Forums

USB to Serial adaptor

Started by Oppie October 9, 2013
Does anybody have a favorite USB to RS-232 adaptor that they can recommend? 
Evidently some are not created equal and have had some issues.

A colleague in France reported that he had found some discrepancies in how 
the grounding was accomplished:
  "It appears that the RS232 plug body is NOT always bound to USB plug body, 
nor to the pin 5 (ground) of the rs232 plug,
so that finally there is not the same ground reference between the PC, the 
RS232/USB plugs and then the RS-232 peripheral.
This is the origin of MANY communication errors."

In his case, the adaptor was run on a laptop that had no connection to 
earth. The RS-232 peripheral (which we produce) was earthed.

tia - Oppie 

I have a Kingwin adapter cable:

- USB shield connects to braided cable shield.
- Not sure if cable shield goes anywhere at the other end.  No apparent 
electrical connection.
- Pin 5: < 1.3 ohms to USB ground pin (I measure ~0 ohms to computer 
chassis once plugged in).


Tim

-- 
Seven Transistor Labs
Electrical Engineering Consultation
Website: http://seventransistorlabs.com

"Oppie" <Oppie@127.0.0.1> wrote in message 
news:BJi5u.21468$z05.4223@fx28.iad...
> Does anybody have a favorite USB to RS-232 adaptor that they can > recommend? Evidently some are not created equal and have had some > issues. > > A colleague in France reported that he had found some discrepancies in > how the grounding was accomplished: > "It appears that the RS232 plug body is NOT always bound to USB plug > body, nor to the pin 5 (ground) of the rs232 plug, > so that finally there is not the same ground reference between the PC, > the RS232/USB plugs and then the RS-232 peripheral. > This is the origin of MANY communication errors." > > In his case, the adaptor was run on a laptop that had no connection to > earth. The RS-232 peripheral (which we produce) was earthed. > > tia - Oppie
"Tim Williams" <tmoranwms@charter.net> wrote in message 
news:l34eel$a4h$1@dont-email.me...
>I have a Kingwin adapter cable: > > - USB shield connects to braided cable shield. > - Not sure if cable shield goes anywhere at the other end. No apparent > electrical connection. > - Pin 5: < 1.3 ohms to USB ground pin (I measure ~0 ohms to computer > chassis once plugged in).
- DE-9 shield not connected. - Voltage levels meet RS-232 requirements: symmetrical (+/-) 6-7V, with some ripple from the charge pump evident. Not good for a really long cable (where the original +/-15V would be better suited), but certainly legitimate, not some phony +3/-0V "compatible" interface. Chip is Prolific (probably PL2303), which has TTL/LVCMOS interface. Don't know what RS-232 driver was used; probably a MAXx232 clone. - Works well in my limited experiments; seems to have sufficient baud rate precision and selection range. Doesn't seem to transmit frames continuously, but that may be a limitation of the program's sending capability, or the drivers or USB. Does correctly recieve continuous frames (i.e., complete stop bit immediately followed by a start bit). Tim -- Seven Transistor Labs Electrical Engineering Consultation Website: http://seventransistorlabs.com
On Wed, 9 Oct 2013 16:09:06 -0400, "Oppie" <Oppie@127.0.0.1> wrote:

>Does anybody have a favorite USB to RS-232 adaptor that they can recommend? >Evidently some are not created equal and have had some issues. > >A colleague in France reported that he had found some discrepancies in how >the grounding was accomplished: > "It appears that the RS232 plug body is NOT always bound to USB plug body, >nor to the pin 5 (ground) of the rs232 plug, >so that finally there is not the same ground reference between the PC, the >RS232/USB plugs and then the RS-232 peripheral. >This is the origin of MANY communication errors." > >In his case, the adaptor was run on a laptop that had no connection to >earth. The RS-232 peripheral (which we produce) was earthed. > >tia - Oppie
We've been using these under WIN7, works on XP. http://www.usconverters.com/index.php?main_page=product_info&cPath=67&products_id=325 The DB9 shell is floating. Pin 5 is connected to the USB shell. I have yet to discover a problem with them. The USB cable is 3 feet long to boot. Cheers
On Wednesday, October 9, 2013 1:09:06 PM UTC-7, Oppie wrote:
> Does anybody have a favorite USB to RS-232 adaptor that they can recommend?
I've always liked the Keyspan adapters; the single USA-19 works well and the dual USA-28 (Macintosh style connections) has full-differential transceivers (RS-422) and works at impressively high data rates (I've tested up to 240 k baud, it claims 1 M bps).
>Does anybody have a favorite USB to RS-232 adaptor that they can recommend? >Evidently some are not created equal and have had some issues.
As a general issue, I prefer the ones which use the FTDI chipsets. The Linux and Windows drivers both seem to be good, and the FTDI chips include unique device IDs masked into the chips that the drivers can use to properly distinguish one
>A colleague in France reported that he had found some discrepancies in how >the grounding was accomplished: > "It appears that the RS232 plug body is NOT always bound to USB plug body, >nor to the pin 5 (ground) of the rs232 plug, >so that finally there is not the same ground reference between the PC, the >RS232/USB plugs and then the RS-232 peripheral. >This is the origin of MANY communication errors." > >In his case, the adaptor was run on a laptop that had no connection to >earth. The RS-232 peripheral (which we produce) was earthed.
As I understand it and read the specs, all properly-defined RS-232 interfaces should be looking at the voltages between the signal lines, and the "common ground" (pin 5 on a DE-9 per TIA-574). I don't believe it's proper for a device to depend on there being any particular connection to the DE-9 metal shell - this may be connected to the host device's chassis ground, or to the "protective ground" RS-232 signal (pin 7 on a DB-25, not defined on a DB-9), or to nothing at all. On a DB-25 serial connection, pints 1 and 7 (signal ground and protective ground) are often connected together but this isn't required. With RS-232, you cannot depend on their being a common chassis-ground reference between the two devices at either end of the line. They may be quite some distance apart physically, may be powered from different AC mains, and there may be significant ground currents between the two. It's even possible for one or the other to be deliberately isolated from ground, for safety reasons, or to be powered like a laptop through an adapter which doesn't carry the building safety ground through to the device. If you've got some sort of terminal device which won't work correctly with even a simple three-wire connection (TxD, RxD, and common ground), then I'd say that the device is probably at fault, and not the USB-to-serial dongle.
On 09/10/2013 22:50, David Platt wrote:

> As a general issue, I prefer the ones which use the FTDI chipsets. > The Linux and Windows drivers both seem to be good, and the FTDI chips > include unique device IDs masked into the chips that the drivers can > use to properly distinguish one
I would avoid FTDI parts like the plague. Often, they work well. Cheers -- Syd
On Wed, 9 Oct 2013 16:09:06 -0400, "Oppie" <Oppie@127.0.0.1> wrote:

>Does anybody have a favorite USB to RS-232 adaptor that they can recommend? >Evidently some are not created equal and have had some issues.
Yep. Anything with an FTDI chip. I'm forced to deal with ancient Motorola radio programming software that talks directly to the serial UART. Without a proper emulator, the Motorola software just doesn't work. I've done battle with a small assortment of USB to serial cables using Prolific, TI, and SiLabs. Each will work with some combination of computah and radio, but only the FTDI chip seems to work with all combinations. I recently purchased several of these: <http://www.ebay.com/itm/370761532032> Of course, the drivers on the CD were somewhat out of date, but that was easily fixed with a download from the FTDI web pile: <http://www.ftdichip.com/Drivers/VCP.htm> Note: I'm using DOSbox to slow down the computer so that the old Motorola software will run. That doesn't work with Prolific cables but works fine with FTDI. -- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
On a sunny day (Wed, 9 Oct 2013 16:09:06 -0400) it happened "Oppie"
<Oppie@127.0.0.1> wrote in <BJi5u.21468$z05.4223@fx28.iad>:

>Does anybody have a favorite USB to RS-232 adaptor that they can recommend? >Evidently some are not created equal and have had some issues.
I have a bunch of Logolink adaptors, those are cheap, small, and so far have no problems.
>A colleague in France reported that he had found some discrepancies in how >the grounding was accomplished: > "It appears that the RS232 plug body is NOT always bound to USB plug body, >nor to the pin 5 (ground) of the rs232 plug, >so that finally there is not the same ground reference between the PC, the >RS232/USB plugs and then the RS-232 peripheral. >This is the origin of MANY communication errors."
I dunno, I killed one by runnig 20A or so from a transmitter 12V supply via the ground. Not all USB cables have screen connected either I found. In his case, the adaptor was run on a laptop that had no connection to
>earth. The RS-232 peripheral (which we produce) was earthed. > >tia - Oppie > >
On Wednesday, October 9, 2013 10:09:06 PM UTC+2, Oppie wrote:
> Does anybody have a favorite USB to RS-232 adaptor that they can recommend? > > Evidently some are not created equal and have had some issues. > > > > A colleague in France reported that he had found some discrepancies in how > > the grounding was accomplished: > > "It appears that the RS232 plug body is NOT always bound to USB plug body, > > nor to the pin 5 (ground) of the rs232 plug, > > so that finally there is not the same ground reference between the PC, the > > RS232/USB plugs and then the RS-232 peripheral. > > This is the origin of MANY communication errors." > > > > In his case, the adaptor was run on a laptop that had no connection to > > earth. The RS-232 peripheral (which we produce) was earthed. >
I have had good experience with this one: http://www.amazon.com/MCT-U232-P9-Adapter-Certified-XP-Windows7/dp/B004V89JMC Cheers Klaus PS Some USB to serial converters do not have good control of the handshake lines, so the f*'' up the timing. AFAIR the MCT behaved well