G6kF6-M.jpg
http://abefm.smugmug.com/photos/238575975_njgyh-M.jpg

http://abefm.smugmug.com/photos/238575658_ksdhE-M.jpg

http://abefm.smugmug.com/photos/238576710_BqKgN-M.jpg

http://abefm.smugmug.com/photos/238576654_vTAJi-M.jpg

http://abefm.smugmug.com/photos/238576093_onX48-M.jpg

http://abefm.smugmug.com/photos/238576308_7B7zT-M.jpg

http://abefm.smugmug.com/photos/238576034_Qb59r-M.jpg

http://abefm.smugmug.com/photos/238576374_HQQNo-M.jpg

http://abefm.smugmug.com/photos/238576075_fR7pH-M.jpg

http://abefm.smugmug.com/photos/238576149_oKipg-M.jpg

http://abefm.smugmug.com/photos/238576149_oKipg-M.jpg

http://abefm.smugmug.com/photos/238576264_av4Xf-M.jpg

http://abefm.smugmug.com/photos/238575982_ns8m8-M.jpg

http://abefm.smugmug.com/photos/238575867_8rQvb-M.jpg

http://abefm.smugmug.com/photos/238576775_nSs5m-M.jpg



Text and images courtest of AbeFM forum member. For image captions follow the following link.

Original Website for mod: http://abefm.smugmug.com/gallery/409...38576775_nSs5m

Heh! Thanks, that's awesome! I'm way too lazy to take the time for that.

I need to add something to that, basically don't get the wrong part. A friend did, and instead of putting out half a volt it was putting out 3 or 4, and his battery was showing over 17V.

It was a similar number, I think it was something like VR38615s. I might be able to find out. It also had a different color plug. When I find out, I'll edit this post.

Nice....looks like i wont be buying an na alternator after all.

sweet. at one point, jason c made me a similar device that was wired in between the alternator field coil and battery voltage. this was pre hydra's ability to reliably control the alternator.

Hmmm, what was in that box, I wonder? Sounds like the same thing. This uses case ground (I think) and those same two wires....

Ok, so hook up the blue wire coming from the regulator pigtail to switched 12V+
Green wire from regulator pigtail to grey/red wire on alternator
And ground case and it charges? That's it? Nothing more or less?

Sober, I would be all over that. But I do remember it being two wires, one to +12V, one to the altenator, and the case I did screw to the body (don't know if it needs ground or not, but even my faith would be stressed to imagine it wouldn't.

My car has been running months on that set up.

Details on my smug mug.

There isn't a 1:1 correspondence between the alternator control input voltage and the output voltage. There is just a minimum reuired *current* into the alternator input pin that is needed for the alternator to give out maximum output (all appliances on, battery not fully charged). The voltage at the input pin for a given input current will vary with temperature.

Interesting! I didn't know that - certainly it was black-box operation to me, I just read somewhere it could be done and worked out how to make it work. Not my favorite way to operate, but certainly it was expedient.

Er, how happy were you with your box, Jason? What goes in it, does it dump a lot of power?

As I'm considering this interposer board, I was just going to leave wires hanging off for the mopar box, but if you think yours would work better, be smaller, or cheaper, I'd like to incorporate it. Tell me a bit more about it?

Just a comment. Since this is for NB installs... maybe make a point of it in the original post or the title. That way the people who don't require it... don't come in here thinking "oh great... something else I have to do".

Jason,
I never asked - you built a circuit to do this? I'd like to see it, for personal reasons just to see it, and to borrow if you don't mind.

Just saw your post. Hmm, I'll have to dig up the schematic and PCB layout, and I'm too sleepy right now to do it. :p

Knock knock - hey Jason, ever find this?? :-)

I just found out, btw, why mine works on my car, and not on anyone else's. I put in my little write up to use pin 1O when it's really 1T, now I got to go fix this on all my drawings. Thanks to pschimt on here for catching this!

Schematic

Give me a quick talk through, since you know, I didn't major in engineering or anything. It sinks current (sources) based on the ratios of the resistors, such that it turns on if the voltage is too low, and off if it's too high?

Does the diode just protect stuff, or is it there as a voltage reference?

I think it'll make sense when I look through it. Have you used it? Does it get hot (are those big transistors?)? What voltage does it "like" in a miata?

It might be a cool thing to make a simple board for.

Those are small transistors.
The TL431 is a voltage reference / error amplifier.
It's configured as a "hysteretic" controller. The setpoint is set by R4 and it's at 14.2V.
I've used it and so have 3 other people.

would you say this circuit is more accurate than the hydra 2D map and PID control?

Faster responding, yes.
More accurate, probably.

dont know why I ever pulled it out and expected to tune the hydra.

Tuning controls is a pain, best avoided. I see now why my buddy was so into analog controls when he was in college. :-)

That's awesome, Jason. Do you care if I use it in a "for sale" if not for profit device? PM if you could.

BTW - which plug on the harness do you use? My car is a salvage title, but I assume the alternator is stock. I have my Mopar controller hooked to the Gray/Red wire (ECU Pin 1-O), but according to the schematic, I assumed it would be the Gray wire (Pin 1-T). The gray wire has a pink dot on it, on both ends, but it's certainly gray.

http://abefm.smugmug.com/photos/454840443_waX4M-L.jpg
It's hard to tell but I carefully verified last night, it has a stripe, the other does not.


http://abefm.smugmug.com/photos/454840442_FLrnm-L.png http://abefm.smugmug.com/photos/454840427_S5cv4-L.png



It sure looks like that control transistor is there to turn the field coil on and off, and the wire I'm using watches the AC generated by the alternator. Or, am I just totally wrong here?

Either way, I've been driving 9 months or something like that on this, and my battery hasn't died, so I keep thinking it's working. The connector is strongly keyed, there's no way it's in backwards...

Matt, any idea what pin you're using? Or Jason? Those four guys, would they know?

I wired it into the hydra harness... but it's the one in the 99 manual called "generator field coil"

This mod is fine and all, but wouldn't it be easier to just pick up an NA alternator and swap pulleys?

Just a thought.

Tony

I can build this circuit in less than 30 minutes and it will cost me maybe $7 in parts. I can't do an alternator swap in less than 30 minutes and it will certainly not cost me $7.

Jim

I have a spare GM alternator control box if anyone wants to do testing.

The circuit doesn't need (more) testing, it's been used in four places.

I'm wondering what wire to put it on. But everything I can see, it looks like you want to put it on the wire I'm not using. I remember being confused at the time, since I had it on the other wire and it didn't work. When I put it on the gray/red wire, BAM, things were working. Maybe the wire colors are wrong in my book. Or... I dunno. I should check which ECU pin corresponds to the wire I'm using. Everyone PM teamplur and make him make me do it.

Abe, do you have a 2000 and a 2000 FSM?
In my FSM the alternator field and neutral wire colors were swapped IIRC.

Can a mod please change the subject to "Alternator..." instead of "Altenator..."?

Jim

Oh good lord, that's what I was thinking... At least I can pretend to have SOME understanding of how the thing works. :-)

Yes, I'm pretty sure that would be the issue - I was going to check it on my car, but it's nice to have someone else verify it. I'll hopefully do that tonight. Sucks, though, since there's more vias now on my board. :-) But I'll take a board with more vias that works over one that doesn't work at all.

o lookit wut i foundz!

sorry about the pictures. but maybe you can reverse engineer it to see if it's the same as jason's schema.

since that sucked, here's something more fun.

http://y8spec.com/miata_tech/nb_alternator_ctrl.gif

I guess since you labeled everything it means it checks out? :-)

It seems I have a strange issue (apart from being a spelling Nazi apparently). The NAPA alternator control box works as long as the stock ECU is in place: I feed B+ to the blue wire, and I get PWM on the green wire (note that the green wire in not connected to anything). I remove the stock ECU, start the car with the MS2 (standalone) and I get nothing on the green wire (voltage is at ~11.5V on the blue wire). What gives?

Jim

you got me. maybe re-verify your power source or tap right into the ig switch "on while running" circuit instead of the ECU pin?

Jason, thx for an other great circuit.
I reordered the components a bit so they fit in the proto area, pretty neat if I say so myself. On the bottom right side of the layout, you'll notice 2 unused spaces. This corresponds with the 2 missing holes in the proto area.
As the circuit only uses 50% of the proto area, there's room enough for other stuff like EBC, Vics and Fan mods.
There's probably not enough room to fit Abe's input circuits, but I built those in the VR area of the board.

re-installed the circuit today and it holds system voltage nicely just under 15V. if I switch on everything (heads, fogs, defrost, heater fan full, hold brake, and hold both window switches) it maintains about 14V. probably less if the fans come on. all at idle. I guess my alt is ok?

LOL that sure looks like my handiwork.

I have another circuit for lighting the warning light when voltage drops < 12.9V (as opposed to the typical light that means the alternator is putting out current, even if it's not enough to keep the battery charged).

Interested to see the schematic Jason.
BTW, is it necessary to use the 2% version of the TL431? Datasheet says 'guaranteed thermal stability over applicable temperature ranges'. Local shop dude looked at me like a cow to an airplane when I asked for a 2% version.
Also, I assume the 48K7 regulates current? I have little electrical stuff in my Westfield so I might need to lower current on my setup.

You can use a 1%.

If you use one looser than 2%, you can trim ithe 48.7k resistor to get the exact voltage if it's off a bit.

The circuit regulates voltage, not current.

I think the schematic is above?

Wow, that's a nice lay out, I like it - and all without resorting to unusual stacking methods. You know, with a couple of those posted, it might be possible to come up with a "standard" NB install.

nothing is stacked weird on the one jason made for me. it's about 3/4 x 1" total size. not bad.

I meant the warning light circuit.

Thx, I'm working on a how-to. It's up on my Westfield site but I'm only halfway done. I'll post it here when it finished. As I have no access to the car right now it might take a month or 2 though.

Jason's is perfect, it's just that I'm trying to install all circuits (alternator, CKP & CMP, EBC, Fans, Vics and WI) on the MS board and often that means stacking components in an unusual way. Not so here :).

I'd have room in mine as well for that, though there's a bit of weird stacking. If I had it to do over, I'd leave off the EBC, which sucks horrifically. Maybe the new PID code on MS-II works, haven't tried it yet, if it did, EBC would be a nice thing to have. Certainly can't complain about spool on my ball-and-spring, though. I don't remember where, but I put mine on the heat sink, replacing some unused thing, threaded it into some of the circuitry at the DB-37 end of the board. Even the trace sizes made sense.

Here's the 12.8V threshold warning light.

edit: a 0.1uF or 0.22 uF cap between top and middle terminals of TL431 is necessary to make the warning light act digitally"

cool, thx

Here's the proto circuit with
- 2 programmable outputs for fans and vics
- knock input for knocksenseMS
- alternator field
- 2 active low inputs for launch control and table switching. Altough the manual says otherwise, JS7/PE0 is for launch control. R1/PE1 is for table switching! You need to free up 2 additional spare ports on the processor if you want to use these options.

There's not enough spare pins on the DB37 for all circuits, so I added an old chassis mount keyboard connector. I labeled the active low inputs AL1 and AL2.

There's not enough room for a 2N2222A (metal can) so use PN2222A (plastic) or ZTX450 instead. These are direct swaps.


http://westfieldmx5.devocht.com/star...lternator1.png

What are y ou using for a knock sensing circuit, where'd you get it from?

I haven't actually decided. I think I'm gonna use the MS circuit with the oem '99 sensor. I have access to a Link Map sensor (with built in knock circuit), so I might reverse engineer that at some point. If all that fails, I can always get KnocksenseMS, but I'm not convinced it's better though.

I have yet another schematic, for a knock sensor amplifier that works with the factory 99/00 knock sensor, and is tuned for the miata knock frequency...

You never fail to amaze me. I'm sure you know what my next question will be :).

I guess my question will be the same as Frank's...

where, how, how much....

can it be used with a parallel install so the stock ECU and whatever else can both see the signal?

With some mods yes.
The circuit is meant to emulate a GM style tuned piezo knock sensor.
IOW you take a 99/00 knock sensor, it filters and amplifies the signal. It is tuned to 13 khz IIRC, which is the peak frequency that I noticed when I did recordings of knock.

Started a new thread in ECU's.

Just found out that I forgot the circuit for the tach out in the circuit above.
Made a new one with tacho out.This one is using JS11 for tach out. It comes out on IAC2A on the DB37.
JS11 was previously used for EBC, but there's just not enough spare outputs to do everything.
FWIW.

http://westfieldmx5.devocht.com/star...lternator2.png

You can change the 3k resistors to 3.3k, and the 30k to 33k.

However the 48.7k needs to remain 48.7k - you can substitute this with a 51k in parallel with 1Meg - but if you use 5% instead of 1% resistors where specified, test that the voltage regulation setpoint is accurate, at 14.1 to 14.4V.

If you do that, you don't bump the 9k to 10?
k?

Also, the ztx690 is a 2 amp NPN - do you need two amps?? Couldn't I just use a 2222A (1 amp)?

What 9k resistor?

Are you looking at the 12.8V warning light circuit or the alternator controller?
The former needs a ZTX690 to drive the warning light because of the heat generated and the gain required, not because of the current.

Jason,

Can I test the alternator circuit out of the car?
I built it, but would like to test on the bench before hooking it up to the car.
When I apply 12V and ground, am I supposed to measure anything on the Field wire?
Increasing power from 12V to 14.4V make a difference?

How to test
 

Place an ammeter between the field output of the circuit to ground.
When the battery voltage is below ~14.2V, the ammeter will see ~13 mA. When the battery voltage is above ~14.3V, the ammeter will see 0.

Jason's circuit never worked for me. I rechecked the pcb I made 100 times and I still couldn't find a glitch. So obviously enough, I did what every geek should do, I made my own PID controller using a microcontroller and a 2x16 LCD. Did I say that it is totally tunable? Pics in the afternoon.

Jim