Just wanted to share something I've been working on recently that some may find useful and even intriguing. As car and trucks have gotten newer the methods we've used for gauges in the past have become almost obsolete. Once you jump into the 4th gen stuff your getting into modules and such to convert the signals to something that an "older" style cluster can work with or if your doing a swap into another type of vehicle you may be looking at needing to replace the instrument cluster with an aftermarket gauge set just to get things working....unless your ready to drop some serous coin for high end conversion modules.


This work was prompted by a guy I know that was looking to put an LS into a 2010 Jeep Wrangler but he was torn on loosing his factory gauges unless he wanted to shell out a couple grand for a conversion kit that would let him retain the factory cluster. He was almost sold on this method until he noticed it REQUIRED a 4th gen LS running an E38 pcm. At that point he bailed and decided to go with a Hemi swap simply due to the cost of getting his gauges to work.

So I decided to see what the big deal was making the gauges work. Before you jump in and start talking smack about every thing being on the data bus bla bla bla hard to do any thing with.....I'm was aware what was needed and I've been working on the 3rd gen data bus for quite some time now and have become fairly familiar with the. The CAN network really isn't much different, things just happen a lot faster and the code structure changed.






So the idea was simple.....Read the bus on a pcm, convert that data into the appropriate format and feed that data back out to the cluster. The first issues to overcome was that the networks would be operating at different speeds based on the type of protocol the Pcm uses. The next thing I had to be able to do was spoof a Vin number to the cluster and pcm so they would talk to each other, but that's not as hard as it sounds. The final hurdle was finding a micro controller that had multiple UART bridges and could handle running two sets of data feeds and handle the conversion of the data in real time. Well after speeding the last couple days hammering out the timing and data conversion I have a algorithm that works. I can take any data message from the data bus regardless of protocol and alter the data into what every format I need. Now this hasn't been done on a cluster yet partly for safety reasons and also because I wasn't sure what cluster I wanted to test it out with to start. I was shocked when I called around at what some of these late model clusters cost....so it'll likely be tested initially with an 06 or 07 NBS cluster since they were the least expensive, but once this works on ANY cluster it's just a matter of sorting out the data from the host vehicles pcm and then the pcm from the instrument cluster of the donor vehicle. My experience with GM is the data used is pretty universal across the years and different vehicle types so it's not like it's unique to one specific vehicle, it would be more like it's specific to that pcm family.

Here's a screen clip where I was taking the serial data read from a pcm on the left; modifying that data and sending it back out but now at 1/2 the speed it was being read.







Updating this post with the link to the "Making Of" video.

 

Quite interesting, the wave of the future!

 

Hmmmm....

 

It took me decades just to figure out how to spell analog....

 

Can you get Holley EFI stuff to work with factory gauges?? I am your ginneu pig !!!

 

I've fiddled around a little with the arduino modules. They also have CAN data arduino "shields" you can add-on. Pretty interesting stuff. I didn't achieve any of my goals with it since it was over my head. Just letting you know the arduino stuff worked well though for getting CAN data if you are interested in the Gen IV stuff.

 

I'm currently working on a swap from a l in my late year 1999 Silverado to a new 6.2l. I've been told by Chevrolet they can supply an ECM that will run the engine and transmission but not the gages. I was planning on using the existing ECM just for the gages and split any input need for both ECM's. Would it work?

 

Very cool project. I will keep watching this for updates. Could solve a lot of problems with LS swaps.

 

The future is now! Best concept I've heard so far, since I did my LS Jeep swap 12 years ago

 

Possibly....PM me full details on your eng/trans/efi system and what type of cluster your using and I can take a look and see what would be involved for something like that and how closely it would be related to my current work.

 

The CAN shield + VPW shield combined are expensive and have a lot of unnecessary extra stuff added to them for other protocols but they were meant to be universal not cheap.

 

Currently there is no "cheap" way to do what you want unless you converted the 6.2 back to 3rd gen pcm. It would not be practical to interface the 3rd gen pcm to the 4th gen pcm since the 3rd gen cluster gets "most" of it's data off the bus network (as shown in the pic below)and is J1850VPW protocol while the 4th gen pcm is going to be CAN based protocol. With enough time and money it could be done the way your wanting to but it's going to be WAY more work then you might think would be involved.




Now you just happen to be in luck.....because what I'm working on is targeted for people doing exactly what you are. A finished version would be something like, Plug connector A into your DLC(or could be hardwired, would require 4 connections) the cable would run to a small box maybe 3x4 inches in size and that would have a plug that you would run up to your cluster and just plug it in and every thing would work.

Now in reality it'd be a long....long...long time until I was able to offer something that simple that that didn't cost an arm and a leg. What I will have up and running in the next month or two will do the exact same thing only larger in size and every thing will have to be hardwired to the cluster/pcm and I will be assembling the pcb's by hand and manually programming all the pic's on the board. At that point I need to get access to a vehicle running the E38 pcm for a good day or so to collect the needed data, analyse it and then verify every gauge is a 1 to 1 reading, that's going to take some time since the 4th gen stuff buffers it's sensors before sending the data to the cluster.......and no real car guy will enjoy or trust buffered data that they know it's true.

Once I'm at that point I'm going to be looking for a one or two people to alpha test the module before I moved the design to a beta build that would resemble the final product.

 

I'm in for a purchase whenever you get it ready!

 

So things are moving along rather well, I have mapped out the 03+ Silverado cluster data stream and can control every thing that runs off the bus from the pcm. It never occurred to me that the tach on the cluster is set as a 4 cylinder(the same way you change the tach output in the pcm from 4>8 cylinder when you do a swap). If I was going to be testing this on an LS pcm that wouldn't be an issue.....but I'll be testing this on a black box vortec pcm and I don't have the tuner cats definition for that pcm and I also don't have an EFI Live license so I can't change the setting in the PCM.

While thinking on that it also dawned on me that the Vortec 5.7 doesn't have oil pressure data on the bus either......

I had been saving this for another thread but since it's now relevant to this project.....I'll post a sneak peak of how I will be getting around the problems I just outlined




The salt packed is being used as a size reference. This was designed to remap an analog sensor's value into something else. Say you want to use an LS sending unit on a ford gauge.....this lets you remap a sensor so that the value read by a sensor corresponds to the value needed by a gauge to give the correct output. Here's an example situation of what this could do. Lets say that you want your LS oil pressure sending unit to read correctly on a 1934 Ford's oil pressure gauge but it were to read backwards(No clue if it does just an example). Well now I can make the sending unit's output value as you would measure in ohms correspond to what the gauge would want to see in the same manner. Now this isn't actually being done in ohms, this is being done digitally and the gauge is being driven with a PWM circuit.

Here's a crude drawing on how this might look on paper.





Now since the signal has already been read into the processor and is assigned a digital value I can also take that value and assign it as a data pid for an instrument cluster. So I can take my plain old SBC oil pressure sending unit and using this convert it to digital values that my "Cluster conversion box" can under stand and send the corresponding data to the cluster to control the oil pressure gauge.

But it get's even better.......I have the opposite of what I posted above as well (just not in on a printed circuit boards yet). It can take data from the bus and convert it into a signal that can run normal gauges. So in the example of the 4th gen, all gauge data can be taken off the network and converted into values that OEM analog gauges can use. Or if you really wanted to you could also have it run your aftermarket gauges as well......maybe for race application?


More to come near the end of next week once a few more of the pieces/parts I'm waiting on for this show up.

 

Well, as soon as you have it nailed down I have dollars burning a hole in my pocket for you!

 

Neat work you are doing. I bought Dakota digital product that pulls data from CAN and outputs to old analog gauges. Tach and speed for example. Pretty sure they have something for temp and maybe oil pressure but installing the right senders for those is easy.

How is what you are working on in the latest iteration different?
I’m a big fan of the notion that hit ridding includes computers and computer controlled engines transmission etc.

Thank you for posting

 

Take the data off the CAN network and convert that to J1850PWM to drive the cluster when swapping to newer electronics. The guy for example that wants to use the E38 to drive the cluster in his 99 Silverado. Most of the gauges in that are feed off the data bus.

This would also work to transplant a newer cluster into an older dash and have every thing still work. Take my 98 Tahoe for example, non of the gauges are driven off the data bus. So if I want to put a cluster out of an 05 Silverado into it most of the gauges won't work.

Now lets say I want to use a 2010 Silverado cluster, none of the gauges can be hardwired.....every thing runs off the data bus. I'll be able to take the data from my 98 Tahoe and feed the 2010 cluster with the CAN data it's looking for.

This also pertains to swaps. The most likely candidate would be Jeep swaps. Yes, you can buy a module that will interface an E38 to the Chrysler data bus....but at $2500 bucks it's not even something most people would think about. If you want to use a 3rd gen PCM your SOL because it only works Can data to Can data. Now if that could be done for say $250 bucks(Just throwing a number out) it just be came a LOT more affordable to most people. Then remove the 4th gen pcm requirement and the cost to swap it just got a whole lot cheaper.

The reason why this likely hasn't been done is it's EXTREMELY time consuming to gather the data with out paying for the OEM information......a one year sub to GM's top secret all inclusive data decoding and breakdown of every line of data on the bus will set you back 50 grand PER MODEL YEAR. So the only other option is to manually extract every line of data and figure out what it does to and what it means. Then designing the hardware to be as cost effective as possible. Sure I could build the hardware with every thing bought off the shelf and then wire half dozen different circuit boards together but it'd be extreme overkill and it wouldn't be cheap.....well it'd still be a whole lot cheaper then current modules like the one's offered for Jeep swaps but not cheap to the point that it's actually going to benefit anyone. The retail cost of most commercial product is accounting for the development cost, only a small portion of the product cost is what it actually costs to build. It's not just pure profit for the company it's just recouping the cost to build it in the first place. I'm doing the R&D in my free time as a "hobby" so there is no cost involved with the time it takes only the cost of materials I use in the process.

Now is there a market for something like this that's worth the cost for a company to design......not likely. But just because it's not something profitable doesn't mean it isn't worth building. There will be a handful of people this will benefit; and to them it's going to be the best thing since tuned port fuel injection.

 

Well, I'm that guy that would benefit greatly from this. I am totally committed to this swap and would be the first in line to purchase it!

 

What you are doing is great. I would think that most people on the forum would want a black box that would take the data and convert it to analog gauges considering that most of the swaps are in much older vehicles. There are a lot of guys that go for dakota digital because of the issue as opposed to using stock or aftermarket analog gauges. An added bonus would be if you can hook up a 12 volt signal to the can bus for the AC signal.

 

This would be very helpful to the SN95 Mustang guys too because most people when doing the LS swaps are keeping both computers, one for the engine management and one to run the gauges.