Pentax SDM Failure Survey Results

My da 50-135mm just suddenly failed to focus last week, so I imagine I have suffered this problem too. I took it into the Pentax shop in Nishi-Shinjuku, Tokyo just today and will have it back in 18 days :(

Moral of the story... don't buy a used out of warranty 50-135 which is what I was very close to doing. Le sigh.

However it is true that people are far more likely to seek out a place to complain about a problem than they are likely to go around posting that everything is A-ok.

Well I just sent my DA★ 16-50mm F2.8, a second time, for repair. I purchased it in 2008; got a first fix in 2010 under credit card warranty (yesss!), but this time it will be out of my pocket. It is in a mint state, I take good care of it and it spent maybe 30 to 35 % of time on my camera, share with other lens. And I don’t even shoot a lot maybe 20-40 pictures a week on average so it is certainly not abuse. I know it is the auto focus but I don’t have the exact cause yet; I sent it yesterday, but I would bet IT IS SDM... I don’t think it is a typical normal failure rate level : too much feedback on it.

@david94903 The first time my 50-135mm's SDM motor failed, the motor simply ceased to function. I could still focus manually. It was sent to CRIS and repaired under warranty. The second time it failed, not only did the motor become completely unresponsive, but the focusing group was jammed in place too. Turning the focusing ring accomplished nothing, the clutch just slipped and slipped. So it's really not obvious what the single point of failure is. And all the random guesses and conjectures posited in previous comments are just that: random guesses and conjecture. They've all been put forward in one way or another during previous years, and they've all been debunked. No-one really knows what the most common point of failure is, except Pentax. And they'll probably never tell us.

After the second failure, I got it fixed again (under the first repair warranty because the original warranty had expired) and I sold it to some other poor sap. Wonder how long it lasted for him...

Okay, so here is the question. What it the point of failure for SDM lenses? Is the motor burning out, is there a bad contact, is there a firmware issue, etc? I've read the DYI post about fixing it yourself and based on the sucess comments of that post, it doesn't seem the motor is actually failing. My 16-50mm SDM failed on my K-5 but the screw drive AF works on my ist-d. AF doesn't work on my K20d but I have the latest firmware on that one, though I've read here in the formum that the screw drive AF will function with older firmware. Would a firmware update with a "SDM switch" fix the problem? I don't have any answers, only questions but would like to know what others experience.

At least there is an emergency option. The screw drive. Be glad its there. If I am reading this correctly, these lenses still work if you disable the pin the controls the sdm motor and they revert to screw drive. Thats pretty handy. I kinda like the screw drive and its loudness. It sounds beefy....robust. It seems to take some abuse too. like when I go to quickshift and tap the shutter and it auto focuses while my hand is still on the ring. I'd say its pretty tough. I just like whatever works personally. These people are right, lenses all do tend to fail. That said, its really hard to say with these flawed statistics how much sdm fails. I'm sure pentax has some excellent figures. If it were more of a widespread problem I think it would be brought up more. That said, a 1 year warranty is not enough on a $2000 lens.

I think the results don't convey the overall real failure rate, but it is indicative of the populace that frequents one website's apparent run of bad luck with the lenses.

If you read the Amazon reviews, there are always at least a few on most or all of these lenses who have had problems with the AF failing. So this survey may inflate the actual failure rate, but the fact of the matter is, the failure rate is way too high and is seemingly not being addressed correctly. Yes it does make Pentax look bad, and that is good because it will force them to address it.

I'm beginning to think that modern Pentax lenses are overrated. The drive motors fail at a relatively high rate, I had a DA 55 - 300mm that jammed after four days of use and the kit 18 - 55mm that came with my KX turns out not to be anywhere near as 'best of breed' as I had read in reviews, certainly not in comparison to the Nikon 18-55mm I changed to later on.

But I have a really old f1.7 50mm lens that is awesomely sharp and with excellent contrast.

Yes folks, they really did build them better in the old days.

@Class A: I'm not trying to downplay SDM issues. I've frequently said that Pentax has done a lousy job of communicating to their customers on the issue - and there is an issue. Class A, I didn't bring up the comparison to HSM, Adam did about 6 months back when he replied to me and said that one couldn't find any HSM failures on the internet. I replied and said there were failures if one searched on x and y terms. You just recently jumped into the middle of that discourse and moved off in your own direction. My only new issue here is that analysis of data from this poll is flawed by self-reporting bias. Nuff said.

At the time of the survey the SDM on my 16-55mm was working fine, then a month or so after the survey it failed. So you can move one OK to FAIL on the 16-55mm chart. My lens was purchased in 2009 and had performed well until recently. I have to say, given the price of the lens, I am hesitant about purchasing another Pentax SDM lens. I have the 50-135mm on my "wish list" radar but I'm seriously rethinking that purchase.

@ClassA
There are two basic classes of ultrasonic motors, micro motors and ring type motors. I know Canon and Nikon use USM and SWM interchangeably for either type of motor. Pentax uses exclusively the micro motor type with their SDM at this time.

If you look at the latest Lens Rentals Repair Survey, the top (worst) 6 lenses all had issues where they were sent in for AF repair, and these are the Sigma 50-500 OS, 120-300 OS, Sony 70-200, Nikon 70-200 VRII, and Sony 16-80. All had repair rates 20% or higher (this includes OS/VR issues and zoom issues too).

@philbaum: The number of HSM lenses sold must surpass sold SDM lenses by a huge margin. Yet, searching for "HSM failure" turns up less than a handful of results. The same search for "SDM failure" has relevant results on five Google pages.

BTW, the only result for "HSM sticky rollers" is your own comment here.

Like anything technical, HSM mechanisms can fail, but trying to downplay SDM issues by pointing to a much better technology (ring motor in lenses comparable to the DA* lenses) that is much more reliable does not cut it.

@philbaum: You write "The probability of me having a failure of any of my 3 sdm lenses is the sum of each individual probability.".
This is wrong. The probability is 100-(100-p1%)*(100-p2%)*(100-p3%). And even if the probability for an SDM failure would be 99%, you could still be the one out of hundred who never sees an SDM failure.

Interesting information in a quantitative manner. However, readers should be cautioned to avoid seriously reading anything into these numbers. The sampling methods are specious at best. I myself have all the DA* zooms since the day each became available. I saw this survey and chose to pass on it as I could not recall off the top of my head which year I purchased the lenses. So there are three lenses that have had no problem at all and have been in continuous use for many years. Enjoy the data but don't allow yourself to be driven to distraction or fear by the numbers. They mean almost nothing.

The folks complaining about sdm failures probably had not had usm/hsm/everything-else failed on them before. The problem is not exclusive to Pentax. Otherwise the other brands won't need service centers. I have had my Canon lenses fail on me before(the af stopped working, the zoom locked, etc), even the flash. My friend's Nikon camera buttons stopped working during an event, he had a backup camera so kept right on working. Plan for failure.
Just get it fix, things breaks down ya know, the more parts it has, the more likely it'll fail.

The methodology of the survey makes the results very easily misinterpreted....as supported by the comments here. Pentax has enough problems without publishing something like this. There will be failures on Pentax lenses as there will be on all others. A more useful survey that would help put things into perspective is a random sample of buyers of all brands....or at least a few others. Obviously that would be a very difficult undertaking if not impossible....but it would actually mean something.

AntiV: Sorry if i came down hard on this poll you apparently wrote. I'm just alarmed at how its being misinterpreted. Already someone has said that the 50% failure rate of the 17-70 in 2012 is pitiful. And there have been similar comments. Its quite silly to think that there is an actual 50% failure rate from this flawed data. I do find the total data interesting, but then i take the "failure rate" with a large, large bit of doubt. But others believe it???
Class A: Yes it does predict it. The probability of me having a failure of any of my 3 sdm lenses is the sum of each individual probability. Just adding up the 2 above equals 54%, i don't see a rate for the 300N for me to add in.
I've previously read a comment on PF that the solution to the SDM problems is for folks to buy HSM lenses by Sigma, because it was stated there are NO reports online of HSM problems. So i checked into it and found all kinds of failure reports for HSM. There's an issue called sticky rollers with HSM that still hasn't been fixed and some very unhappy HSM customers. (i don't doubt there is an issue with sdm, but its nowhere near the issue it was, which means that some work has been done to fix it, imo. But i don't have good statistics either)

Where are you guys coming up with 50%? 35% is the highest I see. And yes, even to someone who is completely unfamiliar with statistics, the bias is obvious, but this is about qualitative, not quantitative analysis. Even with the bias we can infer that certain lenses are more prone to failure than others. The two bits of info that I take away from this survey are: 1) Yes, there does seem to be an issue with the SDM system, and so you are taking your chances when you buy an SDM lens (consumers have the right to know this); and 2) No, it does not seem to be improving. The latter worries me as much as the former.

One real life example. I have two SDMs: 17-70 (since Feb 2009) and 50-135 (since May 2010). At the end of its 2 year warraty the 17-70 stopped frequently focusing when zoomed beyond ~35mm. Even switching to manual focusing did not help - the focus indicator just kept blinking. However, I soon found out that I could reset AF by zooming back to 17mm. The fault was repaired by replacing a chip, so they said. Recently the same fault appeared again, but thanks to the earlier practise I rarely miss a shot any more. If the fault does not get worse, the expensive repair can wait. I've had no problems with the 50-135 so far - knock wood.
Anybody else met with similar problems?

I think response bias could be seriously skewing the results of this survey. I think those who had failed lenses would be more likely to find this survey and respond. A more fair survey would be to randomly choose x people on the forum who identify as having these lenses.

Not QC problem. It is a fundamental design flaw. Scrap it!!!

Not looking good. Even a 10% failure rate is abysmal. Yeah this makes pentax look bad. Even if the results are skewed, it still shows a relatively high rate of sdm failure. Pentax really needs to address qc issues. If they keep up like this, I think I will have to start looking long and hard at nikon. I hear they have a full frame camera or two or three too!

@maxfield_photo: Everyone knows about the QC issues already. PRIC (and Hoya before them) are well aware of them and have taken steps to improve them. And they have much better data on failure rates, which they can project from warranty returns, than does anyone reading this survey.

This survey strongly suggests a 50% failure rate, which isn't remotely honest or fair. Not only are users with broken lenses more likely to answer the survey, they are also more likely to come to this site in general, as it is probably the best source of information on the web for malfunctioning Pentax equipment. As many have already mentioned in this thread, the methodology of this survey is a flat-out joke.

Applaud all you want. The results of this survey are much more likely to get Adam a cease-and-desist letter (and a threatened libel suit) from Pentax than they are to generate positive action.

Useless. Well, not for those who'd gladly use it to spread panic (OMG! 50% failure rate!).
Why do you bother with what you admit being a biased "Failure Rate"? Have you made any effort to skip invalid input? Do you really believe those numbers?
Next time, ask a statistician for help.

Basically useful info. Thank you for doing this survey. At some point, a good second set of eyes and editing of copy would help make this look and read better. Several of the sentences don't read well. Agreed that DC motors may be a good idea for other lenses in future. Best wishes.

@philbaum: I'm sorry, but it is your comment that is of no value. The Survey was for how many lenses out of X experience AF problems. It didn't survey when or if any particular lens should fail based on its age. If you have had no problems, then that's ok. Your copy is just one of the good ones.

Whatever the statistical bias, the survey puts figures on a reality: SDM fails. Pentax Forums, thank you for conducting it.

I hope Pentax is listening and is going to address this serious quality issue by 1) delivering better products and 2) recalling faulty product.

I had to pay for my SDM 50-135 to be repaired. Maybe some lawyers would launch a class action!

I would like to thank this survey in order to make some feedback for Pentax QC and, hopefully, they would hear some. Moreover, I have 3 things I want to say.

1. I don't think the DA 18-135mm uses SDM. At least, I never heard Pentax says that. OK, it's motor and it's pretty quiet, but it might be some different type of motor than the SDM. Also, Pentax call it DC motor(which might stand for Direct Current, I guess). So, it would be better if Pentax, if possible, put this DC motor in DA* lenses in the future.

2. Based on my friends uses, we have observed, we have discussed a lot and we guessed that the SDM failures might come from the lubricant grease is dry. Then the motor is over-load. We found that almost failure lenses usually have been seen in the one that not be used frequently. The lenses were put in dry environment for a long time like in the dry box for 2-3 months and used just a day. On the other hand, the lenses that have been used every day are just fine. I'm not sure if this is correct, but as I mentioned it's our guess.

3. some copies might not reach the time of fail yet. So I think this percentage number of failure is the minimum as we can know now. It will increase in the future.

Pentax please consider this issue seriously and do something before step up to Full Frame world.

This is an absolutely hilarious parody of a scientific survey. Don't you all get the joke?

@dcshooter This survey is an attempt to bring to light a QC issue that has plagued the company since they moved their lens production overseas. I agree it's a bit scathing, but there's an old saying, "If the shoe fits..." I hope the executives in Japan DO see this and decide that they need to take steps to remedy the situation, because apparently they've been in no great hurry up till now. I applaud the efforts of the PF team!

Man, is this forum really hurting for content so badly that within a week, it publishes two front page articles biting the hand that feeds it? First a lens comparison using a suspected faulty lens, one that unfailry makes one Pentax's premiere primes look like a dog. THEN a completely unscientifc "survey" that makes the SDM problem look way worse than it really is?

And we wonder why the brand suffers from such image problems! I thought we were supposed to be Pentax Fans around here. It's one thing to give oneself fair criticism, but this site is can't be making many friends at PRIC when it's coming uncomfortably close to libeling the company.

Any research methods or statistics professor would have a stroke if they saw this survey methodology...even with the "response bias" disclaimer! Trust any conclusions only as much as you're willing to trust the National Enquirer and similar scholarly journals.

My Sigma 50-150mm f/2.8 HSM currently for sale in the marketplace is looking better all the time... :)

@philbaum: No, the numbers do not imply that you should have had a failure by now. Even a high likelihood of failure permits individual copies to stay fine indefinitely. We are talking about probabilities, not certainties.

Surely self-selection is likely an issue in this poll, but it seems valid to conclude that the 16-50 & 50-130 have significantly higher failure rates than other lenses. While the absolute numbers may be skewed, the relative performance should be pretty reliable. Or are we saying that zoom users self-select differently compared to prime users?

@ philbaum. Hence no professional statistician has attached their name to the results. The bias is acknowledged. But the results do give us a general idea of the state of SDM and raise some interesting questions. Particularly about the 18-135 technology and whether it will be used on upcoming lenses as a replacement for SDM.

These results seem to match my own experiences (I've owned the 18-135mm and 50-135mm). Of course, there will always be the doubters who have not been personally affected by a failure and therefore dismiss everyone else's concerns out of hand.

As for "response bias", it would seem to me that if it was statistically significant, it would have affected ALL lens models, not just 3 of them. That is not the case. The vast majority of owners have reported no failures with the 55mm, 200mm, 300mm, 60-250mm and 18-135mm. It would make no sense for "response bias" to affect only 1/3 of lens models.

Those 3 "bad" models are an embarrassment.

Sorry, this survey is of no value. I've had a DA50-135 for 3 years, a DA300 for 3 years and a DA17-70 for 6 months. According to these "statistics", i should have had a failure by now, and that hasn't happened. They've instead been stellar performers. I don't doubt there have been failures, but there is no statistical validity to the method in which this data has been collected, therefore no conclusions can be drawn from it. No professional statiscian would permit their name to be attached to a report such as this.

If you experience an out-of-warranty SDM failure, try putting a small piece of electrical tape over the SDM contacts on the lens. I accidentally shorted out the circuitry on my K20D, and now my DA*50-135 focuses using the screw drive. I think if the camera doesn't detect an SDM lens, it will revert to screw drive.

Still lovin' the 18-135 and this just makes it easier to enjoy and realize that its also a reliable lens.

Regarding the question of "being stored for a while" -- some have also suggested that there is capacitor that loses its charge over time if the lens is not used, and the SDM may be revived by leaving the lens on the body (with the camera turned on and in-use presumably) so that it may again receive power and charge up the capacitor.

Did the survey offer any information regarding how the failures occurred - quickly, or over a period of time? I seem to recall reading several reports of lenses that were stored for a while, then were found broken when they were mounted. My k100dS and DA 17-70 combination seems to be haviing more issues with focusing than it did in the past. While it does not appear to be "failed" at the moment, I just wonder if the lens is in the process of failing. I also wonder whether a "newer" lens in combination with an "older" body is a potentially problematic pair, especially since the body came first.

...
Just bought a DA* 16-50 tonight...

The failure percentage rates for some of these lenses is insanely high. Even a 10% failure rate is a disaster. I see a couple of the lenses have decent rates, but the 16-50mm and the 50-135mm are still awful.

Yes I think there probably is a survey bias - people who have had failures are more likely to respond (want to tell people) than people that have not.

That said these failure rates still look too risky for someone like me (live in NZ and would be importing the lens). Any warranty return is potentially expensive (shipping) and slow.

i think if we wanted "real" statistics we'd have to ask forum members to first register their lens when they receive it new, then survey them after a month/year... survey bias is too high with this polling method to give these numbers any real weight.

sadly the 2 "best" zooms have the worst record, they really should be updated to recover their reputation

A 50% failure rate as of 2012. Pitiful.

Ouch! This is very ugly picture, but it shouldn't be shocking to anybody who hangs out on the forums.