Not really my forté but I’d worry that a higher current load switch for two devices would allow a higher current due to a SEU on a single device. Too much current and device fails before the load switch cuts it off?
On Wed, Jan 25, 2023 at 11:37 AM Jonathan Brandenburg via pacsat-dev < [email protected]> wrote:
On 1/19/23 17:40, Rich Gopstein via pacsat-dev wrote:
Burns might remember the details (or Zach certainly would), but one of those parts was undersized for the RT-IHU and would cut out when we were transmitting with the ax5043. I think we had to back down the xmit power in order for the device not to reset.
It didn't take a whole lot of research for me to figure this one out. The FPF2001 used on the Fox boards has a fixed current limit of 50mA. I can see that could be sufficient for the Fox IHU since it uses a low power processor and doesn't have transceivers or the power amplifier. The direct replacement part for the FPF2001 is the TI TPS22942 and it appears to be pin compatible with a fixed limit of 100mA. Would a 100mA be a problem on Fox? I wouldn't think it would cause failure in the 10ms it takes to blank the current but I make that statement without any real basis. Perhaps the 40mA limit of the TI TPS22942 would be sufficient.
Anyway, back to PACSAT...
With the higher current capacity of the MAX4995, adjustable between 0 and 600 mA, it could be the better choice than the fixed TI TPS22942. I'm not aware of flight experience with either so I tend to look to the MAX4995 with the adjustable and higher current limits. With this decision and looking at the approach employed on the RT-IHU I could see the following:
1 - The TMS570 and associated circuitry get their own MAX4995. As a first order pass contained in Bob Stricklin's spreadsheet the current requirement is 458.3mA, within the MAX4995's capabilities. If we want to further "derate" we could put the TMS570 on one MAX4995 and have a second MAX4995 for the peripherals.
2 - All five AX5043's require 100mA. Would we want to have 2 of the AX5043's on one MAX4995 and the remaining AX5043's on the other MAX4995 in case one MAX4995 fails prematurely? The value of this would depend on all the AX5043's being equivalent in function (i.e., all could be used for command and/or telemetry). This only takes care of this particular failure mode and still leaves a lot of other failure modes in which an early MAX4995 death would leave the system INOP so there may be limited value in using two MAX4995's for the AX5043's.
3 - Finally, the amplifier/pre-amp is an additional 415mA. Again, if we want to "derate" we could put the power amp on it's own MAX4995.
It looks like we need a minimum of 3 MAX4995's and a possible max of 6 MAX4995's. (With further analysis, we might need one more to split the 1.2V and 3.3V power lines to the TMS570.) The MAX4995 are small chips so don't require a lot of real estate. We can discuss the risk vs reward of using more MAX4995's at some point.
I welcome comments
Jonathan
-- Jonathan Brandenburg Radio Amateur Satellite Corporation 1-214-213-1066
pacsat-dev mailing list -- [email protected] View archives of this mailing list at https://mailman.amsat.org/hyperkitty/list/[email protected] To unsubscribe send an email to [email protected] Manage all of your AMSAT-NA mailing list preferences at https://mailman.amsat.org
Now that you mention it, not my forté either! I was working in the abstract without really thinking about what a latchup actually results in… I think I was just assuming “super high” current without any real idea of what that really means.
I conceptually like the idea of each “subsystem” having its own current limiter that is tuned to that particular device. Takes a bit of real estate but these are small chips. Jonathan
Sent from my iPad
On Jan 24, 2023, at 5:43 PM, Burns Fisher (AMSAT) [email protected] wrote:
Not really my forté but I’d worry that a higher current load switch for two devices would allow a higher current due to a SEU on a single device. Too much current and device fails before the load switch cuts it off?
On Wed, Jan 25, 2023 at 11:37 AM Jonathan Brandenburg via pacsat-dev <[email protected]mailto:[email protected]> wrote: On 1/19/23 17:40, Rich Gopstein via pacsat-dev wrote:
Burns might remember the details (or Zach certainly would), but one of those parts was undersized for the RT-IHU and would cut out when we were transmitting with the ax5043. I think we had to back down the xmit power in order for the device not to reset.
It didn't take a whole lot of research for me to figure this one out. The FPF2001 used on the Fox boards has a fixed current limit of 50mA. I can see that could be sufficient for the Fox IHU since it uses a low power processor and doesn't have transceivers or the power amplifier. The direct replacement part for the FPF2001 is the TI TPS22942 and it appears to be pin compatible with a fixed limit of 100mA. Would a 100mA be a problem on Fox? I wouldn't think it would cause failure in the 10ms it takes to blank the current but I make that statement without any real basis. Perhaps the 40mA limit of the TI TPS22942 would be sufficient.
Anyway, back to PACSAT...
With the higher current capacity of the MAX4995, adjustable between 0 and 600 mA, it could be the better choice than the fixed TI TPS22942. I'm not aware of flight experience with either so I tend to look to the MAX4995 with the adjustable and higher current limits. With this decision and looking at the approach employed on the RT-IHU I could see the following:
1 - The TMS570 and associated circuitry get their own MAX4995. As a first order pass contained in Bob Stricklin's spreadsheet the current requirement is 458.3mA, within the MAX4995's capabilities. If we want to further "derate" we could put the TMS570 on one MAX4995 and have a second MAX4995 for the peripherals.
2 - All five AX5043's require 100mA. Would we want to have 2 of the AX5043's on one MAX4995 and the remaining AX5043's on the other MAX4995 in case one MAX4995 fails prematurely? The value of this would depend on all the AX5043's being equivalent in function (i.e., all could be used for command and/or telemetry). This only takes care of this particular failure mode and still leaves a lot of other failure modes in which an early MAX4995 death would leave the system INOP so there may be limited value in using two MAX4995's for the AX5043's.
3 - Finally, the amplifier/pre-amp is an additional 415mA. Again, if we want to "derate" we could put the power amp on it's own MAX4995.
It looks like we need a minimum of 3 MAX4995's and a possible max of 6 MAX4995's. (With further analysis, we might need one more to split the 1.2V and 3.3V power lines to the TMS570.) The MAX4995 are small chips so don't require a lot of real estate. We can discuss the risk vs reward of using more MAX4995's at some point.
I welcome comments
Jonathan
-- Jonathan Brandenburg Radio Amateur Satellite Corporation 1-214-213-1066
-----------------------------------------------------------
pacsat-dev mailing list -- [email protected]mailto:[email protected] View archives of this mailing list at https://mailman.amsat.org/hyperkitty/list/[email protected] To unsubscribe send an email to [email protected]mailto:[email protected] Manage all of your AMSAT-NA mailing list preferences at https://mailman.amsat.org -- 73,
Burns Fisher, WB1FJ AMSAT(R) Engineering -- Flight Software
-
Burns Fisher (AMSAT) -
Jonathan Brandenburg (KF5IDY)