Serial Data speed vs. analog.

[TheMechanic]

Is there really that much difference in speed between the Serial data on a AEM's X-Series OBDII Wideband Gauge (PN 30-0334) and the analog data signal? Getting mixed info. Seems like it would save a little money with a USB to Serial 9 converter and breakout connector over the pro set update of $250. Want to keep the functionality of the CANBUS with the 0334. Heard somewhere that you could increase the update rate also but can't remember where I heard it.

[TheMechanic]

Ah. here we go. What do you think of his logic? https://youtu.be/mGcEw29XZBQ

[dr.mike]

I think I can speak on this


Serial data on the 30-0300 and 30-0334 was only added to maintain compatibility with the old 30-4100 and 30-4110, from dinosaur times. So, it is a bit crippled, vs. what it can actually do.

So, the stock serial output of the 30-0300 and 30-0334 is not great. It has 2 issues.

1. it is only 3 digits of data. i.e. the format is xx.x so... 14.7 14.8 15.1 9.6 8.8 , etc. it does not know anything between 14.7 and 14.8, etc.

2. Update/sampling speed. The default polling/sampling speed for the serial data is 10Hz ( 100ms ). This is 10x slower than the CAN data and 50x slower than the analog outputs.

I do have a FLASH firmware update that increases the resolution to 4 digits ( xx.xx so 14.75 14.76, etc ) and increases the sample rate to 25hz ( 40ms ) which is a more useful rate for working with fast changing signals like AFR.

In general digital data transfer is preferred to analog. It has greater noise immunity. And, it avoids losses ion accuracy, due to D/A and A/D processes. Data is transferred digit-for-digit ( just like copying a file ). AND, CAN transfers even have digital error correction, on top of that.

That said, the analog output of the 30-0300/310 and 30-0334 is VERY good. Since it is a differential output, it compensates for ground offsets, automatically. It also updates at 500Hz ( vs. 100Hz for CAN/OBDII ). So, it is preferred for looking at fast transition events.

As of the moment, I believe the fastest polling rate in HPTuners is 100Hz (10ms). But, you can use a scope to monitor high speed events on the analog line. This is great for balancing carburetors, etc.

Somewhere on this board, a user did an experiment where he hooked up both the CAN and analog outputs and recorded them to the same log. He posted the traces. And, they overlaid each other just about perfectly, within the range of the analog signal.

Some have reported difficulties in keeping the serial connection working properly for all brands of wideband. A bit more with the 30-0334 than the 30-0300. But, that may just be small sample size. With the re-FLASH results are reported as better. But, I still see reports of losing the serial data channel ( again, for ALL widebands; not just AEM ), when starting recording. I think it may have to do with cheap USB/SERIAL adapters.

[TheMechanic]

So Mike. What would be the best solution. I have the 0334. Still in the box. I will be doing vehicles with CAN and some without. I have DSO's as a side note because I do carb work also. Would I best suited to just bite the small bullet and go with the Pro Feature Set? Then go analog on non CAN? What about the issue that GoatRope said about the ability of the scanner? I plan on using Lambda so there are no fuel composition issues as I am here in SoCal and there are a lot of E85 stations here and I have run into some composition issues that threw me for a loop initially before I realized I had a vehicle that was approx 45%. I test on every single on I do now.

[dr.mike]

I think the Pro/prolink is worth it to have a consistent interface. If you do that, you can connect via CAN, over the prolink in all applications. BUT, you would need to have the gauge re-FLASHed back to 30-0300 firmware because the 30-0334 CAN interface "speaks" OBDII rather than AEMNet protocol, which is what the ProLink reads. Technically, at least, this is the best, overall solution.

Failing that, you could have the 30-0334 re-FLASHed with the serial data upgrade firmware. That increases the resolution from 3 to 4 digits and increases the polling/sample rate from 10Hz ( 100ms) to 25Hz ( 40ms ). That is fast enough for most work. Probably not quite fast enough for fast transitions like shifts and recovery from DFCO, etc. You could use OBDII, for 2007 and later vehicles; and, serial, for everything else.

For carbs, the analog output is the same for all x-series. 500Hz (2ms) update. Remember to ground the Analog- ( brown ) wire.

Set the scope to trigger off of the #1 spark. I use an inductive clamp that I cut off of an old broken timing light. Set the timebase to around 1 screen per 2 engine rotations. You SHOULD be able to see variations in AFR between cylinders up to about 2000RPM.

Note: if you have a waste-spark system, this will double-trigger, and make things look confusing. But, that is rare in a carb application.

[TheMechanic]

So I think I would like to stay with the serial data route as It is going to get moved a lot. What is involved in sampling rate updating? What does it run $$? What is your turn around time? Also as to the setup on the laptop. I posted the Goatrope YT video on setting up the serial data. Would it be the same process? Same USB/DB9 connector? Same Baud rate on the windows 10 setup? Ideally I just want something that is easy to transfer between vehicles, easy to transfer to laptop and my tower. Of course easy plug in is necessary and want the gauge functional also because I do have carb stuff that I would need to check WOT Lambda for jetting etc. If you really think the pro feature set is really the professional and update-able way to go in your opinion I will do that. Any downside to the pro feature set?

[dr.mike]

To update the serial data interface on your gauge. you need to send it to me to be re-FLASHed.

The same setup with the DB-9 is used. If you sned it to me, remind me to include a few connector pins; so that you can add the serial data wire to the harness, without hacking up the OBDII lines.


There is no charge for the re-FLASH. But, donations to dr.mike's beer fund are always appreciated