X-band transmitters on the global marketplace
blog.satsearch.co
blog.satsearch.co
Am i reading this correctly?? the x-band cubesat transmitter is the cost of a Telsa???
Many suppliers are struggling with finding the right balance between economies of scale, lead time, modularity/standardization, reliability, and (backwards) compatibility.
As industry volume increases, this should start straightening itself out. The challenge is that there's a bit of a chicken-and-egg problem at the moment: there's a severe lack of transparency, meaning that commercial market forces are largely missing.
So greater transparency = greater competitiveness = better price/value ratio = faster, better, cheaper satellites
Throw in a few mixers for up/downconversion, some expensive ADCs, an FPGA, a CNCed aluminum enclosure, and you quickly end up at around $1000-$5000 in parts. There's definitely demand for these radios, but it's no consumer product, so the markup is going to be painful.
There's definitely opportunity here, but it will be difficult to get the prices down significantly when there are like five semiconductor companies that are doing most of the gouging. They're very happy with their current "sell you a specialty part for $45 that costs them $2 to produce" business.
How difficult is it to design these parts and successfully fabricate them? Are they expensive simply because they're uncommon, or is the R&D process for them nontrivial?
QA processes also add a fair amount to the final price as each part might have several hours of individual attention by a technician to make sure it operates perfectly. Even for something as simple as a strut, they might saw every other one they make in two to inspect it for defects as doing so is vastly cheaper than having something fail in launch or orbit.
The PCBs are aluminum backed. Gold selectively removed from PCB traces to prevent solder embrittlement. Parts sometimes vacuum soldered to PCB (to prevent voiding). X-Rayed to look for voids. Cleaned and vacuum baked. RF-DC burn in. My stuff had the lids welded on. Vibration tested. More electrical testing and thermal cycling. Ping tested to hear for loose parts and solder balls.
All this adds up the cost.
Rework is a nightmare, especially after the lids are welded.
All the other SMT parts are also S-level. A single tube of low-outgassing RTV (Dow 340) is $600.
I’m glad I no longer do that shit; couldn’t pay me enough.
I had nightmares about shit like this:
https://en.m.wikipedia.org/wiki/NOAA-19#/media/File%3ANOAA-N...
But these low volume (both in terms of # of units sold and SW&P) things get really expensive as high R&D costs are mostly amortized by people who are willing to ride the cutting edge. If you don't have high bandwidth requirements you can get away with some $20 transmitters using LORA or some cheapo ~900mhz transmitters that use amateur frequencies
We've mentioned frequencies and transmitter power in ODAR paperwork but it hasn't been an issue (ODAR is a form that the FCC gets which is mostly there to make sure you don't litter while you're in space).
Edit: Actually I looked into this and we had to a bit more paperwork than just the ODAR. The FCC also wanted to know about signal modulation and ground station locations.
"where is the transmitter going to be?"
"where is the receiver going to be?"
"how is the cubesat going to be disposed of?"
"what frequencies will be used?"
"does anyone on the team do DRUGS?"
etc
The whole process took a couple months, with IMO is fairly lightweight for a massive government bureaucracy like the FCC
That's a pretty big IF, no?
VHF systems at 455MHz are about 5K Euros and maxes at 9600 Baud.
There was an article on HN recently about a cubesat mission that failed. And the communication system was one of the critical failure points.
And how far down the dbm sensitivity scale are those TI chips?
You completely missed me pointing out that a cubesat-type project failed miserably for reasons relating to all this (Ah, found the link): https://hackaday.com/2020/01/24/lessons-learned-from-a-cubes...
They call it "rocket science" for a reason.
I think this[1] is the datasheet for the chip that we used (on a custom PCB), I'm not an RF person but maybe that will answer your questions about dbs.
Communication systems are certainly critical, and it's smart to have redundancy like that team did. However it's kinda like backing up your harddrive— redundancy is pointless if you never test it and it ends up not working. Lots of satellites have radio problems. Lots of satellites also don't have radio problems— there's always going to be a small number of random failures, but most of the problems are not insurmountable, especially if you're able to test extensively and iterate
ADC's are probably coming in hot at 300-400MHz. That's Virtex-class land, and for a one-off FPGA at that class it's probably $5K or so.
Everything in aerospace is expensive because it has zero volume. The US DoD is always bitching about this, but it isn't willing to pony up the money to put volume behind their procurement to cause the costs to drop.