Jump to content
Sign in to follow this  
abecedarian

Hercules engine control

Recommended Posts

Well, here's where I'm at so far. :)

(The air-wire is due to those pins having the same assigned function.)

 

post-256-14264605360861_thumb.jpg

 

So, you can see it's roughly the size of two Hercules LP's side-by-side.

 

Ground plane on back. Left is 3v3 plane and right is 5v plane.

 

Injector drivers will go top-right. Ignition coils are logic-level (5v) driven at low current so those drivers can go most anywhere.

 

 

At the connector, the four paired pins are for injectors' signals; I'll probably un-bridge them and use 4 pins for injectors and the other 4 for ignition. Below those, the next row of 4 conductors are for 12v, the 4 below those are grounds, the next 4 are 5v0 out to sensors. The ones air-wired together are as of yet unassigned but will be for sensor signals, fuel pump relay control, and some other as of yet undecided things.

 

The bottom 4 are from two VR sensors, which I'm still working out how to get the MAX9926 connected to the LP. I sort of want it off to the side and away from things since the VR's can output > 300v AC, albeit very low current.

 

The op-amp already there is dividing the 5v signal, MAP, TPS and IAT and sending those to ADC 0-3. There will be a second op-amp for sampling coolant (CLT) and anything else I add, and those will go to ADC 16/17 & 20/21 on JP1- the non-BP header, probably.

 

I need some capacitors on the 3v3 LDO too.

 

I'll post Eagle files if anyone would like to take a look.

Share this post


Link to post
Share on other sites

Thanks for the like @@bluehash. ;)
 
But.. I'm changing things up, as I always do. I think that's why I don't get anything finished. :unsure:

 

I'm moving the Hercules headers to the right side, and the connector to the left side. I realized that trying to get the signals routed would be easier if I took advantage of the wide open space at the top-left of the Hercules board. :D

 

 

I've a few other considerations leading to this and have been spending some time on TI's Webench, so the information from there needs integrated.

 

I'll post a render soon.

 

Share this post


Link to post
Share on other sites

So, here's where I was going, but umm... okay, not finished... no way... no how.

 

post-256-14264605362887_thumb.jpg

 

12v0 plane at the bottom left.

5v0 plane at the top left.

3v3 plane at the right (top to bottom).

Ground plane underneath it all.

 

Reverse voltage (for main supply) down in the lower left corner, with LED's to tell you things are backwards.

"Green" led on means it's correct; "red" led on means it's not. :D Not sure I need a fuse with this either.

 

Have most sensor return signals assigned and routed to pins as evidenced by the wiring to the two op-amps; and 3 "EXT[ernal / ra] I could incorporate into the code later... O2, knock sense, or ...?

 

 

As always, I'm open to suggestions.

Share this post


Link to post
Share on other sites

So....

 

post-256-14264605363704_thumb.jpg

 

As mentioned, have been re-working things. I've altered the planes a bit, and as of this moment do not have a 5v0 plane; 3v3 plane has been shrunk to around J1, J2, J3 and J4. Bottom left is a top-layer ground plane with some vias punching through (I will add a few more ;) ), and 5v0 and 3v3 supplies, mostly. Some of the component values are wrong on the 5v0 side but I was only looking at real estate since the Webench 5v0 and 3v0 supplies TI generated are mostly identical.

 

There is a 12v0 trace running on the bottom layer, near the edge, from left to bottom to right.

 

Top left is reverse power protection. Two LED's, one red and the other green, wired such that if the 12+ and ground are correct, the green LED lights, otherwise the red will. And below that is a p-channel mosfet and Zener diode. The mosfet should only turn on and pass current to the rest of the board if power is correct, otherwise the Zener should block current.... Oh, and a big heat sink, just in case.

 

I've got some 'scrunching' to do to the op-amp bottom-right, and picking pins on the connector for things as well as working out drivers for the injectors and ignition.

Share this post


Link to post
Share on other sites

A little off-topic, but relevant.

 

I suspected it for a while but have confirmed... my fuel pump and fuel pressure regulator are seized.

Pump is $150 on fleebay.

Regulator... no listing. :(

 

One solution is PWM control an appropriate, as in can build greater than 80 PSI fuel pressure, fuel pump.

 

If I do that I need another ADC channel (no big deal- I have 3 unused), but it's getting complicated since I'd have to control a 12v motor's speed in addition to everything else and 'free' Eagle has board size restrictions. I might have to do a separate board.

 

My thinking is do two boards, one with analog and power, and another with digital and this motor control.

 

The idea being that Herc, on the bottom, analog / power above, and digital / motor on top.

 

 

 

Thoughts?

Share this post


Link to post
Share on other sites

I'm not too worried about losing things. I wasn't that far into it.....

 

I did learn that my fuel pump and fuel pressure regulator are kaput! So, not too cheap to repair. Maybe will go with PWM control the pump to maintain fuel pressure as needed. ;)

 

Eagle 'free' has small boards so might need to make two- one for DC-DC regulation and analog interface, and another for digital interface and high-side drivers (for fuel injectors).

 

Don't know.

Share this post


Link to post
Share on other sites

I think -

- one board with the connector, power and analog on it, and this mounts on top of the Hercules board;

- one board with the injector and ignition drivers and motor control, which mounts below the first, adjacent to the Hercules board.

 

The 2nd board has its components on bottom with the ground plane on top so it'd be like: signal layer, ground layer, ground layer, signal layer; that would keep most of the noise the injector drivers generate isolated from the Hercules board.

Share this post


Link to post
Share on other sites

So, here's where I am at now...

post-256-1426460536401_thumb.jpg

 

Still scratching my head though.

 

*edit- chip on the far right is CAN / MAX13054. ;)

And the red polygon down the middle is meant for me to use when I start on the injector driver board.

So I know that I can't go past this area.

Share this post


Link to post
Share on other sites

Switched ECU plugs: don't see the need for all those pins on the one above. I'd need a multi-layer board to take advantage of the pins. So, found an incomplete but apparently accurate Eagle library part for an Ampseal 35 pin plug. I say reasonably accurate because they had photos of the board they etched with one mounted to it.

 

This connector is a little smaller than the other one, but being that I've a two cylinder motorcycle and the stock ECU had 25 pins, it's a win-win. Also it's available in straight and right angle variants so even if the straight is used, just mount it plug-down under the bike seat and it won't get wet. :D

 

It also let me move things around and get some room:

post-256-14264605364275_thumb.png

Share this post


Link to post
Share on other sites

So, the board is set up to sample the following (0-5v signals):

- 5v0 out from the regulator, a.k.a. reference voltage;

- throttle position sensor signal;

- manifold absolute pressure signal;

- coolant temperature signal;

- manifold intake air temperature (IAT) signal;

- O2 sensor (should be able to handle either narrow (0-1v out) or wideband (0-5v out);

- two additional "undefined purpose" sensors.

 

And, I didn't like the above posted layout much so... deleted the 3 post power / ground header (didn't see a need for that)...

... moved 3v3 and 5v0 regulators;

... moved the harness connector;

... re-routed the analog traces to there were ground traces between them for some isolation...

... if it looks weird it's because I'm adding polygon fills as I go:

post-256-14264605366116_thumb.png

 

 

Still contemplating if the analog needs some sort of filtering or clamping... I think I have some room to drop some R/C; and digital-type I/O signals which may need to be dealt with.

Share this post


Link to post
Share on other sites

Well, got distracted a bit, and in the process stumbled on:

http://www.infineon.com/dgdl/TLE6368_G2_DS_rev2+32.pdf?folderId=db3a30431400ef68011421b54e2e0564&fileId=db3a30432239cccd01225a67e02b6c60&ack=t

 

5v out @ 800 mA

3.3v or 2.6v out @ 500mA

3.3v or 2.5v out @ 350mA

- the latter two are hardware selectable by routing a trace on the board.

 

And...

6 x 5v outputs @ 17mA for sensors too.

... and tolerates short-circuit on all outputs.

 

I'm liking this chip, even though it can consume 2x2" of board space. At least that could be constrained to a portion of the board far away from other components.post-256-14264605370153_thumb.png

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×