Roomba 780 right wheel resistance with battery

The very latest news and updates for the iRobot Roomba robotic vacuum, the Scooba Robotic Washer and the Dirt Dog workshop sweeper. All discussion and troubleshooting questions go here.

Roomba 780 right wheel resistance with battery

Postby hotdog88gt » January 31st, 2014, 1:45 pm

A few weeks ago our Roomba 780 started giving us "error 5 right".

So we flipped it over, pulled both wheels and cleaned them. But I noticed something odd with the right wheel. When the battery is removed both wheels rotate OK, but when I put the battery back in the right wheel becomes harder to turn.

I ordered a new right wheel but it did not fix the problem, so I sent it back. iRobot wants $400 for a replacement 780 body only. Is there something else I can do to resolve this right wheel issue?
hotdog88gt
 
Posts: 2
Joined: January 31st, 2014, 12:47 pm

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » January 31st, 2014, 2:19 pm

hotdog88gt wrote:...our Roomba 780 started giving us "error 5 right". ...I noticed something odd with the right wheel. When the battery is removed both wheels rotate OK, but when I put the battery back in the right wheel becomes harder to turn. ...a new right wheel ... did not fix the problem, ... Is there something else I can do to resolve this right wheel issue?
Yes, there is another option that involves locating a faulty mobo component, heating up your soldering iron and replacing it.

Great trouble report, and fix-attempt given in your OP!

Your Wheel-Module-in-robot finding (more effort to turn wheel) implies its motor is being lightly powered by the System. There is insufficient voltage being applied to cause rotation of the motor's shaft, is why I say "lightly powered". However, that does suggest there has been a failure of (at least one) power transistor in what is called the motor's H-bridge (of four power xstrs). Search this site for bridge and read about it. All that is said about the 500 series, specific to H-bridge, should also apply to your 780.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Roomba 780 right wheel resistance with battery

Postby hotdog88gt » January 31st, 2014, 3:47 pm

You say the "motor's H-bridge". This is external to the wheel module?

And thanks for the info.
hotdog88gt
 
Posts: 2
Joined: January 31st, 2014, 12:47 pm

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » January 31st, 2014, 4:06 pm

Yes, external. All motors' H-bridges are located on Roomba's main-board = mobo = main_PCA; and for wheel-motors and the main_brush motor they are easily seen on 5XX mobos. However, the 7XX series has muddied that main-board name (due to additional large PCAs that are close-coupled/connected to what is actually its main-board), and it appears that bridge transistors may have been changed to other packing forms.

vic7767 posted a picture showing inside a 7XX chassis. I'll try to find a link to it. Meanwhile, follow my recommended search for "bridge" to learn about the 5XX Roomba work that has been done. I do not know of any repair studies for the 7XXs to reference--you are the pathfinder!

Added: This is vic7767's image:http://www.robotreviews.com/chat/download/file.php?id=4756&mode=view
Source: -robotreviews.com/chat/viewtopic.php?p=100808#p100808
This is an annoying .GIF format with three images in it. You want to study the image showing much of Roomba's chassis. I expect the largest board to be the "Main_PCA". Most likely, the motor-controllers will be mounted to its bottom face.


I should point out that I know nothing about the 7XX-series, except what I have read in these forum posts; AND, only a short time passes before the memory of having read them vanishes.
{Edit of 140201 basically redacts (via dim font color) vic7767's .GIF file, since he later posted superior 7XX_PCA images.}
Last edited by Gordon on February 1st, 2014, 4:46 pm, edited 1 time in total.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Roomba 780 right wheel resistance with battery

Postby vic7767 » January 31st, 2014, 6:26 pm

Gordon wrote:This is an annoying .GIF format with three images in it


Here are some better images of a 760 PCA including the underside. I had never disassembled a 7XX down to where the underside could be viewed. To my surprise I did not observe any of the larger TO-220 style H-bridge components.
Attachments
760 PCA 001.JPG
760 PCA 002.JPG
760 PCA 003.JPG
Roomba and Neato Mods, come visit: http://www.vic7767.com/

http://www.Robot-Doc.com/
User avatar
vic7767
Robot Master
 
Posts: 14386
Joined: January 14th, 2006, 8:31 pm
Location: Haughton Louisiana - USA

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » January 31st, 2014, 6:55 pm

Thnx vic, I will attempt to outline suspected SMD PWR xstrs that provide H-bridge service. What I say in this post is all tentative, since I am writing about what I think I see in your two views (upper face and bottom face) of the 7XX main_PCA. I do not plan to say anything about the smaller User-Interface PCA (UI_PCA). No longer can we spot motor H-bridges by looking for the H-shaped layout of power transistors, and four adjacent power diodes that dissipate generated energy when power to a motor is abruptly switched off.

The 7XX's main_PCA appears to be using two each 8-pin SMDs for each large-motor bridge. Examples are: Q83 & Q84 (main_PCA, upper face, front/left, looking like Lft Wheel Motor). Since only two packages per motor are seen on the board, each package contains two xstrs, BJTs rather than FETs, one PNP and one NPN with common collectors. All pins on on side of the SMD package are soldered to healthy copper heat sinks expanses. Also, very likely, we will learn that each transistor has its collector and emitter shunted with a suppressor diode that is ordinarily seen separate.

I copied your main_PCA pix and subdued their resolution a bit, and then superimposed colored rectangles around what must be the H-bridges, and added arrows pointing to module connectors. In your view of the upper face of the main_PCA I used RED for markings related to the Left-Wheel Motor bridge, GRN for the Right-Motor bridge, and AQUA for the Main Brush Motor:
main_PCA-upperMedRez.png
Discrete power diodes for those three large motors are marked with same colored rectangles on the bottom face of the main_PCA:
main_PCA-bottomMedRez.png
Those diode suppress voltage generated by a motor shaft rotating to a stop.
The Side Brush Motor continues to use four TO-92 cased BJTs, and four discrete power diodes to form that motor's steering bridge. ORG indicates the Side Brush Module's spring contacts and discrete xstrs and diodes, all seen mounted on the bottom face of the main_PCA. Nominally the same layout as seen for the R3 main_PCA.
As I said above, I have not held this hardware in my hands to inspect it, therefore I could easily have made some errors in what I am reporting.
See if you can read ID numbers on Q83 & Q84 (main_PCA, upper face, front/left, looking like Lft Wheel Motor).
Last edited by Gordon on February 1st, 2014, 11:00 pm, edited 1 time in total.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » February 1st, 2014, 2:47 pm

Edited to remove the PNP/NPN example half-bridge that was shown in this space prior to obtaining the actual STMicroelectronics part number of the board mounted components.
Last edited by Gordon on February 2nd, 2014, 1:15 am, edited 1 time in total.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Roomba 780 right wheel resistance with battery

Postby vic7767 » February 1st, 2014, 3:27 pm

I will look for the markings on the Qs when I get a chance to open up the 760 again. I took some image shots and put the Roomba back together and then discovered that the pictures do not reveal the detail quality I anticipated.
Roomba and Neato Mods, come visit: http://www.vic7767.com/

http://www.Robot-Doc.com/
User avatar
vic7767
Robot Master
 
Posts: 14386
Joined: January 14th, 2006, 8:31 pm
Location: Haughton Louisiana - USA

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » February 1st, 2014, 4:21 pm

vic7767 wrote:I will look for the markings on the Qs when I get a chance to open up the 760 again. I took some image shots and put the Roomba back together and then discovered that the pictures do not reveal the detail quality I anticipated.
Great! Those images are high enough rez to read some board markers, but not to resolve very many of the IC's package/case markings.

My present feeling regarding failure of any of these SOIC-8 (H-bridge halves) that iRobot's EE-packaging engineer has applied to 7XX's main_PCA is: Failure due to board-design and/or component specifications seems highly remote; and, if any were to fail, I would have to ask if the duff device had been formed at a 3rd-party silicon-foundry (i.e., a low reliability 'knock-off').

My current level of confidence in 7XX's H-bridge semiconductors (including their good-looking PCB heat-sinks) translates into a higher EE-qualification level for troubleshooting OP hotdog88gt's (possibly) powered wheel motor!
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Roomba 780 right wheel resistance with battery

Postby vic7767 » February 1st, 2014, 9:05 pm

Here is another image of an area of the PCA where the Left drive motor and Brush deck motor H-bridge components are mounted. A little lighting and lens angle trickery was needed in order to capture the stamped part numbers.
Attachments
004.JPG
Last edited by vic7767 on February 1st, 2014, 9:50 pm, edited 1 time in total.
Roomba and Neato Mods, come visit: http://www.vic7767.com/

http://www.Robot-Doc.com/
User avatar
vic7767
Robot Master
 
Posts: 14386
Joined: January 14th, 2006, 8:31 pm
Location: Haughton Louisiana - USA

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » February 1st, 2014, 9:18 pm

vic7767 wrote:... A little lighting and lens angle trickery was needed in order to capture the stamped part numbers.
That oblique LOS & lighting is frequently needed to resolve IC markings.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » February 2nd, 2014, 2:23 am

vic's HiRez shot of 7XX's main_PCA (its front/left corner/top-face) gives a clear view of the mfr's marking of Q66. It is an ST (STMicroelectronics) component marked: "8(5H30L) | EP044". After milling around http://www.st.com for half an hour, looking / searching for the component I finally decided to try a wild-card search using the characters that were most certain, namely *5H30L*, <Ctrl>, and bam! Out it came as: STS8C5H30L, with no competition! Its a DUAL, Nchan/Pchan MOSFET, just two chummy FETs in the same package--no internal connection between the pair! Here is a view of the upper portion of its datasheet, page one:
7XX_half-bridge.png
RdsOn values of 0.018 ohms and 0.045 ohms, both at MAX currents that would not normally be reached in Roomba, insures much lower RdsOn values when operated way below five or eight amperes!
Pairs of this half-bridge device are grouped in three places around the board to provide H-bridge service to both wheel motors, and to the CHM's main-brush-motor.

vic's image also provides locations for pairs of protective (bridge FETs' protection) power diodes that are required to snub generated voltage when a motor's speed quickly slows. When a bridge is built with bipolar xstrs, each xstr is shunted with a protection diode, totaling four diodes per bridge. On this 7XX board the designer decided to use the FETs' "body diodes" (they are shown in the above data-sheet snippet) in one bridge-half for its protection. But, in the opposite half, the body diodes are supplemented with parallel connected discrete diodes (mounted to the PCB's underside). Curious doings.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Roomba 780 right wheel resistance with battery

Postby mfortuna » February 2nd, 2014, 8:21 am

I believe I see a total of 8 dual diodes, each connected to two drain pins?
Mike
Reds x 3, Dirt Dog, Disco (now a parts bot), Create, Scooba 350, and Security Dawg
Evolution Mint
Neato XV-11
User avatar
mfortuna
Robot Master
 
Posts: 5771
Joined: February 5th, 2006, 10:35 am
Location: NH

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » February 3rd, 2014, 1:56 am

Mike, I take it you are pointing to the SOT-23 style diode packages such as D207 & D209 with "Z6" marking. If so, those dual diodes connect power FETs' gates to sources.

To see the power diodes that handle motor generated voltage look inside the small colored rectangles that I marked on a copy of vic's main_PCA bottom-face image, here: viewtopic.php?p=121853#p121853.

Two cases sizes continue to be in use: 1) Diode-Outline size 0.1"-dia x 0.2" LG (like a 1N1007), for small motors; and 2) Diode-Outline size 0.13" x 0.25" LG (between 1N1007 and 1N5401 outlines) for the CHM Brush-motor. Thus, 2 of O/L(2) for the CHM, and 4 of O/L(1) for the wheel motors is what I see.

The Side Brush bridge appears to be identical to that used for 500-series, hence makes use of four O/L(1) diodes--which I did not include in my report of "six" in association with the new MOSFET half-bridge, SO-8 package devices. The Side Brush is out of scope for this thread.

Somewhere on the board will be located a transistor switch (no bridge reversing required) for the fan motor, and with it there will be an O/L(1) diode around the fan motor. The Bin Fan is out of scope for this thread.
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Re: Roomba 780 right wheel resistance with battery

Postby pawel2m » February 14th, 2014, 3:34 pm

Hello, guys, I have currently two opened 780, both after water damage, first has dead CHM, second, dead CHM and right wheel, 7xx board is hard to measure due to heavy tin oxidation, I have to srub pins to measure anything. I believe it's lead free tin - oxidates as hell.

After hours of measuring and tracing I managed to replace some dead transistors and resistors and one roomba woke up asking to repair her right wheel, but she forget to mention that her CHM is not spinning too. I about to replace both Q59 and Q60 MOSFETS for right wheel and both Q66 and Q67 for CHM as this is my last hope....
All four are marked as on attached photo C3035LD.

1.jpg
User avatar
pawel2m
 
Posts: 14
Joined: November 16th, 2013, 9:04 am

Re: Roomba 780 right wheel resistance with battery

Postby TechGuy » February 14th, 2014, 3:50 pm

pawel2m wrote:Hello, guys, I have currently two opened 780, both after water damage, first has dead CHM, second, dead CHM and right wheel, 7xx board is hard to measure due to heavy tin oxidation, I have to srub pins to measure anything. I believe it's lead free tin - oxidates as hell.

After hours of measuring and tracing I managed to replace some dead transistors and resistors and one roomba woke up asking to repair her right wheel, but she forget to mention that her CHM is not spinning too. I about to replace both Q59 and Q60 MOSFETS for right wheel and both Q66 and Q67 for CHM as this is my last hope....
All four are marked as on attached photo C3035LD.

1.jpg

Have you confirmed that the CHM motor is working by using a 9V battery?
If not, test it first.
If the CHM motor is not turning, turn the motor by hand to clean the motor brushes and rotor contacts.
Make sure the CHM motor is working before changin components.
Charging battery directly: 400 Series, 500 Series,. How to Desolder
For Sale: Black CHM 2 Ball Bearings set
User avatar
TechGuy
Robot Addict
 
Posts: 1501
Joined: February 17th, 2009, 11:52 pm

Re: Roomba 780 right wheel resistance with battery

Postby pawel2m » February 14th, 2014, 4:01 pm

Of course, CHM works fine in other Roomba. I have compared attached photos of 760 board with my 780 and I can confirm that they share the same mobo, the only difference is in add-on scheduling board where 770 has no Wifi module and no soft touch buttons.
User avatar
pawel2m
 
Posts: 14
Joined: November 16th, 2013, 9:04 am

Re: Roomba 780 right wheel resistance with battery

Postby TechGuy » February 14th, 2014, 5:53 pm

Good luck with your repair with the CHM driver
Charging battery directly: 400 Series, 500 Series,. How to Desolder
For Sale: Black CHM 2 Ball Bearings set
User avatar
TechGuy
Robot Addict
 
Posts: 1501
Joined: February 17th, 2009, 11:52 pm

Re: Roomba 780 right wheel resistance with battery

Postby pawel2m » February 14th, 2014, 6:14 pm

Thx, when I buy and replace the C3035LD / STS8C5H30L I will let you know if it helped.
Last edited by pawel2m on February 15th, 2014, 7:35 am, edited 1 time in total.
User avatar
pawel2m
 
Posts: 14
Joined: November 16th, 2013, 9:04 am

Re: Roomba 780 right wheel resistance with battery

Postby Gordon » February 14th, 2014, 7:08 pm

pawel2m wrote:Thx, when I buy and replace the C3035LD / STS8C5H30L I will you know if it helped.
Hello pawel2m. I tend to think that replacement of the dual MOSFETs bridge xstrs is a hasty decision. IOW, those dual FETs should not have been damaged by water, but possibly their Gates could be shunted by corrosion products that precipitated around their pads and over to nearby pads(?). Hence, replacement of the FETs' package may not overcome Roomba's problems.

You should be able to make a series of powered electrical measurements that verify FET Gates are HI when need be and LO when need be. If they don't change when a motor is commanded (by the MCU) to operate, you would shift your points of checking to the BJT xstrs that drive the FET Gates. If their Base terminals are not active (following a command change) you would shift attention further upstream toward the MCU, and so on.

After reading this msg of yours:
... I managed to replace some dead transistors and resistors and one roomba woke up asking to repair her right wheel, but she forget to mention that her CHM is not spinning too. I about to replace both Q59 and Q60 MOSFETS for right wheel and both Q66 and Q67 for CHM as this is my last hope...
...it appears to me that you are not taking advantage of Roomba's Built-in-Tests, (BiTs). Use of the BiTs will give you the comfort of turning ON a motor is test mode and keeping it commanded ON while you make voltage measurements at selected test points. If you need to learn what it takes to put a 7XX-Roomba into BiTs' Mode, read about it in this thread (found in the READ FIRST sticky): viewtopic.php?f=1&t=15192&p=99276#p99276

The 7XXX Roomba is too new for anyone to have reverse-engineered any schematic diagrams for it, however, we have observed that some sub-function circuits have a lot in common with predecessor Roomba-R3 (the 500s, & 600s). You may be able to sketch a schematic diagram of your CHM-motor driver H-bridge by using the information in this thread: viewtopic.php?p=112075#p112075
...as a template. Of particular importance in that R3-circuit is the use of a 74HC02 Quad NOR-gate to interface MCU ports to each motor's bridge inputs. Try to locate that device, or a similar quad NOR-gate IC near your brush-motor's H-bridge--this is a CMOS device that could be susceptible to shunt resistances formed on the PCB by drying of liquid contaminants.

What? You don't know what the signals should look like on that troubled bridge circuit? Find out by doing the measurement(s) of interest on a motor-bridge (use a wheel-motor's bridge to aid learning) that does work!
Gordon
Robot Master
 
Posts: 4304
Joined: April 6th, 2005, 2:02 am
Location: Santa Ynez, CA USA

Next

Return to iRobot Roomba and Scooba Chat

Who is online

Users browsing this forum: 911medic, vic7767 and 467 guests