mesa 5I20 with 7i48 & 2x 7I37

Hi

trying to setup a 5i20 with
1x 7I48 on P2
7i37 on P3 & P4

i have config as
CONFIG="firmware=hm2/5i20/SV12_2X7I48_72.BIT num_encoders=6 num_pwmgens=4 num_stepgens=0"

to use the 6 encoder inputs & 4 analogs out

when starting i get
pastebin.com/U4P7m3HE

i see
[ 8560.650653] hm2: loading Mesa HostMot2 driver version 0.15
[ 8560.652862] hm2_pci: loading Mesa AnyIO HostMot2 driver version 0.7
[ 8560.652882] hm2_pci: discovered 5i20 at 0000:03:05.0
[ 8560.652944] hm2_5i20.0: firmware: requesting hm2/5i20/SV12_2X7I48_72.BIT
[ 8560.831572] hm2/hm2_5i20.0: WARNING: this firmware has Muxed Encoder v2!
[ 8560.831576] hm2/hm2_5i20.0: WARNING: velocity computation will be incorrect!
[ 8560.831578] hm2/hm2_5i20.0: WARNING: upgrade your firmware!

when i do a
dpkg --status hostmot2-firmware-5i20 | grep Version
Version: 0.8

but i see V0.7 listed and above error of 0.7 is firmware an old version some how?

also using
2.5 build of linuxcnc from build bot

thanks
rob

freeby.mesanet.com/5I23_SVSS6_8.ZIP

freeby.mesanet.com/5I20_SV12IM.ZIP

Note the 5I20 bit/pin files have a slightly different name than what you had

both have the .PIN files that should be read to check that the match your connectors

.BIT and .PIN files should be installed in /lib/firmware/hm2/5i23 and /lib/firmware/hm2/5i20

thanks for that and thanks for sorting it so quick. much appreciated

all working fine

rob

while trying to think why my machine would not work yet my test station would

i found the following questionable setup after awhile of prodding

using the 7i48 (on the 5i20 or 5i23)
i would get no Aouts unless i enable pwmgen.00.enable
soon as i enable this i get analog outs on all chanels as requested from hal etc.

if i disable pwmgen.00 and leave say pwmgen.05.enable true this turns off pwmgen.05 (plus all the others)
soon as pwmgen.00 is true things work as exspected as long as pwmgen.00 is always true

did log this on IRC @ 23:56 GMT

rob

Yes, the 7I48 does not have individual enables (there are not enough pins),
it has a single gated enable(both need to be true to enable all outputs).
Normally these enables are connected in the FPGA to the first and last
PWM channel enables (either ENA0,ENA5 or ENA6,ENA11)

Heres the manual section about the enables:

ENABLE INPUTS
There are two active low enable inputs on the 7I48 from the FPGA card. When
either of these enable inputs is high, all AOUTs are forced to 0V, and all enable outputs
are turned off. Pullup resistors keep the enable inputs high if the controller connection is
lost.

Hi,

almost one year later: firmware for a signle 7i48 still contains the bug. Additionally the firmware for 2x7i48 will render the card useless... Will there be a fix or do I have to build my own bitfile?

Please help!

what bug?
The "pwmena00 enables all pwmgens" behaviour is correct as there is only a
single PWM enable on the 7I48 (and 2 enable inputs that must be true to enable (all) the 7I48 PWMGens)

andypugh wrote:

Actually, it is more than 4 years later...

you are right, no idea what happend.
This was not the thread i wanted to post a comment in

If I can find the right one I will re-post, it was about the 5i24 and 7i37 serial bug... Anyway got it running now