Jump to content
  • Announcements

    • Steve Herschbach

      Archives Closed - New Forums   10/16/2017

      The old archive system has been closed and the threads moved to new forums. See the full forum listing here. Detailed explanation here.

Recommended Posts

51 minutes ago, Jennifer said:

In all fairness to Minelab, we have no idea how old this unit is or how rough it was treated. 

[...]

Dont judge the build quality without knowing the history of the unit.... :smile:

We have no need to know how old the unit is or how rough it was treated because in the video we can see the history we need.  I.e. how it was designed.  Which is poorly when it comes to this wire connection.  The ML design here connects some discrete wires to a PCB by use of a simple through hole and a blob of solder.   That's pretty much a design fail for reliability and ruggedness in gear that will be constantly be moved / handled.

There are lots of types of discrete wire to board connectors available and ML used none of them here.  Such connectors exist because otherwise stuff likes this happens. 

Given the label for the connection, "TXA"- transmit A, my (limited) understanding of the Russian narration, and the fact the box is open and under repair, it seems the failed / intermittent connection is pretty important.

Share this post


Link to post
Share on other sites

2 hours ago, MDdetector said:

We have no need to know how old the unit is or how rough it was treated because in the video we can see the history we need.  I.e. how it was designed.  Which is poorly when it comes to this wire connection.  The ML design here connects some discrete wires to a PCB by use of a simple through hole and a blob of solder.   That's pretty much a design fail for reliability and ruggedness in gear that will be constantly be moved / handled.

There are lots of types of discrete wire to board connectors available and ML used none of them here.  Such connectors exist because otherwise stuff likes this happens. 

Given the label for the connection, "TXA"- transmit A, my (limited) understanding of the Russian narration, and the fact the box is open and under repair, it seems the failed / intermittent connection is pretty important.

Through hole soldering is quite common/the industry norm and the strongest method to attach wires to a board when there is no to remove the wires or use a pin header, molex etc.

It's not simply a wire through a hole and a blob of solder as you put it. The wire is tinned and inserted in the hole which is lined with a pre-tinned copper trace which has a round tinned solder tab on the other side and then it is soldered, usually via moving the board above a sheet of molton solder, this is called wave soldering. The surface area of the copper inside the hole and the solder pad give great mechanical strength once cooled. If moved during cooling, a week bond or cold solder joint (indicated by a dull grey finish, not shiny as preferred) can result, as it appears happened here. The above mentioned traces are chemically etched into the printed circuit boards and normally last the life of the board and do not lift.

In relation to our detectors moving, the relationship between the board and connection for the coil that this wire uses is relatively non moving, housed in the same case, they are static to each other, other than jostle from the wire shaking and as I'm sure they used stranded core and not solid core, again sufficient.

These two board photos are the solonoid boards from one of my M5's earlier mentioned. They adjust the engine VANOS timing by allowing or disallowing engine oil under pressure to change the camshaft position. Notice anything in the photos? Exactly the same wire attachment method... (on the little green printed circuit boards, orange arrow. The main board has pins but the bridging boards between the wires and solonoid leads are straight common through hole soldered to solder trace tabs).

I reckon Minelab thought if it was good enough for a $100,000.00 German BMW in the engine compartment, under the constant vibration of a V-8 engine, it was good enough for them, and I concur.

PS: I do concur it appears to be a cold solder joint that opened up the connection but that speaks more to an isolated one off faulty solder as the board transitioned over the wave solder table, not indicative of a bad wiring connection design.....

Note: There's a good chance that the soldering of these wires into the already populated (with components) and wave soldered board is a manual process and as a result was simply human error.

Just a guess but TX A is most likely transmit coil A winding.

Jen

IMG_2436.PNG

IMG_2437.PNG

 

 

  • Like 3

Share this post


Link to post
Share on other sites

I have had detectors made by every brand fail on me. The more anything costs, the higher my expectations get for quality design and durability. Yet as a person who ran a service dept I can testify that is not reality. The best companies make good products, then use a feedback loop based on warranty submissions to improve those things that fail. As a rule this means buying first year product makes you part of the final bug check, with models made in the second and third years of manufacture becoming more reliable as engineering design faults and manufacturing errors are identified and corrected.

Loose intermittent connections and other intermittent electrical issues were usually the worst. Typical failure report by customer and works fine in shop. We sometimes spent days tracking down an issue that took five minutes to fix. You ever try billing a customer ten hours for a five minute fix? Fun stuff! :smile:

Interesting look under the hood - thanks Jen. Welcome to the forum MDdetector.

  • Like 2

Share this post


Link to post
Share on other sites

Notice anything on the photos ?

Yup  looks like ya had a couple of resistors removed ! 

Share this post


Link to post
Share on other sites
20 minutes ago, mungass said:

Notice anything on the photos ?

Yup  looks like ya had a couple of resistors removed ! 

They're not actually resistors but it's an easy mistake to make, notice that the stripes are all the same color, not different per a resistor color code/chart. These are MOV's (Metal Oxide Veristors" and for the repair in question, were being replaced by 1N4148 diodes.:smile:

And the reason I fixed it was because take off VANOS boards range from $2,300 per side (the engine has two, one per bank) up to $2,700.

http://www.ebay.com/itm/Genuine-BMW-E39-E52-M5-Z8-Solenoid-for-Vanos-Unit-NEW-11367831450-/122159471884?epid=1122097616&hash=item1c71459d0c:g:~BgAAOSwZJBX~-Fw&vxp=mtr

http://www.ebay.com/itm/BMW-E39-E52-M5-Z8-Solenoid-for-Vanos-Unit-Genuine-Brand-New-11-36-7-831-450-/361183151555?epid=1122097616&hash=item54183185c3:g:KOYAAOSwxYxUs8~7&vxp=mtr

 

Jen

Screen Shot 2017-08-08 at 4.26.41 PM.jpg

  • Like 1

Share this post


Link to post
Share on other sites

  Yup Jen do notice , a lot more obvious on the pc compared to the 8x4 tablet , and why wouldn,t you .

Many moons ago when I was learning the poker machine repair trade or slot machines as you call them over there , I was with a small private business , and we repaired every board we could , money was money , and a jumper wire didn,t cost more than a couple of beers  unlike the Aristocrat board jockeys who were our competition ,  rip it ,  swap it ,  send it back to factory when box was full . ha ha 

was checkin a couple of curley chords off the 3500 a few weeks ago that didn,t  work , checked continuity and both had none at same pin , took a punt and  cut 3 inches off the most stretched end    and   bingo  problem solved , resoldered ends and when wires of offcut  were stripped back just for interest ,  both white wires in the leads  had breaks   :blink:   ,mmmm  go figure , :cool:

  • Like 1

Share this post


Link to post
Share on other sites
On 8/6/2017 at 10:42 AM, Jennifer said:

It would be interesting to see this compared to a CTX3030.

Jen

Ask and ye shall recieve, inside a 3030 for comparison (he's re attaching his GPS module).

Jen

 

Share this post


Link to post
Share on other sites

Jen, thanks for the cool videos and thoughtful posts in this thread and the forum.

All, on the subject of "good design" or "poor design" to directly solder wires through thru holes I opine this is not a black and white true or false issue.   I started the first 13 years of my electrical engineering career designing airborne electronics for the USA (and friends') Dept of Defense.  In this environment, such a termination was forbidden.  (I recall faintly that waiver could be had, however if the circumstances denied any other type of connection and the soldered wire received additional strain relief than just the soldered connection.)

Having said that, in many years of industrial and consumer design there are always cost, performance, size, reliability, ... tradeoffs.  The inherent reliability weakness of this termination can be mitigated as Jen noted above if other mechanical aspects of the design prevent the wire at the PCB interface from moving under the duress of shock and vibration.   

So, are the BMW and Zed "poorly designed" because they don't meet the standards of Mil-Std 810?   No, not necessarily given cost, size, manufacturability ... constraints.  As Steve notes on product evolution, if ML has seen a noteworthy failure rate due to the design of this termination, then YES.  Else, NO.   

 

  • Like 2

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


  • Similar Content

    • By jasong
      Does anyone have a vid link to a nugget going low-high (filmed after the update only) that shows how deep it was and how big the nugget was?
      On nuggets up to 40 grams I still cannot get a low-high signal with just one exception and that is squirrly little wiry or thin flaky pieces that flip between high-low and low-high, and with some regular pieces when they get really close to the coil they will flip between the two as well. 
      I had a 65 gram slug of solid smelted gold and it was going high-low on the GPZ. It was low-high on my 45 though.
      What I'd really like to see are the types of gold that are low-high from beginning to end, no matter how far from the coil. If anyone has some links or has such a nugget and could toss it on the ground and film it I would be very grateful.
      One reason I'm wondering (ok, don your tinfoil hats) is low-high is so often iron now that I can't help but wonder if discrim might be possible with ZVT tech and there may be some kind of "discrimination update" coming in the future.
      Also, bonus question: can anyone show me a case where General is outperforming High Yield in depth and sensitivity and what exactly the type of gold (weight/sponginess/etc) is that causes such a thing to occur? Or ground type if that is causing the difference. We see it on the ML chart (or it would appear that way but we don't know for certain due to omission of actual sample defintion), but I haven't found a case with my gold where it happens yet so I'm curious what to look out for or if I should just give up on going back with General after HY unless I have gold past a certain size threshold.
    • By Steve Herschbach
      Hopefully this clarifies things for people!
      http://www.minelab.com/__files/f/274444/KBA_26-1%20GPZ%207000%20Tips%20for%20Better%20Ground%20Balance.pdf
    • By kiwijw
      Hi guys, I have done a goggle search on this & have come up with nothing. I thought I saw an article on here about it but a search showed nothing here either.  I thought there might have been something printed on the box to help out.....but no.  Is there any thing I need to know or is it all pretty easy really? Thanks for any help or advise. 
      JW  
    • By russ
      I've been back to 5 old patches with the new detector but I'm limited with a health condition as to how long I can swing the detector but It might be up to 40% less   I've got about 20 more old patches to re-hunt that will take many more days.
       
      So far I've picked up 6.2 dwt's that were missed by several other Minelabs and a gold bug II and different operators over several years of hitting the same patches.
       
      I don't think I need to say much about the detector because the testers Steve, Chris, and JP have well written reviews about the GPZ . The biggest nugget so far is a .8 dwt and a .9 dwt nugget. I'm still waiting for the bigger gold....   but in all fairness to the GPZ the ground is mostly deep where the big gold was found and I used big round mono's up to a 25" nuggetfinder and a 24" Coiltek to find the big gold. 

    • By Rob Allison
      Hello All,
         Just a FYI, make sure you understand the settings and what is working best.  I just realized the first day out, even though I found 8 nuggets, I was running settings that were no better than the GPX 5000 after some further testing.  What I would consider a large nugget, the GPZ considers it a "small" nugget, which is best for High Yield over General.  
      I also realized "Difficult" soil really makes the unit purr, but takes a lot of performance away on the ground I was hunting.  I was trying to find a happy medium from a noisy Normal to a Difficult soil setting with a higher gain.  
      Experiment is the key.  What I originally thought was good, wasnt ..... 
    • By russ

       
      Posted by gullyhunter.
       
      That's very strange... unless he tracked out the target by pumping the coil?
×