BAE Systems has completed the first phase of a DARPA programme aimed at removing heat from radio frequency electronics and has been funded to continue into a second phase, work the company says could nearly triple the range of RF systems, the company stated.
The programme is called Technologies for Heat Removal in Electronics at the Device Scale, or THREADS, and is run by the Defense Advanced Research Projects Agency. It targets the temperature limits that constrain gallium nitride devices, the semiconductors that sit behind most modern military radar and electronic warfare transmitters.
Heat is what caps the output of those systems. A gallium nitride transmitter converts electrical power into radio energy and loses a proportion of it as heat within a device measured in fractions of a millimetre, and once the junction temperature rises beyond a certain point the device degrades or fails. Designers have therefore run transmitters below their theoretical maximum output, because the alternative is a shorter service life. Extracting heat closer to where it is generated would allow more power out of the same device.
Isaac Wildeson, principal investigator at BAE Systems’ FAST Labs, said: “We look forward to advancing to Phase 2 of the THREADS program. The progress we’ve made during Phase 1 validates our approach to material and process enhancements and brings us closer to unlocking the full potential of RF-based systems for our warfighters.”
According to BAE Systems, successful thermal management developed through the programme would nearly triple the range of RF systems, improving safety and engagement distances for military personnel.
Radar range scales poorly with transmitter power, detection range rises with the fourth root of the power radiated, so doubling the range of a radar requires something in the order of sixteen times the transmitted power, which is why incremental gains at device level matter more than they might appear to. The company has not set out how the range figure it cites was derived.
Work is being carried out at the company’s Microelectronics Center in Nashua, New Hampshire, a facility accredited as a Category 1A Trusted Supplier, which builds gallium nitride and gallium arsenide integrated circuits. Partners on the programme include Modern Microsystems, Penn State University, Stanford University, the University of Notre Dame and the University of Texas at Dallas.
Gallium nitride has displaced older gallium arsenide technology across radar and electronic warfare over the past fifteen years because it handles higher voltages and power densities, allowing smaller and more capable arrays. British systems using it include the Sampson radar on the Type 45 destroyers and the European Common Radar System Mark 2 being fitted to RAF Typhoons, while the United States has moved its Patriot and Aegis radars onto the material.
“detection range rises with the fourth root of the power radiated, so doubling the range of a radar requires something in the order of sixteen times the transmitted power, which is why incremental gains at device level matter more than they might appear to. The company has not set out how the range figure it cites was derived.”
That is something a basic radar textbook would tell you.
Detection range rises with the third root of the power radiated – that is the inverse cube relationship which is schoolboy physics and is simply energy dispersed over volume. This is unfixable as it is a physical reality. IRL there is a difference between full volume scanning which is fully 3rd power and more pinpoint methods which are less wasteful of brute RF power.
The fourth power if the dissipation in the junctions. Obvs going from 3rd-4th power is a very big deal in energy terms.
There are three main impacts of heat on Gallium Nitride radars.
Thermal Throttling & Power Derating
To protect transmit/receive (T/R) modules from permanent damage or premature failure, radar control units use thermal management algorithms to automatically dial back RF output power (derating) or reduce the radar’s operational duty cycle.
Because maximum radar range is governed by the radar range equation. Any thermal throttling that forces a drop in average transmitted power directly translates into a shorter detection envelope.
Elevated Noise Floor
Detection sensitivity depends on the ratio of target signal to background thermal noise.
Reduced Efficiency & Gain Compression
As the junction temperature of a GaN High Electron Mobility Transistor (HEMT) climbs:
Electron mobility and saturation velocity decrease.
The amplifier suffers from gain compression, causing a higher percentage of input DC power to convert into waste heat rather than radiated RF energy, triggering a feedback loop of thermal loss.
Removal or heavy control of all three factors can result in significant gains however obviously anything a defence contract says on these documents will exaggerate real world performance.
This is more of your crap – it’s pure AI. If you have to run google AI to post here, you really are a useless little faggot
How about you shut the fuck up with your insults. To counter an argument is fine, to call someone out is fine but the insults make you look like a completely deranged bell end and it’s unnecessary.
Not been the same since you gave up tennis have you.
I don’t understand that. Surely a third root relationship would mean a huge increase in power output is required to increase range?
Up to the third power is well understood as the inverse cubed relationship.
The fourth power is the energy dissipation. The thing is that the fourth power is more important as it is a cliff edged limit to what the radar can do.
No, detection range does not rise with the fourth root of the power radiated. The transmitted signal range does, detection range is something completely different, as the detection range is dependent on the wavelength being used, atmospheric attenuation, the target’s radar cross section, the receiver’s sensitivity, the gain of the antenna, how good the filtering and amplification chain in the receiver is at removing received signal noise (clutter) and not inducing its own noise. But also today the development of the software used to analyse the received signal, is also a critical part of the “detection range” chain, that when added together, determines how far a target can be detected. You may very well significantly increase the signal’s transmitted range. But if the receiver can’t detect the minute returning signal over the received noise, what’s the point? Both the Transmitter and Receiver must be at least balanced in performance, where in reality, you want the receiver to have the better performance.
Electronic component cooling has been a thing since the 1940’s. Where from the 1960’s cryogenic cooling was shown to be the answer. Which works fine in a lab, not so much in the field. As you don’t have the space to put all the gubins needed to continuously run components at temperatures below -100C. Which proved to be completely unfeasible when mounting a cryogenic system to a fighter jet. From the 1990’s immersing circuit boards in a non-conductive oil bath along with its associated cooling gubins, was made compact enough, that amplification boards in particular could be cooled a lot more effectively than using forced air cooling. But even then it was pretty heavy and bulky. A lot of work has been done exploring Peltier based cooling. Which are terribly power hungry devices, but can cool circuits down to well below zero C and are significantly lighter that the oil cooled systems. Be interesting to see if this is the route BAES have gone down?
Yes, the build up of heat in a component can and does shorten its lifespan. But heat build up also significantly affects the amount of noise a component generates. This has been the main problem with using Galium Arsenide (GaAs) components, particular for power amps. As these components cannot handle long periods of generating high power. As they get hotter, their ability to generate power degrades, but conversely the amount of self-induced noise generated increases. Newer Gallium Nitride components can be used to generate high power outputs for much longer before degrading, plus they don’t generate as much noise. Therefore signal filtering becomes much easier.
The radar range equation is only part of the story, as it primarily deals with an ideal model, For example, it doesn’t really cover what is going on within the radar, only expecting the radar to have losses and not gains. Which was kind of fine for the older legacy radars, but misses a lot of context with digital based radars.
It was a discussion of the odd quote from BAE?
And yes, thermally driven junction noise has always and will always be a thing.
I’d not be so negative about cryogenic cooling as I was doing that 25yrs ago with spectacular results.
Very interesting Davey. Thanks.
@DaveyB
An excellent and knowlegeable post, accurate and informative. Very useful, tnx
Yes, a top read. Thanks DB.
@DaveyB,
Thank you for your insight and edification, it always helps having things explained.
Appreciated.
(thank you David Loyd for the ‘@‘ I’ll be using that in future … if i remember.)
I don’t think they’ve actually cited a range figure have they? The ‘4th power’ bit is presumably just based on the simple observation that signals attenuate in accordance with the inverse square law, and a reflected signal from a typical target also tends to attenuate the same way, but that’s obviously a rather simplistic assertion that doesn’t account for the particular characteristics of the transmitter or target.
That is the thing.
It might as well be a response to
‘Ask our favourite AI agent for something mangles but true about radar’
For someone like me, with limited knowledge/experience in this field, I have learned a lot from your comments. Fascinating. This is the kind of thread that I will reread a few times to fully understand. Thank you for your clear explanations!
👍
Great SME knowledge from Davey, as always.
So we can finally match Russia and china just at forty times the cost.
Bae is not our friend nor is it British
Where did the B come from then 😂
One of the conditions for BAE becoming a major US government supplier was that the US arm is completely siloed from the rest of BAE. No IP created by BAE Systems Inc (the US arm) is allowed to be used by BAE Systems without US government approval (and I’d assume financial kickbacks). BAE Systems Inc is British in name only. It’s a US firm, headquartered in the US, with a US board that just happens to be owned by a foreign company and the only flow of information to the parent is financial.
Sums it up pretty perfectly, great limitations on what even the board here is allowed to know, just hope the financial aspect make it all worthwhile otherwise investment here suffers even more than we know of it.
@PaulB2,
BAE Systems Inc (the US arm) is owned 100% by BAE Systems plc and any net profits from BAE Systems Inc (the US arm) go to BAE Systems plc.
Following the 1999 merger that created BAE Systems plc in the UK, BAE ‘realised’ that the British domestic defense market was too small for its growth ambitions. They launched an aggressive buying spree in America, acquiring major US defense contractors like United Defense; makers of the Bradley Fighting Vehicle. The Pentagon soon sounded the alarm. A foreign company was rapidly buying up vital pieces of the US military supply chain, and under American law; foreign ownership, control, or influence (FOCI) over domestic weapons manufacturing is strictly illegal unless mitigated.
To mitigate and close these multi billion-dollar deals, BAE Systems signed a Special Security Agreement (SSA) with the US government. The absolute core condition of the SSA was the creation of a completely separate, firewalled entity … BAE Systems, Inc.
To prove this firewall was real, the US government required the creation of an autonomous, all-American board of directors that we see today. It was the only legal compromise that allowed the British parent company to own the assets while ensuring the US military kept absolute control over the technology.
This is the exact legal playbook that ALL foreign companies must follow if they want to own sensitive US defense contracts /contractors. The US government uses a strict regulatory framework managed by the Committee on Foreign Investment in the United States (CFIUS) and the Defense Counterintelligence and Security Agency (DCSA).
Any foreign company buying into the US defense sector faces a spectrum of strict security arrangements depending on what technology they are buying.
Obviously, BAE Systems plc does not operate the US arm out of charity or an unusual desire to support American foreign policy. They went through the massive administrative headache of setting up an all American board; dealing with the Pentagon’s strict firewalls and enduring intrusive ‘Special Compliance Officers’ for one reason … the US defense market is a f’n goldmine. The US arm of BAE plc generates roughly 40% to 50% of the entire global group’s revenue. Without those American profits, BAE Systems plc would be a fraction of the corporate giant it is today.
Thus, I repeat, BAE Systems Inc (the US arm) is owned 100% by BAE Systems plc and net profits from BAE Systems Inc (the US arm) after paying ‘All US Fees’ go to BAE Systems plc in the UK … let’s be honest about BAE Systems plc’s raison d’être is to generate profit / money.
The dividends flow this way as well, of course, Paul.
Detect the lack of background signal, not the signal return.
Nothing new in this technology, unless l am missing something.
What there do is basically stick the whole thing in a fridge. What will happened as well is, the life of the electronics will be extended before it fails. Heat shortens life of electronics. Anther advantage of cooling.
I found the article a little misleading. It seems to claim that range increases are due to an increase in max power and also imply that they weren’t: “incremental gains at device level matter more than they might appear to”. For triple the radar range you’d need about an eightyfold increase in max power, which I found hard to believe was available from cooling.
On delving further, it turns out that the gains are indeed due to increases in max power, but not from cooling. It’s through not generating the heat in the first place. THREADS is about increasing the conductivity of the elements near the junctions inside the transistors, partly through a change in substrate and partly through reducing resistance at the transition layers. By increasing effective conductivity eightfold in X-band elements, thermal management is similarly boosted. That’s nearly a threefold increase in range, but not of radar. It applies to one-way applications like communications and weaponry where distances scale with the square root of power. It still gives a very nice boost to radar distances, but not nearly threefold.
I can almost hear Davey B wanting to chip in on atmospheric degradation in X-Band signals, but I’m ignoring that and I expect the BAE spokesperson was too. Apparently they are currently testing at fivefold power increases, with eightfold their target. So still a work in progress; a very useful one.
Oops. I misread heat conductivity as electrical. It is cooling after all. The diamond substrate conducts the energy away using phonons. Apologies. I was wondering how the semiconductors worked. The rest of my post was right, though. It’s an eightfold increase in power, from an eightfold increase in thermal conductivity.
So given that this work is sponsored by DARPA and carried out by BAE Systems US this technology will most likely be covered by US ITAR rules. As such will the we have access to this technology for developing our own radars in the future? Also, given we appear to be looking to buy the Australian SEAFAR radars, which by all accounts seem to be pretty good and available, what is the future of UK radar development..?
I do hope that we do not neglect our radar industry, it would be a serious mistake given the shifting global geopolitical situation we face.
Cheer CR
What makes you think ISANKE, the GCAP “radar” doesn’t already have this? It’s an area the UK is already very strong in. Maybe it’s DARPA playing catch up?
Just asking the question, mate. May be the ISANKE system does have this or similar tech or perhaps it achieves it’s cooling using a different approach but I doubt we will find out anytime soon. I am vaguely aware that UK radar research is on-going in a number of areas and that cooling is one area being looked into, but with the moves to procuring the SEAFAR radar is there sufficient service demand to justify investment beyond Tempest which obviously is a trinational program.
I also note that the Tempest system can apparently makes use of off platform distributed transmitters and also has transmit and receive elements spread around the airframe so it not relying on a nose mounted radar. It all sounds very clever and if it works well enough it will be a significant step forward for the nations involved.
Of course, details are hard to come by.
I suppose my point was more about cross over between BAE System US and BAE System UK. There will be a firewall between the two divisions so could similar tech have been developed in parallel and what are the political implications, if any..?
Roll on Tempest.
Cheers CR
GCAP is fundamentally ITAR free, which might mean no US tech, or it might mean none that isn’t immediately substitutable. My guess is it’s the former. I think the US is likely to try to nobble GCAP if it can. Could the same tech have been developed in parallel? Sure. Why not? Japan has been into GaN for as long as the US and is good on material science. The UK has a diamond substrate manufacturing plant (De Beers?), and a lot of the fundamental thermoreflectance work comes out of Bristol University, which has had years of work on GaN on diamond. Both countries are strong on RF engineering.
I’d guess the underlying material science is probably swapped between UK and the US through the usual academic publications. The application work less so and maybe not at all. So aspects of the UK-Japan work (Leonardo UK and Mitsubishi) that began with Jaguar many years ago could easily be parallel to the BAE/Raytheon/Northrop-Grumman work on THREADS. I don’t know where BAE UK is on all this.
Thanks for the informative reply Jon, much appreciated.
Cheers CR
“I think the US is likely to try to nobble GCAP if it can”
Gimme a break dude.
Okay. I now no longer believe that the tens or even hundreds of billions of exports that are expected to go with 6th generation fighters is of any interest to the American military industrial complex and that the US government won’t pressure/threaten allied states to take an American fighter over GCAP. Like they have been trying on Canada over F-35. Nor do I believe they would ever use ITAR to steal a march if they could.
There’s your break. I hope you enjoy it. Of course the British, Japanese and Italian governments are expending a lot of effort making sure that the system is ITAR free. They aren’t as trusting as I am.