Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Frederick Gamache]

Ok thanks.

I was able to start camogm manually but it wasn't starting automatically at
boot time like before.

What I did is I re-wrote the files on the SD Card and now camogm is called
automatically from the webpage link as before.

The problem I had was that the webpage using  camogmgui.php was
continuously refreshing without starting the GUI...

Frederick

On Thu, Mar 15, 2018 at 12:16 PM, Oleg 
wrote:

> Hi,
>
> 1. To check if camogm is alive:
>
>> root@elphel393:~# ps | grep camogm
>
> 3059 root  267m Scamogm -n /var/state/camogm_cmd -p 3456 -s
>> /mnt/sda1/camogm.disk
>> 3081 root  2764 Sgrep camogm
>
>
> 2. Check camogm status:
> http://192.168.0.9/camogm_interface.php?cmd=status
>
> 3. Start manually (Normally it's started by camogmgui.php (linked from the
> main page)):
>
>> root@elphel393:~# camogm -n /var/state/camogm_cmd -p 3456
>
>
> When started from a command line there should be some output when it
> receives commands.
>
> Regards,
> Oleg
>
> On Thu, Mar 15, 2018 at 9:50 AM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
>> Frederick,
>>
>> A link to what?
>>
>> How were you starting camogm? Did it work before? How were you using it?
>> How did it fail to load?
>>
>> Andrey
>>
>>  On Thu, 15 Mar 2018 08:31:29 -0700 *Frederick
>> Gamache>*
>> wrote 
>>
>> Another thing,
>>
>> After I tried pluging a USB drive through an hub, attached to the 10389B
>> board, it seems that the camogm is failing to load, even after re-booting.
>>
>> Is there a link?
>>
>> Frederick
>>
>> On Thu, Mar 15, 2018 at 10:53 AM, Frederick Gamache <
>> frederick.gama...@gmail.com> wrote:
>>
>> Andrey,
>>
>> The supply current drawn is nearly 2.0 Amps @ 3.3V, once the FPGA is
>> loaded.
>> I have the 10389 interface board but no SSD mounted and no external SATA
>> drive attached at the moment.
>>
>> When I received the camera kit, I found that the unit was especially hot
>> when touching it.
>> I don't have the camera enclosure, but only an aluminium frame on which
>> the boards are mounted. I guess that the frame serves as a heat sink at
>> same time?
>>
>> For the power board failure, maybe I accidentally created a short when I
>> plug/unplugged the wall adapter from the power board connector while the
>> power adapter was plugged into the wall...
>>
>> It is hard to tell what component failed on the power board because many
>> components are mounted underneath the Traco DC-DC converter itself...
>>
>> When I plug the 48 V source, no 3.3V appears at the output.
>>
>> Frederick
>>
>>
>> On Thu, Mar 15, 2018 at 10:29 AM, Elphel Support <
>> support-list@support.elphel.com> wrote:
>>
>> Hello Frederick,
>>
>> Do you know power consumption on the ports? Do you have 10389 interface
>> board with the SSD? If you run camera from the 3.3V - can you measure the
>> total 3.3V current? Do you have sufficient fan cooling or heat sink?
>>
>> We never had 10385 failures before - Traco DC-DC converter seems to have
>> all required protection - current, temperature so we had the system shut
>> down when overloaded, but not failed. We would like to try to find out -
>> what component actually failed - is it DC-DC or something else.
>>
>> We do have a new power supply design (https://wiki.elphel.com/wiki/10386
>> - temporary images are for 10385B), but it is not built yet., we just
>> released the PCB design a few days ago.
>>
>> Andrey
>>
>>  On Thu, 15 Mar 2018 06:38:07 -0700 *Frederick Gamache
>> >* wrote 
>>
>> Hi,
>>
>> Its seems that my power board (10385-B power board - 48V) is out of
>> service and I'm now powering the 3.3V  from an external power supply.
>> Do I really need the 10385-B power board to supply the SATA adapter board
>> (10389-B) or should it works ok from my external 3.3V power source?
>>
>> Frederick
>>
>>
>> On Wed, Mar 14, 2018 at 11:10 AM, Frederick Gamache <
>> frederick.gama...@gmail.com> wrote:
>>
>> Hello Andrey,
>>
>> I owe at least this for your great support.
>> By the way, congratulations for your great products!
>>
>> Best regards,
>> Frederick
>>
>> On Wed, Mar 14, 2018 at 10:53 AM, Elphel Support <
>> support-list@support.elphel.com> wrote:
>>
>> Frederick,
>>
>> Thank you for letting us know it did work. Sometimes the only indication
>> of the problem being solved is the lack of the continuation in the thread.
>>
>> Andrey
>>
>>  On Wed, 14 Mar 2018 07:12:13 -0700 *Frederick Gamache
>> >* wrote 
>>
>> Hello,
>>
>> Just to let you know guys that my circuit is now working since I mounted
>> my DC-DC converter on my PCB.
>>
>> Many thanks for your support!
>> Frederick
>>
>> On Tue, Mar 13, 2018 at 12:18 PM, Elphel Support <
>> support-list@support.elphel.com> wrote:
>>
>> Frederick,
>>
>> Yes - your last problem (system getting stuck after 

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Oleg]

Hi,

1. To check if camogm is alive:

> root@elphel393:~# ps | grep camogm

3059 root  267m Scamogm -n /var/state/camogm_cmd -p 3456 -s
> /mnt/sda1/camogm.disk
> 3081 root  2764 Sgrep camogm


2. Check camogm status:
http://192.168.0.9/camogm_interface.php?cmd=status

3. Start manually (Normally it's started by camogmgui.php (linked from the
main page)):

> root@elphel393:~# camogm -n /var/state/camogm_cmd -p 3456


When started from a command line there should be some output when it
receives commands.

Regards,
Oleg

On Thu, Mar 15, 2018 at 9:50 AM, Elphel Support <
support-list@support.elphel.com> wrote:

> Frederick,
>
> A link to what?
>
> How were you starting camogm? Did it work before? How were you using it?
> How did it fail to load?
>
> Andrey
>
>  On Thu, 15 Mar 2018 08:31:29 -0700 *Frederick
> Gamache>* wrote
> 
>
> Another thing,
>
> After I tried pluging a USB drive through an hub, attached to the 10389B
> board, it seems that the camogm is failing to load, even after re-booting.
>
> Is there a link?
>
> Frederick
>
> On Thu, Mar 15, 2018 at 10:53 AM, Frederick Gamache <
> frederick.gama...@gmail.com> wrote:
>
> Andrey,
>
> The supply current drawn is nearly 2.0 Amps @ 3.3V, once the FPGA is
> loaded.
> I have the 10389 interface board but no SSD mounted and no external SATA
> drive attached at the moment.
>
> When I received the camera kit, I found that the unit was especially hot
> when touching it.
> I don't have the camera enclosure, but only an aluminium frame on which
> the boards are mounted. I guess that the frame serves as a heat sink at
> same time?
>
> For the power board failure, maybe I accidentally created a short when I
> plug/unplugged the wall adapter from the power board connector while the
> power adapter was plugged into the wall...
>
> It is hard to tell what component failed on the power board because many
> components are mounted underneath the Traco DC-DC converter itself...
>
> When I plug the 48 V source, no 3.3V appears at the output.
>
> Frederick
>
>
> On Thu, Mar 15, 2018 at 10:29 AM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
> Hello Frederick,
>
> Do you know power consumption on the ports? Do you have 10389 interface
> board with the SSD? If you run camera from the 3.3V - can you measure the
> total 3.3V current? Do you have sufficient fan cooling or heat sink?
>
> We never had 10385 failures before - Traco DC-DC converter seems to have
> all required protection - current, temperature so we had the system shut
> down when overloaded, but not failed. We would like to try to find out -
> what component actually failed - is it DC-DC or something else.
>
> We do have a new power supply design (https://wiki.elphel.com/wiki/10386
> - temporary images are for 10385B), but it is not built yet., we just
> released the PCB design a few days ago.
>
> Andrey
>
>  On Thu, 15 Mar 2018 06:38:07 -0700 *Frederick Gamache
> >* wrote 
>
> Hi,
>
> Its seems that my power board (10385-B power board - 48V) is out of
> service and I'm now powering the 3.3V  from an external power supply.
> Do I really need the 10385-B power board to supply the SATA adapter board
> (10389-B) or should it works ok from my external 3.3V power source?
>
> Frederick
>
>
> On Wed, Mar 14, 2018 at 11:10 AM, Frederick Gamache <
> frederick.gama...@gmail.com> wrote:
>
> Hello Andrey,
>
> I owe at least this for your great support.
> By the way, congratulations for your great products!
>
> Best regards,
> Frederick
>
> On Wed, Mar 14, 2018 at 10:53 AM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
> Frederick,
>
> Thank you for letting us know it did work. Sometimes the only indication
> of the problem being solved is the lack of the continuation in the thread.
>
> Andrey
>
>  On Wed, 14 Mar 2018 07:12:13 -0700 *Frederick Gamache
> >* wrote 
>
> Hello,
>
> Just to let you know guys that my circuit is now working since I mounted
> my DC-DC converter on my PCB.
>
> Many thanks for your support!
> Frederick
>
> On Tue, Mar 13, 2018 at 12:18 PM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
> Frederick,
>
> Yes - your last problem (system getting stuck after bitstream) is caused
> by the wrong power sequencing. But as 1) you have 10338 working and 2) your
> board is tested w/o 10393 it should be working together eventually.
>
> Andrey
>
>  On Tue, 13 Mar 2018 07:48:42 -0700 *Frederick
> Gamache>* wrote
> 
>
> Hello Andrey,
>
> Normally I have a dc-dc converter chip on my pcb which translate 3.3V to
> 2.8V.
> The 3.3V source is provided by the cable attached to the 10393.
>
> I'll wait to receive my chip today and will install it tomorrow. I'll see
> if things will get better with it.
>
> 

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Elphel Support]

Frederick,

A link to what?

How were you starting camogm? Did it work before? How were you using it? How 
did it fail to load?

Andrey

 On Thu, 15 Mar 2018 08:31:29 -0700 Frederick 
Gamachefrederick.gama...@gmail.com wrote  

Another thing,

After I tried pluging a USB drive through an hub, attached to the 10389B board, 
it seems that the camogm is failing to load, even after re-booting.


Is there a link?


Frederick


On Thu, Mar 15, 2018 at 10:53 AM, Frederick Gamache 
frederick.gama...@gmail.com wrote:
Andrey,

The supply current drawn is nearly 2.0 Amps @ 3.3V, once the FPGA is loaded.
I have the 10389 interface board but no SSD mounted and no external SATA drive 
attached at the moment.


When I received the camera kit, I found that the unit was especially hot when 
touching it.
I don't have the camera enclosure, but only an aluminium frame on which the 
boards are mounted. I guess that the frame serves as a heat sink at same time?


For the power board failure, maybe I accidentally created a short when I 
plug/unplugged the wall adapter from the power board connector while the power 
adapter was plugged into the wall...


It is hard to tell what component failed on the power board because many 
components are mounted underneath the Traco DC-DC converter itself...


When I plug the 48 V source, no 3.3V appears at the output.


Frederick




On Thu, Mar 15, 2018 at 10:29 AM, Elphel Support 
support-list@support.elphel.com wrote:
Hello Frederick,

Do you know power consumption on the ports? Do you have 10389 interface board 
with the SSD? If you run camera from the 3.3V - can you measure the total 3.3V 
current? Do you have sufficient fan cooling or heat sink?

We never had 10385 failures before - Traco DC-DC converter seems to have all 
required protection - current, temperature so we had the system shut down when 
overloaded, but not failed. We would like to try to find out - what component 
actually failed - is it DC-DC or something else.

We do have a new power supply design (https://wiki.elphel.com/wiki/10386 - 
temporary images are for 10385B), but it is not built yet., we just released 
the PCB design a few days ago.

Andrey

 On Thu, 15 Mar 2018 06:38:07 -0700 Frederick Gamache 
frederick.gama...@gmail.com wrote  

Hi,

Its seems that my power board (10385-B power board - 48V) is out of service and 
I'm now powering the 3.3V  from an external power supply.
Do I really need the 10385-B power board to supply the SATA adapter board 
(10389-B) or should it works ok from my external 3.3V power source?


Frederick




On Wed, Mar 14, 2018 at 11:10 AM, Frederick Gamache 
frederick.gama...@gmail.com wrote:
Hello Andrey,

I owe at least this for your great support.
By the way, congratulations for your great products!


Best regards,
Frederick


On Wed, Mar 14, 2018 at 10:53 AM, Elphel Support 
support-list@support.elphel.com wrote:
Frederick,

Thank you for letting us know it did work. Sometimes the only indication of the 
problem being solved is the lack of the continuation in the thread.

Andrey

 On Wed, 14 Mar 2018 07:12:13 -0700 Frederick Gamache 
frederick.gama...@gmail.com wrote  

Hello,

Just to let you know guys that my circuit is now working since I mounted my 
DC-DC converter on my PCB.


Many thanks for your support!
Frederick


On Tue, Mar 13, 2018 at 12:18 PM, Elphel Support 
support-list@support.elphel.com wrote:
Frederick,

Yes - your last problem (system getting stuck after bitstream) is caused by the 
wrong power sequencing. But as 1) you have 10338 working and 2) your board is 
tested w/o 10393 it should be working together eventually.

Andrey

 On Tue, 13 Mar 2018 07:48:42 -0700 Frederick 
Gamachefrederick.gama...@gmail.com wrote  

Hello Andrey,

Normally I have a dc-dc converter chip on my pcb which translate 3.3V to 2.8V. 
The 3.3V source is provided by the cable attached to the 10393.


I'll wait to receive my chip today and will install it tomorrow. I'll see if 
things will get better with it.


Currently I'm using an external power supply to source the 2.8V voltage so I 
guess the timing for the power sequence is not ok.


Frederick


On Fri, Mar 9, 2018 at 5:51 PM, Elphel Support 
support-list@support.elphel.com wrote:
Fred,

Just to clarify: 10393 can drive up to ~1.0A, but there will be voltage drop on 
the cable (Elphel cables have thick power traces). Connector is rated 0.4A/pin. 
The sensor power is controlled by the script (same one that loads bitstream), 
power status/control is available over the sysfs interface.

You may also put a buffer on all signals from the sensor, FPGA has flexible 
means to compensate pin delays if needed after the buffer insertion.

Andrey

 On Fri, 09 Mar 2018 14:26:50 -0800 Elphel 
Supportsupport-list@support.elphel.com wrote  

Frederick, you need to power you board either directly from the 10393 3.3V on 
the cable, or (if you board has other components that need more power than 
10393 can 

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Frederick Gamache]

Another thing,

After I tried pluging a USB drive through an hub, attached to the 10389B
board, it seems that the camogm is failing to load, even after re-booting.

Is there a link?

Frederick

On Thu, Mar 15, 2018 at 10:53 AM, Frederick Gamache <
frederick.gama...@gmail.com> wrote:

> Andrey,
>
> The supply current drawn is nearly 2.0 Amps @ 3.3V, once the FPGA is
> loaded.
> I have the 10389 interface board but no SSD mounted and no external SATA
> drive attached at the moment.
>
> When I received the camera kit, I found that the unit was especially hot
> when touching it.
> I don't have the camera enclosure, but only an aluminium frame on which
> the boards are mounted. I guess that the frame serves as a heat sink at
> same time?
>
> For the power board failure, maybe I accidentally created a short when I
> plug/unplugged the wall adapter from the power board connector while the
> power adapter was plugged into the wall...
>
> It is hard to tell what component failed on the power board because many
> components are mounted underneath the Traco DC-DC converter itself...
>
> When I plug the 48 V source, no 3.3V appears at the output.
>
> Frederick
>
>
> On Thu, Mar 15, 2018 at 10:29 AM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
>> Hello Frederick,
>>
>> Do you know power consumption on the ports? Do you have 10389 interface
>> board with the SSD? If you run camera from the 3.3V - can you measure the
>> total 3.3V current? Do you have sufficient fan cooling or heat sink?
>>
>> We never had 10385 failures before - Traco DC-DC converter seems to have
>> all required protection - current, temperature so we had the system shut
>> down when overloaded, but not failed. We would like to try to find out -
>> what component actually failed - is it DC-DC or something else.
>>
>> We do have a new power supply design (https://wiki.elphel.com/wiki/10386
>> - temporary images are for 10385B), but it is not built yet., we just
>> released the PCB design a few days ago.
>>
>> Andrey
>>
>>  On Thu, 15 Mar 2018 06:38:07 -0700 *Frederick Gamache
>> >* wrote 
>>
>> Hi,
>>
>> Its seems that my power board (10385-B power board - 48V) is out of
>> service and I'm now powering the 3.3V  from an external power supply.
>> Do I really need the 10385-B power board to supply the SATA adapter board
>> (10389-B) or should it works ok from my external 3.3V power source?
>>
>> Frederick
>>
>>
>> On Wed, Mar 14, 2018 at 11:10 AM, Frederick Gamache <
>> frederick.gama...@gmail.com> wrote:
>>
>> Hello Andrey,
>>
>> I owe at least this for your great support.
>> By the way, congratulations for your great products!
>>
>> Best regards,
>> Frederick
>>
>> On Wed, Mar 14, 2018 at 10:53 AM, Elphel Support <
>> support-list@support.elphel.com> wrote:
>>
>> Frederick,
>>
>> Thank you for letting us know it did work. Sometimes the only indication
>> of the problem being solved is the lack of the continuation in the thread.
>>
>> Andrey
>>
>>  On Wed, 14 Mar 2018 07:12:13 -0700 *Frederick Gamache
>> >* wrote 
>>
>> Hello,
>>
>> Just to let you know guys that my circuit is now working since I mounted
>> my DC-DC converter on my PCB.
>>
>> Many thanks for your support!
>> Frederick
>>
>> On Tue, Mar 13, 2018 at 12:18 PM, Elphel Support <
>> support-list@support.elphel.com> wrote:
>>
>> Frederick,
>>
>> Yes - your last problem (system getting stuck after bitstream) is caused
>> by the wrong power sequencing. But as 1) you have 10338 working and 2) your
>> board is tested w/o 10393 it should be working together eventually.
>>
>> Andrey
>>
>>  On Tue, 13 Mar 2018 07:48:42 -0700 *Frederick
>> Gamache>*
>> wrote 
>>
>> Hello Andrey,
>>
>> Normally I have a dc-dc converter chip on my pcb which translate 3.3V to
>> 2.8V.
>> The 3.3V source is provided by the cable attached to the 10393.
>>
>> I'll wait to receive my chip today and will install it tomorrow. I'll see
>> if things will get better with it.
>>
>> Currently I'm using an external power supply to source the 2.8V voltage
>> so I guess the timing for the power sequence is not ok.
>>
>> Frederick
>>
>> On Fri, Mar 9, 2018 at 5:51 PM, Elphel Support <
>> support-list@support.elphel.com> wrote:
>>
>> Fred,
>>
>> Just to clarify: 10393 can drive up to ~1.0A, but there will be voltage
>> drop on the cable (Elphel cables have thick power traces). Connector is
>> rated 0.4A/pin. The sensor power is controlled by the script (same one that
>> loads bitstream), power status/control is available over the sysfs
>> interface.
>>
>> You may also put a buffer on all signals from the sensor, FPGA has
>> flexible means to compensate pin delays if needed after the buffer
>> insertion.
>>
>> Andrey
>>
>>  On Fri, 09 Mar 2018 14:26:50 -0800 *Elphel
>> Support

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Elphel Support]

Frederick,

2A is not too much, Traco can provide 3.5A. How did you test 10385 - separately 
or when connected to 10393? 10385 itself will not work - there is enable signal 
coming from the 10393, so 10393 can shut itself down and then wake up with the 
clock/calendar chip.

Heat frame by itself is not enough, for development systems we use 1/2 of the 
normal camera body that looks like U-channel and slide the heat frame into it. 
for higher power we then put this half-body on a larger aluminum frame. The 
design is made so that air flow is not required, heat is transferred through 
the heat frame and should be handled externally. For most outdoor applications 
nothing more is needed, but if it is placed in vacuum camera boards will still 
work if heat is extracted through the metal body.

Andrey

 On Thu, 15 Mar 2018 07:53:46 -0700 Frederick 
Gamachefrederick.gama...@gmail.com wrote  

Andrey,

The supply current drawn is nearly 2.0 Amps @ 3.3V, once the FPGA is loaded.
I have the 10389 interface board but no SSD mounted and no external SATA drive 
attached at the moment.


When I received the camera kit, I found that the unit was especially hot when 
touching it.
I don't have the camera enclosure, but only an aluminium frame on which the 
boards are mounted. I guess that the frame serves as a heat sink at same time?


For the power board failure, maybe I accidentally created a short when I 
plug/unplugged the wall adapter from the power board connector while the power 
adapter was plugged into the wall...


It is hard to tell what component failed on the power board because many 
components are mounted underneath the Traco DC-DC converter itself...


When I plug the 48 V source, no 3.3V appears at the output.


Frederick




On Thu, Mar 15, 2018 at 10:29 AM, Elphel Support 
support-list@support.elphel.com wrote:
Hello Frederick,

Do you know power consumption on the ports? Do you have 10389 interface board 
with the SSD? If you run camera from the 3.3V - can you measure the total 3.3V 
current? Do you have sufficient fan cooling or heat sink?

We never had 10385 failures before - Traco DC-DC converter seems to have all 
required protection - current, temperature so we had the system shut down when 
overloaded, but not failed. We would like to try to find out - what component 
actually failed - is it DC-DC or something else.

We do have a new power supply design (https://wiki.elphel.com/wiki/10386 - 
temporary images are for 10385B), but it is not built yet., we just released 
the PCB design a few days ago.

Andrey

 On Thu, 15 Mar 2018 06:38:07 -0700 Frederick Gamache 
frederick.gama...@gmail.com wrote  

Hi,

Its seems that my power board (10385-B power board - 48V) is out of service and 
I'm now powering the 3.3V  from an external power supply.
Do I really need the 10385-B power board to supply the SATA adapter board 
(10389-B) or should it works ok from my external 3.3V power source?


Frederick




On Wed, Mar 14, 2018 at 11:10 AM, Frederick Gamache 
frederick.gama...@gmail.com wrote:
Hello Andrey,

I owe at least this for your great support.
By the way, congratulations for your great products!


Best regards,
Frederick


On Wed, Mar 14, 2018 at 10:53 AM, Elphel Support 
support-list@support.elphel.com wrote:
Frederick,

Thank you for letting us know it did work. Sometimes the only indication of the 
problem being solved is the lack of the continuation in the thread.

Andrey

 On Wed, 14 Mar 2018 07:12:13 -0700 Frederick Gamache 
frederick.gama...@gmail.com wrote  

Hello,

Just to let you know guys that my circuit is now working since I mounted my 
DC-DC converter on my PCB.


Many thanks for your support!
Frederick


On Tue, Mar 13, 2018 at 12:18 PM, Elphel Support 
support-list@support.elphel.com wrote:
Frederick,

Yes - your last problem (system getting stuck after bitstream) is caused by the 
wrong power sequencing. But as 1) you have 10338 working and 2) your board is 
tested w/o 10393 it should be working together eventually.

Andrey

 On Tue, 13 Mar 2018 07:48:42 -0700 Frederick 
Gamachefrederick.gama...@gmail.com wrote  

Hello Andrey,

Normally I have a dc-dc converter chip on my pcb which translate 3.3V to 2.8V. 
The 3.3V source is provided by the cable attached to the 10393.


I'll wait to receive my chip today and will install it tomorrow. I'll see if 
things will get better with it.


Currently I'm using an external power supply to source the 2.8V voltage so I 
guess the timing for the power sequence is not ok.


Frederick


On Fri, Mar 9, 2018 at 5:51 PM, Elphel Support 
support-list@support.elphel.com wrote:
Fred,

Just to clarify: 10393 can drive up to ~1.0A, but there will be voltage drop on 
the cable (Elphel cables have thick power traces). Connector is rated 0.4A/pin. 
The sensor power is controlled by the script (same one that loads bitstream), 
power status/control is available over the sysfs interface.

You may also put a buffer on all signals 

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Frederick Gamache]

Andrey,

The supply current drawn is nearly 2.0 Amps @ 3.3V, once the FPGA is loaded.
I have the 10389 interface board but no SSD mounted and no external SATA
drive attached at the moment.

When I received the camera kit, I found that the unit was especially hot
when touching it.
I don't have the camera enclosure, but only an aluminium frame on which the
boards are mounted. I guess that the frame serves as a heat sink at same
time?

For the power board failure, maybe I accidentally created a short when I
plug/unplugged the wall adapter from the power board connector while the
power adapter was plugged into the wall...

It is hard to tell what component failed on the power board because many
components are mounted underneath the Traco DC-DC converter itself...

When I plug the 48 V source, no 3.3V appears at the output.

Frederick


On Thu, Mar 15, 2018 at 10:29 AM, Elphel Support <
support-list@support.elphel.com> wrote:

> Hello Frederick,
>
> Do you know power consumption on the ports? Do you have 10389 interface
> board with the SSD? If you run camera from the 3.3V - can you measure the
> total 3.3V current? Do you have sufficient fan cooling or heat sink?
>
> We never had 10385 failures before - Traco DC-DC converter seems to have
> all required protection - current, temperature so we had the system shut
> down when overloaded, but not failed. We would like to try to find out -
> what component actually failed - is it DC-DC or something else.
>
> We do have a new power supply design (https://wiki.elphel.com/wiki/10386
> - temporary images are for 10385B), but it is not built yet., we just
> released the PCB design a few days ago.
>
> Andrey
>
>  On Thu, 15 Mar 2018 06:38:07 -0700 *Frederick Gamache
> >* wrote 
>
> Hi,
>
> Its seems that my power board (10385-B power board - 48V) is out of
> service and I'm now powering the 3.3V  from an external power supply.
> Do I really need the 10385-B power board to supply the SATA adapter board
> (10389-B) or should it works ok from my external 3.3V power source?
>
> Frederick
>
>
> On Wed, Mar 14, 2018 at 11:10 AM, Frederick Gamache <
> frederick.gama...@gmail.com> wrote:
>
> Hello Andrey,
>
> I owe at least this for your great support.
> By the way, congratulations for your great products!
>
> Best regards,
> Frederick
>
> On Wed, Mar 14, 2018 at 10:53 AM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
> Frederick,
>
> Thank you for letting us know it did work. Sometimes the only indication
> of the problem being solved is the lack of the continuation in the thread.
>
> Andrey
>
>  On Wed, 14 Mar 2018 07:12:13 -0700 *Frederick Gamache
> >* wrote 
>
> Hello,
>
> Just to let you know guys that my circuit is now working since I mounted
> my DC-DC converter on my PCB.
>
> Many thanks for your support!
> Frederick
>
> On Tue, Mar 13, 2018 at 12:18 PM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
> Frederick,
>
> Yes - your last problem (system getting stuck after bitstream) is caused
> by the wrong power sequencing. But as 1) you have 10338 working and 2) your
> board is tested w/o 10393 it should be working together eventually.
>
> Andrey
>
>  On Tue, 13 Mar 2018 07:48:42 -0700 *Frederick
> Gamache>* wrote
> 
>
> Hello Andrey,
>
> Normally I have a dc-dc converter chip on my pcb which translate 3.3V to
> 2.8V.
> The 3.3V source is provided by the cable attached to the 10393.
>
> I'll wait to receive my chip today and will install it tomorrow. I'll see
> if things will get better with it.
>
> Currently I'm using an external power supply to source the 2.8V voltage so
> I guess the timing for the power sequence is not ok.
>
> Frederick
>
> On Fri, Mar 9, 2018 at 5:51 PM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
> Fred,
>
> Just to clarify: 10393 can drive up to ~1.0A, but there will be voltage
> drop on the cable (Elphel cables have thick power traces). Connector is
> rated 0.4A/pin. The sensor power is controlled by the script (same one that
> loads bitstream), power status/control is available over the sysfs
> interface.
>
> You may also put a buffer on all signals from the sensor, FPGA has
> flexible means to compensate pin delays if needed after the buffer
> insertion.
>
> Andrey
>
>  On Fri, 09 Mar 2018 14:26:50 -0800 *Elphel
> Support>*
> wrote 
>
> Frederick, you need to power you board either directly from the 10393 3.3V
> on the cable, or (if you board has other components that need more power
> than 10393 can provide) use external power switch driven by the cable power.
>
> The problem is that the cable goes directly to the FPGA port that should
> be powered _before_ any non-zero voltage appears on the FPGA pins. This is
> a feature of the 

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Frederick Gamache]

Hi,

Its seems that my power board (10385-B power board - 48V) is out of service
and I'm now powering the 3.3V  from an external power supply.
Do I really need the 10385-B power board to supply the SATA adapter board
(10389-B) or should it works ok from my external 3.3V power source?

Frederick


On Wed, Mar 14, 2018 at 11:10 AM, Frederick Gamache <
frederick.gama...@gmail.com> wrote:

> Hello Andrey,
>
> I owe at least this for your great support.
> By the way, congratulations for your great products!
>
> Best regards,
> Frederick
>
> On Wed, Mar 14, 2018 at 10:53 AM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
>> Frederick,
>>
>> Thank you for letting us know it did work. Sometimes the only indication
>> of the problem being solved is the lack of the continuation in the thread.
>>
>> Andrey
>>
>>  On Wed, 14 Mar 2018 07:12:13 -0700 *Frederick Gamache
>> >* wrote 
>>
>> Hello,
>>
>> Just to let you know guys that my circuit is now working since I mounted
>> my DC-DC converter on my PCB.
>>
>> Many thanks for your support!
>> Frederick
>>
>> On Tue, Mar 13, 2018 at 12:18 PM, Elphel Support <
>> support-list@support.elphel.com> wrote:
>>
>> Frederick,
>>
>> Yes - your last problem (system getting stuck after bitstream) is caused
>> by the wrong power sequencing. But as 1) you have 10338 working and 2) your
>> board is tested w/o 10393 it should be working together eventually.
>>
>> Andrey
>>
>>  On Tue, 13 Mar 2018 07:48:42 -0700 *Frederick
>> Gamache>*
>> wrote 
>>
>> Hello Andrey,
>>
>> Normally I have a dc-dc converter chip on my pcb which translate 3.3V to
>> 2.8V.
>> The 3.3V source is provided by the cable attached to the 10393.
>>
>> I'll wait to receive my chip today and will install it tomorrow. I'll see
>> if things will get better with it.
>>
>> Currently I'm using an external power supply to source the 2.8V voltage
>> so I guess the timing for the power sequence is not ok.
>>
>> Frederick
>>
>> On Fri, Mar 9, 2018 at 5:51 PM, Elphel Support <
>> support-list@support.elphel.com> wrote:
>>
>> Fred,
>>
>> Just to clarify: 10393 can drive up to ~1.0A, but there will be voltage
>> drop on the cable (Elphel cables have thick power traces). Connector is
>> rated 0.4A/pin. The sensor power is controlled by the script (same one that
>> loads bitstream), power status/control is available over the sysfs
>> interface.
>>
>> You may also put a buffer on all signals from the sensor, FPGA has
>> flexible means to compensate pin delays if needed after the buffer
>> insertion.
>>
>> Andrey
>>
>>  On Fri, 09 Mar 2018 14:26:50 -0800 *Elphel
>> Support>*
>> wrote 
>>
>> Frederick, you need to power you board either directly from the 10393
>> 3.3V on the cable, or (if you board has other components that need more
>> power than 10393 can provide) use external power switch driven by the cable
>> power.
>>
>> The problem is that the cable goes directly to the FPGA port that should
>> be powered _before_ any non-zero voltage appears on the FPGA pins. This is
>> a feature of the FPGA.
>>
>> Andrey
>>
>>  On Fri, 09 Mar 2018 13:47:51 -0800 *Frederick Gamache
>> >* wrote 
>>
>> I use TeraTerm.
>>
>> I receive something else than 0x, but it seam to be the same value
>> for any register...
>>
>> Maybe I should check my power on my sensor chip.
>>
>> FPGA bitstream stalls, it doesn't initialize if I power on my sensor PCB.
>>
>> Fred
>>
>> On Fri, Mar 9, 2018 at 4:37 PM, Oleg 
>> wrote:
>>
>> Hi,
>>
>>
>> Question: does the software needs to reach the  34VL02 EEPROM at any
>> point?
>> Because I don't have one on my own sensor board...
>>
>>
>> It should work fine without eeprom.
>>
>> Did it read chip id? or is it 0x? Do you use minicom or putty or
>> other terminal program?
>>
>> Regards,
>> Oleg
>>
>>
>> ___
>> Support-list mailing list
>> Support-list@support.elphel.com
>> http://support.elphel.com/mailman/listinfo/support-list_supp
>> ort.elphel.com
>>
>>
>>
>>
>> ___
>> Support-list mailing list
>> Support-list@support.elphel.com
>> http://support.elphel.com/mailman/listinfo/support-list_supp
>> ort.elphel.com
>>
>>
>>
>>
>>
>> ___
>> Support-list mailing list
>> Support-list@support.elphel.com
>> http://support.elphel.com/mailman/listinfo/support-list_supp
>> ort.elphel.com
>>
>>
>>
>>
>>
>> ___
>> Support-list mailing list
>> Support-list@support.elphel.com
>> http://support.elphel.com/mailman/listinfo/support-list_supp
>> ort.elphel.com
>>
>>
>>
>>
>>
>
___
Support-list 

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Frederick Gamache]

Hello,

Just to let you know guys that my circuit is now working since I mounted my
DC-DC converter on my PCB.

Many thanks for your support!
Frederick

On Tue, Mar 13, 2018 at 12:18 PM, Elphel Support <
support-list@support.elphel.com> wrote:

> Frederick,
>
> Yes - your last problem (system getting stuck after bitstream) is caused
> by the wrong power sequencing. But as 1) you have 10338 working and 2) your
> board is tested w/o 10393 it should be working together eventually.
>
> Andrey
>
>  On Tue, 13 Mar 2018 07:48:42 -0700 *Frederick
> Gamache>* wrote
> 
>
> Hello Andrey,
>
> Normally I have a dc-dc converter chip on my pcb which translate 3.3V to
> 2.8V.
> The 3.3V source is provided by the cable attached to the 10393.
>
> I'll wait to receive my chip today and will install it tomorrow. I'll see
> if things will get better with it.
>
> Currently I'm using an external power supply to source the 2.8V voltage so
> I guess the timing for the power sequence is not ok.
>
> Frederick
>
> On Fri, Mar 9, 2018 at 5:51 PM, Elphel Support <
> support-list@support.elphel.com> wrote:
>
> Fred,
>
> Just to clarify: 10393 can drive up to ~1.0A, but there will be voltage
> drop on the cable (Elphel cables have thick power traces). Connector is
> rated 0.4A/pin. The sensor power is controlled by the script (same one that
> loads bitstream), power status/control is available over the sysfs
> interface.
>
> You may also put a buffer on all signals from the sensor, FPGA has
> flexible means to compensate pin delays if needed after the buffer
> insertion.
>
> Andrey
>
>  On Fri, 09 Mar 2018 14:26:50 -0800 *Elphel
> Support>*
> wrote 
>
> Frederick, you need to power you board either directly from the 10393 3.3V
> on the cable, or (if you board has other components that need more power
> than 10393 can provide) use external power switch driven by the cable power.
>
> The problem is that the cable goes directly to the FPGA port that should
> be powered _before_ any non-zero voltage appears on the FPGA pins. This is
> a feature of the FPGA.
>
> Andrey
>
>  On Fri, 09 Mar 2018 13:47:51 -0800 *Frederick Gamache
> >* wrote 
>
> I use TeraTerm.
>
> I receive something else than 0x, but it seam to be the same value for
> any register...
>
> Maybe I should check my power on my sensor chip.
>
> FPGA bitstream stalls, it doesn't initialize if I power on my sensor PCB.
>
> Fred
>
> On Fri, Mar 9, 2018 at 4:37 PM, Oleg 
> wrote:
>
> Hi,
>
>
> Question: does the software needs to reach the  34VL02 EEPROM at any point?
> Because I don't have one on my own sensor board...
>
>
> It should work fine without eeprom.
>
> Did it read chip id? or is it 0x? Do you use minicom or putty or other
> terminal program?
>
> Regards,
> Oleg
>
>
> ___
> Support-list mailing list
> Support-list@support.elphel.com
> http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com
>
>
>
>
> ___
> Support-list mailing list
> Support-list@support.elphel.com
> http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com
>
>
>
>
>
> ___
> Support-list mailing list
> Support-list@support.elphel.com
> http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com
>
>
>
>
>
___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Elphel Support]

Frederick,

Yes - your last problem (system getting stuck after bitstream) is caused by the 
wrong power sequencing. But as 1) you have 10338 working and 2) your board is 
tested w/o 10393 it should be working together eventually.

Andrey

 On Tue, 13 Mar 2018 07:48:42 -0700 Frederick 
Gamachefrederick.gama...@gmail.com wrote  

Hello Andrey,

Normally I have a dc-dc converter chip on my pcb which translate 3.3V to 2.8V. 
The 3.3V source is provided by the cable attached to the 10393.


I'll wait to receive my chip today and will install it tomorrow. I'll see if 
things will get better with it.


Currently I'm using an external power supply to source the 2.8V voltage so I 
guess the timing for the power sequence is not ok.


Frederick


On Fri, Mar 9, 2018 at 5:51 PM, Elphel Support 
support-list@support.elphel.com wrote:
Fred,

Just to clarify: 10393 can drive up to ~1.0A, but there will be voltage drop on 
the cable (Elphel cables have thick power traces). Connector is rated 0.4A/pin. 
The sensor power is controlled by the script (same one that loads bitstream), 
power status/control is available over the sysfs interface.

You may also put a buffer on all signals from the sensor, FPGA has flexible 
means to compensate pin delays if needed after the buffer insertion.

Andrey

 On Fri, 09 Mar 2018 14:26:50 -0800 Elphel 
Supportsupport-list@support.elphel.com wrote  

Frederick, you need to power you board either directly from the 10393 3.3V on 
the cable, or (if you board has other components that need more power than 
10393 can provide) use external power switch driven by the cable power.

The problem is that the cable goes directly to the FPGA port that should be 
powered _before_ any non-zero voltage appears on the FPGA pins. This is a 
feature of the FPGA.

Andrey

 On Fri, 09 Mar 2018 13:47:51 -0800 Frederick Gamache 
frederick.gama...@gmail.com wrote  

I use TeraTerm.

I receive something else than 0x, but it seam to be the same value for any 
register...


Maybe I should check my power on my sensor chip.


FPGA bitstream stalls, it doesn't initialize if I power on my sensor PCB.


Fred


On Fri, Mar 9, 2018 at 4:37 PM, Oleg support-list@support.elphel.com 
wrote:
Hi,
 
Question: does the software needs to reach the  34VL02 EEPROM at any point?
Because I don't have one on my own sensor board...

 

It should work fine without eeprom.


Did it read chip id? or is it 0x? Do you use minicom or putty or other 
terminal program?


Regards,
Oleg

 

 


 ___ 
Support-list mailing list 
Support-list@support.elphel.com 
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com 






___ 
Support-list mailing list 
Support-list@support.elphel.com 
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com 












 ___ 
Support-list mailing list 
Support-list@support.elphel.com 
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com 





___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Elphel Support]

Fred,

Just to clarify: 10393 can drive up to ~1.0A, but there will be voltage drop on 
the cable (Elphel cables have thick power traces). Connector is rated 0.4A/pin. 
The sensor power is controlled by the script (same one that loads bitstream), 
power status/control is available over the sysfs interface.

You may also put a buffer on all signals from the sensor, FPGA has flexible 
means to compensate pin delays if needed after the buffer insertion.

Andrey

 On Fri, 09 Mar 2018 14:26:50 -0800 Elphel 
Supportsupport-list@support.elphel.com wrote  

Frederick, you need to power you board either directly from the 10393 3.3V on 
the cable, or (if you board has other components that need more power than 
10393 can provide) use external power switch driven by the cable power.

The problem is that the cable goes directly to the FPGA port that should be 
powered _before_ any non-zero voltage appears on the FPGA pins. This is a 
feature of the FPGA.

Andrey

 On Fri, 09 Mar 2018 13:47:51 -0800 Frederick Gamache 
frederick.gama...@gmail.com wrote  

I use TeraTerm.

I receive something else than 0x, but it seam to be the same value for any 
register...


Maybe I should check my power on my sensor chip.


FPGA bitstream stalls, it doesn't initialize if I power on my sensor PCB.


Fred


On Fri, Mar 9, 2018 at 4:37 PM, Oleg support-list@support.elphel.com 
wrote:
Hi,
 
Question: does the software needs to reach the  34VL02 EEPROM at any point?
Because I don't have one on my own sensor board...

 

It should work fine without eeprom.


Did it read chip id? or is it 0x? Do you use minicom or putty or other 
terminal program?


Regards,
Oleg

 

 


 ___ 
Support-list mailing list 
Support-list@support.elphel.com 
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com 






___ 
Support-list mailing list 
Support-list@support.elphel.com 
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com 





___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Elphel Support]

Frederick, you need to power you board either directly from the 10393 3.3V on 
the cable, or (if you board has other components that need more power than 
10393 can provide) use external power switch driven by the cable power.

The problem is that the cable goes directly to the FPGA port that should be 
powered _before_ any non-zero voltage appears on the FPGA pins. This is a 
feature of the FPGA.

Andrey

 On Fri, 09 Mar 2018 13:47:51 -0800 Frederick Gamache 
frederick.gama...@gmail.com wrote  

I use TeraTerm.

I receive something else than 0x, but it seam to be the same value for any 
register...


Maybe I should check my power on my sensor chip.


FPGA bitstream stalls, it doesn't initialize if I power on my sensor PCB.


Fred


On Fri, Mar 9, 2018 at 4:37 PM, Oleg support-list@support.elphel.com 
wrote:
Hi,
 
Question: does the software needs to reach the  34VL02 EEPROM at any point?
Because I don't have one on my own sensor board...

 

It should work fine without eeprom.


Did it read chip id? or is it 0x? Do you use minicom or putty or other 
terminal program?


Regards,
Oleg

 

 


 ___ 
Support-list mailing list 
Support-list@support.elphel.com 
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com 





___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Frederick Gamache]

I use TeraTerm.

I receive something else than 0x, but it seam to be the same value for
any register...

Maybe I should check my power on my sensor chip.

FPGA bitstream stalls, it doesn't initialize if I power on my sensor PCB.

Fred

On Fri, Mar 9, 2018 at 4:37 PM, Oleg 
wrote:

> Hi,
>
>
>> Question: does the software needs to reach the  34VL02 EEPROM at any
>> point?
>> Because I don't have one on my own sensor board...
>>
>
> It should work fine without eeprom.
>
> Did it read chip id? or is it 0x? Do you use minicom or putty or other
> terminal program?
>
> Regards,
> Oleg
>
___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Oleg]

Hi,


> Question: does the software needs to reach the  34VL02 EEPROM at any point?
> Because I don't have one on my own sensor board...
>

It should work fine without eeprom.

Did it read chip id? or is it 0x? Do you use minicom or putty or other
terminal program?

Regards,
Oleg
___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Oleg]

Frederick,

1. Here are the files: https://community.elphel.com/f
iles/393/20180308_test_0x5d/

2. Before copying - boot from the card and check the linux version:

> root@elphel393:~# uname -r
> 4.9.0-xilinx-v2017.3


* If it's 4.9.0-xilinx-v2017.3 then copying only devicetree.dtb and uImage
might be enough - they are on the FAT partition of the card.
* In case of fail if you use minicom (or putty or other serial terminal
program) there will be some kind of a kernel panic.

3.
My bootlog for 0x5d:

> ...
> [   21.692018] {0} Setting i2c_addr to 0x5d
> [   21.696039] Read i2c (port = 0, sa7=0x5d, reg=0x0) chip ID=
> ...
> [   21.747074] {1} Setting i2c_addr to 0x5d
> [   21.756495] Read i2c (port = 1, sa7=0x5d, reg=0x0) chip ID=
> ...
> [   21.807904] {2} Setting i2c_addr to 0x5d
> [   21.817282] Read i2c (port = 2, sa7=0x5d, reg=0x0) chip ID=
> ...
> [   21.868874] {3} Setting i2c_addr to 0x5d
> [   21.878291] Read i2c (port = 3, sa7=0x5d, reg=0x0) chip ID=

...


bootlog for 0x48 (2 sensors connected):

> ...

[   21.607985] {0} Setting i2c_addr to 0x48
> [   21.612007] Read i2c (port = 0, sa7=0x48, reg=0x0) chip ID=1801
> [   21.617888] Found MT9P001 2592x1944 sensor, chip ID=1801
> [   21.674977] {1} Setting i2c_addr to 0x48
> [   21.684454] Read i2c (port = 1, sa7=0x48, reg=0x0) chip ID=1801
> [   21.690690] Found MT9P001 2592x1944 sensor, chip ID=1801
> [   21.747575] {2} Setting i2c_addr to 0x48
> [   21.757033] Read i2c (port = 2, sa7=0x48, reg=0x0) chip ID=
> ...
> [   21.808291] {3} Setting i2c_addr to 0x48
> [   21.817738] Read i2c (port = 3, sa7=0x48, reg=0x0) chip ID=
> ...


Best regards,
Oleg Dzhimiev
Electronics Engineer
phone: +1 801 783  x124
Elphel, Inc.
___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Oleg]

Hi,


> Can you send me an updated version of the firmware using the 0x5D addr
> please?
>
Ok, I will send you a link once I prepare it later today. I will turn the
kernel debug on.

Will I be able to copy the files you will send me on the camera micro SD
> card so I can boot from it?
>
Do you have linux? There will be a script *write_bootable_mmc.py* in the
archive that can format it, provided it's discovered in the system as
/dev/sdX (if it's discovered as /dev/mmcblk the script will need a minor
modification)
The instructions how to format it manually can be found here
.

Best regards,
Oleg Dzhimiev
Electronics Engineer
phone: +1 801 783  x124 <(801)%20783->
Elphel, Inc.
___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Frederick Gamache]

Hello Oleg,

Thank you for your fast response.

I'm not able to build the firmware at the moment. Following the spec,
changing the SADDR pin to a HIGH value should translate the device ID to
0xBA, so *0x5D* as the 7-bit address.
I'm able to read the device registers on your sensor board using the
print_sensor_i2c command.

Can you send me an updated version of the firmware using the 0x5D addr
please?
Will I be able to copy the files you will send me on the camera micro SD
card so I can boot from it?

Best regards,
Frederick




On Wed, Mar 7, 2018 at 12:35 PM, Oleg 
wrote:

> Hi,
>
> 1.
> I made change so it could be set int device tree (0x48) - just noticed a
> bug and had to fix it:
>
>>  ...
>
> elphel393_sensor_i2c: elphel393-sensor-i2c@0 {
>> compatible = "elphel,elphel393-sensor-i2c-1.00";
>> /* Add known devices: name, slave address (7-bit), number
>> of address bytes, number of data bytes, SCL frequency (kHz) */
>> elphel393-sensor-i2c,i2c_devices = "mt9f0020x10
>> 2 2 500",
>>"mt9p006*0x48*
>> 1 2 500",
>>"el103590x08 1
>> 2 500",
>>"el10359_32 0x08 1
>> 4 500",
>>"pca9500_eeprom 0x50 1
>> 1 100",
>>"sensor_eeprom  0x50 1
>> 1 100",
>>"sensor_temp0x18 1
>> 2 100",
>>"cy223930x69 1
>> 1 100";
>> } ;
>
> ...
>
>
> Commands to recompile devicetree:
> decompile:
>
>> $ dtc -I dtb -O dts devicetree.dtb -o devicetree.dts
>
> compile:
>
>> $ dtc -I dts -O dtb devicetree.dts -o devicetree.dtb
>
>
>
> 2. Are you able to build the firmware or should I upload it to somewhere?
>
>
> 3. Before any updating - can you read sensor's registers?
>
>> ~$ ssh root@192.168.0.9
>> root@elphel393:~# cd /usr/local/verilog/
>> root@elphel393:~# test_mcntrl.py @hargs-after
>> x393 --> r
>> x393 --> print_sensor_i2c (without args it will print usage)
>>
>> *...*
>> *print_sensor_i2c *
>> *...*x393 --> print_sensor_i2c 0 0x3000 0xff *0x10* 0
>> *   Result: 0x2e01*
>
>
>  - 10393's port the sensor is connected to. Can be: 0..3
>  - I tried on a sensor that has 16-bit addresses (0x3000), for
> MT9P006 it can be 0x0 (ID register addr)
> * *- set 0xff
> * -* SADDR
>
>
> 4. Side note (sensor power and clock):
> At boot autocampars.php calls: test_mcntrl.py @hargs
>
> test_mcntrl.py (located at /usr/local/bin/)
> hargs (located at /usr/local/verilog/)
>
> Inside hargs file there are commands for:
> - which bitstream to load
> - setting sensor power: for MT9P006 it's PAR12 which translates to
> {"mv":2800, "freq":24.0,"iface":"2V5_LVDS"}
> - setting sensor clock frequency: for MT9P006 it's 24.0 "2V5_LVDS"
>
> Best regards,
> Oleg Dzhimiev
> Electronics Engineer
> phone: +1 801 783  x124 <(801)%20783->
> Elphel, Inc.
>
___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com

Re: [Elphel-support] Script for changing the SADDR bit for the MT9P031 sensor

[Oleg]

Hi,

1.
I made change so it could be set int device tree (0x48) - just noticed a
bug and had to fix it:

>  ...

elphel393_sensor_i2c: elphel393-sensor-i2c@0 {
> compatible = "elphel,elphel393-sensor-i2c-1.00";
> /* Add known devices: name, slave address (7-bit), number
> of address bytes, number of data bytes, SCL frequency (kHz) */
> elphel393-sensor-i2c,i2c_devices = "mt9f0020x10 2
> 2 500",
>"mt9p006*0x48*
> 1 2 500",
>"el103590x08 1
> 2 500",
>"el10359_32 0x08 1
> 4 500",
>"pca9500_eeprom 0x50 1
> 1 100",
>"sensor_eeprom  0x50 1
> 1 100",
>"sensor_temp0x18 1
> 2 100",
>"cy223930x69 1
> 1 100";
> } ;

...


Commands to recompile devicetree:
decompile:

> $ dtc -I dtb -O dts devicetree.dtb -o devicetree.dts

compile:

> $ dtc -I dts -O dtb devicetree.dts -o devicetree.dtb



2. Are you able to build the firmware or should I upload it to somewhere?


3. Before any updating - can you read sensor's registers?

> ~$ ssh root@192.168.0.9
> root@elphel393:~# cd /usr/local/verilog/
> root@elphel393:~# test_mcntrl.py @hargs-after
> x393 --> r
> x393 --> print_sensor_i2c (without args it will print usage)
>
> *...*
> *print_sensor_i2c *
> *...*x393 --> print_sensor_i2c 0 0x3000 0xff *0x10* 0
> *   Result: 0x2e01*


 - 10393's port the sensor is connected to. Can be: 0..3
 - I tried on a sensor that has 16-bit addresses (0x3000), for
MT9P006 it can be 0x0 (ID register addr)
* *- set 0xff
* -* SADDR


4. Side note (sensor power and clock):
At boot autocampars.php calls: test_mcntrl.py @hargs

test_mcntrl.py (located at /usr/local/bin/)
hargs (located at /usr/local/verilog/)

Inside hargs file there are commands for:
- which bitstream to load
- setting sensor power: for MT9P006 it's PAR12 which translates to
{"mv":2800, "freq":24.0,"iface":"2V5_LVDS"}
- setting sensor clock frequency: for MT9P006 it's 24.0 "2V5_LVDS"

Best regards,
Oleg Dzhimiev
Electronics Engineer
phone: +1 801 783  x124
Elphel, Inc.
___
Support-list mailing list
Support-list@support.elphel.com
http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com