Jeep Liberty Driving me Nuts!

[Squirrel]

I have a dry cell reactor
A tank with the three hose outputs and the cap mounted above the reactor
Big wires and relays
I've tried DC and using a PWM
Im using potassium hydroxide which can make the thing draw 45 amps if I want.
I have a MAP sensor "enhancer" installed (a potentiometer)

I get 17.4mpg from tank fulls, exactly what I got with none of this.
I believe in the technology (especially after flicking a lighter over the tank with the cap off) so besides nearly blowing myself up, what am I doing wrong? Any ideas?

Its as if the EICU has a minimum injector duty cycle dependant upon something other then the MAP....I dont get it.

[myoldyourgold]

Give us some more details. Year and engine specs. A description of your reactor like how many LPM your reactor is making at what amps, size etc. Pictures will be good too.

[lhazleton]

Its as if the EICU has a minimum injector duty cycle dependant upon something other then the MAP....I dont get it.

A MAP/MAF enhancer will help, but you haven't mentioned anything about altering the signals from the O2 sensors. Without an EFIE, the ECU will sense the additional oxygen & inject more gas.
As Carter mentioned, we need to know more about your vehicle & system.

[Squirrel]

Hey thanks for the quick responses guys!

Its a 3.7 liter 2003. Automatic 4wd, lifted with slightly larger tires (not important I'm sure).

Its a 7-element 5" dry cell from Ebay.
I dont know the LPM but it sends a steady stream of bubbles up the line to the tank.
I can get the current up to 45 amps but the wiring gets warm.
I ran 2 40 amp relays in parallel so the heat develops after that point (im considering bigger gauge if needed)
I have a PWM that I bought on Ebay where the groundplate of the reactor is seperated from ground on vibration isolators and the PWM completes the ground path.
I have the output of the tank injecting just on the engine side of the air filter (K&N replacement)
I have the MAP sensor installed and if I adjust it (up) the engine seems so hesitate for a second and then compensate quickly which fits with the statements about the O2 sensors.

So based on your replies I have these questions
1) EFIE? I've heard of these but a bit of a run-down would be good. (I thought the O2 feedback could only swing like 10% and the map overcame that but I guess this is wrong)
2) what can be done to the 02 sensors? (I think I have 4 of them)
3) Is there an easy way to measure LPM?

(love the democracy analogy)

[myoldyourgold]

First do you have a bubbler? If not do not use your system anymore until you add one!!!! You need a reservoir and then a bubbler. The reservoir can not double as a bubbler. Without a good bubbler you can do some damage that will cost $$$.

At 40 amps you are way over the limit for a 5x5. You need to keep that down to around 10 amps. .5 amps per square inch max of active area of one side of one plate. After subtracting the gasket area ports and dead area on the top you will have 20 square inches give or take a few.

Use the search feature and learn about bubblers, EFIE's, and how to measure the output etc. Lots of good info with pictures.

You will not need more than 1.75 LPM if that.

Before you go out and buy an EFIE make sure you know what kind the 2 front ones are. Narrow or Wide band. There are some very good EFIE's on the market but the good ones are not cheep. I would stay away from the cheap ones because they are a waste of money. The good digital EFIE's control all 4 O2's.

[Squirrel]

Sounds great, I was using a check-valve only so I'll take care of all those items and then gat back to you.


Thanks tons!
Squirrel

[lhazleton]

By "7 element" I'm sure you mean 7 plates (Damn Carter, ya didn't catch this?). This means it's (-NNNNN+). Adding another neutral plate would be much better.
Like Carter said, you can't run that reactor over 10 amps. It's active surface area is very small & you'll destroy the plates very quickly.
As he also mentioned, don't run without a good bubbler!!! I'll post a picture once again of what will happen very quickly to any aluminum parts if a good bubbler isn't used.

[myoldyourgold]

By "7 element" I'm sure you mean 7 plates (Damn Carter, ya didn't catch this?)

Lee I did not miss that at all. I have found that plates that are media blasted, ported right and with the right volume of electrolyte so the electrolyte coming in has a temperature that is just slightly less than the temperature of the reactor, helps to make it possible to run 5 neutrals 6 cells which will give you 2.3 volts per cell which is still under 2.5. There is a very fine balance here, to cold is bad. The duel tank reservoir which I have been using for a long time but Ben made public is the way to go. This will result in higher MMW's because the over all temperature drops and when you lower the electrolyte concentration and maintain the same amps or less and find you have the same or more production, this is good. This also adds to other evidence that lower concentrations of electrolyte are more efficient. You have to throw all of this out if you live in the frozen north during the winter though and just plan to get less efficiency. In fact I have tests that have been running at 2.7 volts per cell and have found some very interesting things when set up right. Now you are right as a rule but I think the rules are going to change as reactors will be forced to change to keep up. With less current leakage, with weldon 16 and other innovative things the reactors are getting more and more efficient in a flow through sealed series reactor. I know that this is debatable but tests seem to back it up. In my tests when temperature and concentration are brought down and the amps are slightly less but production is the same the MMW goes up but only within a certain range. There are limits which might change even more. In a normal reactor the effect is a lot less and the range is narrower.

[Squirrel]

Gah....I'm sorry.
Its 9 plates
-nnn+nnn-

That would put 3v across each junction right?
The surface area is a 4" diameter due to the o-ring size.

It turned winter here last night so I wont mess with it again until this weekend.

[myoldyourgold]

Gah....I'm sorry.
Its 9 plates
-nnn+nnn-

That would put 3v across each junction right?
The surface area is a 4" diameter due to the o-ring size.

It turned winter here last night so I wont mess with it again until this weekend.

Squirrel, that is 3.45 volts per cell/pair and you are just putting a lot of moisture and very little gas out of the reactor. Reconfigure it to -nnnnn+ or -nnnnnn+ which I think would be better in this case. This will at least give you a chance but this type of reactor with an O ring is not an efficient reactor. You need to post some pictures of it and maybe you can change it over to square gaskets if they are square plates and get more surface area involved. You should never exceed 2.5 volts in a reactor like this and better to keep it around 2 volts. The 8 plate configuration at 1.97 volts per cell/pair will be the best.