Had a look at Danfoss PDF you linked to.
Page 29 in the 2021. Pdf file.
I think I found they maybe only use about 30 bar pilot pressure. (you could measure pressure on A or B from orbitrol when turning to max)
If you use 6/2 valve you block the open to tank situation they require in neutral as written below.
On page 28 or 29 in the 2021.pdf file I found this, so perhaps (when autosteer is active) a small constant “leak” from A and B to tank can solve your problem.
The OSQ flow-amplifiers require special pilot steering units of the
type
• OSPBX LS
• OSPLX LS or
• OSPCX LS
which are all load-sensing steering units whose L- and R- connections
are open to tank in neutral position.
I also thought about it and made a test with a valve that in neutral position connects ports A and B to T. It did not work and the tractor continued to swing. But I think your idea of adding leaks to L and R may actually help, it cannot be ruled out that the splitter connected L and R to T with some delay.
I think a good solution would be a constant value subtracted from the main PWM fill. We would then obtain a delay in the operation of the second valve.
Yes. The more I think about it, I am convinced that @Larsvest’s idea is brilliant in its simplicity and must work. I will use regularized throttle valves and try to control via QSQA again.
It seems to me that controlling the current is not necessary. The hydraulic configuration that I have been using recently on conventional tractors works very well in a fairly large range of gain settings, so I think any small adjustments will be imperceptible in practice.
