Date: Sun, 23 Sep 2007 01:54:05 +0200
Reply-To: tinker man <tinkerman007@GMAIL.COM>
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: tinker man <tinkerman007@GMAIL.COM>
Subject: Re: ECU firmware location
In-Reply-To: <059701c7fd71$72626010$57272030$@net>
Content-Type: text/plain; charset=ISO-8859-1
Yep, I know those, but usually prefer the solder wick for delicate
jobs because that mechanical sucker sometimes sucks also the thin,
over-heated tracks...
The solder wick actually is safer and doesn't spread the solder
particles around like the mechanical ones. Newly induced shorts aren't
fun at all (like a grain of solder bridge underneath an IC and
shorting two pins, but not visible...
But I'll definitely get myself a solder sucker too, for those
occasions I need it.
On 9/23/07, Leslie Schwartz <lhs_emf@pacbell.net> wrote:
> The mechanical solder suckers - mechanical single trigger action plunger work pretty well.
>
> -----Original Message-----
> From: Vanagon Mailing List [mailto:vanagon@gerry.vanagon.com] On Behalf Of tinker man
> Sent: Saturday, September 22, 2007 6:22 PM
> To: vanagon@GERRY.VANAGON.COM
> Subject: Re: ECU firmware location
>
> On 9/17/07, David Kao <dtkao0205@yahoo.com> wrote:
> > Hi TM:
> >
> > What you are doing is very challenging.
>
> Yup, you can say that again...
>
> > Another suggestion, if you really want to skip all the headache in
> > dealing with those linear IC's. LM2902 and 2903 are very cheap.
> > Just buy enough of them and replace them from your board. It is not difficult
> > to replace them.
>
> Nope, not so easy with my board because it's dual sided (as opposed to
> the US versions) and it's quite difficult to clean the holes with a
> solder sucking wick, even if I cut those parts off. The solder cools
> too fast, and the holes seem pretty tight too. I don't have a vacuum
> type yet. The board is also covered with lacquer, making soldering and
> cleaning pretty difficult, heck, even measuring voltages requires
> penetrating the lacquer with the probes!...(on both sides of the
> board, even the part pins).
>
> > It would be extremely difficult if you were dealing with
> > surface mount devices.
>
> Yep, lucky it's not.
>
> > Just replace them one at a time. There are only
> > a few of them. I used to do a lot of that. You will need a deslodering tool,
> > a vacuum tool to suck the solder off when it is melt by your solder iron.
> > The trick in doing this is you need to do it fast to prevent heat damage
> > to other components not being removed.
>
> Yep, BTDT. Removed the power transistors only to find they were OK...
> Actually, I'm pretty experienced in that area too, but I'm afraid that
> all the fighting with board might damage it, and it's also hard to see
> shorts because all the shiny gunk left after soldering - I don't have
> a cleaning agent to clean the board after it.
>
> > My estimate is it will take about one hour the most to replace all the IC's
> > on the board. It will take a minute to solder a replacement chip back in.
>
> Yes, but with the danger of damages and shorts to my only board...
>
> > I would go this way than troubleshooting the circuit. It's a dumb blind
> > approach but it is faster.
>
> I can understand this approach and it works sometimes, but I usually
> prefer analysis before I use the cutting knife...
>
> Actually, I've made good progress and have even verified that some of
> the IC's and transistors are working without removal, but tomorrow I'm
> supposed to get two working units and then comparing signals will make
> it much easier.
>
> Anyway, thanks for the good suggestions!
>
> cheers,
> TM
>
>
> > --- tinker man <tinkerman007@GMAIL.COM> wrote:
> >
> > > Hi David, seems this post is on the wrong thread (should be on ISCU debug).
> > >
> > > Anyway, I agree with all your recommendations and have actually
> > > followed just that path.
> > > Seems that in my ISCU the output transistors are actually OK because I
> > > can inject a signal and get an output (and I've removed them for
> > > testing too).
> > > Also, even with analog circuitry not all signals affect all others,
> > > since certain signal paths can stay unaffected until they reach a
> > > problematic junction (where they meet a bad signal). Also, analog
> > > circuits can be easier to debug with simple tools like a DVM than
> > > digital ones that require storage scopes and logic analyzers...less
> > > timing issues, clocks, etc. So it's not all so bad.
> > >
> > > And thanks for your complements, and although this crusade started
> > > because I'm struggling to fix my van's multiple FI problems (I got
> > > sick of waiting for the dealer to solve them, wasting my time and
> > > money...) I hope this is going to be more of a group project (already
> > > have gotten some preliminary qualitative results from a few members,
> > > and lots of useful info anyway) for the future benefit of all list
> > > members.
> > >
> > > And regarding possible damage on the test bench, I think I warned
> > > enough about that...
> > >
> > > Cheers,
> > > TM
> > >
> > > On 9/17/07, David Kao <dtkao0205@yahoo.com> wrote:
> > > > Linear circuit like this without a schematic is very difficult
> > > > to troubleshoot. Unlike digital circuit, a bad component that
> > > > causes a signal (voltage) to drip off will be propagated to
> > > > everywhere. That makes all signals all over the circuit board to
> > > > goo wrong. It then becomes extremely difficult to locate the
> > > > component that causes everything to go wrong. Digital circuit
> > > > on the other hand is extremely easy to isolate the faulty device.
> > > >
> > > > I would take a dumb approach to troubleshoot such a circuit board.
> > > > Check all cable connection first. Check the power transistors
> > > > by checking the P-N or N-P junctions to see if any of them becomes
> > > > short or open. Check all resistors, especially power resistors, to
> > > > see if any of them is burnt out or changed value because of heat
> > > > over time. Look for any other components for sign of damage.
> > > > This approach usually can be faster and sufficient to locate the
> > > > bad component and have it replaced.
> > > >
> > > > If you want to check out the active components then use an oscilloscope,
> > > > preferably a digital scope that can memorize the measured signals.
> > > > Sync and hold the signal with traditional scope is almost impossible
> > > > for this task.
> > > >
> > > > LM2902 can be configured as amplifier, filter, analog buffer, etc. Single
> > > > supply voltage makes their circuit more complex than those with dual
> > > > differential supply. But LM2902 and LM 2903 operating at such low power
> > > > level makes them extremely durable. Troubleshooting with live voltage on
> > > > can be more damaging than actual use on the Vanagon.
> > > >
> > > > Your effort in troubleshooting it is admirable. Good luck.
> > > >
> > > > David
> > > >
> > > >
> > > > --- tinker man <tinkerman007@GMAIL.COM> wrote:
> > > >
> > > > > On 9/17/07, Bob Donalds <donalds1@verizon.net> wrote:
> > > > > > TM
> > > > > > I did not mean to exclude the list
> > > > > > I'm kind of out of my league with the tech part of this
> > > > >
> > > > > Maybe, but you have lots of important info backed up by lots of experience...
> > > > > And yes, those microprocessors ARE my league...:-)
> > > > > (but engines, distributors, dynos and mechanics aren't...).
> > > > > If my Garage mechanics (not only VW) were competent at fixing fuel
> > > > > injection systems I wouldn't get into it, but it seems when it gets to
> > > > > electronics they get a dumb look on their faces and start replacing
> > > > > very expensive parts unnecessarily...
> > > > > So I've learned to diagnose before I let them get near. They do the
> > > > > mechanical stuff, I do the thinking...:-) (With the help of the list
> > > > > and Bentley).
> > > > >
> > > > > > so share anything I write you with the list
> > > > >
> > > > > Great. This will stir much interest, I'm sure.
> > > > >
> > > > > > the CPU is again not something I looked at I was getting this looked at with
> > > > > > SRS in Walpole MA Dick Shine
> > > > >
> > > > > What's "SRS in Walpole MA Dick Shine"?
> > > > >
> > > > > > He was taking out the EPROM's and reading them for me that's a far as I got
> > > > > > so that's why I have what's on the EPROM's
> > > > >
> > > > > Like I said, good for backup and verification.
> > > > >
> > > > > > and as I said Daryl told me the Euro has no hookup for 02s
> > > > >
> > > > > That's interesting, because in the US Digijet (I think you told me) do
> > > > > support O2 sensors, and someone on this thread reported that US
> > > > > Digijets use a completely different CPU (Z80). So maybe US and Euro
> > > > > Digijets are completely different animals. I'd like to get an internal
> > > > > picture of all those. Currently I have just a 1990 Euro Digijet and
> > > > > Jon Kenneke has posted on his website
> > > > > (http://kenneke.com/vwfi/fi.html) a US Digifant.
> > > > > I need more internal pictures of all models available (1983.5-1991, US
> > > > > and EURO) to clarify these issues.
> > > > >
> > > > > > can you tell me what dist you have is it the 2.1 or the 1.9 with the vac can
> > > > >
> > > > > Already did in previous thread, it seems it does have a vac can.
> > > > > TM
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > ____________________________________________________________________________________
> > > > Be a better Globetrotter. Get better travel answers from someone who knows. Yahoo! Answers -
> > > Check it out.
> > > > http://answers.yahoo.com/dir/?link=list&sid=396545469
> > > >
> > >
> >
> >
> >
> >
> > ____________________________________________________________________________________
> > Yahoo! oneSearch: Finally, mobile search
> > that gives answers, not web links.
> > http://mobile.yahoo.com/mobileweb/onesearch?refer=1ONXIC
> >
>
>
>
|