Posted 2008.12.31
NOTE: This article was updated in April of 09 after Mark at Precision Camera, our favorite repair shop, provided us with some additional information.
This is probably the most boring article I’ve written for LensRentals. (Personally, I like Smashed Front Element the best.) But, I love a good mystery, and I love debunking DSLR “urban legends”. Every so often I run across an online forum where someone makes broad statements about Error 99 which I know are incorrect or at least incomplete. As is my practice, I’ve boldly charged into these online gunfights devoid of intellectual ammunition (i.e. facts). And predictably, the intellectual level of the discussions quickly spirals down from “Is not”, “Is so” into the traditional online-forum sign-offs of “You get Err99 because you’re a bad photographer” and “if you’ve never gotten Err99 you’re obviously not taking many pictures”.
Because Canon Corporate apparently believes that releasing no information about a problem makes it go away, there is little factual information to debunk the online myths regarding Err99 unless you really do a lot of digging. Googling Err99, Canon EOS error codes, etc. brings up several dozen pages of links most of which are the above mentioned “discussions”. There are a few nuggets of truth out there, though. The most complete of these is a multi-year 2,300 post discussion of Err99 problems at Richard’s Notes. There are also a few thoughtful and factual discussions that have taken place in some of the better forums online. After spending far more hours than I intended looking through these sources to settle an online debate, I thought I’d write a summary of what I found and what we’ve experienced here— LensRentals has over 700 Canon lenses and over 50 Canon bodies, so we have a bit of Err99 experience.
Some of the most common Err99 myths are listed below. Strictly speaking, they are not myths; almost every one is true. The myth part comes from thinking that any one of them is actually the cause of Err99. So:
There are a lot more. Almost all of them are true for at least some cases of Err99. The best myth, though, is that Canon purposely created error 99 to prevent the use of third party lenses. As best I can tell, that one isn’t really true, but it does make fun speculation. And, of course, Canon’s nearly total silence on error 99 and other problems certainly helps feed the conspiracy theorists among us.
Canon no longer officially comments on Err codes, but if we go back to the golden days when they did, we can unravel a lot of the Err99 mystery almost immediately. Back in 2000 Canon released its first mainstream DSLR, the 3.1 megapixel D30. The manual contained a helpful list of the camera’s built in error codes:
In 2002, the D60 was released. It had a reduced set of error codes:
Key point: Error 99 is a catch-all which can mean almost anything went wrong.
As far as we can tell, the Canon error codes have remained the same through the 50D and 5DMkII camera bodies, at least nominally. The more recent bodies have added an Err 06 code for ‘sensor cleaning unit malfunction’, and there are now ERR 10, 20, 30, 40 . . . 80 codes on 5D Mk II cameras (with the useful message ‘Shooting is not possible’. Duh!). Also some more recent manuals now define ERR99 as “an error other than one of the above (ERR1-ERR06)” has occurred. The only semi-official statement from Canon in recent years is one from Chuck Westfall in TheDigitalJournalist saying “[ERR99] is a non-specific error code which can be caused by a wide range of malfunctions. … a variety of problems can be caused by the use of non-Canon accessories such as lenses, memory cards, battery packs, electronic flash units, etc.”
We know from the above that the ERR99 code has existed since the D60, but most of the current ERR99 online discussion and speculation started around 2003-2004. A large part of this is for obvious reasons: the number of Canon SLRs in service exploded around 2003 and 2004 with the introduction of the Digital Rebel and the 20D cameras. There are some other factors that may have contributed to the marked increase in ERR99 reports around this time. Several changes that occurred, but probably did not have much to do with the ERR99 increase include:
There are a few changes, however, that logic suggests might have had some causative effect on ERR 99 messages.
Although Canon hasn’t said so specifically, pretty strong circumstantial evidence indicates that the electronic connections between lens and camera were changed at least once and probably twice since 1998. The first change is probably better documented and seems to have occurred first with the EOS 3 and EOS 1V film cameras, which introduced the 45-point autofocus system later used on the 1D series digital SLRs. A number of third party lenses (mostly Sigma) would not communicate autofocus information with these cameras, and required re-chipping by the manufacturer to regain compatibility. The same problem occurred with the introduction of the 10D digital camera, which increased prosumer autofocus points from 3 to 7 and introduced the Digic processor. Of note, those incompatible third party lenses gave an ERR99 message, not ERR01, when used with the 10D. As best I can find, the first widespread ERR99 reports occurred when third party lenses couldn’t communicate electronically with the new 10D camera, and the soon-to-follow Rebel and 20D. This is the source of many people’s partially incorrect belief that ERR99 always means a miscommunication between camera and lens.
The second change is less clearly established. Some sources state that lenses with IS systems have higher electric current transmitted from the camera than other lenses do, which makes sense, considering that they have more work to do. In-lens image stabilization first appeared in 1995 with slight improvements in 1997 and 1999. A major improvement was made in 2001 with the faster IS system used in the 70-200 f/2.8L IS and again in 2006 with the new four-stop system in the 70-200 f/4L IS. The newest IS systems are more powerful and stabilize more quickly (0.5 seconds as opposed to 1 second with older systems), so it’s logical to assume they draw more current across the connections, although this is not documented anywhere that I can find.
Several lenses with newer IS, including the 70-300mm f/4-5.6 IS (2005), EF-S 17-55mm f/2.8 IS (2006), and 70-200mm f/4L IS (2006) became very popular with photographers shooting EF-S mount cameras. We know that malfunctions in some of these lenses, most commonly reported with the 17-55 f/2.8 IS, cause ERR99 (and not ERR01) on EF-S mount cameras. Cleaning the electronic contacts on the camera and lens will often fix, or at least improve the problem. There are a few reports that the problem is more common with original Digital Rebel and 20D cameras, and less common with newer cameras; our data supports this too. Some knowledgeable people have speculated that there was a change in contact alloy, a thinner layer of gold plating, or other electrical contact issues with the XT and 20D cameras that make it more difficult for these cameras to deliver the required current to the newer IS lenses. On the other hand, the problem may simply be more common with older cameras because the lens contacts are more likely to be worn.
Another theory that has some factual basis was reported several years ago on DPReview. A tester found that Canon 20D cameras would display ERR99 if the camera voltage fell below 7.3 volts. The BP511 battery used in all prosumer cameras prior to the 5DMkII should deliver a bit over 8 volts in fully charged state, but will fail to deliver sufficient voltage in certain conditions: dirty contacts, failure of a cell within the battery, age, rapid power consumption, or some combination of the above. This certainly would explain the ERR99 problems occurring with bad batteries or bad battery contacts. Again, just speculating, but I would suggest that a fall in voltage across just part of the camera circuitry would also cause ERR99— for example, across dirty or corroded electrical contacts, across a cracked ribbon cable, or perhaps a slightly corroded circuit board connector. I’ll come back to this idea later.
I’m listing these in my own perceived order of frequency, combining our experience at LensRentals with reading countless online ERR 99 reports. The list has been revised after further information from Precision Camera about actual causes they see during repair. One important thing they reminded me of: when we see ERR99 on the screen, the camera actually has more information about the cause of the error internally. A repair shop can read this information from the camera and often determine the cause with complete accuracy.
Most of us cannot define in-camera circuitry failure more specifically than “it went to Canon for repair”, but a few skillful individuals have disassembled their ERR99 frozen cameras and reported things like loosened solder splats that were causing shorts, ribbon cables not firmly seated in their connectors, loose metal shields in the camera that could move enough to short or ground an electrical part, etc. If you’ve never seen the insides of a digital SLR, there are an amazing amount of electronic connections packed into basically no space in there.
There are a number of “what to do when ERR 99 occurs” lists that I’ve summarized below. Before we jump to them, though, I think its important to try to narrow down the ERR 99’s cause, rather than to go through the list and hope everything is better. I realize “finding the problem” sounds remarkably obvious, but bear with me for a second. With hundreds of lenses mounted to hundreds of different cameras at any one time, we deal with a lot of ERR 99 problems. Some ERR 99 issues are specific and immediate: suddenly the camera starts giving ERR99 with almost every shot. Resetting the electronics helps for a few shots— or doesn’t— but the problem recurs and the camera is basically useless. These “catastrophic” ERR 99 episodes usually means there’s been a major injury in the camera or lens: shutter failure, circuitry burnout, etc. Sometimes the fix is easy— new battery, smoothing a jammed shutter curtain, changing lenses, etc.— but most often a trip back to the mothership for either the camera or the lens is in order.
Other ERR 99 problems are more subtle: ERR 99 shows up after a lens change, lets say. Cleaning the contacts makes it go away, but it comes back a week later, getting more frequent over time. It may just be with one lens at first, but may start occurring with other lenses. In these cases I think it important to remember the point about electrical voltage made above: if voltage drops below 7.3V in the camera’s circuitry, ERR 99 is likely to occur. Voltage drops across different connectors are cumulative, and batteries produce less voltage as current increases. What may seem a case of dirty contacts may really be a narrow power margin, due to oil on the lens contacts, an old battery that’s not producing its rated voltage when fully charged, a lens that’s sucking down power to run the IS servos, which finally drops below the lower operating threshold when autofocus is activated. Cleaning the lens contacts might help, but that doesn’t make it “the problem”. So, be careful when diagnosing an intermittent problem. Its also important to do everything you can to narrow down the problem. Sending the camera to Canon for “intermittent ERR99” without more information is likely to lead to “can’t reproduce problem” at the Canon Service Center.
For example, once or twice a month we’ll have a customer tell us “the lens is causing ERR 99 on my camera, none of my other lenses do that”, so we send them a replacement. We have the luxury most individuals don’t get in that situation: the ability to test that lens on multiple cameras, plus the customer will try another copy of the same lens on their camera. In some cases, the customer will tell us the second lens is the same as the first on their camera; meanwhile, the first lens seems fine when tested on other bodies. Here, the problem is a weakness in the customer’s camera body that became apparent only when a lens with heavier power requirements was used. In other cases, the second lens works fine for the customer and the first lens, when returned, gives ERR 99 on other bodies. Again, problem obvious, the lens had internal damage to the electronics or aperture system.
In a lot of cases though, the problem is less obvious. The new lens works fine for the customer, the old one seems to work well on other bodies. At first, we just shrugged our shoulders and said “one of those things” but over time, as we track the problems that occur with various copies of lenses and cameras, something became apparent to us. Unless we found the specific cause of ERR 99 and corrected it then the problem, while intermittent, would recur. We’ve worked on developing an ERR 99 stress test for lenses that only show the problem intermittently: we use an older camera body, halfway charged battery, and take up to several hundred shots being sure to change the aperture, zoom, and focal distance frequently. Doing this we’re sometimes able to reproduce the problem in a lens that otherwise seems to have just had some isolated ERR 99 reports.
That being said, there are also circumstances where ERR 99 has occurred and then never, ever happened again:
This is based on Canon’s technical support suggestions, with a couple of additions we’ve made as we gained experience.
First, we pull a “tech support” reboot.
Sometimes that works. If not then a full reboot:

If there’s no ERR 99 after the above, the next step is to check the storage system:
Finally, check the lens:

Note #1: Some ERR 99 problems occur only with the camera in certain settings: i.e only in Av mode, or only in AI servo. In these cases, repair is almost certainly needed.
Note #2: I know first-hand that Canon Service Techs use the “gently rubbing a clean pencil eraser” technique of cleaning the electrical contacts. I also know that knowledgeable electronic engineers state this is a bad idea, that the friction could wear out the gold plating on the electronic connections, leading to corrosion. Radio shack and other electronics stores sell electrical contact cleaning solution that can be used with a Q-tip or soft cloth to clean the contacts as an alternative. I’m still using the eraser; I figure if rubbing metal contacts across each other every time I change a lens isn’t wearing out the coating, the pencil eraser sure isn’t.
I know this has been an overly long and probably not-very-useful essay, but it’s a topic I really got into. I certainly will have made some omissions or mistakes in something this long and complex. I welcome corrections and suggestions from any of you with different experience in this area and plan on updating and upgrading this piece as I get more input.
Roger Cicala
LensRentals.com
More Lensrentals Articles