[karro]

overall, the K3 is definitely more power hungry. the same mirror - regardless of the shooting mode, I don't think it's lifted any slower than in "full throttle 8.3fps mode", meaning, it is lifted with the same speed; I've forgotten all the formulas, but it requires more torque to make it move faster. Metering requires more calculating power, white balance analysis has been improved (and requires more cpu), data writing/processing speed has been increased, I think, this may put a serious strain on the battery, which was perfectly fine and capable handling something less demanding. just a theory, though.

[DRabbit]

overall, the K3 is definitely more power hungry. the same mirror - regardless of the shooting mode, I don't think it's lifted any slower than in "full throttle 8.3fps mode", meaning, it is lifted with the same speed; I've forgotten all the formulas, but it requires more torque to make it move faster. Metering requires more calculating power, white balance analysis has been improved (and requires more cpu), data writing/processing speed has been increased, I think, this may put a serious strain on the battery, which was perfectly fine and capable handling something less demanding. just a theory, though.

Not a bad theory... But what would be the proposed fix? A newly designed battery?

[karro]

Not a bad theory... But what would be the proposed fix? A newly designed battery?

I don't know. if the theory holds true: either that OR the firmware, which "underclocks" the machine, but that would be cheating of sorts (all claims about the enormous speed would need to be reviewed), and I'm not sure, if the firmware actually controls the speed of the mirror mechanism.

[awaldram]

I don't know. if the theory holds true: either that OR the firmware, which "underclocks" the machine, but that would be cheating of sorts (all claims about the enormous speed would need to be reviewed), and I'm not sure, the firmware actually controls the speed of the mirror mechanism.

It's a tough one even if true as I doubt they can lower the IR of the battery so it can absorb load fluctuations better.
Though given the issue seem to be a minority maybe it not so much fix the battery as grade them prior to use.

I used to work for a very large computer manafacturer we had a Mother board design that used to actually hit the cas and ras refresh specs of the memmory.

The result was first shipments had unacceptable failure rates with crashes and lockups common. We were already using the fastest memmory of the day so the solution was to limit suppliers to those who could build memmory to the spec they claimed (about 30% of our approved suppliers) and set up in house quality testing rather than trust 3rd parties.

Failure rates went from 8% to 1% inside 12 months.

It's an expensive option and Ricoh would need to be certain that was the only issue and that it was viable to grade the batteries.

Even if it was the issue it might be cheaper to fix on fail much to the customers frustration.!

I still think there must be a k3 body component as if I use some known worn out batteries from 2010 in my k3 it can be pursuaded to lock up but refuses to flop.

---

Last edited by awaldram; 04-25-2014 at 10:45 AM.

[DRabbit]

I don't know. if the theory holds true: either that OR the firmware, which "underclocks" the machine, but that would be cheating of sorts (all claims about the enormous speed would need to be reviewed), and I'm not sure, if the firmware actually controls the speed of the mirror mechanism.

I'm not sure it explains why the issue can happen even with AC adaptor though? What do you think?

[karro]

AC adapter should be able to provide enough "juice" for the camera to operate. I could, of course, say, that probably it's not always the case, and may vary from adapter to adapter, but that would be bending the facts to fit the theory, so -10 points to the "brownout"
awaldram may be right about a faulty component in the camera, in that case - Pentax should fix it for free and do it quite fast

DRabbit, you've been looking into this this since the day one - is there a correlation in faults/camera serial number ranges? If there is - could indicate a faulty batch of component X. if there isn't and it's completely random - this is a bit worse, may indicate a design flaw somewhere, since the components supposedly behave normally on their own. for that we'd need to know the actual range of camera SNs to date.

[DRabbit]

AC adapter should be able to provide enough "juice" for the camera to operate. I could, of course, say, that probably it's not always the case, and may vary from adapter to adapter, but that would be bending the facts to fit the theory, so -10 points to the "brownout"
awaldram may be right about a faulty component in the camera, in that case - Pentax should fix it for free and do it quite fast

DRabbit, you've been looking into this this since the day one - is there a correlation in faults/camera serial number ranges? If there is - could indicate a faulty batch of component X. if there isn't and it's completely random - this is a bit worse, may indicate a design flaw somewhere, since the components supposedly behave normally on their own. for that we'd need to know the actual range of camera SNs to date.

Based on all the reports I've heard, it would seem that it's more than just a batch, only because there's been a wide variety of purchase dates and geographical locations, and of the serial numbers that are there, I don't see a pattern. In this thread, the issue has been reported from serial numbers anywhere in the range of 482XXXX to 491XXXX.

At least compared to the database of serial numbers here on PF, that would seem to extend to virtually the entire range of serial numbers except the absolute latest cameras (492XXXX)... at least from what I can tell from the information available.

---

Last edited by DRabbit; 04-25-2014 at 12:18 PM.

[cfraz]

I would seriously like to know what is the difference between the K5-II/s bodies and the K3 as to show this fault in the K3 but not the K5-II/s bodies.

The K-3 uses three separate motors for shutter, mirror and diaphragm compared to a single motor on the K-5, IIRC. That may or may not address the power question, but certainly the control mechanisms for all major moving parts in the two cameras are quite different.

I was told today to send my K-3 to the Swedish repair shop and they would fix the problem. I will call them on monday and ask if they have a fix or if the sales rep for Pentax just wanted me to move along...

[gryhnd]

Just experienced this for the first time over the weekend. K3's firmware is up to date. Batteries in camera and grip fully charged (grip with a Wasabi battery in it). A couple hours in to shooting a relay race the mirror freaked. Thank god I remembered seeing these posts or I too would have freaked. Turning off the cam and grip solved it, but whoa...not fun!

The only anomalous thing for me was shooting in the rain. I had a camera condom on though (my Sigma 70-200 is of course not WR) so everything remained dry, humidity not withstanding (temps in the 40s).

It didn't happen again for the remained of the event.

[SteveB]

AC adapter should be able to provide enough "juice" for the camera to operate. I could, of course, say, that probably it's not always the case, and may vary from adapter to adapter, but that would be bending the facts to fit the theory, so -10 points to the "brownout"

Good batteries can provide more current than the majority of AC adapters, at least for a while. AC adapters' main advantage is their constant performance, and often they put out a slightly higher voltage than batteries but not always. Simple enough to measure.

[rodeknyt]

Has anyone given thought to the possibility that Pentax got a bad batch of IC chips from one of their suppliers? Those things are probably in many bins along more than one assembly line. Some cameras get a bad one and others don't. Wouldn't even be traceable to any specific assembly run or series of serial numbers.

[SteveB]

Has anyone given thought to the possibility that Pentax got a bad batch of IC chips from one of their suppliers? Those things are probably in many bins along more than one assembly line. Some cameras get a bad one and others don't. Wouldn't even be traceable to any specific assembly run or series of serial numbers.

It could happen but it's unlikely in my (long) experience of the electronics industry. ICs tend to be faultless when new, unless they're fake ones which is rare apart from memory chips.

[jhmos]

For those curious about the K-3s power requirements, I have started getting some results with a load test rig I've been putting together. What I found is that the power drawn from the external power supply is very spiky. I have not identified yet what all the spikes are caused by but it looked like the biggest lump of power happens when the mirror is being moved.
In live mode, the K-3 draws between about 300mA and 600mA just idling.
In normal mode the draw is typically 150 to 200mA.
Autofocus I have not tested much yet. A DA 18-135 DC lens didn't seem to draw all that much though.
Here are some oscilloscope screen shots I took while the K-3 was doing continuous shooting in live mode. I only have an old non-storage oscilloscope, so I had to photograph the screen with my K100. The oscilloscope is set up for 1ms per division and vertical scale is about 400mA per division. I also will try to log some data using an Arduino in the next couple of days.

The first two screen shots are the most typical, with peaks a little over 2 amps. The third is the most extreme I captured, getting a spike up to 3 amps (the trace was not quite adjusted right for this one, so its not 3.2 amps like it looks). The forth is captured with a 2 second exposure, trying to record more shots. It looks like the average draw during continuous shooting is 1.2 amps or so but any power supply better not collapse under a 3 amp spike. There may be higher current spikes still that I have not captured, possibly ones that lead to mirror-flap etc.

[DRabbit]

Jhmos - it would be interesting to test this against other Pentax cameras and even other brands, to see if that is more typical than not?