Page 1 of 2

Sequential transmission problem

Posted: Sat Mar 21, 2009 3:24 am
by Basek20a
Hi Everybody,

I am running a Sadev st75-14 transmission on my engine. With that i have a paddle shift system that makes it possible to shift up and down clutchless. The paddle shift system has a GCU, as soon as it changes gear 1 wire will give ground as long as the shift takes. After that the engine can pick up again.

Iév seen the seq options in the Hondata but it only gives me the choice of "exact times". Is there a possobility to cut as long as the ground in on the wire and after that recover?

Kind regards,

Bas van Melis
The Netherlands

Posted: Sat Mar 21, 2009 8:58 am
by Hondata
Use the full throttle shift input. Wire the transmission output to one of the clutch switch inputs (power steering pressure is probably the best). Keep the fuel & ignition retard activation rpm high, set the shift rpm limiter to 3000-5000 rpm.

Posted: Sun Mar 22, 2009 3:31 am
by Basek20a
If i understand it correctly i will need to clutch below 3000 rpm (as it won't cut the engine below that) And i have to inverse the input? So ground is "activator of the full thrttle shift" ?

Kind regards and thanks for your help!

Bas

Posted: Sun Mar 22, 2009 5:56 am
by Hondata
You should not need to invert the input when using the power steering input, as it is active when grounded.

You can use a lower shift rpm than 3000 rpm, but it must be high enough that the engine keeps running when you select first gear.

Posted: Sun Mar 22, 2009 9:07 am
by Basek20a
Is it possible to add a re activation delay with the full throttle shift? Are is this not necesary?

In the menu it says [gnd = off] so i should interpit it as GND = activates no GND = nothing?



Regards Bas

Posted: Sun Mar 22, 2009 11:23 am
by Hondata
Yes, GND = off so you'll need to invert the input.

Posted: Mon Mar 23, 2009 6:55 am
by Basek20a
Is it possible to get a set recovery time after a cut? So the shifts go a bit smoother? In other words can you add a gnd cut function in the seq shift menu??

Kind regards,

Bas

Posted: Mon Mar 23, 2009 8:44 am
by Hondata
The correct approach is to describe and document the problem you are having with the full throttle shift, and then we can make the decision as to what is the best way to help you.

Posted: Mon Mar 23, 2009 9:50 am
by Basek20a
The gearbox controller is an intelligent closed-loop system that only cuts the engine for as long as is necessary. This means that every shift will require a different cut time. The cut time must be determined dynamically by the GCU and not be for a fixed duration set in the ECU. The cut must be enabled at all engine speeds and throttle angles. There is no need to set a minimum rpm because the system 'knows' when it's selecting 1st gear and so doesn't send the cut signal. All the logic that decides when a cut is required or not is contained within the gearbox controller.

In summary the cut needs to work like this:

1. driver requests an upshift via the paddle
2. GCU pulls the cut signal to ground
3. GCU senses when the next gear is engaged and lets go of the signal (weak pull-up to 5v)
4. Engine ECU returns to full power over a period of approximately 40ms after the signal is released.

Ideally, the cut is not a complete turning off of the sparks, but a severe retard of approximately 60 degrees ATDC. This is enough to keep the engine turning, but not to produce any significant torque.

Posted: Mon Mar 23, 2009 3:44 pm
by Hondata
I did not realize that there needs to be a delay after the transmission signal has finished signaling that the gearshift has finished.

What you have described would require a new type of shift cut being added to KManager.

Posted: Tue Mar 24, 2009 2:20 am
by Basek20a
Point 4 is not absolutely necessary for this application, but the essential point is that the engine must remain cut for exactly the same time as the signal remains present.

Is it possible to add a new cut feature to the Hondata software?

Posted: Tue Mar 24, 2009 7:32 am
by Hondata
I believe that the existing full throttle shift will cut the engine output correctly for you. It will not delay after the signal goes high for 40 ms, but if that was vital you could wire up a simple RC network to provide the correct delay.

Posted: Tue Mar 24, 2009 9:02 am
by Basek20a
The 'delay' after the cut signal is removed is not as simple as you think - what we need (in the ideal world) is a progressive return to full power after the shift is complete. So, for example, if the cut was done by retarding the ignition to 60 ATDC, we would gradually advance the ignition by about 10 degrees per 4 milliseconds until we were back to full advance. Assuming full advance was about 30 degrees BTDC this would progressively return the power over a period of about 40ms. The reason we do this is to make the shift smoother and cut down the shock loadings in the transmission. If we simply cut the ignition dead and then suddenly return it to full power, the shift can be a bit agressive.

Posted: Wed Mar 25, 2009 7:48 am
by Hondata
A new type of shift cut would need to be created.

Posted: Wed Mar 25, 2009 10:56 am
by Basek20a
Could you consult your development team if this is possible?

Kind regards,

Bas