I would love to always test my stuff but im not as learned as you, nor do I have an Oscilloscope. Btw, awesome channel! I needed a good fix since Photonicinduction stopped posting videos and your videos are well presented and very informative. Kudos.
Most of the cheap devices don't make a legit Tx enable for the TTL to RS485 driver. They just tie the TTL Tx data line to both data input and Tx enable input on RS485 driver chip. This means RS485 driver tri-states during logic '0' transmitted, relying solely on termination load resistor on twisted pair to pull a RS485 logic 'zero' on the pair. It greatly reduces maximum baud rate possible and reduces maximum line length possible.
Hi Mike, I just wanted to confirm, are the RS485 ports on that full duplex which is why you suggest (and I totally agree) they should also have a ground connection?
I can't agree with you anymore. Always test them! That thirty party chunk is my headache at work, because I used to believe in manufacturer specifications.
@simontay1984 RS485, its similar to 232 but it has noise cancelling, sort of like USB but at a slower speed, it was used in old macs/apple2GS's for printers and modems also known as appletalk
That driver enable is weird; perhaps it was designed to also do CAN bus and they've messed up the configuration, mixing the CAN style driver enable with the RS485 style driver data?
I'm pretty sure that almost no-one posting in the comments here (aside from the actual author of the video) actually understands what the CE marking means. You'll find a CE label on everything from a beer glass (to verify that it will actually hold a "pint" or other specific amount of fluid) to devices that are covered by the low-voltage directive (LVD), R&TTE and the EMC directive. None of these directives necessarily mean that a product has to be "awesome" or have truthful marketing about signal integrity, baud rate, distortion metrics or anything of the sort. CE certification is also only necessary for devices/products offered for sale/trade in the EU. Otherwise it's pointless. Additionally, the CE marking is NOT comparable to UL, TUV, ETL or any other quasi-regulatory certification lab marking.
essentially, CE marking is self-deceleration. So, if the manufacturer has no idea what he is doing, the CE mark is as reliable as the specification: Non at all. The CE mark is - more or less - just a way to get CEOs in charge for the products. Not fulfilling the relevant regulation is very technical - but slapping the CE mark on when not fulfilling the associated regulation: That is simple forgery. Something a judge understands. Seeling a product without CE mark: Not allowed in the EU. Makes the life of a judge easy and the CEO responsible, regardless of "technical what evers" he is willing to not understand.
It looks like it is a hub or a center point in a star configuration?! If you are not cnnecting any transmission line and use just a single termination resistor you might get really bad response .. and in a star configuration terminating the center point always gives less good results .. the termination only works well when there really is a transmission line and the resistors are at the both ends ..read for example (TI's - AN-979 The Practical Limits of RS-485 - which is available in the Internet) - cheers!
Says nothing about the product actually worked, but it is electrically safe and all that stuff. :-) I like your vids, mike. You take a lot of unusual stuff apart, so I don't have to. :-)
CE mark is almost worthless. At the place I used to work for we needed to make some UL approved stuff. UL is the American code, and they are tough. We had a external consultant travelling from Holland to check if the product met the requirements. He got a workbench in our shop and was there two days to examine our product. We had to change a lot and he had to come again etc.. If you want something that are actually checked ok for safety, go for the UL symbol, or at least UL listed symbol.
Oh yeah, it's like one of those Magic Eye things! Lemme just cross my eyes... Okay, I think it's... oh! It's an illusion of a high quality RS485 converter! Just like the device creating the waveform!
Good review. :) The isolation is indeed total crap. I wonder if that ground break is enough even for 10kv. Seems a little narrow. But yeah, that's not a real show stopper. The signal integrity is total shit, though. I also hate equipment that tries to set itself up. Very error prone, never seems to work properly for me. No manufacturer should ever include such functionality, not even on scopes. I wouldn't buy the thing...
I know. But to put a CE mark on a product the producer declares that the product is produced to meet some criteria. It is illegal to mark it with CE if it does not work as intended or is expected to fail during normal operation.
The whole problem here is the CE certification, CE of course standing for Chinese Engineering... Mounting the wrong connectors and not working to specifications is a general rule with cheap Chinese stuff.
RS-485 does NOT measure the signal voltage from the GND but between D+ and D- (no reference to GND/RTN). As long as all systems are earth-grounded, there should be no problem (furthermore, not having those extra GND connections rather results in eliminating possible ground loops candidates)
I love the rainy background videos. So soothing, even with a rant!
I would love to always test my stuff but im not as learned as you, nor do I have an Oscilloscope.
Btw, awesome channel! I needed a good fix since Photonicinduction stopped posting videos and your videos are well presented and very informative.
Kudos.
Would I be right then in assuming the cheapo ebay adapters @ 115200 baud would be atrocious?
Most of the cheap devices don't make a legit Tx enable for the TTL to RS485 driver. They just tie the TTL Tx data line to both data input and Tx enable input on RS485 driver chip. This means RS485 driver tri-states during logic '0' transmitted, relying solely on termination load resistor on twisted pair to pull a RS485 logic 'zero' on the pair. It greatly reduces maximum baud rate possible and reduces maximum line length possible.
Hi Mike, I just wanted to confirm, are the RS485 ports on that full duplex which is why you suggest (and I totally agree) they should also have a ground connection?
I thought for the rs 485 there is no ground, it was for the rs 232, please correct me if I'm wrong?
Not true. CE mark only indicates it meets applicable standards for safety, EMC etc. Not that it actually works or does anything useful.
Also, CE is self certified!!!!!
This! Its self certified!
[Always] insightful. Thank You
I can't agree with you anymore. Always test them!
That thirty party chunk is my headache at work, because I used to believe in manufacturer specifications.
@simontay1984 RS485, its similar to 232 but it has noise cancelling, sort of like USB but at a slower speed, it was used in old macs/apple2GS's for printers and modems also known as appletalk
love your site and I only just found your youtube channel!
That driver enable is weird; perhaps it was designed to also do CAN bus and they've messed up the configuration, mixing the CAN style driver enable with the RS485 style driver data?
CE marking has nothing to do with whether things actually work.
I'm pretty sure that almost no-one posting in the comments here (aside from the actual author of the video) actually understands what the CE marking means. You'll find a CE label on everything from a beer glass (to verify that it will actually hold a "pint" or other specific amount of fluid) to devices that are covered by the low-voltage directive (LVD), R&TTE and the EMC directive. None of these directives necessarily mean that a product has to be "awesome" or have truthful marketing about signal integrity, baud rate, distortion metrics or anything of the sort. CE certification is also only necessary for devices/products offered for sale/trade in the EU. Otherwise it's pointless. Additionally, the CE marking is NOT comparable to UL, TUV, ETL or any other quasi-regulatory certification lab marking.
essentially, CE marking is self-deceleration. So, if the manufacturer has no idea what he is doing, the CE mark is as reliable as the specification: Non at all.
The CE mark is - more or less - just a way to get CEOs in charge for the products. Not fulfilling the relevant regulation is very technical - but slapping the CE mark on when not fulfilling the associated regulation: That is simple forgery. Something a judge understands. Seeling a product without CE mark: Not allowed in the EU. Makes the life of a judge easy and the CEO responsible, regardless of "technical what evers" he is willing to not understand.
It looks like it is a hub or a center point in a star configuration?! If you are not cnnecting any transmission line and use just a single termination resistor you might get really bad response .. and in a star configuration terminating the center point always gives less good results .. the termination only works well when there really is a transmission line and the resistors are at the both ends ..read for example (TI's - AN-979 The Practical Limits of RS-485 - which is available in the Internet) - cheers!
@ahessenflow I wouldn't credit them with that much intelligence. My guess is a mixture of recycling an old RS232-485 design and skimping on optos.
No idea what you are saying, to a degree. But I like listening to your videos.
Great deconstruction of a crappy product. Thanks for the detailed analysis.
Says nothing about the product actually worked, but it is electrically safe and all that stuff. :-)
I like your vids, mike. You take a lot of unusual stuff apart, so I don't have to. :-)
Is your workshop in a shed?
CE mark is almost worthless. At the place I used to work for we needed to make some UL approved stuff. UL is the American code, and they are tough. We had a external consultant travelling from Holland to check if the product met the requirements. He got a workbench in our shop and was there two days to examine our product. We had to change a lot and he had to come again etc.. If you want something that are actually checked ok for safety, go for the UL symbol, or at least UL listed symbol.
I love you Mike
Anyone notice the optical illusion at 4:00?
Oh yeah, it's like one of those Magic Eye things! Lemme just cross my eyes... Okay, I think it's... oh! It's an illusion of a high quality RS485 converter! Just like the device creating the waveform!
Cool shirt!
Great video! More penny pinching crap hardware. Hopefully that unit was not from Black Box. They usually make good stuff.
Good review. :)
The isolation is indeed total crap. I wonder if that ground break is enough even for 10kv. Seems a little narrow. But yeah, that's not a real show stopper. The signal integrity is total shit, though. I also hate equipment that tries to set itself up. Very error prone, never seems to work properly for me. No manufacturer should ever include such functionality, not even on scopes. I wouldn't buy the thing...
Check out spreadshirt - create your own easily
Lets see what happons!
We had something similar with harddrives. Bought a whole load.. and they all died within a few months.
Mike being british, I'd say it's rain.
I know. But to put a CE mark on a product the producer declares that the product is produced to meet some criteria. It is illegal to mark it with CE if it does not work as intended or is expected to fail during normal operation.
"About Mike's Electric Stuff
Video department of electricstuff.co.uk
If you want to contact me do it via the website above, not TH-cam messaging"
4:27 it turned into lvds 😂
Haha, startech fail :)
The whole problem here is the CE certification, CE of course standing for Chinese Engineering...
Mounting the wrong connectors and not working to specifications is a general rule with cheap Chinese stuff.
Doesnt it mean China Export? CR would also be nice - China Replica.
You should report the company. If the product is CE-marked they have violated the terms and should get a smack on the hand.
RS-485 does NOT measure the signal voltage from the GND but between D+ and D- (no reference to GND/RTN). As long as all systems are earth-grounded, there should be no problem (furthermore, not having those extra GND connections rather results in eliminating possible ground loops candidates)
Is this guy speaking English? He speaks real fast so he must know what he's talking about, I guess.