Problems in Configuration of my L6482 board

val_ltd · ‎2014-03-23

Posted on March 23, 2014 at 11:37

I have designed a board using a L6482 chip.

My L6482 has started, and I can send and receive messages a través de SPI bus, but now I have a problem:

My STATUS register says:

1110 0000 0000 0011

(Get Status Command = 0xE003).

It means:

OCD

= 1 --> overcurrent

BUSY

= 1 --> Stopped

Hiz

= 1 --> Bridges in High Impedance State.

The problem occurs

even

without connecting the

stepper

motor.

...!

Just turn on the power, and my L6482 gets quite hot, but I have measured the total current of the board, and it is only 0.1 Amp.

Configuration:

I have read AN4354 and UM1685 Application note and User Manual, in order to calculate config values of my board:

OCD_TH = 1000mV

TCC = 1200 nSg

IGATE = 64 mA

TBLANK = 500 nSg

TDT = 250 nSg

I use 3.3v external power supply: VCCREG, VREG and VDDIO are connected to this 3.3v external power supply.

I use VSREG and internal regulator for obtaining VCC=15v. VCC pin is unconnected.

I am using Mosfets STB75NF75L with D2PAK package, of ST Microelectronics manufacturer.

My Rsense =

200mΩ

, is it a too high value?

So I think the problem could be:

Bad values in registers of configuration. (AN 4354, chapter 2.4: Too low gate charge to complete the Miller plateau region...?)
Bad connections of: Vs, VSreg, Vccreg, Vcc, Vdd, etc.......

Could anybody put here the right connections?

And the configuration values?

Thanks in advance,

val_ltd · ‎2014-06-12

Posted on June 12, 2014 at 21:09

I have put BAR43SFILM diodes, and afterwards I was checking voltages.

With VS = 24v, I have tested them with a Fluke multimeter :

In EAVAL6482H R1.0 board:

VCCREG = 7,59v

VREG = 3,23v

In my board:

VCCREG = 6,86 -- 6,97v

(too low, it is in the limit of UVLO alarm)

VREG = 4,62v

(too high, it should be 3,3v.....)

I see that in my board VCCREG is in the limit of UVLO alarm, and VREG is too high, but I do not understand what coud be the reason, because they are internal regulators...

val_ltd · ‎2014-06-14

Posted on June 15, 2014 at 01:36

I am thinking in testing VREG and VCCREG with a board before mounting their 8 Mosfets, in order to know more about so strange values of voltages.

Is it possible, or L6482 microcontroller could be damaged?

val_ltd · ‎2014-06-15

Posted on June 15, 2014 at 09:40

I have also bought

STB80NF10T4

(D2PAK)

Mosfets. That way, at this moment, I have two boards:

Board with STB75NF75L (D2PAK)
Mosfets, and Vccreg= 7,5v
Board with
STB80NF10T4

(D2PAK)
Mosfets, and Vccreg= 15v

Both with

BAR43SFILM

low drop diodes.

I have tested them, and obtained the same results.

In this project I am using bigger Mosfets than in EVAL6482H example board. so, my question is:

Is

L6482H

able to drive

these

bigger

MOSFET

s?

In other case,

Could be a problem in configuration values, or

the

order

in which

the configuration sequence

is done

?

Comparing, typical values of ''Total Gate Charge'' for

STD25NF10T4

is 55 nC with 10 Vgs = 10V, whilst

STB75NF75L

needs 75 nC and Vgs = 5V

, and

STB80NF10T4

needs

135 nC a Vgs = 10V.

I am asking about it, because I always found problems about OVERCURRENT, THERMAL_WARNING, THERMAL_SHUTDOWN......

odstrcilik · ‎2014-06-19

Posted on June 19, 2014 at 15:09

I have the some problems like ramirez_prieto.valen. I was not able to run L6482. The communication is good (checked on osciloscope). I read all of parameter back to uC and the numbers are the same. But I am still not able to does RUN AT SPEED command for motors 0,5 Amps, 1,5 Amps, 3 Amps - Nothing. Status is all of time 0xE003. The L6482 didn't set just one MOSFET Gate pulse but still claim OCD. I presumed that the driver will generate at least one Gate pulse and after that it can claimed the OCD. Is it true?

val_ltd · ‎2014-06-22

Posted on June 22, 2014 at 21:29

I have found the solution of one of my (small) problems

:

In my board, some voltages were wrong:

VCCREG = 6,86 -- 6,97v

VREG = 4,62v

I had forgotten

to put the condenser

CVREG=22uF

Now all voltages are exactly equals to voltages checked on the EVAL6482H board.

When I was thinking about VREG stabilization, I checked differents documents of ST (Application Notes, Manuals of Boards, etc), and all of them give different solutions :

UM1685 (pg.17)

VREG uses two condensers: 22uF and 100nF

VDD uses one condenser: 100nF

AN4355 (pg. 5)

VREG uses two condensers: 47uF and 100nF

VDD uses two condensers: 10uF and 100nF

UM17656 (pg.9)

VREG uses one condenser: 220nF

VDD uses one condenser: 100nF

Which one is the right one? or at least Which one is better?

val_ltd · ‎2014-06-22

Posted on June 22, 2014 at 23:18

Afterwards, I have mounted two boards, for comparing their behaviors:

One board with
STD25NF10T4
Mosfets, Vg=15v, (DPAK), as EVAL6482H board.
One board with
STB80NF10T4 Mosfets
, Vg=15v, (D2PAK, a bigger mosfet).

I have made several tests, and:

1) Board equivalent to EVAL6482H runs with 1Nm motor connected.

(BSY flag

is

successfully activated

, motor runs, and FLG is not activated).

(Led yellow On, Led Red Off)

2) Board with

STB80NF10T4 Mosfets

runs fine when there is not any motor connected.

(BSY flag

is

successfully activated

, and FLG is not activated, Led yellow On, Led Red Off).

But when I connect a PaP motor and I send a Move command, FLG signal is activated.

(Led yellow Off, Led Red On)

Then GetStatus() returns this

Error: 0x00000603

Overcurrent event is right away activated, but there is no reason for this happen. What is happening?

odstrcilik · ‎2014-06-23

Posted on June 23, 2014 at 11:51

I have used

VREG uses two condensers: 22uF and 100nF

VDD uses one condenser: 100nF

like is meaned in UM1685 (pg.17)

and I measured

VCCREG = 7,52 V

VREG = 3,30 V

val_ltd · ‎2014-06-23

Posted on June 23, 2014 at 18:12

As is shown in this image, the

test

environment

is very simple: Two STM32F4 DISCOVERY boards, connected to this PCB boards. However, the results

are totally different.

..

I ask to ST the same question once again:

Is L6482H able to drive bigger Mosfets than STD25NF10T4 model?

Has this situation

ever

been tested

by ST?

Maybe I should stop trying to control these MOSFET, and use these smaller STD25NF10T4 Mostfet....

odstrcilik · ‎2014-06-24

Posted on June 24, 2014 at 12:56

Can you please post your drivers number on both boards?

I have 7B262 L6482 V5 PHL 11220

It could be the different silicon version and any bug there, but in errata is nothing.

val_ltd · ‎2014-06-24

Posted on June 24, 2014 at 19:44

My board with

STB80NF10T4 Mosfets, Vg=15v, (D2PAK, a bigger mosfet).

started to run, when I have modified the next parameters, and I have put very low values.

CSPIN_SetParam(CSPIN_TVAL_HOLD, 0x08);

CSPIN_SetParam(CSPIN_TVAL_RUN, 0x05);

CSPIN_SetParam(CSPIN_TVAL_ACC, 0x05);

CSPIN_SetParam(CSPIN_TVAL_DEC, 0x05);

In board with:

STB75NF75L (D2PAK) Mosfets, and Vccreg= 7,5v

I have reduced gate current = 64mA to 32mA and a longer TCC value, and then board started to work,

Movement is not perfect and some times the test secuence of movements stops suddenly and unexpectedly.