Bluetooth connecting to an Innovate LC-1

This is the place for questions about things electrical and mechnanical...or any other automotive-related hardware issue.

Moderators: cgrey8, EDS50, Jon 94GT, 2Shaker

Post Reply
User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Tue Jul 17, 2018 4:49 pm

This topic was split from:
Bluetooth adapter for the Moates Quarterhorse
Enough information accrued that it really deserved to be its own topic. So I split off the posts from the above thread into this thread that related to Innovate Bluetooth connectivity.

_____________________________________________________

To follow up on this rather old post, I've been using Bluetooth for quite some time now so I've got some use time under my belt to share.

First off, the changes Clint made to BE work great. It simply performs "retries" when it gets back messages it doesn't like (i.e. NACKs).

The downside is Bluetooth is SLOW compared to being connected via USB. For those that remember using the TwEECer, write times are about the same. Write times were one of the "upgrades" the QH brought to the table. Bluetooth brings wireless, but sets us back some in that area. Although once the tune is written, datalogging seems to work just fine...and I datalog MOST of the payload items listed in GUFB.

Now one of the things I had high-hopes on was the Superlogger. I wanted to wire my Innovate into the Superlogger and use the same Bluetooth connection to datalog it via the Superlogger. But that doesn't seem to work. I've worked with Clint for quite some time just to realize, it's just not how the Superlogger works. The Superlogger WILL datalog the QH's payload items right along its own, but it wants to store its datalog locally. That works fine for people that drive, then look at logs later. But I want to see both QH and SL values in real-time...and I don't seem to be able to do that. So I'd been suffering through just havnig 1 USB connector connecting to the laptop, but I think I've found a solution to that...and it's FAR FAR cheaper than the Superlogger!!!

The heart of this is the Bluetooth HC-06 slave module, powered by a 12v-to-5v voltage converter, and the Innovate LC-1's RS-232 serial connection line-converted to TTL via a converter.
Bluetooth_HC-06.jpg
Bluetooth_HC-06.jpg (76.88 KiB) Viewed 1745 times
voltage_converter.jpg
voltage_converter.jpg (122.01 KiB) Viewed 1745 times
D9-pin_serial-to-TTL.jpg
D9-pin_serial-to-TTL.jpg (53.92 KiB) Viewed 1745 times

If you don't mind them being slow-boated from China, these will cost you less than $10 from eBay. If you buy from US sources, you can have them in less than a week for less than $20. Once they get delivered, I'll find out if they are a viable solution and report back my take. If anybody's having trouble finding these on eBay, let me know, and I'll post a link to the ones I bought.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by cgrey8 » Tue Aug 07, 2018 5:25 pm

I got all 3 of my parts from China. And I realized I got the wrong gender RS232 DB9 connector. So if you attempt this, make sure to get the male DB9 connector, not the female. I have a gender-changer so no biggie for me. But if you are buying new, be aware you can buy the correct gender adapter. You just have to pay attention...unlike I did.

I don't have them setup and installed yet, but I don't expect it'll be difficult. What I don't know is if this is going to work. I know BE is finacy about what USB-232 adapter you use. The IOGear's GUC232A works, but the Prolific adapters don't. I might get this setup powered and paired just to find this solution is like the Prolific adapters...a no-go.

Stay tuned. I'll know shortly.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by cgrey8 » Tue Aug 14, 2018 11:37 am

Most of this is lessons-learned and ramblings I want to get recorded somewhere before I forget them just in case I need them for myself later. And the hope is, that these details help someone else in the future which is why I'm talking about all this. That said...I fabri-cobbled these electronic lego-pieces together, tried to connect it up in the truck and FAIL. Just fail! And without more diagnostic ability, I had to set the truck back up the way it was and go to plan B. All my diagnostic tools were at the office, and this was going on over the weekend. So Plan B would have to wait until Monday.

Plan B was to get the RS232 and Bluetooth devices to my work where I have a USB-TTL device that I can confirm the two separately and eliminate one from being the problem. Both devices had a problem.

The simplest to figure out was the RS232 device's Tx and Rx pins are labeled backwards!!! That's a nice feature. But once I figured that out, it seems to work no problem. Optimistic at that finding, I was thinking that this was the culprit. But no.

Moving onto the Bluetooth device (an HC-06 Bluetooth slave module), I found the only baud I could speak to it was 9600. Well, the LC-1's serial connection is hard-coded to 19,200 baud. So then I got to thinking, maybe I just bought a cheap chinesium knock-off that simply wouldn't work at other bauds. So I took an extra Moates HC-05 Bluetooth module I had and started testing with it. I couldn't get it to work at all!!!

But I could at least get it to behave when I monitored it at 115,200. The behavior was very reminiscent of a baud-rate mismatch, which I've seen many times over the years. So I decided to test this theory. A diagnostic loopback test. Wire the module's Tx right back to it's Rx proved that whatever character I sent it from my PC (connected via Bluetooth) would receive the character I sent. This proved-out the Bluetooth pairing. It also deepened my suspicions that I was seeing a baud rate mismatch. At the very least, this eliminated some things even if it didn't get me all the way there. Doing this on both devices proved that whatever I was sending could be received on both ends. Now it was just a matter of getting them to see each other at the right speed.

So this got me to researching the HC-05 (master/slave capable) and HC-06 (slave only) modules a little closer. And this is when the Ah-Ha!-moment came. These modules require their local TTL-UART baud be set manually. Simply capturing the Bluetooth COM port at the desired baud rate does nothing to control the baud rate of the TTL-side of the Bluetooth module. I'd messed with the Moates HC-05 modules years ago when I first started dabbling with bluetooth on the Quarterhorse. So I was aware of the AT-commands that set them up. But I'd forgotten that baud was one of the things that has to be set via AT-commands. At the time, I wasn't interested in changing the baud, only the module's Bluetooth-presented name and pairing-PIN number.

What I didn't realize is that despite the HC-05 and HC-06 looking so similar, they are actually very different in non-obvious ways. About the only thing they have in common is their appearance and the fact that they are Bluetooth devices! Here's how they are different:
  • The AT commands used to configure them are different
  • The default AT-command mode configuration baud rates are different. HC-05 is hard-coded to 38,400 and the HC-06 wants 9600.
  • The HC-05 requires a pin-shorting procedure at powerup to put it in AT-command mode, where the HC-06 does not.
  • The HC-05 requires every AT-command be line-terminated with CR/LF characters. But the HC-06 doesn't want any line-termination AT ALL! As a result, the HC-06's command-characters have to be sent "fast." This poses a problem with terminal programs like Hyperterminal or GTKTerm which send the characters as you type them instead of letting you type the entire command and send it at once like other terminal programs (e.g. Termite). To get around this using these other terminal programs, the entire AT-command you are trying to send needs to be copied to the clipboard and paste-sent via the terminal program to minimize inter-character delay...or just use Termite.
  • There's mis-information out there about these issues on the Internet that ends up being VERY high on Google's search results of these issues thus leading you to do incorrect things. Thanks Google!


Anyway, after all that, I got my HC-06 device's TTL UART configured for 19,200 baud and confirmed that from my PC, connected via Bluetooth, to my work computer connected to the RS232-TTL device at 19,200 the way the LC-1 will be, I can send and receive characters on both ends. I haven't gotten this into the truck to test it, but I have a lot more confidence that this will work this time. It doesn't mean it will. It just means I'm closer.

As for why the Moates Bluetooth device also wasn't working, I found it's default baud rate is 921,600...WAY above what most PC-based serial com ports will support. When I reset it down to something lower, it also worked. However if I want it to ever with with a Moates device, it'll need to remain at 921,600 baud.

Another thing peripheral to all this that I stumbled across is that Prolific, the competitor to FTDI in USB-serail chips, has been a victim of Chinese knock-off chips nearly since their inception. It's a fair assumption that there are more knock-offs out there claiming to be genuine Prolific chips than there are genuine ones. To fight this, Prolific updated their Windows drivers to disable the clones and prevent them from working. As it turns out, I have 2 different devices that are affected by this (a USB-232 adapter and a super-cheap USB-TTL adapter). What makes it worse is these devices worked on my Win7 machine, but do not work on my Win10 because Win10, by default has the nasty habit of updating software without asking and without giving you the ability to disable this easily. There are ways to configure Win10 to do this, but they are not easy either. Anyway, if you have devices affected by this, there's a program out there that "fixes" this.
Prolific USB-to-Serial Fix (Official Solution to 'Code 10 error')

Hopefully, I'll get this electronic-menagerie installed and tested out again with a more-positive result this time.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by EDS50 » Wed Aug 15, 2018 11:43 am

Why run the lc1 through BT? You can hook up the analog out from the lc1 to the SL and log it on the BE dashboard. The BT connection is a little slower than usb or opto isolator but works fine with the SL. Only draw back so far is updating tune changes while logging. Simply connect Sl1 to QH and SL2 to BT dongle. The 4 pin wire combination os critical for proper function and communication. I would recommend you use the Moates BT dongle if you are not.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by cgrey8 » Wed Aug 15, 2018 1:06 pm

EDS50 wrote:
Wed Aug 15, 2018 11:43 am
Why run the lc1 through BT?
If it works, only because I already have the hardware to do it and I don't have to worry about translation error between the Innovate outputting voltage and the SL seeing a slightly different voltage and thus my recording the AFR differently than Innovate initially intended. If the values stay "digital" from the Innovate to the laptop, the value logged is as the Innovate controller intended, guaranteed.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by EDS50 » Wed Aug 15, 2018 3:34 pm

There is no discrepancy in output values. Since you have the SL just run analog out to the SL and use the same transfer function in the SL as you would use for QH/BE.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by cgrey8 » Wed Aug 15, 2018 7:03 pm

Well followup. It doesn't work. At least not with Innovate devices. It connects, it attempts to talk, but the Innovate driver is rather impatient and expects the COM ports to connect and immediately have the device available. But when the COM port represents a Bluetooth device, opening the port is when the PC pairs with the Bluetooth module which does not happen instantly. It takes a few seconds for the pairing to complete. By that time, the Innovate driver has deemed the device nonexistent. And there's nothing I can really do about this.

There are still possibilities, but beyond the scope of what I was wanting to do. The possible solution at this point would be to write my own COM port application that hosts its own COM port while capturing the Bluetooth COM port. With that, my application could connect to the Bluetooth COM, get that connection established. Once the Bluetooth connection is established, then I could be BE to connect to my app's COM port and play man-in-the-middle passing whatever's said from one to the other. Even that may have too much latency, but would certainly have less than the pairing time currently does.

That would be an interesting project, but if the SL can be made to work as I want it to, then I'm more motivated to try and get that to work.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by EDS50 » Wed Aug 15, 2018 7:41 pm

Im not sure how your connectivity between devices is but I assure you it all works on my end. Let me know if I need to go more in depth on how to set it all up.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by cgrey8 » Wed Aug 15, 2018 8:24 pm

EDS50 wrote:
Wed Aug 15, 2018 7:41 pm
Im not sure how your connectivity between devices is but I assure you it all works on my end. Let me know if I need to go more in depth on how to set it all up.
In my last post, I wasn't talking about using the Sl. I was talking about testing my fabri-cobbled Bluetooth solution to connect the Innovate's serial port to the computer via Bluetooth.

In the case of using the SL, the Innovate driver isn't used. The LC-1's 0-5v output wires to the SL and all coms are done with the SL so the Innovate driver never comes into the picture.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth adapter for the Moates Quarterhorse

Post by EDS50 » Wed Aug 15, 2018 8:46 pm

I understand now. Have you tried the moates bt dongle? I would scrap the serial connection and run 1 wire from one of the 2 analog outs from the lc1 to the sl. Then you can log it with BE.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Sat Sep 15, 2018 8:36 pm

I finally got this working! This afternoon, I literally datalogged both my Moates QH and my Innovate from my living room via Bluetooth.

But there were numerous non-obvious issues I had to work through that I couldn't have possibly foreseen without actually attempting this.

One of the most recent things I had to realize, and it's an old-school serial connectivity 101 issue that I simply forgot about, is that when you connect two RS232 DB9 serial devices together, you cannot simply use a gender-changer. In the serial DB9 days, you had master/client devices and slave/server devices. The master always had a male DB9 connector. The slave always had a female DB9. But even more importantly than this, their Tx/Rx connections are opposite. This is necessary so that the master's Tx goes to the slave's Rx on the same pin number. And vice-versa for the master's Rx.

Well this RS232-to-TTL device I have has a female connector which implicitly meant it was wired to be a slave...and so is the DB9 connection on the Innovate. So with a simple gender-changer, I was literally connecting both my adapter and Innovate's Tx's together and Rx's together. What I needed was a crossover gender changer. A standard DC99 cable used to serve that purpose for connecting two master's together in the old days. However a DC99 cable has 2 female connections. What I needed is the same cable but with male connections...something not generally done. So I had to make one with old serial DB9 parts laying around the office that were doomed to never be used again since our product has long since moved away from supporting serial devices like modems. So I commandeered them for this project.

For reference, here's a pinout of a DB9 connection:
D-sub 9 Connector Pinout
Image

Only 3 of the pins are actually needed, the Tx (pin 3), Rx (pin 2), and ground (pin 5). All those other pins were needed back in the modem-days where the modems needed to indicate whether they were dialing, whether dialing failed, when a connection was established, etc etc. But for directly connected serial devices, only the 3 pins are required. Also note those pinouts are for a master. But no matter, my crossover consists of two male DB9s with the Pin 2 of one wired to the Pin 3 of the other...ditto for the first's pin 3 to the second's pin 2. Pin 5 is joined for both. And that's it for the crossover.

The only other caveat I had to figure out is my RS232-to-TTL adapter's silk screen has the TTL pinout for Tx and Rx backwards. Fortunately I have a USB-TTL device I could test with and figure this out. So in my setup, I actually had to connect the Tx of the HC-06 bluetooth module to the Tx of the USB-TTL device...and ditto for the Rx pins. I've got a different RS232-TTL device being slow-boated over from China right now. I'm hoping it is both labelled correctly AND eliminates the need for the crossover adapter. That would certainly be nicer if it did...

While this seems like it should work, the Innovate's driver, as mentioned earlier, doesn't like "waiting" on the Bluetooth connection to establish. It times out before the connection is made. So to solve that problem, I made use of 2 solutions. First, there's an open source project called com0com which creates Virtual COM port pairs for testing and experimentation. Basically all it does is create two COM ports and whatever is said over one, is repeated to the other, and vice-versa. Although what it cannot (or does not) do is capture a "real" COM port, like my Innovate's Bluetooth COM port, and bridge it to a virtual COM port.

To solve that problem, I downloaded Visual Studio 2017 and wrote a rather simple COM relay program to do just that. To do a simple proof of concept, I hard-coded a bunch of stuff just to get the connection to work and proved the logic out. There were some learning opportunities there too I could elaborate on, but not really relevant to this discussion. Suffice to say, I got that working and then proceeded to make the program a bit more useful and configurable. That's where the VAST majority of my time on that program has been and I still don't have it setup to where it "remembers" your settings on reload. But it at least does default to the right baud rates for the Innovate. You just have to select the virtual and real COM ports each time you relaunch the thing.

If there's anybody out there interested in trying this for themselves, I'll be glad to write up some kind of step-by-step doc of what's necessary to get this to work. There are a number of non-obvious little things you need to be aware of. However I don't want to take that kind of time if nobody's using the LC-1 anymore or nobody has a need or desire to wirelessly connect to the Innovate LC-1.

BTW, I keep saying LC-1, but I believe this would work with any RS232 based device. It actually gets MUCH more complicated to attempt to Bluetooth expose USB devices.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by EDS50 » Tue Sep 18, 2018 8:40 am

Very interesting and inspiring work. Have you noticed any display delay between what the gauge is displaying vs what the BE dashboard is displaying? I know it is currently an issue with the mtxl wired through the evp port displaying values up to .6 difference between gauge and BE displays.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Tue Sep 18, 2018 10:57 am

My LC-1 is monitor-only...so I don't have a gauge to compare to. Also note, I haven't extensively tested this yet as I'm waiting for the new RS232 device to come in before I take the time to install it permanently. I only confirmed that I can successfully connect to it...again with the truck in the basement/garage and my laptop above it in the living room directly above the garage.

I'd like to come up with some kind of plastic enclosure to put it all in just so it's all held together and protected since it'll be at the passenger side kick panel...right where passenger's feet will be. If I can't come up with something, I'll wind up hot-glueing the pieces together and then zip-tie the whole thing under the dash close to the EEC somehow to keep it out of the way. Again, before I take the time to figure all that out, I wanted to see if the alternative hardware would reduce the physical size of this concoction by eliminating the crossover gender changer.

Once I get things settled, I'll be sure to report back details like that. Although based on my testing with just dumb serial data, the thing doesn't seem to introduce any noticeable delay or lag. And based on my experience with Bluetooth on the QH, I don't expect connectivity issues either. With the exception of the connection just being noticeably slower than USB, the QH wireless connection is solid!
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by EDS50 » Tue Sep 18, 2018 11:00 am

You don't have a gauge for your lc1, how did you monitor wideband afr before?
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Tue Sep 18, 2018 11:08 am

I've always datalogged it via BE...using the RS232 serial connection plugged up to a USB-RS232 adapter. This wireless setup will take the place of the USB-232 adapter.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by EDS50 » Tue Sep 18, 2018 11:15 am

Interesting, you would be surprised to see how far off the BE display is from an actual gauge.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Tue Sep 18, 2018 11:39 am

At one point, I was datalogging the WB via both serial AND a DataQ, which wires to the LC-1 via one of the Innovate's 0-5v analog output. And I noticed that the DataQ was able to log the values noticeably faster. So that doesn't surprise me. It's for this reason that I believe using the Superlogger will be superior. This was just something I was trying given the SL and BE just weren't able to do what I needed at the time.

I've got Clint's latest BE changes which is supposed to improve the SL logging as well as work with Win10. I just haven't spent time to test with them yet. It's on the to-do list.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

jsa
Tuning Addict
Posts: 519
Joined: Sat Nov 23, 2013 7:28 pm
Location: 'straya

Re: Bluetooth connecting to an Innovate LC-1

Post by jsa » Tue Sep 18, 2018 11:49 am

Double post
Last edited by jsa on Tue Sep 18, 2018 11:56 am, edited 1 time in total.
Cheers

John

95 Escort RS Cosworth - GHAJ0 / ANTI on a COSY box code
Moates QH & BE
ForDiag

jsa
Tuning Addict
Posts: 519
Joined: Sat Nov 23, 2013 7:28 pm
Location: 'straya

Re: Bluetooth connecting to an Innovate LC-1

Post by jsa » Tue Sep 18, 2018 11:49 am

EDS50,

A difference between the duel converted analog signal and displayed data value is par for course. Grounding, wiring and DAC/ADC errors all have an effect on accuracy. In the instrumentation world calibrating out the errors is normal practice.

The most accurate signal will be the serial data value. The egr lambda transfer function in BE should be adjusted so that output matches the serial data.
Last edited by jsa on Tue Sep 18, 2018 11:57 am, edited 1 time in total.
Cheers

John

95 Escort RS Cosworth - GHAJ0 / ANTI on a COSY box code
Moates QH & BE
ForDiag

jsa
Tuning Addict
Posts: 519
Joined: Sat Nov 23, 2013 7:28 pm
Location: 'straya

Re: Bluetooth connecting to an Innovate LC-1

Post by jsa » Tue Sep 18, 2018 11:55 am

Cgrey8,

Beyond TX/RX I'm interested to hear what other hurdles you encountered?
Cheers

John

95 Escort RS Cosworth - GHAJ0 / ANTI on a COSY box code
Moates QH & BE
ForDiag

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by EDS50 » Tue Sep 18, 2018 12:17 pm

jsa wrote:
Tue Sep 18, 2018 11:49 am
EDS50,

A difference between the duel converted analog signal and displayed data value is par for course. Grounding, wiring and DAC/ADC errors all have an effect on accuracy. In the instrumentation world calibrating out the errors is normal practice.

The most accurate signal will be the serial data value. The egr lambda transfer function in BE should be adjusted so that output matches the serial data.
There should be absolutely no difference in display values when logging the analog output from the gauge itself. The current issue which Clint and I have been working on is the sample rate and display speeds currently offered by Innovate. Innovate is aware of the issue but not too concerned with fixing it.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

jsa
Tuning Addict
Posts: 519
Joined: Sat Nov 23, 2013 7:28 pm
Location: 'straya

Re: Bluetooth connecting to an Innovate LC-1

Post by jsa » Tue Sep 18, 2018 12:35 pm

EDS50 wrote:
Tue Sep 18, 2018 12:17 pm

There should be absolutely no difference in display values when logging the analog output from the gauge itself. The current issue which Clint and I have been working on is the sample rate and display speeds currently offered by Innovate. Innovate is aware of the issue but not too concerned with fixing it.
So BE and the gauge display the same reading at or around a steady lambda value?
Cheers

John

95 Escort RS Cosworth - GHAJ0 / ANTI on a COSY box code
Moates QH & BE
ForDiag

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by EDS50 » Tue Sep 18, 2018 12:38 pm

No, BE displays a value anywhere between .3-to.8 afr leaner than what is displayed on the gauge.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Tue Sep 18, 2018 12:46 pm

jsa wrote:
Tue Sep 18, 2018 11:55 am
...Beyond TX/RX I'm interested to hear what other hurdles you encountered?...
Here's a brief list:
  • The RS232-TTL adapter's silkscreen has the Tx and Rx pins labelled backwards
  • I couldn't find a lot of info on the Internet about the Innovate protocol and what baud it runs at. I eventually found someone mention it in a forum and Clint confirmed my finding.
  • The HC-06 (or HC-05) Bluetooth module's TTL UART must be configured for 19.2kbaud/8/N/1
  • I forgot about the need for the crossover adapter.
  • The Innovate driver is impatient and won't wait for the COM port to be ready before giving up on it.
  • To work around the Innovate driver, I had to research ways to solve that Innovate driver issue (more on this below).
  • The com0com driver's defaults don't allow buffer overruns.
Interestingly unless you want to use LM Programmer to configure the LC-1, no transmit support is actually necessary to datalog the LC-1 over it's serial connection. Under normal conditions, the LC-1 is constantly chattering what the AFR is over the RS232 connection even if nothing is connected or listening which is an interesting strategy. That was a surprise. Also the data is binary, not ASCII based. So the characters that come across isn't human readable in a terminal. But because of this, the Innovate Windows driver is just opening the COM port and listening to whatever comes across when used with BE. The driver evidently assumes the COM port will open nearly immediately, which it usually does in the case of pure RS232 or USB-based COM ports. But that's not the case with COM ports that trigger a Bluetooth pairing between the PC and Bluetooth device. The best thing I could come up with is the idea of bridging COM ports so the Innovate driver would have a quick-open port that would feed it LC-1 traffic. So I researched how to create virtual COM ports and that's when I ran across the com0com open source project.

Initially, the com0com driver looked like it was perfect right out of the box being it creates virtual COM port pairs. But it kept locking up once I connected it to the LC-1 because of the constant chatter the LC-1 performs. That's when I learned about the constant chatter behavior of the LC-1 and once I saw that, I realized it was overrunning the buffers of com0com. And sure enough there are settings in the com0com interface to tolerate buffer-overruns which solved that problem.

Next, I thought I had a nice quick solution with a rather interesting port-forwarding feature in the Termite terminal program...if it didn't crash very shortly after capturing LC-1 traffic for some unknown reason. Even when not port-forwarding, it crashes. So because of that, I couldn't rely on Termite despite it's port forwarding being exactly what I needed.

This is what caused me to write my own C#.NET program to perform the Bluetooth port-forwarding. My C# program captures the actual Bluetooth COM port and thus suffers the delay it takes to connect while the Bluetooth pairing occurs. Once that is complete, it's relaying the LC-1 traffic to com0com. This allows me to configure BE to use the other com0com port which opens immediately.

And with all that, everything works! It's just far more convoluted than I'd hoped for or expected when I first considered this solution.
EDS50 wrote:
Tue Sep 18, 2018 12:17 pm
...There should be absolutely no difference in display values when logging the analog output from the gauge itself...
Theoretically, that's right. But in actuality, there will be error between the value the LC-1 outputs to it's Digital-to-Analog converter and the AFR gauge OR SL's analog-to-digital converter. If the LC-1 outputs 3.20v, but the gauge/SL registers this as 3.21v, then there's error. The question isn't whether there'll be error. The question is if the error will be negligible or significant enough to throw your interpreted value...it's just the nature of Digital-to-Analog-to-Digital conversions. To illustrate the point further, find 2-3 volt meters and read the voltage on a AA battery and see how different each meter's value is. I've got 2 VOMs at my desk and reading a 3v battery, one reads 3.01v, the other 2.99v. That's analog-to-digital error. For the purposes of reading the voltage on a battery, it's negligible. But if the voltage were being translated into some other value, a 0.02v difference could be enough to significantly alter the value. Over a 0-5v to 7.35-22 AFR conversion, it's probably still acceptable and negligible. Heck there's probably more error in the actual WBO2 sensor than that. But that said, I remember when I logged via the DataQ simultaneously with the LC-1, the error was noticeable. Note this is not error due to propagation delay.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

jsa
Tuning Addict
Posts: 519
Joined: Sat Nov 23, 2013 7:28 pm
Location: 'straya

Re: Bluetooth connecting to an Innovate LC-1

Post by jsa » Tue Sep 18, 2018 12:52 pm

Ok, then I have reached a different conclusion to you.

My reasoning is that a reading delay error would be cleared out in time, if actual lambda remained steady for a period.

IMHO If an error remains while actual lambda is steady for a sufficient period for BE to catch up, then the error must be calibration of the transfer function.
Cheers

John

95 Escort RS Cosworth - GHAJ0 / ANTI on a COSY box code
Moates QH & BE
ForDiag

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Tue Sep 18, 2018 12:59 pm

jsa wrote:
Tue Sep 18, 2018 12:52 pm
...IMHO If an error remains while actual lambda is steady for a sufficient period for BE to catch up, then the error must be calibration of the transfer function...
Conversion error vs Propagation delay error.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

jsa
Tuning Addict
Posts: 519
Joined: Sat Nov 23, 2013 7:28 pm
Location: 'straya

Re: Bluetooth connecting to an Innovate LC-1

Post by jsa » Tue Sep 18, 2018 1:04 pm

Cgrey8,

The protocol baud and behaviour detail is available in the freely available innovate protocol SDK. Sorry, info is after the fact.

The latest version of Termite has crashed on me twice while watching 32 channels of innovate data. Single channel has not presented an issue so far.
Cheers

John

95 Escort RS Cosworth - GHAJ0 / ANTI on a COSY box code
Moates QH & BE
ForDiag

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Tue Sep 18, 2018 1:08 pm

jsa wrote:
Tue Sep 18, 2018 1:04 pm
...The protocol baud and behaviour detail is available in the freely available innovate protocol SDK. Sorry, info is after the fact...
That's what Clint informed me of. The only thing I really needed was to know the baud rate so I wasn't google-searching for how to interpret the protocol as much as just what the baud rate was. But I searched enough to get my answer when I needed it even if it wasn't immediate.
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

User avatar
EDS50
Administrator
Posts: 3651
Joined: Sun Apr 05, 2009 9:17 am
Location: Tampa Bay, FL
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by EDS50 » Tue Sep 18, 2018 1:11 pm

The biggest issue/difference is the fact that the lc1 was capable of outputting instant display speeds via lm programmer and their xd16 gauge and was in sync with BE. If there was an error as cgrey8 mentioned it was so minimal that it was unnoticeable. The newer mtxl series through innovate and updated lm programmer firmware have discontinued use of instant speed display and only allow 1/3, 1/6 or 1/12 second display speed. After working on this issue Clint contacted Innovate to find out:

"The MTS network sends data every 83.33 ms. That is the fastest any innovate device can report on the MTS network. Even the LC1. Now the sampling from the O2 sensor and the electronic in the gauge is much faster. The speed at which the hardware samples are put on the network is what the user can define in the config of LM programmer. Hence the 1/12 speed being the fastest and that is .083 ms. Since most digital gauges only refresh at about 100 ms due to most human eyes can not see faster than that without flicker and so on. I have asked them several time is the hardware averages the point in between samples being placed on the MTS network. I was told NO it only place the last known value on the networks. That sucks I told them. It should average between sample. This helps get rid of noise."


Add in the slower responding BT connectivity and you can see why the BE display is up to .8 afr leaner than the digital gauge display. When I had the lc1/xd16 and connected via usb cable I didnt have the display issues but I did have a ton of connectivity issues due to emi. The BT connection has eliminated all of the emi issues I was having and now the AFR display is off but I would rather that than the connectivity. I understand there will always be an ever slight error between displays but should not be .8 afr error. One shouldnt have to manipulate the 5 volt transfer function to gain a fairly accurate display.

After looking into things further, the lc1/xd16 and lm1 had instant speed, mtx-l and dlg1 1/12 sec and mtxl-plus has instant speed available again. So it seems like those of us with the mtx-l are the ones stuck with the display offset issue. I also noticed the gauges have different display min/max settings available also. :confused: It might be time to either upgrade or revert back to my lc1/xd16.
1992 LX - 25.1c Chassis, Vortech Blown Dart 332, Lentech Trans, TRZ Backhalf, A9L, Moates QH/SL v1.9, BE, EA, TunerView

User avatar
cgrey8
Administrator
Posts: 10564
Joined: Fri Jun 24, 2005 5:54 am
Location: Acworth, Ga (Metro Atlanta)
Contact:

Re: Bluetooth connecting to an Innovate LC-1

Post by cgrey8 » Tue Oct 09, 2018 6:27 pm

I was perfectly fine with getting one from eBay since you can get them slow-boated over from China for about $1 including shipping. But any box I would've bought would have required some Dremel work. And I was perfectly ready to do that, but when a colleage a work told me he has a 3D printer and could print me anything I need, I took him up on the offer. Here's how it turned out.

Here's the side view of all the components:
ViewOf232-BT-PS.jpg
ViewOf232-BT-PS.jpg (124.56 KiB) Viewed 609 times
The component on the far back wall is the DC bucking power supply that drops 12v down to 5v which is what the other two components want.

He guessed at a width for the RS232 connection based on the width of a null modem cable's width. But it was a bit much. So once I got all the components hot-glued in place, he took caliper measurements and made the lid to fill the gaps.

Here's the box with the lid in place:
BoxWithLid_SideView.jpg
BoxWithLid_SideView.jpg (73.63 KiB) Viewed 609 times
...Always Somethin'

89 Ranger Supercab, 331 w/GT40p heads, ported Explorer lower, Crane Powermax 2020 cam, 1.6RRs, FMS Explorer (GT40p) headers, Slot Style MAF, aftermarket T5 'Z-Spec', 8.8" rear w/3.27s, Powertrax Locker, Innovate LC-1, GUFB, Moates QuarterHorse tuned using BE&EA

Member V8-Ranger.com

Post Reply

Who is online

Users browsing this forum: No registered users and 10 guests