L6472 / L6474

[hof]

Posted on August 27, 2013 at 10:05

I am experiencing problem with the current regulation of the L6472. We made 2 series of boards, one was fitted with L6474s, the other one with L6472s. The board has 8 channels, supply voltage is 24V / 3.3V, timing parameters are reset default, internal oscillator. The motor windings have 1.7Ohm, 3.6mH. The board has 4 layers of 35�m including solid planes for 24V and GND. Everything is fine for the L6474. With the L6472 the phase current jumps wildly close to the zero crossing of the microstep sequence. All channels are affected, but usually only one phase of each channel. No difference between predictive and peak current regulation. I have several ideas what could be wrong:

1) Current measurement influenced by some ground shift effect 2) Supply voltage too noisy 3) Suboptimal timing parameters I am wondering where to look first. #l6472-current-regulation #l6472-l6474-current-regulation

[Enrico Poli]

Posted on September 09, 2013 at 16:17

Hello,

This seems to be an offset effect on the current sensing circuitry. You can try the following solutions: - slightly increase the target current, this way the sensing circuitry should work better - decrease the fast decay duration (the fast decay causes the system to go out of control) - decrease the slew rate By the way, what is your current control setup (fast decay, minimum on time, tsw, etc.)? Regards Enrico

[brett2]

Posted on September 26, 2014 at 00:10

I'm having a similar issue. We had problems with overheating of the IC, so I increased the slew rate and that helped. But now, we're seeing random over-currents and am wondering if dropping the fast decay duration alone will be enough. We have several motors in the 3 ohm/ 3mH range, My current parameters are what was recommended for the L6472 Eval board:

POW_SR = 00ton_min = 4ustoff_min = 21ustoff_fast = 10usfast_step = 16ustsw = 48usI tried dropping fast_step to 10us, and it seems to work ok, but the error is intermittent, and I'd like to have some margin. Any suggestions?Also, I wonder if you can explain why adjusting the SR or fast decay helps.