SGII Constant speed MPG

Just finished my first tank with the SGII to calibrate it. 46 mpg calculated, driving very easily. And now I am seeing much lower than expected mpg figures at constant speed on level ground.

My question is am I either A) getting crappy mileage at speed or B) is the SGII not all that accurate on the ALH or C) do I just need to run another tank through to get it calibrated. or D) A combination of the above


I am getting the following #'s
80 mph 27 mpg
75 mph 30 mpg
70 mph 33.5-34.5
65 mph 36 mpg
60 mph 39 mpg
and 55 (no return run, just on side roads) 43 mpg

My semi-scientific test consisted of a 5 mile stretch of (Seemingly very flat) interstate, 1 mile to get up to speed, start test @80 mph Cruise set for .8 miles watch the SGII MPG figures, decel to 75 for mile 2, decel to 70 for mile 3, decel to 65 for mile 4, and decel to 60 for mile 5. I reversed the procedure for the return run, so each speed took place on the equivalent piece of pavement on the return trip, all figures were essentially the same (either direction).
All figures are kind of un-scientifically scientifically averaged, but the margin of error is > 2 mpg

All speeds indicated; to get corrected speeds subtract 3~4 mph and you will be very close; in this speed range; in my car.

Clifs Got first full tank on SGII 46 mpg calculated, but was hyper miling extensively, now SGII is indicating poor MPG's that don't change very much with speed.

It takes a few calibrations to get it dialed in. Mine continues to bounce between 25% and 35% correction. It seems to vary depending on your driving. Pulling a trailer on the interstate for 500 miles will yield a different correction factor for that tank than my misc suburban drives to work. If you keep your driving consistant then the calibration will "home" in after a 3-4 tanks.

Awesome!
That is exactly what I needed to know.
I guess I can quit driving 80 now (just did 70 miles with cruise set at 80 :-( )
What do you think of doing a few 1/2 tanks to dial it in faster? or would the added uncertainty of gallons kill any potential benefits to this idea?

It becomes a point of diminishing returns the more gallons you fill with. Resolution on most modern fuel pumps is 0.001 gallons, the SG's resolution is 0.1 gallons. I would say that fills of at least 10 gallons should get overall accuracy in the ~1% range. If you only fill with 7 gallons for example your potential error would be ~1.5%. If you fill with 2 gallons it will be ~5%, and so on. You need to balance accuracy with patience kind of. I'd go with at least 1/2 tanks.

Put the SG on 'CURRENT' average and attempt to run the numbers up to a higher figure.

Reading the MPG 'live' is to jumpy and doesn't help as much.

Try closing the throttle all the way whenever possible (slight downhill..) and your MPGs will soar.

Bill

I do get the jist of it. I drove it un-calibrated as though it was calibrated. I made 10 days in a row with "indicated" 80 + MPG, 2 of which were 92.x mpg.
Of course my uncalibrated SGII said my tank average was 57.x and I got 46

However I must note that that is my daily route, on which I could reliably get 43 mpg in the Vibe (auto, 34 mpg EPA) or 32 MPG in the supra (19 mpg EPA + 295 wide tires), so I am somewhat versed in art of hypermiling, but certainly am no expert, this is basically my first experience with instant MPG #'s while driving, so it's just extra data for me.

Clarification on why I asked this question...
I got 48 driving it home and am willing to gamble that I had less than 30 miles on that tank under 70mph, so I was thinking that "at that time" I must have been getting 40+ at 70. Of course my logic might be totally shot to ****.

It seems that when you "calibrate" the SG at fillup it overcompensates the percentage too much. I have noticed (but have not paid enough attention to be 100% sure) that if for example I fuel up with 14 gal and SG had the useage at 13.5 gal (aprox 3.5% difference) the SG will compensate more than this percentage and report lower numbers until the next "calibration". Usually it gets my tank mileage within 1-2 mpg but it will report low or high mpg numbers after a fillup that differs from what it "recorded" you used. (yes, I do fill/vent to the top each fill) I am going to pay attention this fillup/calibratation and see for sure because last tank I was off more than usual after installing new Sprint 520 nozzles and used a bit more than SG recorded so this tank all the mpg numbers are lower than I'm actually getting. (overcompensating) Mileage is up a bit from my stock holes. In order for it not to swing too much optimisticly I plan on entering a number of gal. half way between what I actually put in and what it records and see how that goes. I am very happy with the SG as it is quite accurate and it is fun to compare how weather/wind/temp/fuel blend/other factors affects the mileage on my commutes. It sort of helps me compete against myself to get as good of mileage as I can each day while still having to stay within the same timeframe (speed average). Having said that I just realized how boring my life is!

LOL (how boring your life is) I've spent virtually all day fixing cold solder joints in door mechanisms, how is that for excitement?

That does sound likely that SGII overcompensates a little. I talked to a 240 buddy about it and he agreed with your hypothesis, he hasn't tested it either. And it would also explain why my mileage appears on the low side after calibrating it 20% lower.

I recently developed an interest in the SGII as well, read all I could about it but haven't pulled the trigger on it yet. Theoretically speaking, it seems to me that this small inaccuracy could be reduced (if not eliminated) by using the metric input system (liters instead of galons). Perhaps someone with an SG using the metric units could chime in on this, I'm curious to find out.

I've had the SGII for 4 or 5 mos now and have not been able to calibrate it in a way which makes sense. The readings I get depend heavily on the speed I'm driving (but the real MPG do not). I keep the reading on TANK for these trials

I've compared freeway driving with local as follows:
If I creep around locally using side roads so I can avoid stops and other traffic I can nurse the MPG up to 58, 59 or even the low 60's but when I fill up it's 51-52 MPG calculated. When I'm on the expressway doing 70 the MPG creeps down to 39-40. When I fill up it calculates to 49-50 MPG.

This boils down to a real 2-3 MPG difference but showing a nearly 20 MPG difference on the gauge. I have totally no explanation.

2002, 5 sp Golf

tgiaTDI said:

I recently developed an interest in the SGII as well, read all I could about it but haven't pulled the trigger on it yet. Theoretically speaking, it seems to me that this small inaccuracy could be reduced (if not eliminated) by using the metric input system (liters instead of galons). Perhaps someone with an SG using the metric units could chime in on this, I'm curious to find out.

Click to expand...

Two things affect the apparent MPGs on my scangage: Fill gallons if different from scan gage (SG) indicated gallons, and miles since last fillup. Miles are only adjustable on the SG through the 'speed' correction feature. If your 'tank' miles differ very much from the car instrument you need to adjust the speed readings. You can also find out how accurate your 'tank' readings by comparing with a hand held GPS if you have one [Or can borrow one].
The tank fill gallons is critical. you MUST fill to exactly the same point on the tank EACH TIME. The only way I have found to do this reliably is to vent the tank - or do a vent-ectomy - and fill TO THE BRIM with NO foam. Got to wait for the foam to settle. Ideally, the car should also be in the same spot at the same pump each time - ie "level".
Eventually the number of gallons on the SG and shown on the pump will agree quite closely.
I have PP520s and a mild Aligator tune and have gotten 51.8 + - .2 mpg consistently the last three fillups. By the way, the fillups do NOT have to be full 14 gals. 4-5 gallon fillups also work although small pump errors and rounding errors in the SG have a slightly larger effect.
With only a few miles on the 'current' setting it is remarkable to watch the mpg increase on a reasonably long downhill coast with no throttle applied. The TDI engines really do NOT use ANY fuel when coasting in gear with clutch engaged (rpm must, or course be above idle). I also see a SHARP drop in mpg whenever I have to use the brakes. If you can bleed speed by just backing off the throttle, do it. Using the brakes only throws kinetic energy away as heat, energy you have already 'paid for' by burning fuel to get up to speed. Get all the distance you can out of that energy by letting speed off without braking.

syrjohn: That sounds exactly like what I'm noticing. I wonder if anyone else is noticing the same?

I am going to fill tonite and will note the percentage of correction versus the actual fill difference (if any) percentage when calibrating.

jettawreck said:

I am going to fill tonite and will note the percentage of correction versus the actual fill difference (if any) percentage when calibrating.

Click to expand...

OK, here is what I noticed after/during refueling last night. SG had 14.7 gal used after 783 miles for 53.1 MPG. It actually works out a little different (53.3) but due to SG rounding to one decimal that's what it reported. The VW odometer had 770.3 miles and I filled 14.509 gal for 53.1 MPG. (exactly what the SG had for the tank) The odd thing is that even though it was right on the money for this tank, when I imput the 14.5 gal (down 2 clicks from 14.7) it adjusted +6.7%. Even if I had left it at 14.7 like it recorded it still showed +6%. I cann't fault it for accuracy but I don't understand the compensating percentage. The actual difference was only less than 1.5%, so now its reporting MPG numbers a bit inflated. It's a great little device with a lot of information that seems to be quite accurate, but it seems to swing from high to low when "calibrating".

The correction is a factor; the first fillup you do gives a number and each fillup after that gives only a slight difference as all the constants have been worked out.

DPM said:

The correction is a factor; the first fillup you do gives a number and each fillup after that gives only a slight difference as all the constants have been worked out.

Click to expand...

That is the way it is supposed to work, but this is after many, many fillups.

Fourdiesel said:

Two things affect the apparent MPGs on my scangage: Fill gallons if different from scan gage (SG) indicated gallons, and miles since last fillup. Miles are only adjustable on the SG through the 'speed' correction feature. If your 'tank' miles differ very much from the car instrument you need to adjust the speed readings. You can also find out how accurate your 'tank' readings by comparing with a hand held GPS if you have one [Or can borrow one].
The tank fill gallons is critical. you MUST fill to exactly the same point on the tank EACH TIME. The only way I have found to do this reliably is to vent the tank - or do a vent-ectomy - and fill TO THE BRIM with NO foam. Got to wait for the foam to settle. Ideally, the car should also be in the same spot at the same pump each time - ie "level".
Eventually the number of gallons on the SG and shown on the pump will agree quite closely.
I have PP520s and a mild Aligator tune and have gotten 51.8 + - .2 mpg consistently the last three fillups. By the way, the fillups do NOT have to be full 14 gals. 4-5 gallon fillups also work although small pump errors and rounding errors in the SG have a slightly larger effect...

Click to expand...

I don't have any problems with my SG. I had the SGI in my 2004 F250 Diesel and now the SGII in my Jetta TDI. I would suggest re-reading Fourdiesel's post (quoted above).

The largest difference I have aver seen between the SG gallons used and the number of gallons displayed on the pump is 0.3 gallons and that was on a 27 gallon fill up on the F250. Typically on my Jetta, I am never more than 0.1 gallons difference (over 12.7 gallons).

I calibrate after every fill up. The time where I see the biggest difference is if I drive differently from normal. For example, I usually drive 90% freeway, if for some reason the tank has mostly city miles on it, there will be a larger difference.

Also, the % is a fudge factor. It is not a raw % increase/decrease for mpg. For example the SGII says you used 12.1 gallons, the pump says you used 12.3 gallons. The delta is 0.2 gallons, so the fudge factor needs to be increased by 1.65% to compensate for the extra fuel. This 1.65% is added to the current fudge factor, so if it was at 5.5%, now it will be at 7.15%. You don't look at the SG calculated mpg, and your hand calculated mpg, and figure out the % fudge factor and then adjust to that. Just adjust the number of gallons as reported by the SG to match what the pump displays.

That 7.15% is a correction factor for total amount of fuel used. So the larger the fill up (in gallons) the better the resolution of the fudge factor.

If your speed on the SG is calibrated correctly, and you consistently fill up the tank to the same point of reference, your SG will be quite accurate.

As an update
I just filled up Correction went for -10% to +3%
I've only driven a few miles so far, so I don't know what the affect is thus far.

That's a MASSIVE swing. Did you vent both tanks? Spill some? Leave the car idling for an hour?

Vented both tanks (well all tanks thus far) ALL the way

Probably idled more on the second tank, but not much

The big difference was that the second tank had a lot more highway I drove it to Crusin on the coast ( http://www.cruisinthecoast.com/ ) about 150 miles each way at 75~80 mph.

Calculated mpg was 45.7 for the second tank

I'm now thinking that my SGII has issues.
The screen just went all black (each character block) I fiddled around with it for a few minutes (no result), then unplugged it from the OBDII connector. It appears to work now, but all of my trip data is gone.

Any ideas?

I'm considering sending it in for upgrade and hoping that resolves any issues.