A project I am working on requires a reliable current limiting solution for a pocketqube picosatellite.

Currently a Texas TPS2553 is being employed which will cut power once the limit is hit. We are concerned that this could enter the power supply into a never ending reset loop due to damage from radiation.

The maximum expected current, during a signal transmission is 135mA. The sleep current is 7uA.

We have come across depletion mode p-channel MOSFETS such as the Infineion BSS126 (application notes) which shows:

Infineon BSS126 Application Notes

1) Is there a more reliable solution than the MOSFET Current limiter shown above?

2) Do these depletion mode p-ch devices have particularly high or low resistance to cosmic radiation?

  • \$\begingroup\$ What is your load? \$\endgroup\$ – Voltage Spike Feb 28 '19 at 15:54
  • \$\begingroup\$ The ATMEGA328 and SX1278 are the largest loads. \$\endgroup\$ – MIL-SPEC Feb 28 '19 at 15:57
  • \$\begingroup\$ You should very seriously contact others who have developed pocketcube satellites or those who fully understand the environment. You haven't said exactly where this is going (LEO or ??) or how long you expect it to operate. Environment matters. Large feature size devices (like the 4000 series) may be helpful. But you should contact Goddard Space Flight Center and see if they can put you in touch with scientists and engineers who may give you valuable pointers to consider. Ask them about NASA's open source core Flight System, too. \$\endgroup\$ – jonk Mar 1 '19 at 3:26
  • \$\begingroup\$ @jonk Target 385km SSO. We will pass through the South Atlantic Anomaly a few times per day, which is expected to be the greatest threat. Flight time is estimated a 11 months - although our solar panels are not expected to survive that long. We will contact Goddard and see what they say - thanks. \$\endgroup\$ – MIL-SPEC Mar 1 '19 at 8:04
  • \$\begingroup\$ Do you have a battery? \$\endgroup\$ – Voltage Spike Mar 4 '19 at 20:49

1) Is there a more reliable solution than this?

A part like this could become a problem with a Single Event Upset (SEU), it isn't rad-hard (resistant to radiation). If the SEU causes a problem in the circuitry of the TPS2553, then it could cause power issues for your system.

It really depends on the system and how this is connected as a whole if there is something upstream that can shut the power off (which there probably is because you have a battery and it needs a DC to DC converter).

With latch ups and satellite design it's beneficial to have:

  • A system that can detect latch ups (over current) and shut down that IC. This means monitoring the current from the battery and solar cells (if you have that in your mass/volume/power budgets). A watchdog timer (external to the microprocessor) is usually employed that needs to get updated from the processor, if it doesn't get updated, then it switches everything off and resets the system power (and the TPS2553). The main thing is preventing the battery from draining to a critically low voltage, if you get the voltage too low, the battery dies (with most chemistries) and the mission is only working in the sun or dead.
  • Radation hardened parts, if you really want a robust system, then for the power circuitry use rad hard parts which are increasing in availability (and also are more expensive). Generally pico and nano sats use commercial parts and deal with the fact that they could latch up more often since the mission budgets are lower.
  • Use a flash based FPGA or anti fuse for mission critical circuitry

    In contrast, the configuration of antifuse and flash-based FPGAs is immune to SEUs because of their non-volatile structure. Figure 4 shows the typical flash structure with a floating gate located between a control gate and the metal-oxide semiconductor field-effect transistor (MOSFET) structure below, encased in good dialectic. The bit value is stored as a charge on the floating gate. A charged gate represents a zero value for NOR flash cells. Source: http://archive.cotsjournalonline.com/articles/view/102279

Interestingly enough, TI has a recent write up on latch ups and how they are prevented in IC's. It is not known if the TPS2553 has been built and tested to the JEDEC JESD78 standard for latch ups. (and good luck finding out, over the years TI has been poor at communicating with engineers, so I only use them if I absolutely have to). But since 2010 TI has some designs conform to this standard, and mitigating latch ups. Below shows a design that is more immune to latch ups with guardlines

enter image description here

2) Do these depletion mode p-ch devices have particularly high or low resistance to cosmic radiation?

The problem is mostly related to having pairs of transistors (which the TPS2553 has many):

Latch-Up is a condition where a low impedance path is created between a supply pin and ground. This condition is caused by a trigger (current injection or overvoltage), but once activated, the low impedance path remains even after the trigger is no longer present. This low impedance path may cause system upset or catastrophic damage due to excessive current levels. The Latch-Up condition typically requires a power cycle to eliminate the low impedance path. CMOS and BiCMOS circuits use NMOS and PMOS transistors to create the circuit functions. In the design of the CMOS integrated circuit, the proximity of the PN junctions that form the NMOS and PMOS transistors create inherent parasitic transistors and diodes. These parasitic structures create PNPN Thyristors, also called silicon-controlled rectifiers (SCRs). Excursions (overshoots and undershoots) outside the normal operating voltage and current levels can trigger PNPN Thyristors and may cause Latch-Up. Latch-Up is not a risk if the voltage and current levels applied to the device adhere to the absolute maximum ratings. Source: Latch ups SCAA124 TI app note

So it would be more concerning with the driving circuitry of the TPS2553 and having something latch up there and either not being able to control it, or having it self destruct from overcurrent (For example the TPS2553 has an error amplifier (op amp) and many other places which has an output stage that has two transistors in series from VCC to ground. If both of these turn on then you have a short essentially, if the TPS is the thing that is suppoesed to be monitoring current then you have a problem. I would probably go with a circuit that is not integrated with a single high side switch for the load with a pull up and a watchdog circuit on it for resets.


1) Is there a more reliable solution than the MOSFET Current limiter shown above?

A single switch with a watchdog would be better. Odds are the system being designed doesn't have power to waste. Putting a resistor in series with one of the largest loads in the system is not a good idea. This paper has information on the power systems for a nanosat, it would be a good idea to use (or even contact the authors) for information on how they designed their power systems

enter image description here Source: Nanosatellites in LEO and beyond: Advanced Radiation protection techniques for COTS-based spacecraft

2) Do these depletion mode p-ch devices have particularly high or low resistance to cosmic radiation?

This is unknown, but it doesn't matter what the device is, it will most likely latch up. The way to design spacecraft power systems (or any electronic system) is to plan on the device being fully off or fully on from radiation and then allowing the system to clear the fault (by power cycling). No devices is resistant to latch ups completely but some more than others.

Device resistance is determined by many factors including the materials and construction. The only way one really knows is if the device has been designed and tested against radiation. If you want radiation tolerance, then buy rad hard parts like these MOSFETS. But you still have to plan for latchups due to SEU's. There are some very very high energy cosmic rays out there, and they can ionize anything.

| improve this answer | |
  • \$\begingroup\$ I have edited my question as you interpreted it wrong and mostly wrote a great answer about the TPS, when I actually wanted an answer about the MOSFET solution. Not your fault. This is a single junction - so should it not be less vulnerable than the TPS? \$\endgroup\$ – MIL-SPEC Mar 5 '19 at 9:16
  • \$\begingroup\$ Yes a single fet will be more reliable. Current limiting is not a great idea for a satellite. A high side switch would be best, one that is fully on, and a way to reset it after a latch up has been detected \$\endgroup\$ – Voltage Spike Mar 7 '19 at 4:09
  • \$\begingroup\$ Thank you for the helpful edit. +50 awarded \$\endgroup\$ – MIL-SPEC Mar 7 '19 at 10:55

If you can afford a few volts of insertion loss, a simple and reliable current limit with LM117: SG117 current source The resistor should be fixed 8.2R (for ~152mA limit) or 9.1R (for ~137mA limit), and able to dissipate at least 0.25W even at the highest operational temperature. If the radiation is extreme, you can use a radhard version of LM117, like SG117AR. https://www.microsemi.com/existing-parts/parts/56283

| improve this answer | |
  • 1
    \$\begingroup\$ A linear regulator is undesirable in satellite applications due to losses \$\endgroup\$ – Voltage Spike Mar 5 '19 at 19:55
  • \$\begingroup\$ As laptop2d says, we cannot use such an inefficient device, sorry. \$\endgroup\$ – MIL-SPEC Mar 7 '19 at 10:53
  • \$\begingroup\$ A low Vadj radhard LDO is more efficient and stable than the mosfet solution. \$\endgroup\$ – g77 Mar 8 '19 at 11:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.