July 19, 2018, 08:48:03 PM

Author Topic: 580EX II Faulty by Design?  (Read 75286 times)

DetlevCM

  • Guest
Re: 580EX II Faulty by Design?
« Reply #45 on: February 21, 2011, 11:30:28 AM »

That "door slamming comparison" doesn't hold - I don't think you're sending electric signals out of your fingertips.
But look at a stereo:
Assume a car comes with one, and then some adolescent idiot installs a stereo system which causes excessive noise at an incredibly low quality - same as using a pocketwizard so far, isn't it? 3rd party add-on.
Now, let's assume that thirds party sound system fries the cars electronics - some part somewhere between the connectors and the battery doesn't like the power draw - would you complain that the car is badly manufactured?

Actually, I would complain. There should be a fuse to prevent that from happening. Now if the idiot tampered with the fuse..... he would get what he is asking for.

In any case, I do not expect Cannon to admit any design faults, so we will never find out. PW's investigatoin, unfortunately, is not likely to provide any more insight either, given the direction they are taking. As scalesusa pointed out they need an expert in high voltages and arcing.

I do expect the 580 EX III, whenever it comes, to be designed such that is does not arc ;-)

Except that the stock fuse could be in the stereo itself ;)

-> the key point here is, you have a third party add-on manufacturer pointing a finger at Canon and shifting the blame to them when THEY make the equipment that breaks the flash.

Canon does not have to care about what other companies do - they only have to design their equipment to work within it's own specs.

canon rumors FORUM

Re: 580EX II Faulty by Design?
« Reply #45 on: February 21, 2011, 11:30:28 AM »

SouthXylophone

  • Guest
Re: 580EX II Faulty by Design?
« Reply #46 on: February 21, 2011, 12:21:21 PM »
When this happened to my only 580x2, I was ... sad, angry, vengeful, morose, apoplecticly ragey, and other emotions that require special characters.  When I contacted PocketWizard, they said the same thing - make sure to get the head repaired properly because of arching (WTH is "properly?")  My flash was repaired under warranty by Canon: "battery box replaced" and "flash head assembly replaced" so I hope I'm okay - would have been nice to have the pictures from this report, though.  I just hope they did it "properly" but I'm not voiding the rest of my warranty or risking my life to take it apart and find out.   If it blows again, it'll go in with a printout of this report (saving a copy in case PW pulls it for some reason).  PW said it would blow again if it wasn't fixed right.  Great.  Just ... great.

Their explanation seems to make sense, but like y'all I wanted evidence that others had the same issue WITHOUT the Mini/Flex to prove it wasn't a PW problem.  I found some.  See below.

Even without evidence there is one thing that sticks in my craw on this.  The PW talks to the flash through the shoe.  I assume there's a Canon CPU *between* the shoe and the rest of the flash guts.  Canon's CPU runs the flash guts based on what it hears from the shoe.  Why did the CPU let it happen?  Isn't it the CPU's job to operate on the guts properly?  If it heard something it didn't like from the shoe, should it even fire?  PW says "everything they say to the flash they learned from a camera" so how could it be out of range?  PW radios move *information* around.  The Canon CPU moves *power* around.  Who is responsible for damage that appears to be totally power based (granted, according to PW only)?

The analogy I thought of is car stereo-based, too.  It's like putting a CD in the STOCK player and then some bass note in the music blows your speakers.  The CD can handle the bass note.  So can the stock player and the stock amp (bass is part of the CD spec).  But the stock speaker blew and you weren't even playing it that loud.  Is it the CD's fault?  Hardly - it's just information.  The stock speakers were crap, obviously.  If all you played till now was Joni Mitchell and you dropped in some Eminem for the first time, it's probably not the rappers fault your speaker blew.

But in the PW report it seems like they even tried to "sound" exactly like Canon (page 15), but the problem stayed until they fixed the tube alignment.  No amount of firmware updates to the Mini/Flex will move the tube around for you.  I really want to hate PW because then it would be easier to love my no-longer-perfect 580EX II, but they make a compelling argument.


EVIDENCE without PW (I registered just to share my research on this which I did because I wasn't convinced at first that PW wasn't part of the problem, still don't know for certain - anyone know anyone at Canon to ask because only they would know?):

http://www.fredmiranda.com/forum/topic/904307

A whole topic about it.  PW never mentioned.


http://www.flickr.com/groups/pocketwizards/discuss/72157623362956049/72157623898130266/

"im having the same problem on my 580exii. It exclusively fires full power at any mode. no more ettl for me for 4 months now.  ...  however, there's a little difference in my case. i haven't used my flash on the new PocketWizards that supports ETTL. I have been using the old PlusII's for 3 years with this flash."

P2 can't do HSS or pre-flash, so that's interesting.  Maybe it was used on a camera, too?

 
http://photography-on-the.net/forum/showpost.php?p=11857089&postcount=15

"Both of my 580EXIIs failed as described. One failed again when being tested on a camera when it returned from CPS. I had a 580EX exhibit the same failed behavior long before I bought my first FlexTT5.  And I use the 580EXs all the time on the FlexTT5 without a problem. The problem is unique to 580EXIIs for me. And its happened in the hot shoe."

Underlining mine - gee, I hope *my* flash was repaired better than that!

http://photography-on-the.net/forum/showpost.php?p=11859199&postcount=21   

"There actually have been reports of people frying the 580EX's and never even owned a TT5. It happens far less but the potential of the problem is still present without the use of the TT5. So in the end this issue falls back on Canon."

No links, though.  The plural of anecdote of is ... rumor?   :P


AND RIGHT HERE IN THIS THREAD:

http://www.canonrumors.com/forum/index.php/topic,615.msg7997.html#msg7997

"My original 580 EX II failed with a blown flash tube.  After sending it to CPS 5 times for repair, they finally gave me a new "refurbished" 580 EX II.  I EXCLUSIVELY use HSS without any third-party intermediary device.  I'm not sure I can fault Canon on the first flash blowing out because I probably shot 10,000 frames on HSS- I think it was just it's time."

Old flash, maybe.   I've heard of wedding guys triggering 3,000 flashes at one wedding (probably not all HSS, granted), so 10k doesn't seem like a lot.  Hard to tell if this is the same problem, though.  If they were all TTL, then the pre-flash *is* HSS, so ...


http://www.canonrumors.com/forum/index.php/topic,615.msg7928.html#msg7928

Canon Rumors administrator - "There has been a lot of people writing in over the years about their 580EX II’s constantly failing. People became more vocal around the time the PocketWizard FlexTT5 and MiniTT1 launched..."

Care to elaborate on that?  Is it *this* failure, full dumps all the time?  Or is it bad shoe contacts or something else which is different?

DetlevCM

  • Guest
Re: 580EX II Faulty by Design?
« Reply #47 on: February 21, 2011, 01:12:47 PM »
Yay :)

Finally someone who provides source links for non-Pocketwizard issues so that we (readers) can evaluate it by ourselves.

Not just anectdotal evidence.

SouthXylophone

  • Guest
Re: 580EX II Faulty by Design?
« Reply #48 on: February 21, 2011, 01:33:22 PM »

http://www.fredmiranda.com/forum/topic/904307

A whole topic about it.  PW never mentioned.

 ??? This one is not as clear cut as I thought.  BubbaJon owns the FlexTT5.  We just never know if it was involved with the failed 580EX II.  BillyBuff claims the same problem, but never mentions PW. 

Canonista

  • Guest
Re: 580EX II Faulty by Design?
« Reply #49 on: February 21, 2011, 06:54:37 PM »
I don't get it. Why is Canon taking the blame here for what appears to be a simple case of equipment incompatibility? I've used the 580ex for years without any problems, and I've never known anyone in my community of photographers to have that flash fail like this.

It's situations like this that cause all manufacturers of electrical products including Canon to advise against using third-party components, particularly where it may affect the electrical integrity of the product. Canon haven't tested and aren't responsible for determining whether those third-party components are compatible with Canon equipment, and hence the warning.

A good example of this is the typical mfr warning against using third-party batteries. Of course, many of us do it for our cameras and laptops to save a few dollars, but should I blame Canon when I am knowingly taking the risk that the battery may be outside Canon's spec or made to poor quality standards and fries my camera?

By the same logic, if I buy third-party lenses, and they later turned out to be incompatible with future generations of Canon cameras, I have no one to blame but the third-party lens mfr or myself.

We can complain all we want about the cost or the limitations of what Canon offers versus third-party mfrs, but there's probably a reason for both.

Caring

  • Guest
Re: 580EX II Faulty by Design?
« Reply #50 on: February 21, 2011, 10:27:20 PM »
I don't get it. Why is Canon taking the blame here for what appears to be a simple case of equipment incompatibility? I've used the 580ex for years without any problems, and I've never known anyone in my community of photographers to have that flash fail like this.

I totally agree.  My work is primarily on the IT vendor side, and it is an industry standard that vendors do not accept responsibility for failure of products when used in conjunction with unapproved third party products.  Until PW can conclusively demonstrate that the sole cause is the 580EX II design, then Canon should be presumed innocent.

I may not be a technical guru, but I identify a number of issues with the report:

1.  The only firm conclusion in the report is that the IGBT was damaged in each of the failed units ("IGBT Damaged Units").  Otherwise, the report is riddled with 'weasel words' and opinions, so it makes a mockery of a scientific study (eg. "We have no knowledge of", "It appears", etc).

2.  The report is only talks about arcing in IGBT Damaged Units.  There is no statement in the report that arcing occurs in all 580EX IIs, particularly those which have not been used in conjunction with PWs.  I guess the IGBT has something to do with regulating flash output, and as such I would not be surprised if there was damage caused to the tube when the flash is used with a damaged regulator component.

3.  The Preflash Boost or anything else that modifies the pre-flash output is essentially a 'hack' as it alters the normal Canon operation of the flash.  This is already using the flash outside of Canon specifications.  The report's statement that they have defaulted the Preflash setting to 'Off' for 5 months is hardly any form of reliable methodology.  If this 'hack' is damaging the flash, then even using it once or twice may be overloading the IGBT or other components, and does not necessarily need to be 'On' at the time of flash failure.

Thanks,

Caring
« Last Edit: February 21, 2011, 10:31:23 PM by Caring »

Flake

  • Guest
Re: 580EX II Faulty by Design?
« Reply #51 on: February 22, 2011, 05:43:46 AM »
The analogy I thought of is car stereo-based, too.  It's like putting a CD in the STOCK player and then some bass note in the music blows your speakers.  The CD can handle the bass note.  So can the stock player and the stock amp (bass is part of the CD spec).  But the stock speaker blew and you weren't even playing it that loud.  Is it the CD's fault?  Hardly - it's just information.  The stock speakers were crap, obviously.  If all you played till now was Joni Mitchell and you dropped in some Eminem for the first time, it's probably not the rappers fault your speaker blew.


Many years ago a company called Telarc recorded the 1812 overture with live Canon and the Cincinnatti orchestra the final canon had a double charge and blew out many windows when fired.  The recording was High Fidelity, and carried a warning that it could damage equipment low level stuff couldn't even play the record & even the high end gear struggled.  Now it's released on CD and still carries a warning, it has destroyed audio equipment and in some cases peoples hearing too.

Your analogy is flawed therefore it can be the fault of the CD and in much the same way if the software in the pocket wizzard causes the 580EX II to operate outside of its design limits then it's the pocket wizzards fault, and it does appear this is what is is doing with the HSS.

canon rumors FORUM

Re: 580EX II Faulty by Design?
« Reply #51 on: February 22, 2011, 05:43:46 AM »

SouthXylophone

  • Guest
Re: 580EX II Faulty by Design?
« Reply #52 on: February 22, 2011, 09:24:39 AM »
Your analogy is flawed therefore it can be the fault of the CD and in much the same way if the software in the pocket wizzard causes the 580EX II to operate outside of its design limits then it's the pocket wizzards fault, and it does appear this is what is is doing with the HSS.

(I guessed Telarc discs would be brought up.  I have an Erich Kunzel from the 80's with that warning on it)

Analogies are always flawed - they try to map something unknown to something known, but they aren't the actual thing therefore imperfect. 

That said, let's play with the analogy for a moment.

PW says they use commands they learned from the cameras.  This isn't a new master recording.  They are sampling the music they've already heard and are playing it back.  They are looping Joni Mitchell, not creating Erich Kunzel. 

The Telarc thing was certainly *very* true back in the early days of CDs being played on equipment designed for a limited analog range.  Nowadays every CD is compressed for MAXIMUM LOUDNESS, and stereo components are made of better things.  If a stock car stereo from the last 10 years can't handle any CD you'd buy at WalMart ... well.   ::)   

New Canon gear is supposedly state of the art.


1.  The only firm conclusion in the report is that the IGBT was damaged in each of the failed units ("IGBT Damaged Units").  Otherwise, the report is riddled with 'weasel words' and opinions, so it makes a mockery of a scientific study (eg. "We have no knowledge of", "It appears", etc).

What you call weasel words are also used in science.  Some tests were done.  There is more to do as they said.  They may learn that they are in fact the problem, but so far they have a strong theory as to why they are not the problem.  Do you take issue with the work, or just the use of careful words?

They concluded that in flashes with poorly aligned tubes, the IGBT gets hosed easily.  They say with certainty that every flash they've found with a bad IGBT *also* has charring.  Align the tube properly (or replace the head), and the IGBT does NOT blow.  This is all on page 2.  Not weasely at all.

2.  The report is only talks about arcing in IGBT Damaged Units.  There is no statement in the report that arcing occurs in all 580EX IIs, particularly those which have not been used in conjunction with PWs.  I guess the IGBT has something to do with regulating flash output, and as such I would not be surprised if there was damage caused to the tube when the flash is used with a damaged regulator component.

You've got it backwards.  The poorly aligned tube damages the IGBT, not the other way around.  Page 2.  Fix the tube and the IGBT does not blow AND the arching stops.  That's pretty clear.

And they do things with "good working units," too.  Figure 10 says it all for me - a *good* flash just doesn't arc even on a PW.  This all makes sense when you look at the numbers - the grand majority of flashes work just fine and less than 1% fail.  There is going to be more a lot more "good" data then bad.

3.  The Preflash Boost or anything else that modifies the pre-flash output is essentially a 'hack' as it alters the normal Canon operation of the flash.  This is already using the flash outside of Canon specifications.  The report's statement that they have defaulted the Preflash setting to 'Off' for 5 months is hardly any form of reliable methodology.  If this 'hack' is damaging the flash, then even using it once or twice may be overloading the IGBT or other components, and does not necessarily need to be 'On' at the time of flash failure.

Only Canon knows if it is outside specification and they are silent.   

The point of turning Boost OFF for 5 months of firmware was to prove the failure rate didn't change.  Presumably they had at least as many new customers in those 5 months as they did in the previous 5 months.  If the rate went down they would have learned something different.  That's just logic.  It is highly unlikely that *new* users all went out of their way to enable it once just to keep the fail rate the same (and internet scuttlebutt says "TURN IT OFF" all over the place, so why would they turn it on?).

I'm just pulling stuff from the report and here: http://www.flickr.com/groups/pocketwizards/discuss/72157623362956049/72157625949175952/

If you are in that less than 1% group, and your flash is not in warranty, you are screwed.  You can choose who you think is screwing you, but PW has shared how they think they are not.  Canon has shared nothing.  For all we know, they have piles of flashes with this same problem, else why would they repair it under warranty?  That bugs me a little.

CR Backup Admin

  • Administrator
  • EOS 5DS R
  • *****
  • Posts: 836
Re: 580EX II Faulty by Design?
« Reply #53 on: February 22, 2011, 01:13:26 PM »


You've got it backwards.  The poorly aligned tube damages the IGBT, not the other way around.  Page 2.  Fix the tube and the IGBT does not blow AND the arching stops.  That's pretty clear.


While it is true that you can modify the flash to give a wider potential arc gap, that requirement itself makes me suspicious that the PW is using a faster rise time to the signal pulse. 

IGBT's are notorious for inducing arcing due to their very fast rise times of the output.  If you modify that, then that is very likely to cause the problem. 

Its a simple impedance issue, the fast rise time creates a impedance mismatch and power is reflected at the connection of the tube.  The reflected voltage adds to the incoming voltage, and you suddenly have twice the voltage, and it arcs over.  It has happened in many applications.  These guys might be good logic and low voltage engineers, but they apparently fail to see the obvious that someone working in high voltage pulses would look for first.

smeggy

  • Guest
Re: 580EX II Faulty by Design?
« Reply #54 on: February 22, 2011, 02:01:33 PM »
There is a lot of unreasoned guesswork going on in this thread.

While it is right to initially reasonably suspect PW, PW have done some reasonable testing that suggests the problem is not down to their product; so it is now wrong to assume PW is the sole suspect.

The point about risetimes is unreasonable as it is unlikely that mere control signals from PW can affect it.
All the talk of risetimes, incompatibilities and whatnots have already been cast into doubt by one of PW's simple tests: removing the front panel. Doing so has absolutely no effect on the electrics, yet seemed to have stopped the symptoms completely. This is why I cannot discount their tentative conclusion of O3.

PW’s testing is far from conclusive, but assuming they’re not fibbing, it does show that no one has any reasonable cause to lay blame at PW only.

The fact that 430s (and other flashes) don’t blow does rather suggest the 580s genuinely are unusually susceptible to something. Perhaps it is merely the RF from nearby PWs, I don’t know. We know the 580s radiate more than most, so it stands to reason they could also be susceptible to RF (bad shielding, poor step-up design).
The flashes being damaged by a nearby RF source (not necessarily electrically connected) is whose failure?

Speculation is fine, but I think folks shouldn’t jump to any conclusions, at least not until someone does a proper follow-up analysis.

I own PWs and the 430EX II. Fingers crossed!
(I am not affiliated with PW, or Canon, or any other related company, in any way)

Lawliet

  • EOS 7D Mark II
  • *****
  • Posts: 549
Re: 580EX II Faulty by Design?
« Reply #55 on: February 22, 2011, 03:04:27 PM »

While it is true that you can modify the flash to give a wider potential arc gap, that requirement itself makes me suspicious that the PW is using a faster rise time to the signal pulse. 


The communication between flash and the camera is SPI based, you request a pulse of a certain power level and tell the flash when to fire without anything analog involved. Driving the high voltage part is the responsibility of the controller in the flash. I can't see how the could make the flash do something it is not designed to do, at least if you're willing to see the boosted preflash as a normal flash, similar to the stroboscope mode.

Caring

  • Guest
Re: 580EX II Faulty by Design?
« Reply #56 on: February 22, 2011, 04:00:37 PM »
Quote

They concluded that in flashes with poorly aligned tubes, the IGBT gets hosed easily.  They say with certainty that every flash they've found with a bad IGBT *also* has charring.  Align the tube properly (or replace the head), and the IGBT does NOT blow.  This is all on page 2.  Not weasely at all.

If the flash has bad IGBT, then it is already damaged, and why is it being used for further testing.  This does not equate to a controlled environment nor does it confirm whether the chicken or the egg came first.


Quote
You've got it backwards.  The poorly aligned tube damages the IGBT, not the other way around.  Page 2.  Fix the tube and the IGBT does not blow AND the arching stops.  That's pretty clear.

And they do things with "good working units," too.  Figure 10 says it all for me - a *good* flash just doesn't arc even on a PW.  This all makes sense when you look at the numbers - the grand majority of flashes work just fine and less than 1% fail.  There is going to be more a lot more "good" data then bad.

I do not know what an IGBT is, but I do know how to read.  Page 2 clearly states "Below are examples of 580EX II flash heads that were causing IGBT failures to repeat".  PW purchased Damaged IGBT Units, then PW repaired the IGBT, only to find further failure of the tube (possibly because it was already damaged) in those repaired Damaged IGBT Units.  I don't know whether the damaged tube comes first or the damaged IGBT comes first.

All of these fancy oscilloscope graphs show absolutely nothing.  There is no methodology, or an explanation around the controlled environment in which they should have been conducted (eg. is it a new 580ex unit?  is it a repaired Damaged IGBT Unit?).

 I only need to see one oscilloscope graph - that is the one of a Canon 580EX II flash straight out of the box, firing without a PW unit.   Strangely, this one was omitted?   :D     No scientist would test an alleged 'design fault' solely with graphs using a unit tempered/tainted by an unapproved third party product.


Thanks,

Caring
« Last Edit: February 22, 2011, 04:11:44 PM by Caring »

SouthXylophone

  • Guest
Re: 580EX II Faulty by Design?
« Reply #57 on: February 22, 2011, 09:50:52 PM »
I only need to see one oscilloscope graph - that is the one of a Canon 580EX II flash straight out of the box, firing without a PW unit.   Strangely, this one was omitted?   :D     No scientist would test an alleged 'design fault' solely with graphs using a unit tempered/tainted by an unapproved third party product.

You're right in that they could have done a better job showing us the control.  But consider this:

  • What if the control looks exactly like Figure 10?  It seems pretty damn likely.
  • They say it is less than 1%.  They might have to buy 200 new flashes to find 1 that exhibits the issue.  So they found a way to get failed flashes and did a lot of analysis on them.  That seems logical and very practical.
  • The failures they collected aren't "lab experiments," they are the real deal.  By changing ONE aspect in these failed flashes, they can repair it once and get it to not fail again.   The argument against that is they weren't "scientific" enough?  Okay, but that does not invalidate their finding of FIXED TUBE=FIXED FLASH.  I'm open to being convinced otherwise, but that statement is pretty well backed up by their doc, at least to me.
  • I can't be led to believe all the flashes PocketWizard owns are previous failures.  Since the problem has low manifestation (less than 1% they say), they must have had flashes in house that have never failed (they test their own circuitry, presumably) or they would have known about this 2 years ago.  Just like 99+% of their customers, they probably just used the "good" flashes they own as a baseline.  Who knows or CARES if every flash they own has been used with a radio - a good flash doesn't fail regardless (strongly suggesting that radio usage is NOT the cause) and that evidence is overwhelming (99+%)).  I'll take a risk and say they presented the evidence from failed flashes mostly because that's where the rubber meets the road on this issue.
  • Flashes repaired BY CANON can fail again.  I linked that earlier.  The IGBT is probably on the "battery box" board and if that is the only thing Canon repaired, the flash can fail again.   CANON.  Not PW.   What PW is saying is that *only* fixing the tube *stops* IGBT failures.  You don't need to replace anything else in the flash after that.
 

I had to look "SPI" up on Wikipedia.  Lines up with what I was told: the CPU in the flash controls the high voltage stuff. There is nothing PW could say over a serial port that will mis-align the tube.  PW may or may not have "high voltage engineers."  I'd have to say the same for Canon because why would a high voltage engineer design a system where a serial command causes an unwanted electrical arc ... like EVER?  It's like a hacker's dream - a command you can give your modem to blow up the other guy's computer.

You either believe PW when they say tubes are misaligned, and are the primary source of failure and simplest successful repair, or you don't.  I kinda sorta do.  I'm an easy skeptic, but nothing said here has been extremely compelling against PW, at least in my estimation so far.  They could have done a better job with the control data.  They could tell us their high voltage cred or not.  But all of that seems like peanuts compared to what they DID find pretty clearly: FIX THE TUBE = FIX THE FLASH.

canon rumors FORUM

Re: 580EX II Faulty by Design?
« Reply #57 on: February 22, 2011, 09:50:52 PM »

DavidWhelan1

  • Guest
Re: 580EX II Faulty by Design?
« Reply #58 on: February 23, 2011, 03:44:51 AM »
Humm... sounds like PW is trying to head-off a class-action lawsuit... or the demands of everyone who's bought them from demands to have them returned....

Glad I'm not them... and yeah, good point, why didn't they include the results from 'non-contaminated' units that haven't been attached to PW?

I own two of these units and am quite concerned - at this point, no problems with my 580EX II...

-David


Boogietek

  • Guest
Re: 580EX II Faulty by Design?
« Reply #59 on: May 26, 2011, 05:20:17 PM »
This was a very useful conversation that I had first with Canon Support and then with Pocketwizard Support...

I called Canon support regarding an issue with my flashes 580EX II that doesn't regulate the power, what I mean is that they shot on 1/1 even if I set them to 1/6, 1/16 or any other value, and the customer support told me that Pocketwizard had recent reportedly an issue with the communication with the flash and was blowing up the circuits and blah blah blah... I said ok! perfect, I'm gonna send you my all four flashes because I need them fixed...  Then this morning I send the following email to Pocketwizard and here is what they explain me:

------------------------------------------
Message:
Hi, I own seven (7) Pocketwizard's FlexTT5 radio slaves with AC5 RF Soft
Shield (all of them with the accessory), and I'm having a problem
working with my four (4) Canon 580EX II flashes. I have called Canon
Support to report that my all four flashes aren't regulating the potency
of the light, they only shot at highest 1/1 even if you set them to any
lower setting like 1/8, 1/16 that are the most usable settings in my
case. Then, when Canon support made me a lot of questions about what is
all the equipment, settings, and habits handling my cameras (two Canon
7D bodies and one Canon t2i body), and they concluded that the problem
is that the Pocketwizard's FlexTT5 are burning the flash circuits in
certain shots. As Canon Support received a communication letter from
Pocketwizard explaining some communications problems and causing this
failure to the Canon 580EX II flashes. Now I want to send you all my
seven (7) radio slaves to fix this problem that have caused me a lots of
inconvenience including loss of money and my client's time and the setup
of all scenarios situations in my photo sessions of weddings and model
portraits and what means that my business is loosing advantage and
having a bad reputation because of this. I want a quick solution with
this because I have trusted my photography business in Pocketwizard
since many years ago and I can't loose anymore clients with this
problems. Regards, Boogietek.

Answer:
Thank you for your inquiry.  I am sorry that you have experienced this frustrating issue.  Our research indicates that this is a mechanical issue in the 580EX II that can happen with High Speed Sync.  It does not require our radios to happen, though a lot of people use High Speed Sync or E-TTL features of their Speedlites for the first time, or more heavily, when using our radios.

Everything we "say" to the flash is an E-TTL command we learned from the E-TTL system.  It is highly unlikely that our saying “Hey flash, please fire at one of your known levels” is the problem.  There is a microprocessor inside the flash between the flash’s hot shoe and its triggering/tube components.  We tell the microprocessor, in its own language, to flash at pre-defined levels.  What happens with communications from the flash’s microprocessor to the triggering/tube circuits is not within our control.

I'm not sure what document your Canon Support contact was referring to, but we have never written anything with that description.  We firmly believe the issue is in the tube/reflector alignment and there is nothing our radios can do to move that tube around.  This issue does not occur exclusively with PocketWizard radios, but does occur exclusively with the 580EX II.

If a 580EX II is under warranty, then we recommend that warranty options be pursued.  If not, read on:

We have just started offering a repair service to help people out of a bind with this issue, and only this issue.  We won't repair any other kinds of flash failure, only this one. It is expressly for the problem of "BLOWN IGBT."  The primary symptom is that the flash still fires, but you are unable to control the output - it is always a full power flash.  When in ETTL mode directly mounted on the camera, you get a full power pre-flash and no flash in the exposure.  In manual mode, all settings from 1/128 through 1/1 are full power dumps.  Otherwise the flash appears normal (LCD and buttons work, display changes when used on a camera, hot shoe is intact, etc.).

We charge $30 USD per flash to repair the blown IGBT and tweak the flash tube alignment so it is much less likely to ever fail again.  While we have the flash open, we would also offer to do the RF mod to the flash (so it doesn't need the AC5) for $15 USD, if you wanted.  Again, this special offer for the RF mod is only if we have the flash open expressly for IGBT repair.  Other parts would be at cost.  For example: sometimes the reflector assembly is so charred we cannot adequately clean it, so a new tube/reflector assembly is required for an additional ~$40.  If the flash reveals some other mechanical failing that is outside the scope of the service we provide, we can work with our local Canon-authorized repair center to facilitate those repairs (you would have to pay for these repairs) - an example would be the internal optical fiber being kinked, which, when paired with a tube misalignment, leads to faster IGBT failure.

Our repair fee covers slow surface return shipping.  If you want the flash back faster, shipping charges would be at cost. No repairs would be performed without your authorization first.

Here are some images showing the tube misalignment in detail:

http://www.lpadesign.com/Tube%20metal,%20none%20-%20580EX%20II.jpg
http://www.lpadesign.com/Tube%20metal,%20a%20lot%20-%20580EX%20II.jpg
http://www.lpadesign.com/Tube%20metal,%20a%20little%20-%20580EX%20II.jpg


The repair process is pretty simple.  You just send the flash to my address below.  Include phone contact information and when we receive it, we'll call you for credit card or PayPal information.  We tend to turn these around very quickly.

If you have other 580EX II Speedlites and you are worried about them, you can send them to us.  We will look inside and tweak the tube as needed.  Note - we only offer this service to folk who have actually experienced an IGBT failure.  When looking, we might discover that your flash tube and reflector are still working, but in poor condition or uncleanable - we could then replace that assembly for ~$40 as I mentioned before.  This service could void your warranty.  While we do a very clean job, there may be evidence visible to a trained technician that we've been inside the flash.


To clarify:

$30 each = IGBT repair, tube tweak+reflector clean, return ground shipping

+ $15 each gets you the RF MOD, but only if that flash is also getting an IGBT repair

+ ~$40 each if the tube/reflector carriage needs to be replaced (so far, > 50% need this)

$0 = Validate tube on other flashes, but $0 charge is only for folk like you that have experienced IGBT failure



If you do decide to get rid of your 580EX IIs, you might want to consider replacing them with 430EX IIs. Learn more about the 430EX II versus the 580EX II here:

http://www.robgalbraith.com/bins/multi_page.asp?cid=7-10050-10598-10599

We have never seen evidence of a 430EX II experiencing this issue.  Our analysis of this flash shows that the tube spacing to the reflector is adequate, and the area is vented.  This is a great flash with our system and we hope all future Speedlites are designed as well as this one.

If you have further questions, please reply to this email and keep the case number in the subject line.

Take care,

Patrick Clow
PocketWizard Tech Support Manager

Then I replied:
Thank you for this good news, I'm gonna wait for Canon to return all my 580EX II's and send it to you to do the IGBT repair, tube tweak+reflector clean, and the RF MOD to all four flashes so I don't have to experience this again... trust me, I've been in serious trouble in the middle of outdoor shootouts with my clients.  By the way, I'm gonna send you in the same package all my Pocketwizard's FlexTT5 so you can check them for any issues (just in case), because I'm really afraid of having problems again in the middle of a work with any client and I want to be sure that everything is working in perfect conditions.  First I was thinking in what you told about changing all my flashes for the 430EX II, but the deal is that I have plenty accessories and adapters, and stuffs that only fits to the 580EX II, and I have invested much money in all that...

Thanks again.
---------------------------------------------
I hope that my case helps anyone because I really know what a mess this could be!

canon rumors FORUM

Re: 580EX II Faulty by Design?
« Reply #59 on: May 26, 2011, 05:20:17 PM »