Use all EPWMs on TI Launchpad F28377s

Stefano,

Perhaps you can depopulate R36, R37, R40, and R41 from the Launchpad, and tap the ePWM signals at the resistor pads.

Regards,

David

David,

Thank you for your reply. I could indeed short R and unsolder C, that would theoretically be correct althoug kind of a hack.

What concerns me is that in the PWM mux configuration (Mux = 1) it is the EQEP1A channel for GPIO20 instead of DAC1 selected. And I am a bit lost with what is inputs and what is outputs on this board ... Cannot i select the mux for each single signal?

David,

Thank you for your answer. I though the Launchpad expressively limited the number of mux to 0,1,2,3. So you're basically saying that the Launchpad is not a limited version of the TMS320F2837xS and that it's just that the Launchpad datasheet is not complete?

Regards,
Stefano.

Stefano,

The Launchpad has regular F28377S silicon on it. Where are you seeing that the Launchpad limits the mux configurations to 0,1,2,3 options?

- David

David,

From the datasheet of the Launchpad: www.ti.com/.../sprui25a.pdf (Table 1,2,3,4 of pages 8-9).

Anyway you replied my question perfectly. I'm looking forward to experiment with the launchpad capabilities ;).

Have a nice day!
Stefano

HI,

where do you see that all PWMs can be routed to pins?

From what I can tell there is a typo and GPIO2,3 are EPWM2A,B, so EPWM1 is not routed to pins.

So you've got EPWM2,6 and 10 on J8 and you've got EPWM 7,8,9 and 11 on J4

And EPWM 11 and 10 are routed twice. Once directly and once through RC circuits David has mentioned

All in all you can get 7 complementary pairs, which is great, but could be better as there are couple of empty pins left.

And routing could be better as a lot of existing code uses EPWM1,2,3

stefano65 said:

 
From the datasheet of the Launchpad: www.ti.com/.../sprui25a.pdf (Table 1,2,3,4 of pages 8-9).

Yes, those tables list only mux options 0,1,2,3.  But it also says in the text of section 6.1 "Table 1 through Table 4 lists the pin out and pin mux options for the C2000 LaunchPad. Additional muxing options are available and can be found in the TMS320F2837xS Delfino Microcontrollers Data Manual (SPRS881)."  So there you have it.  The tables are not showing all options, and that is stated.

Regards,

David

Mitja,

Mitja Nemec said:

HI,

where do you see that all PWMs can be routed to pins?

From what I can tell there is a typo and GPIO2,3 are EPWM2A,B, so EPWM1 is not routed to pins.

So you've got EPWM2,6 and 10 on J8 and you've got EPWM 7,8,9 and 11 on J4

And EPWM 11 and 10 are routed twice. Once directly and once through RC circuits David has mentioned

All in all you can get 7 complementary pairs, which is great, but could be better as there are couple of empty pins left.

And routing could be better as a lot of existing code uses EPWM1,2,3

Actually, I realized just now that the Launchpad has the PZ 100-pin QFP package of the F28377S on it.  This package does NOT have all the signals available.  For example, GPIO0 and GPIO1 are not available (meaning, ePWM1A and ePWM1B are not available).  So, soldering wires, as I suggested earlier, won't help.

The reason this package was used was cost.  This is the lowest cost package.  The Launchpad is a low cost development board.  They go together.

Regards,

David

David,

Thank you for this answer. As I was still trying to work with the datasheet I couldn't understand how come EPwm2 was routed to EPwm1 pins ...

However I can only work with EPWM2 for now. EPWM1 is not routed as stated by you and EPWM3(A) is a DAC I guess or some kind of strange signal (partially sawtooth where the PWM is low, square where PWM is high). I guess some kind of filtering or high impedance state instead if being pulled low.

A screenshot of what I previously mentioned (problem: the triangular / sawtooth shape instead of square)

Furthermore the PWM frequency seems to vary according to the CMPA value which is very weird