Radiation realities

[ka9q]

The other week I attended a meeting with quite a few NASA people in attendance, both from JPL (which specializes in unmanned planetary exploration) and Johnson Space Flight Center (which handles manned space flight).

I asked some of the Johnson guys about some pictures of the ISS that were taken from a departing Soyuz while the last Shuttle was still docked. It seemed to me a real waste to leave the cameras in the Soyuz orbital module to burn up on re-entry. They explained that Soyuz down-mass was severely limited, which I already understood, but I wasn't really thinking of the value of the cameras themselves. Their real value came from their presence in space, given that launches even into low earth orbit typically cost tens of thousands of dollars per kilogram. Launching replacements greatly exceeds the cost of the cameras themselves.

Their answer was interesting: these cameras have to be periodically replaced anyway because their CCD sensors accumulate permanent damage from radiation hits in the South Atlantic Anomaly.

I thought this interesting because most people (including hoaxers) assume that humans are much more sensitive to radiation than inanimate hardware. This may be true for some kinds of hardware and for short exposures to high radiation rates, but radiation in manned space flight tends to be at low rates for long periods of time that give the body a chance to repair any damage. Unlike living things, semiconductors can't repair themselves so the damage steadily accumulates over time.

[Rob260259]

I heard that HST regularly is shut down. Is it because of the SAA?

[ka9q]

I've heard that too, but I don't know for sure. Sounds plausible.

[cjameshuff]

I heard that HST regularly is shut down. Is it because of the SAA?

Wikipedia says it is. I don't think it's to protect the hardware though...the damage occurs because atoms get kicked out of position in semiconductors whose properties depend on specific types and concentrations of defects in otherwise extremely regular crystal lattices. It could protect against corruption of data, however...fewer tasks running mean fewer targets in memory that can cause a chunk of code to go writing garbage over important buffers.

[Noldi400]

According to the document Instrument Science Report WFC3 ISR 2002 -01 from the HST site:

HST’s orbit intersects the “South Atlantic Anomaly (SAA)”, a region of high density and high intensity particle radiation. For HST instruments and all other low-earth-orbit satellites, the SAA threatens both instrument lifetime and exposure quality. Images are compromised by cosmic rays and high dark and background levels. Detectors exhibit both short and long term effects as a result of SAA interactions. Charge Transfer Efficiency (CTE) problems accelerate as an instrument’s exposure to radiation increases. Residual glow and high dark counts are measured after an SAA passage and the number of hot pixels may increase. The deposition of high energies into a material by a single particle can damage electronics, flip memory bits, compromise instrument operations and sometimes alter material properties. Therefore, the SAA contours and the associated operational procedures are both clearly and uniquely defined for each instrument onboard the HST.

Source: http://www.stsci.edu/hst/wfc3/documents/ISRs/2002/WFC3-2002-01.pdf

  ^{I don't know what all that means but you gotta admit it sounds bad.}

[ka9q]

There is one way that shutting down electronics before a high radiation event can help protect it. It's the phenomenon called "latch up". Depending on how an integrated circuit is fabricated, there can be stray "PNPN" paths between the power rails. The PNPN structure is the same used in the "silicon controlled rectifier" (SCR), a once-common high power semiconductor device that has since been largely replaced with more modern and flexible transistor designs like power MOSFETs and IGBTs.

The main property of the SCR is that it latches. Give it just a little pulse on its gate terminal and it turns on hard. The only way to turn it off is to remove power. Normally this pulse is an external electrical signal, and since the PNPN structure is not actually desired in most ICs there is no counterpart to the gate connection in an SCR. But triggering can also come from an energetic charged particle barreling through the PNPN structure. Since most integrated circuits are not designed to withstand direct shorts across their power supplies, they can burn up very quickly, usually starting with the fine gold wires used to bond the connecting pads on the chip die to the pins of the package.

Electronics for space applications can be designed to be latchup-free, but it's not always possible to find the part you want in this flavor; you may be forced to use an off-the-shelf commercial part. In that case, you can pre-emptively shut down the circuit before you enter a high radiation environment like the SAA, and/or you can implement a "latchup detector", like a fast-acting circuit breaker, that detects the spike in supply current and interrupts it before it can do damage.

[cjameshuff]

Ah, yes, forgot about latchup...it could well be to protect the hardware. Perhaps especially in scientific instrumentation where parts are already highly specialized, often with only a single source...