722.6 build - Page 4 - PeachParts Mercedes-Benz Forum

whipplem104 · #46 11-30-2015, 04:06 PM

You could use map voltage to skew the shift tables. It adds a layer of complexity but would give you the best of both worlds. Add an additional input to the map sensor signal voltage. Set this up as shift trim vs voltage. You can then skew the whole shift table by the mph amount you want vs the voltage. You can also do just downshifts or upshifts.
Also you can use your WOT table to help in this area. Set your WOT value to a value that you would be at while pulling a mountain pass. Around say 70% and then you would just go fixed to this table above that point. Really the thing is that you should probably just pull the shifter into 3rd for the pass. It is hard to get all the dynamics of all conditions.
Just a few ideas.

duxthe1 · #47 11-30-2015, 08:49 PM

You've given me a lot to ponder. I looked at the analog input options and there is a lot of stuff buried in there. I hadn't much thought about the analog inputs once I got TPS and MAP working. I still don't even have ECT hooked up yet.

I had hoped that the WOT tables would handle the mountain passes but I found it would go into the WOT table and downshift to 4th and hold it. To get a downshift to 3rd I had to let up and get out of the WOT table and floor it again to get the WOT 3rd gear downshift. A 104.980 at 9-10K ft altitude with 3 adult passengers and loaded to the hilt can't afford to let off the throttle and hope to hold speed.

Reverting to TPS shift tables I set up the 3rd gear downshift to occur at 75MPH at 100% throttle. This got her back over the passes pretty well. I think I'm going to try to work on getting the TPS shift tables really dialed in around town and then maybe giving the shift trim v/s voltage a try. I was pretty happy with TPS upshifts but the downshifts based on the same tables seemed a bit off.

whipplem104 · #48 11-30-2015, 10:27 PM

In the WOT setting it is purely a mph lock out on the downshifts. So you set what the mph is that it will not shift above. So say you are in 4th gear and it is set to 45mph. If you are above 45 mph it will not down shift. The fact you are having to let off means this number is set lower than your regular tables.
The rpm setting is for just upshifts. Just be careful not to set the mph to high as to cause to high rpm on the downshift. You can use your shift tables to see what rpm you would be at any given rpm or gear. Or your gear ratios table in transmission setup.

duxthe1 · #49 11-30-2015, 10:42 PM

Ok thanks for clarifying. I was a bit confused about the WOT settings description in the manual. The learning curve is pretty steep on the first one. By the time I talk my buddy into swapping a .6 into his wagon I'll be a pro.

whipplem104 · #50 11-30-2015, 11:00 PM

The manual can be vague. Also remember that in order to make the tcm something people can use it has to be limited. Honestly the tcm-2000 is about as complicated as you can make something like this for the general public. The shift tables are very limiting but to make them correct you would need acceleration rates and a dozen other things to calculate skews. Add a brake pressure sensor for downshifting under heavy breaking. Your tps/map blend and then remember which one is active when adjusting a shift. But you can do it if you want to spend the time.
The new tcm that I am working on is just that much more complicated and it is killing me. There are so many settings and different tables to adjust stuff. We have not added temp comps yet which should be pretty simple. But you really have to think for a while before adjusting. I am going to have to limit what people have access to.

duxthe1 · #51 12-01-2015, 01:00 AM

I'll constantly tinker with it, for sure. Heck in a few days I had it acting near enough to stock that the untrained eye would think its always been. However to me after 20 years working on these things I'm thinking "that shift was a smidge late and a little hard for its load point" I'd like to take a day and put my buddy in the driver's seat while I tune it. So far my tuning efforts have been mostly based off daily driving it, certainly not the most effective approach. Drive it to work and note what shift bothered me the most. Maybe check a datalog, make a small change, and load it for the drive to lunch. Rinse and repeat.

vannatta20 · #52 12-15-2015, 12:28 AM

Awesome thread! I learned alot and hope to try this swap sometime when my trans takes a dump.

duxthe1 · #53 01-19-2016, 08:57 PM

Been a while since an update to this thread. I've been working on nailing down a great tune for the trans. I've got it where I've been really happy with its shifting but there has been a nagging problem since I installed it. My EZL ignition module gets pissed off that there is no transmission overload protection signal coming in. This manifests itself as severe hunting after a high RPM upshift. If I wound out first to 7K it would shift into second and buck violently until letting off the gas. This is just the normal behavior for the EZL when the overload protection signal is missing.

The EZL uses this signal to cut timing to 5 degrees btdc during upshifts over 4K RPM. Pretty cool to have a OE engineered shift cut back in '91. A switch on the side of the original transmission closed when the B2 band was applied. This signaled for the shift cut in 1st and 3rd. Some transmissions also had a second switch that closed with brake B1 that signaled for shift cut in 2nd.

The EZL holds 5 volts on the signal line and when the first switch closes it pulls the signal line down to 2.5 volts. On cars equipped with the second switch the signal line is pulled to 1.67 volts when the switch is active.

Needing to duplicate this function I was able to use the shift cut feature in the TCU. It has programmable PWM outputs that can be configured for ground instead of 12Volts. Used this way the PWM'ed ground signal can simulate the O/L protection signal. At about 26% PWM the 5 volt signal is pulled down to 2.5 volts and at 47% PWM the signal is pulled to 1.67 volts.

While having an active shift cut is pretty cool the real impetus was fixing the bucking. I may be able to further fine tune the shifting using the shift cut. The EZL cuts timing for about 400 milliseconds on upshifts and 50 ms on downshifts. This doesn't line up as well as I would like because it is triggered with the shift command and it takes a little bit of time for the hydraulics to catch up. The original trans triggered the function with the hydraulics as it was actually shifting. I've only got a few miles datalogged but I'm going to keep playing with it and see how well I can use the feature. In the future when I go to a standalone fuel and ignition I won't be handicapped by the 400 ms window.

Stay tuned....

whipplem104 · #54 01-19-2016, 10:30 PM

Cool,
I have been asked about this in the past and just never really had the right car to do testing on. To set up a delay you can use a PWM shift solenoid instead of the shift cut feature. It takes a little setting up but you can do some really cool stuff using this output function. You just have to set it up in the advanced solenoid section to when you want it active. You can delay and over run if you need to.

duxthe1 · #55 01-19-2016, 10:50 PM

Hmmmmm The shift solenoid idea is an intriguing one. Right now I really like the table setup in the shift cut function but wish there was a programmable delay instead of rigidly synching with the shift command. There's a couple tenth's to be had there if it were a scenario where tenths mattered. Maybe something to throw at the PCS guys... jus sayin'

Right now I'm using two different pwm's. One for 1,3 and another for 2. This is to emulate how MB did it originally. I really doubt the EZL cares. My car originally didn't have the second switch, though the EZL did recognize it's request on the 2-3 upshift. If I'm to try setting up a solenoid output I'll have to see how the EZL handles an B2 shift cut request on what it would think was a B1 upshift.

I'll probably play with it a bit as is. As I get some miles and datalogs I'll see if it needs more complex control than shift cut currently gives it. As it stands the shifts that could improve with the cut is 1-2 and 3-4. My 2-3 is always pretty smooth. That said I'd like to get it in all gears. With just the little driving so far it did get a bit more of an "OE" feel that I liked with the shift cut. One thing to note is that the EZL doesn't cut timing at low RPM. I may be chasing my tail trying to dial in shift cut with the EZL in charge. Maybe I should be content to fix the bucking for now.

whipplem104 · #56 01-20-2016, 09:12 AM

The shift cut is a nice feature because you can do different values for each gear. The shift solenoid setup would only be one output value but you can select what gears it is active in.
And you can time it however you like in relation to the shift.
One thing to consider though is that reducing torque will also shorten a shift. On my setup with a standalone I run .2s on timing pull. To really do something like the factory does it requires a full table based on input tq to run reduction. You want varying amounts of tq reduction for you specific load and gear.
The timing is actually pretty interesting as well. The peak reduction is of course at the beginning of the shift and then as the shift completes power is ramped back up to no reduction. There is a delay as well but there are pre shift fills happening. So there is a few .1s of stuff happening before the actual shift begins.
Anything below 100nm of tq there is no reduction.

duxthe1 · #57 01-20-2016, 10:28 PM

After Whip put the bug in my ear I figured I'd go ahead and look at setting up a PWM shift solenoid for the shift cut. I'm not gonna lie it was a PITA to get a simple on/off with defined times for delay and active on. After about an hour I had numbers for percentages and slopes that resembled what I wanted to do that the tuning program would accept. I set this up on shift solenoid F and set it to activate on 1st, 2nd, and 3rd.

After datalogging my drive to work, lunch, and home I landed on a 100ms delay before the shift cut. This has the shift cut coming on right in the transition between gears. Too early and you can see the RPM drop before the shift and too late you can see a flat spot after the shift completes. When its right on you can't really see it in the datalogs because it happens as the RPM is falling between the gears.

One thing I do notice in the datalogs is that the duration of the shift cut on time is sometimes longer than defined. It doesn't affect how it works because the EZL is looking for the trigger not the length. I just figure there is some stuff going on behind the scenes with the shift solenoids.

It works pretty nice without any additional tuning. As I get more miles on it I may tweak the tune if it feels like there is any improvement to be had. As is there is a neat guttural growl from the exhaust as the timing comes back after a high load upshift. Kinda makes me smile inside every time.

whipplem104 · #58 01-21-2016, 08:59 AM

Great job. It can take a while to understand what the settings are supposed to be in relation to each other. It just keeps yelling at you if you do it wrong though.
The on time is dictated by shift duration. So you can control the ramp up and the ramp down. But the actual shift time will have an effect on all of it as well. I am talking about how long a shift takes, not any set timers.
Shoot me a datalog sometime. I would love to see all of this in action.

m1tch · #59 01-22-2016, 03:14 AM

Have you installed the Olefejer 722.6 gearbox controller?

whipplem104 · #60 01-22-2016, 07:18 PM

He has the PCS tcm-2000.