evoredy

iTrader: (7)

Will do. I'll snap a picture of the sensor connector with a mic when I get back today.

I've seen knock off arduino-compatible atmels on a pcb on the shelf for 18 bucks (needs rs232 header and power). If the tuned circuit doesn't yield good results, I'll try a pic. At least I have a way to test (until my place catches fire lol)......

evoredy

iTrader: (7)

General connector/element pictures. I can measure if you need.

evoredy

iTrader: (7)

Initial results with Pump 93 and Pump E85:

All tests were done with the same voltage/resistance configuration as my previous post. Different end caps were used for each fuel as well as syringes and graduated cylinders. Fuel was set in jugs at room temperature of 70 degrees over night. The sensor was washed with a stream of 10ml of the incoming fuel before it was filled in a vertical position.

The sensor was brought back to control offset of 50HZ before and between fillups. It basically went to to exactly 50HZ when the fuel drained out. The microprocessor inside the sensor does it's job very well and there are no spikes, sawteeth, etc. Just a super smooth ramp up or down in HZ. Thats why I don't see the need for yet another programmed controller after this but whatevs. When you *think about sneezing, an analog scope pics stuff up (exaggeration) in real time and there was nothing.

Pump 93:

Edit: See attachments: pump93

2 ms/div
8 divisions
62.5 hz

Pump E85

Edit: See attachments: E85

1 ms/div
7.1 divisions
140.625 hz

*Expect a modest 1-5 hz error level. Sensor should be a great replacement for the pricier models. I'm still waiting for my stuff dammit.

EDIT: Added another test.

ROFL Everclear 190 is considered my control as it's e95 plus 5% water [/B]

Edit: See attachments: E95

1 ms/div
6.3 divisions
158 hz (i'm thinking 156.25 at 6.4/div though whatever lol it think it's pegged)
I have my control lol.

"Pump 93 (E10)" 20ml and "Everclear" 16ml

Edit: See attachments: E50

2 ms/div
4.8 divisions
104.166 hz

EDIT2: I did a 50/50 mix test (20ml "pump 93" and 16ml "everclear") and swirled it until completely uniform. Also, made a quick graph. The ramp from E10-E100 is eerily linear and EXACT to my findings! This would mean I do have E85 at the pump and E10 at the normal 93 pump. I'm pretty sure it's spot on. Note E0-E10 is not linear--may have to do with the offset compensation.

Edit: See attachments

STLEVOIX

iTrader: (17)

Looks like ya got e90 there. Good work!

evoredy

iTrader: (7)

I'm gonna go buy some everclear 190 and have a go (it's legal here in Texas lol)....at testing of course... to get a better control lol. It should be 95%.

STLEVOIX

iTrader: (17)

Cool, good stuff! I just got done "making" my harness for my setup. Should have it done soon. I will post up pics of my completed setup when its done.

ast

It seems to be at least very close to the old one, if not exactly the same. Easily in the usable range and it doesn't really matter with the analog converter that is not using the full 0-5 V range anyway.

How much ethanol 93 octane should have in the US? 5% or 10%? From the results it looks like 10.

What diameter is the inlet / outlet pipe in the sensor? I use Dorman Quick disconnect fuel line connectors with the old sensor which has two different size pipes, this seems to have both with same diameter.

evoredy

iTrader: (7)

10%. But the pumps say "up to 10%."

3/8 od on the in/outlets. A 3/8 fuel line slides on perfectly.

thebluesky

iTrader: (28)

Well hold on a sec there. While it says up to 10%, the only way to really know is to test the content. Unfortunately, it's possible to see or start to be seeing soon content as high as 15%. A simple test kit will tell you with some measure of accuracy what you're really dealing with.

evoredy

iTrader: (7)

I needed one "for sure," and I got it with the everclear (still holding myself back from taking a pull from the bottle). E100 is basically E95 with a 3-5% water content hence: everclear and a max reading.

I just needed a few points on the graph with sensor output in hz so I can tune my circuit. Besides, the everclear is melting everything.

As for accuracy, this test is probably as accurate if not more than the readout of a ZT or equiv. gauge even with reading hatchmarks on an analog scope.

With a 1000HP racecar on an agressive tune before track day, I might bust out the chemical test set.

VGergo

iTrader: (3)

it would be so good for Evo 9 too 88580714 EDM V7 and 88590715 USDM V7

99EclipseGSX

We've been using this sensor on standalones as a replacement to the older, more expensive and bulkier cast aluminum GM FF sensors. The Hz output and PW in regards to Ethanol % as well as fuel temperature seem to be the same. There was a few % discrepancy between the two at higher Ethanol contents, but the sensors DO have a rated accuracy (Zeitronix listed this as +/- 5% which seems excessive to me).

Beau

ast

Thank's for the confirmation!

I have seen readings from few older type sensors and I too feel that the +/-5% is excessive, in reality it seems to be less variation between sensors and the reading doesn't wander too much. That's really the important part when sensor is used as an active feedback to control mixture and other parameters. If it's 4% off nobody cares really as the car is tuned to compensate but it should be then always 4% off.

My own old sensor seems to be pretty much spot on, I have 5% with european 98E5 gasoline and I have seen maximum 84% with E85. Most of the time our E85 here is closer to 80-82%, but sometimes higher.

evo8426

iTrader: (27)

Fixed, now it would be good

This is exciting to see this actually happening.

99EclipseGSX

I'd love to try this out; can you please elaborate on the patching process? Jojodiff instructions are almost useless. Does one run the jptch.exe from command prompt? Once I run the exe file, it does not understand the command you provided on the first page. Also, there are no instruction as to where to place the old bin/hex files as well as the patch file you include in relation to the original jptch.exe file. I feel I have a passable understanding of software and using CMD, but this really isn't very intuitive!