I'm sure anyone reading this might think this was an awful lot of work to go though just to build what might seem like a simple little module.....and it was. But it was the only way this was ever going to get done and bench simulations are never the same as real in vehicle testing....

This WILL BE what the future of engine swaps looks like....

 

While I’m not a mother mopar fan, this looks like a great build. No one will expect it. You could have fun with a lot of people once it’s finished.

 

How did you get the factory gauges to work?

Thanks.

 

extremely impressive... great work esp on your conversion module

 

My 2nd to last post on page 2 has pictures of the Gm to Jeep conversion module I build for this. It reads the GM class 2 data and converts it the correct message format and scaling the Jeeps PCI bus uses.

 

I was rather shocked to see an interview I had done with 1320Video at Ls Fest regarding my Jeep was posted yesterday. I have edited their video so this is just the portion that pertains to my Liberty Swap, if would like to watch the original full length video I've posted a link in the description on Youtube.

 

Good job

 

Nice job! Needs a turbo

 

awesome work pete!

 

def needs a turbo and pete's special brewed closed loop boost controller

 

When I saw the video, and they showed a Liberty, I was like "Hey, I wonder if that's Pete." Sure enough...lol

Jay

 

When do you think you will start having a kit or built ones available pete

 

What sort of kit do you mean?

 

A diy kit or completed data bus converter for GM to Chrysler

 

OK, I thought you meant engine swap kit. Hopefully he will tag up here soon and give you a solid answer.

 

Liberate a liberty faceplated t56 while we are at it.

 

Due to the weather I ended up putting the Jeep into storage before I even finished prototyping a circuit board for the cruise control interface using the Jeeps steering wheel controls. I was also in the process of tracking down the auto door locks that are based on speed, the Jeeps BCM was enabling auto locking using the same set of messages that the cluster uses for the cruise status indicator. What was odd is I must have a set of split messages that have an odd interaction because it auto locks when the cruise is on but never unlocks regardless of what cruise control messages are sent.....that may not sound like much of an issue but it was a pain in the *** to deal with.

Probably the biggest thing that's kept me from doing much else with the hardware designs was that the main processor the module runs on(it uses 3 processors total) does not yet have bit locking enabled to prevent someone from reading the program back out of the processor. Now that may seem petty but I've dealt with that once and have no intentions of letting it happen again. There are other processors I could probably use that I'm able to lock right now but nothing nearly as powerful as what I've been using. I don't know exactly how much processing power it would take to keep everything working the way it is right now since the 4 that I've used throughout development are pretty hard to compare.

Using a 16mhz processor was beyond laggy unless you were only doing one or 2 gauges.

Using an 84mhz processor it greatly improved and I could drive all 4 gauges with minimal lag but warning lights were out of the question as they took to much overhead to deal with.

Overclocking the 84mhz processor to 120mhz took care of the gauge lag but adding in control of the warning lights started to make things lag again.

The processor I'm currently using runs at 600mhz and is able to processor data similar to how a dual core CPU works and is able to flat out smash though anything I throw at it with out ever skipping a beat.....and this is the processor that I'm not currently able to lock out from dumping the code off it.

So I would either need to find another processor that was capable of executing code similar to a dual core or I would need to find something significantly faster then a 120mhz chip to use.

Since both the Jeep and Ls computers are both VERY slow protocols by the time the conversion module has read data from the LS computer enough time has past that the data read is already on the edge of lagging so by the time that data is formatted and converted to work with the Jeeps data bus and then transmitted onto the Jeeps bus that by the time the Jeeps cluster/bcm have processed the data it's been long enough that you can clearly tell the data is no longer being displayed in real time. The only way I was able to "make up" for the lack of speed on both sides of the data bus was by throwing a very large processor in the conversion module at the situation. If either the Ls computer of the Jeep was using CAN this would be an whole different situation and processor speed wouldn't be nearly as crucial....and before anyone says something. NO, there is not any version of this that works CAN, I'm just pointing out the fact that these data buses are extremely slow and the reason this has been difficult with out running into any type of lag.

 

The shiftier on the T56 would end up WAY to far back to even think about using one.......I looked into manual trans options a while ago and there really are not any good candidates unless your going to spend several thousand dollars on an aftermarket transmission.

 

Excellent job Pete on everything. I appreciate your approach to trying to keep everything as budget friendly as possible and was wondering once you get this Data Conversion Module finished if you could do another for the PCM's that use a serial data request for A/C request on the I believe 03 to 07 PCM's. RPM Extreme sells one but it appears a bit price to me.

 

Please forgive me if this was answered all ready as I don’t have time now to read The whole thread yet, But what is the profession of the OPof this thread? You seam to haves at more than common knowledge of. The electronics of this swap. This will be a good read when I get home

great work I imagine a lot of BMWswaps could use this CAN conversion