Twisted USB

[tooki]

The connector in your picture above is not a Lemo.
You've already been told what it is in another one of your moronic threads.

Lemo is compatible to Odu. And since Odu is not a clone of Lemo. I refer to it as Lemo just for sake of discussion. This is in contrast to Lemo clone which are brandless.

Though both companies are approximately the same age, ODU didn’t release its push-pull circular connectors until 30 years after LEMO introduced its. So no, LEMO isn’t “compatible [with] ODU”, and ODU is the clone of LEMO. One of many clones, and at least one made by a reputable, quality manufacturer. ODU also makes clones of Fischer connectors, another brand of similar-style circular connectors (which aren’t compatible with LEMO, though).

Note that in connectors, we usually consider something to be a clone if it is fully compatible (= designed to intermate with the original), even if there are differences in design, like how they assemble or the design of grips. (In contrast with second-sourcing, where the parts must be identical and interchangeable.)

[Andy Chee]

USB 2.0 has no error correction and so garbages can occur?

USB has cyclic redundancy checking. It can tolerate a lot of garbage before complete failure. 

[tooki]

What is the impedance of original Lemo vs Lemo clone?

They aren’t impedance-controlled.

[Ogitek]

Please refer to my edited image.  The section marked in blue, is what potentially causes an impedance discontinuity.  This section should be kept very short, which unfortunately doesn't provide much wiggle room and makes hand assembly somewhat tedious.

I don't get it. Should it be very short so it is enclosed by the shield? or because there is less twisting (but then, twisting can occur one turn every 5 inches so it's negligible)? What else?  And how does it differ to the USB cable soldered to a LEMO plug like the following?

The LEMO plug is a metal body, so it shields the wires.  Just make sure the shield is connected to the plug body.

If the plug body was plastic, then yes, it needs to be short in order to be enclosed by the shield.

Ok thanks. I heard that the termination in the USB 2.0 must be designed to have the same characteristic impedance as the twisted pair.  Have you heard of this? What if they are different?

The circuit termination is not really in the cable per se.  The circuit termination is on the PCB next to the USB chip of the PC, and on the PCB next to the USB chip inside your device.  Circuit termination could also be within the chip as well.

If impedances are wildly different, you get signal reflections, which may or may not be interpreted as extraneous bits in the serial data, depending on how well error detection & correction does its job.  Too many extraneous bits and you will lose data.  Recall the discussion on the previous page, USB protocol has no speed reduction when encountering bad data.

This is the deciding factor. You said even with error correction, error can still occurs and data can get lost between wires with subtle crapiness??  If yes, then need the original cable because I'm splitting hairs and detecting on the boundary between silence and sound, or measuring the sound of silence.

[wraper]

USB 2.0 has no error correction and so garbages can occur?

USB has cyclic redundancy checking. It can tolerate a lot of garbage before complete failure.

It can tolerate errors, but just a little bit.

[tooki]

I thought USB 2.0 Hi-Speed (480Mbps) can't run over all sorts of crap cables and connectors and it has to change mode to USB 1.1 Full-Speed (12Mbps)? If I'll try to test my USB 2.0 to my external drive via a USB Hub. Where can I see the speed displayed in 480Mbps or 12Mbps?

Still no.

High-Speed (HS) devices won’t fall back to full-speed (FS) mode when the cable is poor. USB 2.0 devices* negotiate the mode based on the highest mode supported by both host and device, irrespective of the cable used. So if both device and host support HS, then HS is the only mode they will use, and if the cable then fails to make that speed function properly, the connection will simply fail. It will NOT renegotiate down to FS.

In contrast, if there is a mismatch in the highest supported mode, the host and device may be able to use FS mode.
- If the host supports HS but the device doesn’t, it will use FS or LS, as needed by the device.
- If the device supports HS and FS, but the host doesn’t support HS, then it’ll use FS.
- If the device requires HS, but the host does not support it, the device driver won’t load.


*Remember, the USB 2.0 standard includes the USB 1.1 standard, so every USB 1.1 low-speed (LS) or 1.1 FS device is also a USB 2.0 LS or 2.0 FS device. USB 2.0 does NOT require a device to support HS mode.

[tooki]

This is the deciding factor. You said even with error correction, error can still occurs and data can get lost between wires with subtle crapiness??

That’s not what he said. He said error correction, then total failure. Note the absence of anything in between.

USB has error correction and detection, but once there are uncorrectable errors, the connection fails completely. It doesn’t tolerate uncorrected errors, you just get disconnected.

If yes, then need the original cable because I'm splitting hairs and detecting on the boundary between silence and sound, or measuring the sound of silence.

That’s not how digital audio works, nor how USB works, nor how digital audio over USB works. A random error in a digital audio stream can have a completely irrelevant effect, or a horrible one, on a sample, depending on which bit of the sample happens to be flipped. That’s why CDs use multiple layers of error correction, and will interpolate missing samples rather than skip them.

But because of how data is sent over USB, you’ll never have “marginal” data like you seem to think. It’ll either be 100% correct, or it won’t arrive at all, so to speak.

[Ogitek]

I thought USB 2.0 Hi-Speed (480Mbps) can't run over all sorts of crap cables and connectors and it has to change mode to USB 1.1 Full-Speed (12Mbps)? If I'll try to test my USB 2.0 to my external drive via a USB Hub. Where can I see the speed displayed in 480Mbps or 12Mbps?

Still no.

High-Speed (HS) devices won’t fall back to full-speed (FS) mode when the cable is poor. USB 2.0 devices* negotiate the mode based on the highest mode supported by both host and device, irrespective of the cable used. So if both device and host support HS, then HS is the only mode they will use, and if the cable then fails to make that speed function properly, the connection will simply fail. It will NOT renegotiate down to FS.

In contrast, if there is a mismatch in the highest supported mode, the host and device may be able to use FS mode.
- If the host supports HS but the device doesn’t, it will use FS or LS, as needed by the device.
- If the device supports HS and FS, but the host doesn’t support HS, then it’ll use FS.
- If the device requires HS, but the host does not support it, the device driver won’t load.


*Remember, the USB 2.0 standard includes the USB 1.1 standard, so every USB 1.1 low-speed (LS) or 1.1 FS device is also a USB 2.0 LS or 2.0 FS device. USB 2.0 does NOT require a device to support HS mode.

Thanks for that summary. My mixer is 16 channels with 16 ADCs and 24 Bit each so I'm sure it uses Hi Speed of 480Mbps for simultaneous transfer to the PC. Ok. I'll get just the original cable then, just to be sure. After I got the mixer and it's working. I'll tell you all what it is. This is to protect a backup plan at Ebay that I need to acquire in case it's not working. Thanks a lot guys.

[ejeffrey]

I'm googling a certain information I can't find.

If the data lines are not twisted but only very short length like less than one meter. Can it still somehow run USB 2.0 High Speed of 480Mbps?

Maybe only when the cable is beyond certain length that twisting is necessary for High Speed? What is the magic length?

There is no spec on how bad your wires have to be to guarantee the connection won't work.  480 Mbit/s is not terribly fast, it's likely to work over a wide range of crappy non compliant wiring.  Also the wiring requirements are probably as much or more driven by making sure devices have a chance of passing EMI than signal integrity.

[magic]

Well, EMI goes both ways. My aforementioned dodgy cable connecting an internal USB card reader to the motherboard had this interesting property that if the second "lane" in the cable was used simultaneously by a full speed peripheral (12Mbps 3.3V signaling), the card reader would drop off the bus

But it would take deliberate effort to create something worse than a $400 audiophile USB cable

[Ogitek]

https://web.lemo.com/img/resources/catalog/ROW/UK_English/High_speed_data_transfer.pdf

I found some information related to all this (someone shared to me).  LEMO has made special insert configurations just for USB 2.0 and USB 3.1 and they happen to use the 4 wire pin we are talking about. For other LEMO pin configurations like 3 pin or 5 pin. It is not USB 2.0 compliant. This means the China LEMO 4-pin clone doesn't have special pin characteristic so can't do a 480Mbps? (too much impedance mismatch causing so much reflections that can't transmit the 480Mbps)

[ejeffrey]

The USB 3.1 version obviously isn't 4 pin as USB 3.1 requires at least 8 pins. Nothing in this thread so far has talked about usb 3+ so I'm going to ignore that for now.

You would have to make a catastrophically  bad connector to fail to work for usb 2.0.  A bad cable maybe but the connector is just too small and absent deliberate inductors in the signal path or failing to connect a wire altogether its going to work.  That doesn't mean that audiophile idiots making $1000 cables can't fuck up that badly, just that they would have to be dumber than a 2 year old, or more likely deliberately sabotaging it in order to sell a $1000 filter to fix the problem they caused.

LEMO connectors are great when you need super rugged and easy to assemble custom connectors.  They are in no way better performance than a $2 monoprice overmolded USB cable.  So this whole premise is nonsense.

One note: USB connectors are designed with the power and ground pins longer than the data pins.  You can see this quite clearly on a standard type A plug.  This is important for hot plugging bus powered devices, and shorting out common mode voltage to prevent damage.  A non standard connector not specifically designed for USB is very likely to get that wrong.  It doesn't matter for signal integrity but could cause problems with hot plugging.

[Ogitek]

So you are saying LEMO just made it up.. telling people their 4-pin have special impedance that matches that of twisted wires, and it is not true at all.  But then for USB 3.1. It can also run on normal pins? So they made it up as well?

Anyway. I got the following from one of the threads here. The left one is genuine Lemo, the right one is fake or clone. What do you think is the MTBF (Mean Time Between Failure) of the Lemo clones? In your experience. 1/4 that of Lemo? 1/8? or even 1/20th of original?

I need other cables as I have to move my gear to house, studio, work and need to plug and unplug the Lemo (clone) connectors many times a day.

https://www.eevblog.com/forum/chat/a-tale-of-two-lemo-connectors/msg1273862/#msg1273862

[tooki]

https://web.lemo.com/img/resources/catalog/ROW/UK_English/High_speed_data_transfer.pdf

I found some information related to all this (someone shared to me).  LEMO has made special insert configurations just for USB 2.0 and USB 3.1 and they happen to use the 4 wire pin we are talking about. For other LEMO pin configurations like 3 pin or 5 pin. It is not USB 2.0 compliant. This means the China LEMO 4-pin clone doesn't have special pin characteristic so can't do a 480Mbps? (too much impedance mismatch causing so much reflections that can't transmit the 480Mbps)

(Attachment Link)

I saw that document long ago. It’s just their recommendations. As far as I can tell, they aren’t special insert configurations, at least not for USB 2, they’re just their standard 4-pin inserts in the recommended sizes, with a recommended pinout. They probably calculated which of their existing inserts works best for USB 2 and recommend that connector size. Because if you look at their overall catalog, “304” is one of their standard insert sizes, with no mention of USB 2. I think they list that on some SKUs simply as a convenience.

[tooki]

So you are saying LEMO just made it up.. telling people their 4-pin have special impedance that matches that of twisted wires, and it is not true at all.  But then for USB 3.1. It can also run on normal pins? So they made it up as well?

They don’t say any of that. What they say is how to choose LEMO connectors that they know work well.

Note that in that datasheet, when they say “proprietary interface” as the insert type, they mean “LEMO design, not an open standard”, NOT that it was designed specifically for USB!

I think you really need to a) relax before you have a seizure from how much you’re overthinking this, and b) listen to the folks who’ve given you more than enough information to have solved this issue days ago. I’m not really sure why this discussion is even still ongoing.

[Ogitek]

https://web.lemo.com/img/resources/catalog/ROW/UK_English/High_speed_data_transfer.pdf

I found some information related to all this (someone shared to me).  LEMO has made special insert configurations just for USB 2.0 and USB 3.1 and they happen to use the 4 wire pin we are talking about. For other LEMO pin configurations like 3 pin or 5 pin. It is not USB 2.0 compliant. This means the China LEMO 4-pin clone doesn't have special pin characteristic so can't do a 480Mbps? (too much impedance mismatch causing so much reflections that can't transmit the 480Mbps)

(Attachment Link)

I saw that document long ago. It’s just their recommendations. As far as I can tell, they aren’t special insert configurations, at least not for USB 2, they’re just their standard 4-pin inserts in the recommended sizes, with a recommended pinout. They probably calculated which of their existing inserts works best for USB 2 and recommend that connector size. Because if you look at their overall catalog, “304” is one of their standard insert sizes, with no mention of USB 2. I think they list that on some SKUs simply as a convenience.

Yes, there is mentioned of USB 2.0. In fact, only the 304 spec sheet has the following lines added "Differentiated pairs (High Speed) 1  Protocol: USB 2.0 (480Mbps)", the rest doesn't have the lines:



So is it only because there are 4 pins or the pins are some kind of alloys?

In fact, some believe that Lemo connectors are for audio, not half-GHz RF. So you need special material to do that. "Generic" Lemo connectors were originally NOT meant for high frequencies. They were used for DC or low-frequency measurements where you typically don't have to worry about the impedance but you need a rugged, reliable connector. Note that Lemo was selling the exact same type of connector well before USB was even a thing.

So does the 4 pin configuration use special metal that can change the impedance?

I don't just need 1 cable, but many in our production business where we only loan so we want to make sure we made the right investments.

[tooki]

https://web.lemo.com/img/resources/catalog/ROW/UK_English/High_speed_data_transfer.pdf

I found some information related to all this (someone shared to me).  LEMO has made special insert configurations just for USB 2.0 and USB 3.1 and they happen to use the 4 wire pin we are talking about. For other LEMO pin configurations like 3 pin or 5 pin. It is not USB 2.0 compliant. This means the China LEMO 4-pin clone doesn't have special pin characteristic so can't do a 480Mbps? (too much impedance mismatch causing so much reflections that can't transmit the 480Mbps)

(Attachment Link)

I saw that document long ago. It’s just their recommendations. As far as I can tell, they aren’t special insert configurations, at least not for USB 2, they’re just their standard 4-pin inserts in the recommended sizes, with a recommended pinout. They probably calculated which of their existing inserts works best for USB 2 and recommend that connector size. Because if you look at their overall catalog, “304” is one of their standard insert sizes, with no mention of USB 2. I think they list that on some SKUs simply as a convenience.

Yes, there is mentioned of USB 2.0. In fact, only the 304 spec sheet has the following lines added "Differentiated pairs (High Speed) 1  Protocol: USB 2.0 (480Mbps)", the rest doesn't have the lines:

(Attachment Link)

You're reading and quoting me selectively, missing the point.

I said: "...if you look at their overall catalog, '304' is one of their standard insert sizes, with no mention of USB 2."

The overall catalog, not the individual datasheets, not the high-speed data guide, but the general catalog. It lists insert layout 304, but never mentions USB. (Admittedly, "overall" was not the clearest choice of word.)

I checked an old LEMO catalog that predates USB 2.0 and guess what? Insert 304 is already there, because it's been there for decades.

If you check the high-speed data guide, you will notice that only connector sizes 0 and 1 are recommended for USB 2, despite the fact that the "304" layout exists in almost every size of connector they have, too, both larger and smaller than 0 and 1.

So is it only because there are 4 pins or the pins are some kind of alloys?

It's obvious that you don't even know what characteristic impedance is, so there's no point in you speculating on how they achieve it.

Especially since the answer is: they don't do anything special for USB 2. They simply recommend for USB 2.0 the existing 4-pin connectors whose insert dimensions best match the needs of USB 2.0. Characteristic impedance is fundamentally defined by the properties of the dielectric (which is always PEEK in these series) and the dimensions of the dielectric and the conductors. Since the parts they recommend for USB 2 existed long before USB 2, they clearly did not design special inserts because that would have required a time machine.

In fact, some believe that Lemo connectors are for audio, not half-GHz RF. So you need special material to do that.

People can believe whatever they want, that doesn't make them right. LEMO's most widespread use is laboratory equipment, especially in nuclear physics (which is one reason they use PEEK so much: it doesn't degrade nearly as much in radiation as other plastics -- they even talk about this in the catalog). But that isn't to say that LEMO doesn't have "widespread" use in various niches. (I mean, as widespread as extremely expensive connectors can be...)

500MHz is not that high (my boss's boss at the physics department where I used to work always said "to me, anything under around 3GHz is just DC!") in the grand scheme of things. High enough to need to take care, but not as demanding as multi-GHz stuff. (Nuclear physics equipment still uses tons of LEMO coaxial connectors, which are noticeably inferior to the better "mainstream" coaxial connectors like SMA, according to said boss's boss.)

The audiophile world has an uncanny knack for latching onto a kernel of truth about physical or electrical reality and then grossly misapplying it in nonsensical ways. So it's not worth following what they do. There is no performance reason to use LEMO in anything audio; it's just robustness and prestige.

"Generic" Lemo connectors were originally NOT meant for high frequencies. They were used for DC or low-frequency measurements where you typically don't have to worry about the impedance but you need a rugged, reliable connector. 

If the clones match the pin and shell sizes of the originals (which they kinda have to to make them compatible) and used the correct insulator (PEEK) and dimensions, then they'll necessarily be very close to the originals' impedance profile.

Note that Lemo was selling the exact same type of connector well before USB was even a thing.

Yep, that's what I said a while ago.

So does the 4 pin configuration use special metal that can change the impedance?

They don't, and it couldn't affect it in any significant way.

I don't just need 1 cable, but many in our production business where we only loan so we want to make sure we made the right investments.

So you should really be paying a consultant rather than dicking around in a forum where you just talk in circles anyway.

[Ogitek]

Thanks for your convincing exposition.

But what about USB 3.1?  The old ordinary LEMO connectors with the respective numbers of pins matching USB 3.1 can also transmit at 5Gbps without significant reflections that can render it null?

What speed when the average LEMO connectors can no longer catch up?  I want to be ready in case I can get some old used Lucasfilm, etc. gears that may work at USB 3.1 and they used Lemo too.

For reference: