XV-21 Short clean time, getting lost, not docking: SOLVED

News and information about the Neato XV-11 Robotic Vacuum. All discussion and troubleshooting questions go here.
Post Reply
bretzle
Posts: 4
Joined: November 21st, 2012, 5:50 am

XV-21 Short clean time, getting lost, not docking: SOLVED

Post by bretzle »

I found a solution to this problem: Warranty warning!!

Neato XV-21 was working great for a week or so, then I noticed the cat hair was piling up again.

After a few tests it became apparent something was wrong. It would start cleaning, doing short circles not far from the dock, exclaiming it was finished! and was returning to the dock. It would do some clumsy circles again, and die not far from the dock. After a quick search on the net I began to suspect a dodgy wall sensor.

I confirmed this with an IR capable camera, there was no visible light coming from the wall sensor. Confirmed confirmed with a USB debug.

This is what I did:

Plugged it in to my mac, ran a terminal program, in this case CoolTerm (free), selected the right com port, hit connect (typing help should confirm that it's connected).

Testmode on <- fires up the sensors, don't forget to type Testmode off when you're done!
GetAnalogSensors <- gets reading from the analog sensors

My wall sensor reading was -24. It shouldn't be a negative figure. I work with analog sensors and the ones I'm used to work on a range of 4-20mA, 4mA being zero and 20mA being 100, i'm guessing Neato does something similar. If you get a negative figure it means it's getting 0mA through the circuit, ie no signal. Could be a broken return wire but the lack of any IR glow from the sensor suggests it wasn't getting power.

I had to disassemble the entire unit to get at the sensor in the front bumper, don't do this if you have a chance of warranty return, you have to break the seals to get in. I jiggled the wires and checked the readings again. It was good.

I suspect that the wall sensor wires take a bit of a beating when it runs into things as they seem to hang in mid air from the plug to the main body of the Neato. I doubled the wires back on themselves and hot glued them to the sensor body itself to take any strain off of them, then ran them back the way they were in the first place. I put it back together, checked the sensors, slapped it round a bit, checked, slapped it again, checked. So far so good.

It's been working well for almost a week now, I know it's a bandaid fix but I'm in Australia so getting spares or organising a warranty return are a bit of a headache. If it dies in the arse again I'll order a new sensor.

Ps. I've read elsewhere you can substitue a cliff sensor for the wall sensor, you'd probably have to put a resistor in the circuit of the snavelled cliff sensor to fool it into thinking it's ok. But then, I'm not sure how Neato would like a trip down the stairs.
glnc222
Robot Master
Posts: 5490
Joined: January 23rd, 2012, 8:19 pm
Location: The Villages, Florida
Contact:

Re: XV-21 Short clean time, getting lost, not docking: SOLVE

Post by glnc222 »

Wall sensor Sharp GP2Y0A51SK0F data sheet:
http://www.beck-elektronik.de/fileadmin ... 51sk0f.pdf
dpostman
Posts: 7
Joined: November 15th, 2012, 8:23 am

Re: XV-21 Short clean time, getting lost, not docking: SOLVE

Post by dpostman »

Document was signed by Yoda...

"Must work it will"
bretzle
Posts: 4
Joined: November 21st, 2012, 5:50 am

Re: XV-21 Short clean time, getting lost, not docking: SOLVE

Post by bretzle »

Thanks glnc222! I was close.

HA! eagle eyes dpostman, love it!
jgo
Posts: 13
Joined: May 23rd, 2011, 7:00 am

Re: XV-21 Short clean time, getting lost, not docking: SOLVE

Post by jgo »

Bretzle,

Thank you so much for posting this.

I have just enough computer/electronics experience to follow your instructions, but I don't have the knowledge necessary to do this independently.

My neato was exhibiting the same erratic behavior that you describe, and after following your instructions I got the exact same reading on the wall sensor. Now I will follow up with checking the cable. At least the wall sensor isn't a horribly expensive part to replace.

This will make the second repair I have done on the unit. The first one was to create a new anchor point for a wheel spring by using an eye screw and some epoxy. Some minor hassles, but the unit has served me well for the past 2+ years.
TeflonDom
Posts: 3
Joined: March 2nd, 2013, 1:40 pm

Re: XV-21 Short clean time, getting lost, not docking: SOLVE

Post by TeflonDom »

This is exactly why I will not be buying a Neato.
tjhouchin
Posts: 4
Joined: February 17th, 2013, 4:10 pm

Re: XV-21 Short clean time, getting lost, not docking: SOLVE

Post by tjhouchin »

TeflonDom wrote:This is exactly why I will not be buying a Neato.
Where is the button to vote if this post was useful or not?
jgo
Posts: 13
Joined: May 23rd, 2011, 7:00 am

Re: XV-21 Short clean time, getting lost, not docking: SOLVE

Post by jgo »

TeflonDom wrote:This is exactly why I will not be buying a Neato.

I don't know...

Over the years I have had all kinds of electronic gadgets, and for something that operates as actively and often as my robotic vacuum cleaners (1 Roomba, 1 Neato xv-11) I am satisfied with the product's failure rate. Could it have been designed better? Probably so (fatigued plastic wheel spring anchors). However, regarding the Neato in particular: no other robotic vacuum cleaner that I have seen can deal so intelligently and efficiently with the multi-room basement floor of my house.

When my xv-11 becomes irreparably broken I will buy the Kobold Vorwerk version.
Post Reply