Hoover Vacuum Power PCB, Putting it all together

Discussion in 'The Projects Forum' started by Feign, Mar 31, 2009.

  1. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    Threads 1 and 2, in General chat.
    http://forum.allaboutcircuits.com/showthread.php?t=21507
    http://forum.allaboutcircuits.com/showthread.php?t=21511

    Before I heat up the Iron.

    Does anyone have any suggestions of things I should test?

    My inclination is that the protection diode supplied and installed by the hoover contractor(cyntec) exceeded it's thermal dissipation, destroying its self and once it had "let out the magic smoke", the next usage caused the transistor to follow suit. It was brought up that this maybe in error and there might have been a short.

    This is quite the fancy vacuum, it has a main motor current sensor(stall protection?), main motor temperture sensor (over heat protection), several smaller motors, one spins the hepa filter to clean it, one raises and lowers the the ride height, one engauges a roller bearing to tension/start/stop the beater/brush bar, and an unidentified sensor to detect "embedded dirt" (it was glued in behind some foam, possibly an impact/shock sensor). All controlled by a ZiLOG 20pin microcontroller (as of yet not properly identified). With an exposed debug port! Aside from being broken, my only complaint is it is a bagless which means it's a bit dirty inside the housing nothing some stratigicly placed weather stripping wouldn't fix I am sure..

    It has six junction connectors, I haven't tracked down who(thats what wires) goes exactly where yet.

    J6: GND, CHA_LOC, V33(3.3v), CHA_POS, CHA_COL
    J5: TP1, GND
    J4: GND, POWER_SW, HA_KNOB, V33(3.3v)
    J3: HALL, GND, VHIGH(12v), EDF, EDF_GND
    J2: GND, DGB, V33
    J1: FLT_COL, VHIGH(12v)

    So What would you test? I only have one replacement NPN left.

    Please don't let me break anything else, but if I can't fix it I'll be scavenging the board for my recumbent tricycles onboard!

    Google turned up pure "Un-obtainium" for me when it sent me here last night, I think I'm moving in and camping on the front lawn.
     
  2. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Can you post large, sharp (in focus) photos of both sides of the PC board, and explain what wires connect where from the board?

    Going from the explanation is rather difficult.
     
  3. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    I most certainly can (they turn on my dsl 8 hours early YAY me, sorry I've been in dialup land for 3 years)

    Note: The + and - are reversed, permanet marker =)
    And the exposed weave is too much dremel.

    It is fully intact except the mentioned Q and D


    EDIT: All hyped up on adrenalin, Missed the what/where question.
    If the pictures are not clear enough let me know the labels are quite readable here.

    The Silver Pin in the lower right of the PCB with the covered up label is the AC in, the relay connects are both ground, Silver supplys the whole board, and the brass one goes to the motor, if there was a pin next to it it was cut and filled at the factory(Cyntec or OMRONS).
     
    Last edited: Mar 31, 2009
  4. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Check R45 and TR1, as well as the relay coil for too low/too high of resistances, less than a few ohms or several K. R45 might be a "Fuseable Resistor", it looks like the type they use in Compact Flourescent lamps. If it measures to the value shown by color code, shouldn't be any issues.

    On the Diode under the relay, what was there? I only see one cratered solder pad and two that appear unaffected, yet it has a "D3" label. If the soot is wiped away with isopropyl alcohol and a cloth, does anything else show up?
     
  5. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    The U? parts are

    U7: STmicro TL431
    U6: Sharp PC817 C(space)U3
    U5: Sharp PC817 C(rightangle triangle)TN
    U4: STMicro L1117 (33)
    U3: STmicro LM324
    U2: ZiLOG (something) The full text "layout" is in Thread #1 it's the "Mercury" chip

    D7 is Presumable the same as D3 (the carbon stain) I can only read the first digit 2 clearly the second looks to be a 4 but it got a dope/rosin bubble that covered half of it and I can't see anything else on it in full sinlight with the eye loupe


    Wire bundle connections to follow...
    U1: STmicro UC3843B
     
  6. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    R45 is labeled 104, (side note There is a faq/article for those in the volumes?), and measures ~4.5k Ohm in the board. and I get a polarised reaction, only get a reading with the V+ to the right side (away from Tr1)

    You're teasing me about the color code or did you mean R42 that one measured right at 470 Ohm

    R21 and R22 are 104 like R45 and read ~8.83k

    Relay measures 152 Ohm, it is rated at 155 in the sheet

    The intact pads near D3 are the coil inputs, It had a SMT diode similar to the 3 pinned D7 under the Sharpie "Mercury" connected in the same way single pin to the right and it's lower right pin was floating.

    I have a 1N4148 slated for that spot(D3), from Thread #2 above a 1n4k series should provide enough current drain. I estimated the power disipation at almost 1w in Thread #2. I still have not found, or really looked hard, for the formula to calculate the "possible" back EMF.

    My other quandry about power disipation, unvoiced and thus un answered, pertains to the power disipation formula.


    W​
    MAX = (TJ max – TA3)/θJA



    Is there a article here that explains that, I'm most concerned about time to Wmax relationship?
     
  7. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    I meant where D3 was, there only seems to be one solder pad. Where is the other solder pad for it? Is it a very small SMD diode that was the width of the crater?

    --ETA: 104 = 10k (just like color codes). 10 are the significant digits, 4 means add 3 zeroes. So 10,000 Ohms.
    On a capacitor, 104 is similar, 10 are significant digits, and 4 means add 4 zeros, except the answer is in picofarads, so 104 = 100,000 picofarads = 100nF = 0.1uF
     
  8. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    There was a pad there. I used a dremel to remove the burnt Q3, you can still see abit of the trace to the other coil pad if you look really really hard, so the pad you're wondering about is dust on my work bench. i didn't mean to destroy it, it happened really quick. It may have been abit charred, more so than the one that is still there, so came off in a flurry of dust. ;)

    Thank you for the refresher, I remember reading the SMT markings somewhere now.

    I'll go look at the wire bundles right now,
    I expect the CHA_??? Pins goto the brush height controller
    the EDF pins to the "embedded dirt" sensor and led board
    the FLT connector to the hepa filter
    and DBG is a debug port for the Zilog chip.

    *I just hit the wrong button and lost a paragraph, whoops*

    CHA_COL/POS drive the ride height can motor and gearbox CHA_P is a varistor position feedback
    EDF the "dirt finder" leds/transistor/resistor subboard
    FLT the hepa can motor and gearbox
    HAL is a hal effect sensor monitoring the main vac motor's output to a fly wheel (two magnets embedded)
    I left out the HA_KNOB before, varistor that sets ride height..

    the J5 con goes to an unknown board, it's in the dirt path just before the main dust bin. might be temp might be impact, might be both. It is connected thru the R19/C13 to the LM324.

    *shortor version of the lost paragraph*
    There were what might have been drywall debris in the dust bin. They either sucked up something BAD and stalled several things at once. Or ran it way past hot and D3 failed because of excess heat buildup... Remeber I estimated at -40C only ~.8w room temp is like half that for a SO/SOT.
     
    Last edited: Mar 31, 2009
  9. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    Just FYI this thing has more screws than a looney bin!

    I am going to start warming up the soldering iron. Unless someone has questions or comments in the next 5-10mins, while I wash up and pollute my lungs.
     
  10. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    :) Success!

    :( I only know where two of these extras fours screws came from.

    :D Putting the attachements back in, the upholestry brush definatly has water spots, it is NOT a wet dry vac.

    :eek: I got lots more broken stuff...
     
  11. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    What are the existing transistors? Did you get any part of the value from the one that is missing? If the existing are all the same, or two different for PNP and NPN, that will give a good idea of what the working voltages are.

    I ripped the images out of the PDF for you and am posting them here.

    Can you explain where the other end of the surface mount diode was, so an approximate size can be SWAGed?

    Thumbnails are Full Size View. next post has Forum Sized images
     
    Last edited: Mar 31, 2009
  12. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Here are mid-sized images of the above.

    [​IMG]

    [​IMG]
     
  13. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Power question is for later.

    No, not teasing about the resistor, but I meant R46. Some are meant to open quickly if they get "hot", like a fuse, and still work like a resistor.

    On R45, that is the surface mount on the underside marked "104", but you only get a resistance reading on it in one direction?

    Quick Question: What is your meter? Does it put out over 0.4V on resistance check? If not, something is whacky with R45, since 1k or less should appear across it no matter the polarity.
     
  14. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    Answers to post #11:

    About the PDF, the posting guidelines recommended useing PDF as the prefered format, I took the pictures in max resolution TIFF format and converted them to jpeg with photoshop originaly, I did not change the default quality setting from ~8 and they were too big for the forums jpeg limits, so i went with pdf since the dsl was alive when I got home from the store. Applogies I was giddy and size didn't matter right then. The upside down pictures, I should have rotated for clarity or took them the correct way up. I have camera issues one of which is a broken battery door, and was just happy it let me take them. Even on the tripod I was triggering the shake sensor/software.

    It is all screwed together already but I'll answer what I can.

    Q9 Directly above(below in the pictures) the Q1 is a 2N3904 also. I believe I read some of the part number, and extrapolated the portion that was in the damaged area from Q9. If the other Q part numbers are listed in Threads 1 or 2 in the general section they must have been in one of the posts I had to redo due to forum timeouts and my web browser cache settings.

    I need to take it apart tomarrow, to find the wholes for the two extra screws, and install the two I left out semi intentionaly. Two are buried somewhere I some how missed, and Two are part of the four that hold the handle holder in.

    I can try to get readable pictures of the part numbers for the other Q parts then, excluding Q2 you see bent over on the topside I am not brave enough to straighten it, if I break another NPN I will have to remove one from something else or visit the parts store (~20 mile trip), I believe they are all or mostly all 2N3904's.

    The board uses TTL(?) 3.3v and 12v power busses, measured across GDN and the V33 and VHIGH labeled pins. I then traced these lines around the board via the poor mans light table method, sun light thru the window a finger for marking the traces i was tracking and manual rotation. It works well unless the trace runs under a large chip.

    I believe D1 was a sot-23 foot print, one pin to the sharpie negative one pin to the sharpie positive, and the other appeared to be floating. Most of the chip was destroyed, I removed the first portion just brushing it with my finger the pins were melted into the pcb. So ~2.5mmX1.5mm, I measured D7 (bottom side of the board near the the sharpie V. 1.0.0) with my slide caliper. In the high resolution images you can make out a black circle, this is the smoke ring the diode left whe it self destructed.


    Answers to #13

    You asked about R45 and TR1 back in post #4, I never did check TR1 !
    R46 is on the Front Image lower left, between the fuse and TR1
    R45 is located on the Rear image near the mercury sticker (removed from U2)
    R45 connects to TR1 pin 5, D9's (anode?) high side, and C24. I believe the anomolous reading (~4.7k?) is due to the alternate current path via TR1 D9 and C24.

    The one way reading was achieved on R22 or R21, Back image upper under the LM34XX.
    I was takeing these readings to compare with the R45 reading, waiting for the refreshment on SMT markings you gave me. There are alternate pathways in this area also, and a very tight space for my meters terminals. You can see the traces in the larger resolution images that connect, pin "1" of R22 pin "1" of C21 and pin 4 of U3(LM34XX), pin "2" of R22 and pin 5 of U3, and pin "2" of C21 and pin 11 of U3.

    There is some interaction here, which proves why you shouldn't treat in board component testing as 100% acurate, One thing I do remeber from my electronics classes. If you can not isolate the part to test it you can and should not treat them as more tha a reference.

    I also ment to inquire about C33, is that the worlds smallest SMT or why is it not there? What was it's purpose? Was the ST L34XX not the original part choice? Was it "cruft" in the schematic? Good luck getting an answer from hoover or thier pcb contractor. I'm certain it is not un the LM34 or it is again, the worlds smallest most transparent component I've never layed eyes on..


    I am useing a walmart ~10USD Digital Meter, "GB Instruments GDT-11", I am not sure where the booklet is off hand.

    Well I just re read my quoted post getting your post number to seperate these answers and I did say it was the first resistor you asked about that gave the one way reading. there is AC current there, and high voltage judgeing by the slit in the pcb under, I can't see the designation, the large transistor on the Front image to the left of D9 between the resistors. And maybe residual charge in C33 from me flipping meter terminals

    On the front image lower left the connector is the AC input, flip to the Rear image AC_in lower right corner, conenction path is AC_in (you can make out the top and bottom traces here due to "light table effect" ) -> D1(H bridge?) follow the upper trace to TR1 Pin 3, not sure where the path goes from there thats the wrong pin for D9 to conduct... Oh I forgot to short that 250V Polarised Capacitor that explains the alot, I remeber some tingleing aswell now. Leason learned, hopefully..
     
  15. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    Well look at that I wrote a short novel!
     
  16. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    The reason I ask about the resistor is they are not polarity sensitive, and the resistance shouldn't measure any higher than the value of the resistor in either direction.

    Did you determine what caused the diode to evaporate? Did Q9 die first, or the diode?

    Those chicken/egg problems can be nightmares, as once you replace both semiconductors, and power it up, they will simply blow up again if the real issue isn't corrected as well.
     
  17. Feign

    Thread Starter Active Member

    Mar 30, 2009
    50
    2
    Interesting I KNOW I just responded to this, wonder what happened. Some one might of just got a wierd post in thier thread...

    Found the zilog data ... First link I tryed *shrug*

    ZiLOG Z8Encore 8F0223
     
Loading...