I saw the word "shunts" in this answer about DSCOVR's upgraded photovoltaic arrays when it was unmothballed and readied for space. Reading it, I get the idea that shunts are used to dissipate the difference in power between what the satellite needs and what the array is producing.

In this answer which is actually about silicon PVs on the ISS, it is mentioned that when entire arrays are offline, they get hotter, and I'm assuming this is because they are left open-circuit and so all the electron-hole carriers remain in the junction and recombine, returning their energy as heat.

And I'm also wondering that if you need less than the full amount of energy, if you had a modern switching power supply to condition the voltage, wouldn't you just draw less current, instead of full current and shunting some into a big hot resistor?

So to narrow the question down I should ask about modern designs and recently built and launched satellites. I'm thinking communications satellites, but since DSCOVR received a modernized system, that's OK too. Of course if there is something more compelling, that's fine, but the ISS can be a separate question because there are issues of legacy and compatibility.

  • 1
    $\begingroup$ Third option - rotate them to catch less sunlight? This would keep them cooler. $\endgroup$ – MSalters Jul 27 '16 at 23:00

I wasn't sure if your last sentence meant that you didn't want to hear about the ISS, but in case you do, in the case of the ISS, they are shunted. The device that handles this is the Sequential Shunt Unit (SSU). The SSU is the primary power regulation device that controls Solar Array Wing (SAW) output. There is one SSU per power channel/SAW. The function of the SSU is to manage the output of the solar cells to produce 160 V dc. The SSU maintains this voltage by balancing the system demand with the number of connected array strings in the SAW. Each array string can be individually connected or disconnected from the primary power bus, thus the output from the SSU is the sum of all connected strings at any given time. Array strings that are disconnected from the power system are shunted (shorted back to the array itself).

A high level graphic from here.

enter image description here

For what it's worth, the SSU on power channel 3A shorted out in May 2014 leading to the loss of its associated power channel (1 of 8). It was replaced during an EVA in October 2014.

  • $\begingroup$ Sure I love to hear about it! I should have asked a separate question for this, because the ISS power system is unique in its hugeneness and complexity and assembly over time. I had a hunch there were going to be shunts here. I'm wondering if shunts are less frequent with newer, high power satellites. $\endgroup$ – uhoh Jul 27 '16 at 14:58
  • $\begingroup$ So can you reconcile or help me sort out the idea mentioned here that when some panels are totally unused, they get hotter? Because it seems to me that it suggests they would have to be left open circuit for that to happen. Are shunts used only when power is needed but there is too much, and otherwise they are left open circuit? $\endgroup$ – uhoh Jul 27 '16 at 15:08
  • $\begingroup$ If memory serves I don't think "open circuit" is an option in the SSU. The high level descriptions available online in documents like this aren't a lot of help. google.com/… $\endgroup$ – Organic Marble Jul 27 '16 at 15:25
  • $\begingroup$ @OrganicMarble is correct. On a relatively high voltage bus like what the ISS uses, open circuit is a dangerous condition, especially for crew on EVA. Keep in mind that solar cells are current sources, not voltage sources, so (within limits) they will put out as much voltage as they can to deliver a set current. To render them safe, you shunt them so that they can deliver that current at zero voltage, hence zero power. The running hotter part is obvious from a bulk energy balance. Unfortunately, I'm not versed enough in semiconductor micromechanics to explain it at the micro level. $\endgroup$ – Tristan Jul 27 '16 at 15:43
  • $\begingroup$ @Tristan Your reasoning is not correct - the open circuit voltage of a solar cell is only slightly higher than at the optimal working point, 10-20% higher. Shorting them is a good idea nevertheless. $\endgroup$ – asdfex Jul 27 '16 at 17:41

This is just a short answer addressing the headline question. I apologise for the short answer, I don't have time at the moment to dig out references or a full explanation as the answer comes from experience.

Both shunt regulators and series regulators (open circuit) exist in unmanned satellite power system design. I think that series regulators are more common for smaller satellites though I don't recall the design rationale.


  1. Rotation

Whilst I haven't heard of a real design example, the "rotate to catch less sunlight" would improve cooling but would presumably have to be used in conjunction with one of the regulators in order to allow the voltage to remain stable to changing loads.

  1. Benefits

One might go a step further and look into whether we care much about reducing the array temperature. A cooler array is usually more efficient, however this question came up in attempting to understand the oversupply situation and so efficiency is irrelevant. A significantly higher prolonged temperature could in principle lead to increased defects and failures though I haven't seen this cited as a degradation factor for typical array temperatures.

  1. Thermal IR

The video of the shuttle leaving the ISS that is linked to from one of the links in the question is curious. Its really interesting to see that the arrays appear to have a striped pattern. This might indicate sections working at different temperatures on account of regulator action. This could occur in a series regulator but this is inconsistent with the assertion of a shunt regulator from Organic Marble's answer. Of course we don't know where the shunt loads are. Its possible that if the shunt is a direct short then the cells themselves are the shunt load and unused sections could still run hotter than connected sections. On some satellites, perhaps where the dimensions are lesser, shunt loads can provide useful thermal maintenance for other parts of the satellite.

  • $\begingroup$ I appreciate your answer, thanks! Trying to understand the issues and what actually is going on with these systems - this gives more to think about so it's helpful. I don't know if array temperature changes induced electrically were meant to have some purpose, or if it were just an effect that could be seen in thermal imaging as just a way to illustrate an effect in an answer. $\endgroup$ – uhoh Jul 27 '16 at 23:39
  • 1
    $\begingroup$ As far as the design of an array is concerned I think its the latter, recall that the cooling occurs when the string is open circuit and thus its cooling benefit is of no use except possibly by lateral thermal conduction from an adjacent string that is still connected to a load. In this case though it seems like a long way around to get a benefit that could be more easily obtained by keeping all strings in circuit in the first place! $\endgroup$ – Puffin Jul 28 '16 at 10:58
  • $\begingroup$ In the infrared video, the "stripes" may possibly be because of the hinges between the panels. These hinges are aluminum with a stainless steel hinge pin and would certainly have different thermal characteristics from the panels themselves. You can see the hinges at work as the panels unfold in this video of an array being deployed. Around 2:30 into the video there is a nice closeup. youtube.com/watch?v=XRXbi3sQKWc $\endgroup$ – Organic Marble Jul 28 '16 at 17:28

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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