PDA

View Full Version : HydroProp Project



Painless
05-22-2009, 06:53 PM
I've decided to start a project thread for the HHO management unit I am currently working on. I plan to keep a feature list in post #2 which may or may not undergo changes during development.

For those of you that haven't seen my previous posts, I am developing a micro controller based device that will handle PWM and automated vehicle sensor management. Please see post #2 for a full description of the proposed features.

As a basis for the device, I am using the excellent Parallax Propeller Micro controller chip. The Propeller runs at 80mhz and incorporates 8 parallel execution processors (known as cogs) to deliver 120 mips of 32 bit CPU power.

The main aim of the project is to provide a device that does not simply replace the need for a separate PWM, MAP/MAF & EFIE enhancers, but ties all these devices together along with feature improvement and a very affordable package.

Please feel free to post on this thread with any related comments or suggestions, your input will help to shape an end product that will be of benefit to all.

Russ.

Painless
05-22-2009, 06:53 PM
Pulse Width Modulator

The PWM feature will utilize input from the fuel injectors in order to control PWM duty cycle and frequency. This will provide the ability for HHO cell output to match engine fuel requirements.

User tunable options:

- Fuel to duty cycle multiplication factor

MAP/MAF Enhancer

A load sensing Dynamic MAP/MAF enhancer. This will adjust the level of enhancement with respect to the load that the engine is currently under.

User tunable options:

- Enhancement base percentage
- Enhancement to load curve intensity

Oxygen Sensor Enhancer

Provision for automatic control of up to four (two pre cat and two post cat) narrowband oxygen sensors.

User tunable options:

- Additional voltage offset (in addition to automatic tuning)

Later in the project, wideband sensors may be addressed also.

Safety

Exhaust gas temperature monitoring will be incorporated either by an included probe or connection to an external EGT gauges alarm.

User Interface

Output will be provided via a 4x20 LCD with backlight support.

Input will be via a four way directional keypad with an additional 'select' button (5 keys total).

Maintenance

A method will be provided for easy download of new firmware to the device, as updates become available.

Painless
05-22-2009, 07:15 PM
So far, I have completed building the basic development board for the HydroProp. This board incorporates the LCD screen and 5 button keypad along with other hardware to aid in development and debugging.

Work on the user interface (LCD and button) control code and basic menus is about 30% complete. I'm also currently going through a selection process for the external components I will need to add to the system, such as A/D converters. I've already purchased extra chips to allow a testing platform for installation in my truck.

I plan to re-install my HHO cell into my truck this weekend after I've finished upgrading the plumbing to 3/8".

Q-Hack!
05-23-2009, 02:16 AM
Does your EGT have an external output that you can feed to a computer? While mine is a digital model, it is just a basic sensor and meter.

Painless
05-23-2009, 01:05 PM
Yes, it has a 12v line to an optional warning bulb. When I get to that point in the project, I want to look at the thermocouples that are out there and see what's involved in interfacing them. Probably just an A/D converter.

Painless
05-26-2009, 08:55 PM
I made use of an old oversized project box this weekend and built a development unit to use in my truck while I'm developing the system. Its large enough for two breadboards, one for the micro-controller and its eeprom and crystal, another for the interface IC's as I incorporate them.

I ordered some samples today of digital potentiometers, some 8 channel 12 bit analog to digital converters and quad 12 bit digital to analog converters. As soon as I have these items and some voltage regulators I'm missing, I'll be good to dive into the rest of the project. Meanwhile, I'm working on the fuel proportional PWM code and interfacing with the fuel injector signals.
I'm also trying to decide if I should stick with my current direction of using a five key system for interfacing with the user (four directional arrows and a central select button) or to go for a full blown 3x4 keypad that will allow for numbered menu selection and entry of data. In terms of code there really is no difference, in fact, the keypad will probably require less code but I also want to keep the units size down as much as possible.

Russ.

Painless
06-04-2009, 11:26 PM
I've received some of the sample IC's I ordered from maxim, the digital potentiometer and a DAC. I'm working on code for controlling those at the moment.

Painless
06-18-2009, 04:06 PM
I'm having issues with the digital potentiometer sample that Maxim sent me, they just don't seem to want to work as designed. These are step up / step down type devices and I'm thinking that an I2C based design would be better anyway, therefore I going to evaluate other options here.

Due to these setbacks I've decided to re-order the project and work full time on the fuel / load related PWM feature for the moment, with the aim of getting a working prototype to Larry ASAP. Once this is complete, I will return to the mainstream part of the project.

Russ.

Painless
06-29-2009, 03:16 PM
Today hasn't been a very good day, after spending the past days working on code to read pulse widths and translate them into frequency and duty cycle I hit a wall today when trying to interface with my truck.

The wires to the fuel injector are reported by my multimeter to be operating at a hz rate which is what I would expect in relation to RPM, however, the meter reports 99% to 100% duty cycle which is obviously false for one injector.

I went ahead and interfaced to my microcontroller and wrote some code to analyse the signal and feed back to a laptop, however, from all points it appears to the microcontroller as if the pulse never stops.

I suppose I will think on this tonight and try again tomorrow. What I really need is a proper oscilloscope to analyse the signals going to the fuel injector, but they are effing expensive and my PC based scope just doesn't seem to be cutting it.

Russ.

Painless
06-29-2009, 10:26 PM
Well, it seems that I managed to blow up my number one fuel injector today, all by putting a multimeter on it. Looks like I will need to find another way of measuring fuel if this kind of tap is going to produce this kind of issue.

jankovig
07-06-2009, 01:55 PM
Today hasn't been a very good day, after spending the past days working on code to read pulse widths and translate them into frequency and duty cycle I hit a wall today when trying to interface with my truck.

The wires to the fuel injector are reported by my multimeter to be operating at a hz rate which is what I would expect in relation to RPM, however, the meter reports 99% to 100% duty cycle which is obviously false for one injector.

I went ahead and interfaced to my microcontroller and wrote some code to analyse the signal and feed back to a laptop, however, from all points it appears to the microcontroller as if the pulse never stops.

I suppose I will think on this tonight and try again tomorrow. What I really need is a proper oscilloscope to analyse the signals going to the fuel injector, but they are effing expensive and my PC based scope just doesn't seem to be cutting it.

Russ.

Hi new guy here. Electronics/Electrical Engineer by schooling/trade who recently got VERY interested in the whole HHO side of things and would like to offer help.
When using the multimeter to measure the PWM signal, all you're going to end up with is the DC component of the PWM. In other words just plain voltage. To properly evaluate the PWM signal, you need to use an oscilloscope. From my experience, the most you're going to see on the PWM frequency is 100-200 hertz. These days you can rent an oscilloscope. You wouldn't need much. 20MHz, 2-channel would be fine.

Painless
07-08-2009, 05:41 PM
Hi Jankovig,

Unfortunately, it was much worse than just my fuel injector. I replaced the injector only to find that the original was fine and am now looking at finding the money to buy a new ECU for my truck.

Not very happy at all. This project is definitely on hold until I get my truck sorted. It is barely drivable at the moment, just limited to 2500 rpm and lumpy / erratic.

Russ.

jankovig
07-08-2009, 11:23 PM
Here's a stupid question, but did you try resetting your ECU by disconnecting the battery? I know it's a long shot, but it might help.
Any time you want to do any type of voltage/freq. measurements that involve engine ECU, you want to use good quality high impedance probes. Low quality meters just don't have a high enough input impedance and end up loading the ECU output to the point where it can't source any more current it just burns out.

Painless
07-09-2009, 03:18 PM
That was the first thing I tried, I've even removed and opened up the ECU to see if there are any visually blown components.

I've since learnt the lesson of high impedance equipment as you can imagine, shutting the proverbial barn door after the horse has bolted, so the speak.

Russ.

Vulture
10-20-2009, 09:23 AM
Painless, you seem to be much farther along on an idea I've been working on as well. My whole purpose of computer control was for safety, if any one part failed the whole system shut off. I was planning on using an Arduino chip, I'll have to check out the one you're using.

As a possible solution to your injector pulse feed, have you thought about tapping straight into the ECM via the OBDII port? The amount of data you can read is amazing, more than those dinky <$200 scanners you buy at the parts stores. Depending on the vehicle, you should be able to read injector pulses, MAF, O2, EGT (if so equipped), timing, everything else you need. Depending on how gung-ho you get, and your programming skills, you could do all of the necessary data manipulation right in the ECM without the need for EFIEs and the like. The OBDII standard is publicly available.