Help with USB powered cable problem

[lnsstudio]

All,
I’m hoping you can help me with a serious overlook when installing active USB extension cables in my walls. You may want to grab a cup of coffee and laugh off my obvious mistake when installing these cables in my walls. Unfortunately, the room is complete with drywall so there’s no turning back.

Here’s the deal - I installed them backwards. Hard to believe, but I did it

These are type A connections on both ends. One female, the other male. The female end is located where the computer is installed. A wrong move of course.

I did put on wall plates; therefore, there are female A type connectors on the plates. I put a gender changer on the male at the wall plate.

I’m assuming this does not work since the USB extension cables inline booster chip never receives power from the computer’s USB controller. Is this correct, or could there be more to it?

Is there anything I can devise that I could add to the end of the line to power the cable circuit, and are there other considerations?

This is for a small project studio.
Below is the link for the cables.

Thank you,
Steve

https://www.amazon.com/gp/aw/d/B00N7C8DW0?psc=1&ref=yo_pop_mb_yo_pop_mb_pd_t2

 

[dknguyen]

I’m assuming this does not work since the USB extension cables inline booster chip never receives power from the computer’s USB controller. Is this correct, or could there be more to it?

I wasn't able to follow the rest of your post about what the problem is, but what makes you think the quoted section? If the active USB cable does not plug into another power source, it must then be receiving power from the computer. It uses the aformentioned power to repeat/boost the USB signal again.

This is possible because the power that a USB port is expected to provide to power a device dwarfs the power that is used to transmit the USB signal itself.

 

[JonSea]

If the cable expects to be powered by a PC, providing power from the wrong end may not work.

 

[lnsstudio]

JonSea,
Exactly, that’s my issue. My computer’s located on the wrong end.
dknguyen, sorry. I certainly forgot to write my problem statement.
Is it possible to tap into the cable and power it, or will there be a communication issue due to the cable running opposite direction?

It doesn’t work now with the female end at the computer. This is the device end.

 

[dknguyen]

Oh, so the problem is the cable is reversed and it can only take a power input from one end? I don't think there would be a communication issue. But if you aren't sure, you can test by connecting the end that requires power to a powered USB powered hub and then connecting X device to the hub.

 

[lnsstudio]

I did try a powered hub, but I may try it again. It was an old usb hub. Is there one that you know is reliable? Amazon?
Thanks, Steve

 

[dknguyen]

I don't know about reliable...the only ones I ever used are these ones for work and I've only had it for about a couple of months:

https://www.startech.com/ca/Cards-Adapters/USB-3.0/Hubs/4-Port-Black-SuperSpeed-USB-3-Hub~ST4300USB3

Or you could get something like this and plug the power-only end into a 5V supply like a phone charger. But if it isn't working with the hub then it probably won't work like this either (unless a USB hub works differently than I think it does).

 

[rjenkinsgb]

An actual USB "hub" is an addressable device, each port is separately controllable from the host.
It's unlikely to work in reverse.
(If you look in a computer device manager under USB, each hub has its own driver instance).


I'd carve up a couple of passive USB extension cables and make two back-to-back connector gender changer cables & see if that works.

Looking at the specs for a typical USB repeater IC, they do appear to be reversible so it's hopefully just a matter of getting power in the right place.

If the gender changers do not work like that, I'd try disconnecting the 5V link in the one at the PC end and add a 5V PSU connected to the gender changer cable joints at the remote end.

 

[JimB]

You may want to grab a cup of coffee and laugh off my obvious mistake when installing these cables in my walls. Unfortunately, the room is complete with drywall so there’s no turning back.

OK, I had a chuckle at your obvious mistake. We have all been there at one time or another.
But...
I will shake my head at your other mistake...

Installing active devices which cannot be replaced if (when) they fail.

What will you do when one of these cables with a built-in amplifier fails?
How will you replace it?
Rip down your nicely installed drywall?

As a general rule, never, ever, install something which could easily fail, in a position where it cannot be easily replaced.

JimB

 

[lnsstudio]

OK, I had a chuckle at your obvious mistake. We have all been there at one time or another.
But...
I will shake my head at your other mistake...

Installing active devices which cannot be replaced if (when) they fail.

What will you do when one of these cables with a built-in amplifier fails?
How will you replace it?
Rip down your nicely installed drywall?

As a general rule, never, ever, install something which could easily fail, in a position where it cannot be easily replaced.

JimB

Agreed. I certainly thought that as I buried this crap in my walls.

 

[lnsstudio]

An actual USB "hub" is an addressable device, each port is separately controllable from the host.
It's unlikely to work in reverse.
(If you look in a computer device manager under USB, each hub has its own driver instance).


I'd carve up a couple of passive USB extension cables and make two back-to-back connector gender changer cables & see if that works.

Looking at the specs for a typical USB repeater IC, they do appear to be reversible so it's hopefully just a matter of getting power in the right place.

If the gender changers do not work like that, I'd try disconnecting the 5V link in the one at the PC end and add a 5V PSU connected to the gender changer cable joints at the remote end.

Yes, I thought about adding the supply, but was a bit unsure the repeater IC would work in both directions. I still need to think about where to disconnect power at. Everything’s backward so it gets confusing.

 

[lnsstudio]

rjenkinsgb, and anyone who can help,
If you wouldn’t mind help confirm the idea by adding a 5V supply.

1. The computer side of the cable I already have connected the gender changer, so the cables device end is connected to the computer.

2. The computer end of the cord is located near my devices ( drives, card reader etc..). I have gender changed this from male to female.

3. Since the cable is backwards, I need to tap in 5V power on the device side of the cable. Where my devices are located.

4. So without cutting the power lines 1 and 4 and disrupting the power to the devices, I can just connect the 5V supply to lines 1 and 4.

5. This action will provide power to the devices and back up to the repeater chip.

6. Now we have the repeater chip active, should I cut the lines 1 and 4 at the computer’s side? I imagine they’re now not needed since the devices and repeater chip now have power.

Thank you,
Steve

 

[Daveychef]

I dont know about dry wall but I have had success with pulling cables through walls (UK) in the past.
Chop the ends from your current cable and use it to pull the new one through. Tape round the join so it pulls through smoothly.

Sorry just a low tech possibility bit of thinking inside the wrong box.

 

[rjenkinsgb]

should I cut the lines 1 and 4 at the computer’s side?

No! Or rather, only the 5V one. The ground wire needs to remain connected end-to-end.
The data signals need a ground reference to keep them within the common mode range of the drivers and receivers.

I'd check continuity through 1-1 and 4-4 on one of the repeater cables. If the have a direct connection it should not matter which end power is fed from.

 

[Nigel Goodwin]

I dont know about dry wall but I have had success with pulling cables through walls (UK) in the past.
Chop the ends from your current cable and use it to pull the new one through. Tape round the join so it pulls through smoothly.

I imagine he probably build the cable in the wall as he built the wall?, so probably fastened it down to the studs - which makes pulling it through impossible. The obvious method is to fit ducting in the wall, so you can easily pull cables through afterwards - but it's rather late for that now.

 

[lnsstudio]

No! Or rather, only the 5V one. The ground wire needs to remain connected end-to-end.
The data signals need a ground reference to keep them within the common mode range of the drivers and receivers.

I'd check continuity through 1-1 and 4-4 on one of the repeater cables. If the have a direct connection it should not matter which end power is fed from.

rjenkinsgb,
Thanks, I was wondering about the ground at the computer end. As you suggested, I’ll cut only the +5V line at the computer side.
But before I do this, I will check continuity for power lines through the cable. Since the repeater chip appears to be not powered when the cable is ran backwards, it’s probably not straight through. Thanks for the help.
Also for other’s feedback input too. I can’t fish the line, there’s too many bends through studs and a 90 degree turn. It was hard to run without the drywall. Yes I feel like an idiot!

Steve

 

[Nigel Goodwin]

Yes I feel like an idiot!

Perhaps this might 'cheer you up'.

One of our customers (where I used to work) bought a 20m HDMI lead from us - he chiselled down the wall, ripped floorboards up, and chiselled up the wall the other side. He then installed the cable, refitted the floorboards, and re-plastered both walls. When all was dry he redecorated, then fitted his TV and equipment, to find it didn't work.

So he ripped it all out and brought it back, playing absolute hell - so we asked if he had installed it the correct way round? (long HDMI leads have an amplifier built-in, which is powered from one end, and obviously only works one way round - and the leads have directional arrows on them).

Needless to say he hadn't got a clue which way he had fitted it, and as both ends had been under plaster we couldn't tell from that either - so fairly obviously he had ignored the instructions, and fitted it the wrong way round.

 

[JimB]

Some years ago we were having some plasterboard (Drywall) work done.
I came home and my wife showed what had been done that day.

I recoiled in horror when I saw that the little junction box where the incoming phone line connects to the internal house wiring was no longer there.
Where the hell is it! was my reaction.

The guy doing the plasterboard work had buried it behind the plasterboard.
Probably he had been instructed to do this in the past by many fussy housewives.
Not in my house!
Next day it was nicely re-done with the little junction box visible and accessible.

As I said earlier:

As a general rule, never, ever, install something which could easily fail, in a position where it cannot be easily replaced.

And that includes simple junction boxes for the phone line.

JimB

 

[lnsstudio]

Perhaps this might 'cheer you up'.

One of our customers (where I used to work) bought a 20m HDMI lead from us - he chiselled down the wall, ripped floorboards up, and chiselled up the wall the other side. He then installed the cable, refitted the floorboards, and re-plastered both walls. When all was dry he redecorated, then fitted his TV and equipment, to find it didn't work.

So he ripped it all out and brought it back, playing absolute hell - so we asked if he had installed it the correct way round? (long HDMI leads have an amplifier built-in, which is powered from one end, and obviously only works one way round - and the leads have directional arrows on them).

Needless to say he hadn't got a clue which way he had fitted it, and as both ends had been under plaster we couldn't tell from that either - so fairly obviously he had ignored the instructions, and fitted it the wrong way round.

Great story Nigel. I’m still a bit ticked at myself, especially since I’m aware the power comes from the computer’s usb controller, not the device your plugging in. My mistake was not paying attention and was thinking to much on the gender changes to have “A” type female wall plates. Oh well, WTF is one to do but to move on. Since the USB connector is so small, I can’t get my meter’s probe in the connector without touching the damn metal sleeve. I need to trash one of my regular USB cables so I can determine if lines 1 and 4 power connections run straight through the cable with the repeater. I certainly would have thought it did, but now that’s in question.

 

[KeepItSimpleStupid]

We remodeled a Lab and initially I got in trouble for being in the construction zone. If it was done right, I should have been working in one area while the construction crew was there. My work really would not have interfereed, but no, everyone had to wait until the major renovations were done.

1. So "they" plastered the tip of the handle of a shutoff valve into the drywall.
2, They were making the doors open out. Opening so they would not slam into the face of someone in the ghallway in an emergency was one of the specs. I caught this BEFORE they completed the wrong doorway.

Finally, someone wised up and decided that a walk-thru was in order before the renovations were complete.

After they were done, the HVAC guy (an employee of where I worked) diganosed the $5000 air to water heat pump was bad. He asked for my help and I had to help him without getting caught by my micromanaging boss. After some troubleshooting, I told him to replace the tstat wire which he did with some grumbling. AC worked and I didnt rat on him. N one, but me and him knew that the HVAC unit did not have to be replaced.

Because I knew about the renovations, that was part of my inputs. It turns out when remodeling, no grommets were used in the metal studs and a short developed. He didn't argue about by dx.

The contracor tat built the building did some STUPID stuff. 36 air to water heat pumps..
1. With a set-back time and individual timer overrides the t-stats.
2. Building has a massive power factor problem
3. Building had massive outside air leaks.
4. They put a door on backwards which I exploited a few times.
5. To build an isolated concrete pad, they poured everything and then broke up the part the part that needed isolation and then poured that part.
6. They put two heat pumps in one lab with two separate thermostats. Really stupid.
7. They put 3-wire 60 Amp services to about 8 pieces of equipment that needed 4 wire.
8. No one really mentioned that the building was 208 rather than 240. The diffusion pump heating elements had to be changed.
9. We were supposed to have a "cooling loop". A 6" 90 F HVAC loop, as it turns out was suppsed to be the cooling loop. This rain at low pressure. Then they tried to add a booster pump to push this water through a 3/8" line at 80 PSI. The equipment needed 60F water or ground water. So, we ended up dumping lots of water down the drain just for cooling. It gets messyer.
10. People layed out their labs and then found out that there were large support columns where ther equipment was supposed to go.
11. They did the machine shop really close to right. They brought a high voltage feed there and put the breaker box in the shop.
12. The HVAC drain lines and the suspended ceiling conflicted. There wasn;t enough slope, so their was always a water leak.

It says something for HVAC - "Holiday, Vacations And Comp time.

My expensive mistake was not asking for help for a $100,00.00 budjet computer system upgrade from PDP-11's and x-y recorders when there was Windows 3.1 with 8.3 filenames and a crappy memory model fr the PC and the MAC with 255 character filename, a flat memory model. The software chosen was labView which was developed initially for the MAC and was in constant flux during development. I bit off more than I could chew in a linear fashion. I was managing a programmer too.

Both he and I didn;t know Labview. He wrote the software all under instrument simulation since I designed it that way.
I did a lot of proof of concept stuff and the instrument programming. Software worked for 17 years before radically upgraded.
i wanted to use a SMU, but I was told no by he powers that be. so we ended up with a "slower" system.

The SMU and the use of a server for data storage was adopted at the last upgrade that I know of. The PC was the obvious choice too.

The initial design plan was to have real time output. That turned out not to be possible because of set-up time. The printers were Laserjet 4m's and 4m plus which was a nice printer.

The other major issue was that we essentially had an embedded interface to a monochometer with the PDP-11 and there were no monochometers with a controllable interface. During the development of another system which used a DSP lock-in, monochometer and a really nifty I-V converter designed by me and layed out by me. The only in budget choice I had was to try to develop a motion control Labview vi with really cool stepper drivers. I was unable to complete that project.

I would have had to develop motion control vi's and develop an ORG sensor.

I did develop , machined and programmed a filter wheel which we did not have to use. I also developed shutters, of which one I used. So, the $1000.00 USD in hardware that I set aside is controlling one single shutter and I doubt my ex-employer has any clue how to replace it.

An IEEE-488 monochometer with a built-in filter wheel became available for $15,000. Results were inferior to what we had.

I had tow major issues:

1) The MAC RS-422 could not electrically interface to the RS-232 stepper controller without some help.

2) My I-V converter worked perfectly on devices made in house, but would not work correctl with large area calibration devices. I was sweating bullets.
a) My output stages would not drive the capacitance load.
b) I learned the hard way that that a few PA across a few milliohms causes a huge offset.

I was unable to complete the Zero check and zero correct circuitry and I did not allow for potentiometer nulling. I managed a 40 pA DC offset which didn't matter. The primary measurement was AC measured by the DSP lock-in.

I gave my boss, at the time the following "The technology to do the upgrade isn't quite ready yet. He said "tough", the money is. We had to do it.