Advertisement
Advertisement
⚡ Community Insights
Discussion Sentiment
54% Positive
Analyzed from 3160 words in the discussion.
Trending Topics
#labs#fcc#llm#testing#china#more#test#radio#lab#chinese
Discussion Sentiment
Analyzed from 3160 words in the discussion.
Trending Topics
Discussion (82 Comments)Read Original on HackerNews
This move is being executed too broadly, in my opinion, but the “bad labs” problem especially in China is widely known in the industry. If you spend any time in the electronics industry at smaller companies you will encounter people who know Chinese labs that will give your product passing test results every time as long as you’re not so far past the limits that it’s too obvious.
And really, you can't tell. Nobody can tell. Humans write badly and blandly also. It's just a trope at this point.
No, you're comment is an LLM.
We should probably go ahead and get over it.
But you can tell it to use different styles. To be formal or in-formal, to insert colloquialisms or to remove.
People are depending on their own 'gut-sense' a lot, and not realizing they are really not correct.
If you think all it takes is paying attention, then you are missing it. It's both more widely used than assumed, and also now obscuring what is non-AI.
It is obvious, when it is obvious. When it is not, you don't know it.
There are ton more false positives now. Everyone is calling everything 'LLM Slop'.
Because there is a lot of slop. Now every bad human writer is being called an AI just for being human.
And, that is covering that a ton of stuff is LLM and nobody can tell.
People that say they can tell the difference are fooling themselves.
If you personally can't tell then just say that rather than casting aspersions on everyone else by claiming they can't.
Not a dig at this author by the way or saying it applies to this post, just in general.
*or if they did anyway, the result is the same: bad writing.
Idk, I learned a little bit about our regulatory system, that a lot of these labs are in China and that those are now banned (and that the ones in India may be next).
The style is admittedly annoying. But I'm glad the author put in the work to highlight something they, and now I through them, found interesting.
LOL, some of us spent 12 years in public schools refining this very art to perfection.
https://news.ycombinator.com/item?id=47963465
Half the articles are now LLM's.
If he didn't admit it, we'd be arguing over 'style', which itself can be configured.
Prompt> LLM don't use em-dash.
LLM> OK.
I know pie charts are decisive. I thought they were visually helpful in this instance.
If you're the author, can you comment on whether you used AI to write this? (Specifically, the text.)
Where it might be suffering is in its presentation of a list of facts unorganised around any thesis. It took me until your China Question section to see the meat of your piece.
If I had to suggest some edits, they would be making everything above that section more concise (by reducing the number of charts and/or moving them to footnotes) and adding a summarising subtitle.
There are also jargon jumps, e.g. from TFAB to TCB. (I initially assumed the FAA was a TCB, the latter being a generic international term.) This compounds the lack of conciseness presented by the accredition-body breakdown and TCBs vc. test-only labs sections. If those sections were moved after your thesis section, you could dive into whether China's labs differ from the U.S. labs in those respects.
I very much appreciate your feedback. As I look at the article now. I totally see what you're saying. I should have let off what was going on with the vote today since that's what I referenced in the title of the post on here.
If you've only got a paragraph worth of information to share, say it and let us get on with our lives.
As expected this is an attack on China-sourced electronics.
I hope this isn’t the start of a return to the Bad Times when many niche electronics were simply unavailable at any price, and what was available was $1K+ for what should be commodity gear.
What the FCC does is important, but there needs to be a sense of proportionality. I am a ham radio user but I am not particularly bothered if my $30 DVD player has a few spurious emissions, as long as they aren’t egregious. I also don’t mind imperfect but cheap radios like Baofengs if they help get people into the hobby. It’s good to have a box of these to hand out in emergency situations! Can’t do that with Yaesus unless you’re made of money.
I'm bothered when my neighbors turn on their christmas lights, and the whole 40 meter band is wiped out.
Also baofengs are horrible all those regards:
* spurious emissions (thus banned in quite a few countries) * useless in most emergencies (but preppers somehow buy them for some reason... probably due to youtubers shilling for them) * handing them out to whom exactly? You need a ham radio licence to use them, and i'm pretty sure every licenced ham has a radio and doesn't need handouts from others (unless we're talking about baofeng FRS/PMR radios, but somehow preppers never buy those)
Also a yaesu ft65 costs around 100eur over here, you don't have to be made of money to afford a much better radio.
Where I’m at a Baofeng can hit the local repeaters just fine. I handed them out to my family when we had a major multiday communication outage (cellular and internet were down) and set them up to listen to the repeater. I told them if there’s a life threatening emergency they can transmit. It made everyone feel a little safer.
While I personally have a better radio, they are great as cheap backups.
Legally you do, that exception only applies to amateur station, not unlicenced users.
Why not get a gmrs licence instead, and give them gmrs type-accepted radios that they can use and try out and get experienced with even when not in an emergency? It's like buying cheap cars to give to people to drive for the first time in an active emergency... dangerous both to them and to others.
The frontend is horrible, the filtering is horrible, they get easily overloaded, and they're still not emergency radios. People will die because they will rely on them instead of getting a proper radio for emergencies. Garmin inreach will actually get you help when stuck in the middle of nowhere (because no one will be in simplex range then), and a starlink setup is much better for anything at home, becuase you can actually reach someone who can help that way. Baofengs are just something that earns percentages to youtube "preppers" (many of them not licenced hams either... it's like taking car advice from someone who doesn't even have a drivers licence).
Of course initially, expecting the same quality and low price will likely be an issue. But over time it gets better is probably the idea. Will it actually happen, who can say? But I can understand the idea they have here. I'm not saying they'll be successful. Definitely not saying I agree with it. (There are far more effective ways to accomplish a manufacturing ramp up with far less risk.) But I get the idea.
Maybe the FCC lab rules ought to be more selective, allowing consumer goods but not commercial/industrial goods at these labs.
Alternatively, maybe we can subsidize the consumer. What I don’t want to see is everything becoming 10x more expensive or completely unavailable at home while the rest of the world gets to keep the status quo.
For those who don't know, the point of these labs is to generate certified test results. The lab's job is to certify that the test was done correctly with calibrated test equipment. They then give you the results which you then give to the regulating body.
However, you are the one who knows your product and how it operates. You are present in the lab, in the room, doing the testing with them. This introduces a lot of grey area where a lab may or may not go along with what you say. Chinese labs are known for just going along with it. After all, you are the one paying them tens of thousands, and they know you probably have many other products that need certifying. It's mutually beneficial for them to be lax.
Are the engineering firms going to Chinese labs because of their rubber stamp approval or because the process is “tens of thousands” as opposed to “hundreds of thousands”?
The tests can be 100+ parameters and every single one is a hard binary pass/fail. So a mouse fart over the line on one test and you are not compliant. Go home, re-engineer, hope you fixed the issue (it's basically impossible to test for without a full dedicated RF testing chamber), and pay for a whole new testing run.
This is where the Chinese labs come in to play, as upon seeing your .5dB issue, they may feel that maybe their antenna placement was slightly off, or maybe you say that actually the "center of noise" (a bs thing) on your product is 5cm away from the testing center point, and move it back from the antenna with a wink and a nod.
In western labs the people are just as friendly, but that antenna is definitely in the right spot, and your product goes in the same spot as everyone elses.
However you need to go to China to do all this. If you are big megacorp with huge resources, western labs are fine. If you are smaller and possibly will become insolvent if you cannot pass testing, China is the play.
Most people are going to want to sell their products on an international market, which essentially means designing and testing it to the strictest rules any country uses and just having to do a whole bunch of paperwork for the rest: getting a lab to do both FCC and CE is not significantly more expensive than either one on its own. And because FCC requires external testing, that means doing external testing.
Besides, although CE is technically a self-declaration that you follow the relevant rules, it still requires you to be able to demonstrate that you follow those rules - which means you have to test and report on a level comparable to an external lab, which means building a testing lab with a price tag comparable to a very nice home, and doing all the annoying paperwork like having your equipment regularly tested and calibrated. You are allowed to do it in-house, but is it worth it?
In-house emc testing is quite fun and you dont need much more than a spectrum analyzer, antenna and E/H-field probes.
When done just for this purpose, it can be done much more cheaply than at a proper lab, because you do not need very accurate results.
I did not FCC certify my product because the primary focus is for the European market but now I would also have to consider costs.
The first thing I did was build an LLM-maintained wiki about the hardware certification universe - FCC rule parts, equipment authorization types, test standards, the TCB system, international equivalents, all of it. About 30 pages of structured knowledge that Claude could reference when doing the actual enrichment work. Then I ran a loop of subagents over multiple days to enrich the labs - pulling from the Socrata API, cross-referencing TCB registrations to see which labs can certify (not just test), hitting Google Places for websites and coordinates, crawling accreditation body directories to figure out which labs are actually still active. The FCC's own expiration dates are useless for this - tons of labs show 2022 or 2023 dates but are clearly still operating. Claude synthesized the descriptions and capabilities from all the scraped data into structured records, using the wiki as context.
The directory is at markready.io/labs. You can browse by country, US state, and TCB status.
Today's an interesting day to launch this because the FCC is voting to ban all 126 test labs in China and Hong Kong. Not just the government-controlled ones they banned last year - all of them. 21% of the global total gone. 27 of those 126 are Western firms (Intertek, SGS, TUV, UL) operating China offices. I wrote up the full impact analysis at markready.io/blog/fcc-bad-labs-vote.
Full disclosure: I've never actually gotten a device through FCC certification myself. I've been building RF hardware since I was a kid but always on the hobby side. What pulled me into this was the data problem - the FCC publishes all this information but nobody had stitched it together into anything usable, and it felt like a genuinely interesting dataset to enrich and a real gap in understanding the hardware product space.
Built with Next.js + Cloudflare. The enriched dataset covers 28 countries.
Does this increase or decrease demand for your tool? Less fragmentation would be expected to decrease demand from insiders. But more regulatory scrutiny would raise the stakes for outsiders getting it wrong.
It's not like the US has the strictest radio requirements on products, in fact, it's close to the opposite. Both asia and europe have much stricter emissions and sensitivity requirements.
But the big question to me, is what do they intend to do about the overwhelming majority of US consumed electronics being built in China?
The idea of abandoning chinese manufacturing is a delusion. It would take decades.
The idea that we do the "really advanced, complicated" manufacturing in the US, and let the Chinese build the "simple" stuff is even more delusional. China has the most advanced manufacturing tech, and the greatest scale of manufacturing in the world, by far. By very very far.
Sure, the US could catch up, assuming they stop banning Chinese cars and other advanced products and instead start importing and learning from them.
All of this delusional thinking goes back to one common source: elect a clown, expect a circus...