This post is really about 2 things:
- Specifically, I believe the transmission oil weight called for in the Suzukiclone owner's manual is incorrect, and
- An experiment that starkly illustrates the magnitude of energy losses in a manual transmission comparing "thick" vs. "thin" oils.
If you don't have a Suzukiclone, just skip Part 1 and scroll down to Part 2 for the interesting viscosity comparison results.
Why I believe the Suzukiclone owner's manual recommended transmission oil weight is wrong
An acquaintance in British Columbia converted a 1987 Suzuki Forsa to electric drive, and was concerned that he was seeing a lot of energy loss in the drivetrain (based on the amount of energy required to spin the wheels on jack stands).
He mentioned to me that he was thinking of going to synthetic oil in the transaxle, and I told him I believe the specification in the owner's manual (calling for 75w90 gear oil) is incorrect.
I came to this conclusion after having changed the transmission oil in my own 2 Suzukiclones. The first one I changed to 75w90 synthetic soon after buying it. I noticed afterward (but wasn't certian) that it seemed to increase the amount of "grinding" of synchros in quick gear changes - a common complaint about the transmissions in these cars.
When I got Firefly #2 (the Blackfly) with just 2,000 km on the odometer, I also thought I should change its transmission oil, because it sat parked for 7 years. But when I did this, I was careful to pay much closer attention to the viscosity of the fluid I drained from
it - and I saw that it was obviously MUCH thinner than the 75w90 oil called for in the owner's manual. Note this was unquestionably the factory original
tranny oil that was drained from the car.
A bit of investigation on teamswift.net uncovered several people recommending GM's AC Delco Synchromesh brand of semi-synthetic manual transmission oil (it doesn't have a viscosity/weight listed on the packaging). I bought some, but before pouring it in the car, I did a few tests to compare the various viscositis of the oils I had:
- I did a simple timed "pour test" of a measured amount through a cotton filter, and saw that the AC Delco Synchromesh lube was actually closest to 5w30 engine oil in viscosity (based on the time to drain a measured amount through the filter).
- I also pour tested the factory original oil that I drained from the Blackfly (Firefly #2). It also poured through the filter at the same rate as the Synchromesh oil and 5w30 engine oil. Hmm!
This told me that the manual-recommended 75w90 synth oil I put in Firefly #1 was too heavy. Fortunately, I owned both cars simultaneously for a couple of weeks after getting the Blackfly (before selling #1). So I was able to do another test:
- my next test was to drain the 75w90 synth oil from Firefly #1 and replace it with the "old", thinner, factory original oil I drained from Firefly #2.
- After doing this, the amount of "synchro grinding" immediately reduced.
So I returned the 75w90 synthetic gear oil I had bought for Firefly #2 to the store, and instead added the GM Synchromesh oil to that car. I have had almost no trouble with its synchros since then. I also "moved" that fluid from the original transmission to the used "taller" transmission I installed later on. Still little synchro trouble to speak of, even though the used transmission has 160,000 more km on it than the car I put it in.
From all of this, I came to the conclusion that the 75w90 recommendation in the owners manual is incorrect. It's too heavy.
An experiment comparing energy losses from "thick" vs. "thin" transmission oils
Written by Roger - electric Forsa owner.
Apr 12, 2007
RE: Sprint/Metro Drivetrain losses
I had written:
> The tranny is full of fresh 75W90 gear oil, per the manual,
> however, I've been advised that the manual is wrong and a GM
> semi-synthetic lube is what should be in there. I'm
> skeptical that the lube alone could be responsible for this
> sort of loss, but since an oil change is easier than
> rebuilding the tranny, I'll try that first ;^>
As it turns out, it seems (from checking a partial bottle remaining in
the garage) that the tranny is actually full of 80W90, not 75W90 as I
I put the car on stands again (under the frame with the suspension at
full droop since the suspension angle didn't seem to affect energy
consumption in my earlier tests and I feel better with the stands under
the frame when I'm going to be getting under the car).
All following energy consumption observations are based on battery pack
voltage and current as reported by my E-Meter.
I spun the wheels in 2nd for a few minutes to warm things up first, then
measured energy use in each gear at 40kph (except 1st, which was
measured at 30kph):
5th 23.4A @ 123.5V (116.3Wh/mi or 72.2Wh/km)
4th 22.9A @ 123.0V (113.4Wh/mi or 70.4Wh/km)
3rd 25.5A @ 123.0V (126.2Wh/mi or 78.4Wh/km)
2nd 30.8A @ 121.5V (150.6Wh/mi or 93.6Wh/km)
1st 37.5A @ 121.5V (244.5Wh/mi or 151.9Wh/km)
After draining the old fluid, I poured in a half litre or so of Varsol
and spun the wheels for a minute or two to flush things out:
2nd 22.5A @ 123.0V (111.4Wh/mi or 69.2Wh/km)
This was then drained, and the tranny filled with 2.5 litres of the
recommended AC Delco Synchromesh fluid (p/n 89021808, IIRC):
5th 16.1A @ 123.0V (42kph, 75.9Wh/mi or 47.2Wh/km) (34.7% lower
4th 17.0A @ 122.5V (42kph, 79.8Wh/mi or 49.6Wh/km) (29.6% lower
3rd 19.7A @ 120.0V (42.5-43kph, 89.6Wh/mi or 55.6Wh/km) (29.0% lower
2nd 22.4A @ 120.0V (39.5-40kph, 108.2Wh/mi or 67.2Wh/km) (28.9% lower
1st 37.7A @ 118.0V (39kph, 183.7Wh/mi or 114.1Wh/km) (24.9% lower
So, going to the [GM Synchrmesh] fluid definitely seems like a step in the
right direction, however, not nearly as large a step as is required
On the plus side, the 3rd-2nd downshift "crunch" is gone, which had been
one of the touted benefits of using this tranny fluid.
So, some progress, but the hunt for better efficiency continues...
Just a note:
the energy savings seen from the less viscous gear oil doesn't translate directly to an equal decrease in total fuel consumption (or electricity use, in this case). Only the energy difference required to spin the wheels on jack stands was measured - not the total difference to actually drive the vehicle (with all its other associated losses factored in).