[reduno]

I acquired a Pentax K-1 manufactured in September 2016, with a shutter count of only 4,095 (firmware now updated to latest V1.53). The camera was described as shutting off after 5-10 seconds but the issues seem more involved. At times the camera seems to operate normally but pictures appear to have a purple tint that takes over as the ISO is increased. The first two attached photo crops (sorry snapped through a window) were taken at ISO-800 and ISO-1600 respectively. Increasing the ISO further completely washes the photo out with the purple tint. The metering and exposure appear to be correct.

Switching into live view when the camera is working produces a correct display on the rear LCD. However, if the camera is left on for several minutes the display will tend to start breaking up. If the camera 'gains up' the LCD when pointed at at darker scene, the purple tint will often appear on the LCD view. Bands of purple can also appear on the LCD display, as shown in the third attached photo. Eventually, the entire screen can turn purple and the camera will lock up.

When the camera is not working, turning on will allow full menu access and the GPS will operate and display correctly. However, taking a photo using the optical view finder results in either a completely black or completely white photo depending on the ISO. Photos taken at ISO-1600 and ISO-2200 respectively, are on either side of the black to white flip point. Again the metering and recorded exposures appear correct.

If the camera is placed in live view mode when not working, the rear LCD is illuminated but is completely black. The standard text information and level bars are not displayed either. At this point the majority of the buttons are non-functional, although the ISO will change on the top display when the third wheel is turned (if function is set accordingly). The camera will not exit live view and the camera will not turn off. If the camera is left alone after turning off, the LCD will turn to purple (with some white dots) after several minutes. Shortly after that, the mirror cycles back down and the camera does eventually turn off.

I am suspecting the camera may have an intermittent communication issue with the sensor? This may explain the long delay before turning off, assuming the microcontroller is waiting for interface timeouts before completing the cycle. At present, I have tried reseating the flex circuit connectors on the control board but have not dived any deeper. I wanted to see if anyone had seen something similar before (with any model), or if this was likely going to need a trip to Precision, if covered by their flat-rate repair fee?

Attached Images

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

---

Last edited by reduno; 03-04-2021 at 09:30 AM. Reason: Removed non-informative blank photos

[ramseybuckeye]

I wanted to see if anyone had seen something similar before (with any model), or if this was likely going to need a trip to Precision, if covered by their flat-rate repair fee?

Very odd problem, looks very serious, Precision's estimate is not a flat rate fee, you need to read that a little closer.

[DeadJohn]

Magenta is red+blue color channels, so that color suggests the green data is getting lost somewhere. If it was just in saved photos I would say bad memory card, but you are seeing it in liveview, and you are also getting camera lockups.

There are so many things that could be wrong that I don't think it's a DIY repair. Maybe it's a simple broken cable, capacitor, or voltage regulator, but it's infeasible to start randomly replacing parts. The problem could also be the sensor, main circuit board, or multiple parts.

[caliscouser]

The camera appears to be missing some screws. Have you been in there?

If not someone went inside and either tried to fix this or tried to fix something else and messed it up.

Does it have any impact damage?

[Not a Number]

If you bought the camera from eBay return it "not as described".

Cost of repairs plus what you paid already will probably be more than what you can buy a used K-1 in good condition.

[DeadJohn]

Very odd problem, looks very serious, Precision's estimate is not a flat rate fee, you need to read that a little closer.

Agreed. That initial Precision dollar amount is meaningless.

I think they are purposely misleading customers by labeling it a "repair estimate", then bury details in legalese under terms & conditions that says it isn't really an estimate. They want people to send in cameras based on their lowball estimate. I can understand not wanting to commit to an estimate before they see the camera, but they should not call it an estimate then.

Also note that if their real estimate ends up being more than the cost of new camera, so you choose not to proceed with repair, you are responsible for paying shipping plus a "diagnostic fee". They don't state an average or maximum diagnostic fee. So, if Precision incompetently botches their diagnostic work and spends a crazy amount of labor determining that a camera is too broken to fix, you have already agreed to pay as much as they want to charge you, with no upper limit.

Their terms & conditions enable them to engage in shady business practices should they so choose. Maybe they've never hit a customer with a $1000 diagnostic fee before, but that's solely at their discretion.

[reduno]

Precision's estimate is not a flat rate fee, you need to read that a little closer.

I understand that the estimate from Precision is subject to revision once the camera is inspected, and may require my consent for them to continue with the repair. However, if the cost is much higher than the standard estimate, the viability of the repair becomes questionable. Unfortunately, I don't know what their diagnostic fee will be if I decide not to go ahead with the repair.

it's infeasible to start randomly replacing parts.

I agree that randomly changing parts is not the way to go. I was hoping that this was going to be a simple fix and allow me to enter the full-frame world at a reasonable cost. However, without documentation it becomes very hard to diagnose the root cause, which is why I asked if anyone had seen anything similar (I certainly appreciate your comment on the missing green data). My other options is to list the camera again as "parts or repair" and try to reclaim some of my initial outlay.

[Fogel70]

As it seems to happen when the camera been on for a while it is probably increasing heat that triggers the problem, and could be bad solder joints, bad cable connection, or failing component somewhere between sensor and display. I do not think it is the sensor itself that is bad.

But they will most likely at least have to replace the main PCB in the camera, but possible more parts. I may be a problem that is difficult and time consuming to diagnose so it may be more cost effective to replace several parts at once instead of trying to pin point the exact cause and then replace only that part.

I'm not sure what happens if they replace the main PCB on a K1, if they by standard upgrade the camera to K1 II?

[reduno]

The camera appears to be missing some screws.

As indicated in the initial post, I did reseat the flex circuit connectors on the main control board. This required opening the back of the case to gain access. At this time, I have not replaced all the screws that were removed. The camera is in very good cosmetic condition, with only a few minor chips in the paint. I do not see any evidence of impact or moisture damage.

If you bought the camera from eBay return it "not as described".

The listing indicated no returns but I can try opening a case, as these issues were not described (or covered by a blanket 'there may be other issues' statement). I was not able to replicate the 'camera shuts down' issue beyond the expected Auto Power Off and Meter Operating Time menu options. I suspect the seller was not knowledgeable about the camera, and was unable to do more than a simple power on test.

Agreed. That initial Precision dollar amount is meaningless.

That is certainly the risk of sending the camera in. I did find a few posts on the forum where Precision had replaced a control board or shutter for the standard estimate cost. However, with the potential complexity of diagnosing this issue I certainly have my concerns.

I'm not sure what happens if they replace the main PCB on a K1, if they by standard upgrade the camera to K1 II?

I'm going to guess that is probably unlikely. I suspect there are likely a number of K-1 boards floating around form cameras that have been upgraded. However, I'm not certain how ethical it would be to use one of those in the repair of another camera without full disclosure.

[C_Jones]

If you sent it to Precision with a note describing the problem, they would check to see what is wrong with it and tell you the source of the problem and repair cost. I am not sure if or what an associated diagnosis cost may be that might be charged by them. You may want to inquire with them about specifics.

I have gotten some purple effects when leaving my viewfinder uncovered during long exposures (due to light leak), but other than that nothing like you are seeing. If you do not have your camera closed completely, light leak may be partially or fully to blame.

Good luck.

[reduno]

Update #1: I was able to negotiate a partial refund with the seller, based on the camera not being as described. This gives me a little more wiggle room to cover the cost of repairs (time will tell if it is sufficient) . In the meantime, I have been investigating the problem further, and have fabricated a test bracket to support the rear panel (see attached photo). The bracket attaches to the tripod mount and removes any stress on the flex circuits while I probe the main control board. The white clips in the photo secure the rear panel to the bracket, and hold the top cover in place to allow the main power switch to operate.

Using this setup, I was able to determine the issue with live-view appears to be more prevalent when the camera is cold. Enabling live view when the camera is cold results in a completely blank LCD display. If I use a hairdryer to slowly warm the camera, the informational text and level bars will appear on the display, and live view will start operating. I am now trying to identify the specific circuit locations that appear to be temperature sensitive.

I also found a link that shows pictures of the internals of the K-1 duding the upgrade procedure to a K-1 II:
How Pentax K-1 is upgraded to K-1 Mark II | YURA photography
One thing that intrigued me, is that the rightmost flex circuit for the sensor board appears to have some additional connections made by wires. This looks like a power supply connection fed from a separate board in the camera. I suspect that Ricoh may have discovered during the development phase that more current was needed than the standard control board flex connector could provide. Rather than redesign the entire interface, it looks like they may have added the additional connection as a modification.

Interestingly, I am not able to measure any voltage on these pins (I have confirmed one is a chassis ground connection). The separate power supply board appears to incorporate some sort of switching regulator but I cannot determine the component part numbers form the photo. I have been able to determine the battery 7.2V supply is present on the wires feeding this board. If the sensor board is being starved for power because this supply is missing, it may explain the strange operation (semiconductors usually draw less current at higher temperature). I will update once I have more information.

[reduno]

Update #2: The mystery deepens. Without unsoldering any wires, I was able to raise the main control board a little to gain some visibility below. The separate small board underneath appears to match the configuration of that shown in the YURA photography link. Using a small mirror, I was able to read the marking on top of the main components on the board. However, at this point I have not managed to cross reference these markings to any manufacturers data sheets.

Looking closely at the configuration of the circuit, it looks like the topology matches that of a fly-back converter but not all the expected circuity is included on the board. The attached image shows my interpretation of the likely schematic for this board. The main switching component is probably a N-channel MOSFET, and has the top marking ".OM" (. is most likely the pin-1 designator). At first, I assumed this chip was a complete switching converter but there appears to be no direct connection to the input supply. The pin assignment (with 4-pins connected together) matches other MOSFET devices in a 6-pin SC70 package. Assuming this is a fly-back converter, the coil component would need to be a transformer. It does have 4-pins and the top marking is "4tJ", which is different to that expected for a simple inductor. Unfortunately. I can't get a good view of the side of the part to confirm the windings connect to all 4-pins.

For this board to operate as a fly-back converter generating a DC supply voltage, a Pulse Width Modulation (PWM) controller would be needed on the main control board, and a rectifier/filter circuit would be required on the sensor board. I have shown these components dotted on the schematic. Whilst this is possible, I'm surprised Ricoh is routing the switched AC output through the wiring harness, as this would generate some Electro-Magnetic Interference (EMI). It also seems contrary to my original assumption, that this was an add on circuit after the initial main control board design was complete.

When I made my initial voltage measurements, I did set the multimeter to both DC and AC volts but did not measure anything. I may need to go back and try probing with the oscilloscope, to see if there is any switching going on. At this point, I am still uncertain of the intended function of this circuit.

[AfterPentax Mark II]

I saw this once and hope for you that it is not what I was confronted with: a sensor dying. I had this same behaviour with my Fuji digital camera (only not purple but white). As the sensor does not give no or wrong information for the exposure the shutter time will be the longest. If you set the switch to turn the camera off, the camera will end the exposure cycle and after that the mirror goes down and the camera is shut off. But again, I hope it is not this, but be prepared for the worst.

[reduno]

I had this same behaviour with my Fuji digital camera (only not purple but white). As the sensor does not give no or wrong information for the exposure the shutter time will be the longest. If you set the switch to turn the camera off, the camera will end the exposure cycle and after that the mirror goes down and the camera is shut off.

Thank you for providing feedback on a similar issue you experienced. This is the sort of information I wanted to receive, even if the final outcome is not what I am hoping for. It does sound like your symptoms were slightly different, and I understand why the loss of metering information would result in a very long exposure and a totally white picture. I'm not certain which Fuji digital camera you had but I don't believe they have made a DSLR since the 2007 FinePix IS Pro.

On the K-1 I'm seeing issues using both live view (metered off main sensor) and the optical viewfinder (uses a separate light meter). The mechanical sound when taking a photo using the optical viewfinder also suggests the exposure time is correct (not grossly extended).

I'm certainly not ruling out a main sensor problem as the root cause and certainly appreciate your comments. I have not had a chance to pursue the issue much this week but I did order a temperature controlled soldering iron to help with further disassembly (if I go that route). True to my luck, the one that turned up was the wrong item, so I am now waiting on a replacement.

[reduno]

Update #3: This weekend, I was able to re-secure the main control board and take some voltage measurements. This is a double sided board (components on both sides), so I was only able to access the power supply components on the top side. These primarily consist of an Analog Devices ADP5080 integrated supply chip and two Monolithic Power MP8715 switching converters. The ADP5080 contains 2x stay alive Low Drop Out (LDO) regulators, 5x step-down converters, 1x step-up/step-down converter, and 1x high-voltage LDO. The MP8715s each contain a single step-down converter.

These components are located on the left side of the main control board, as indicated on the board plan view in the top right corner of the first photo. For the ADP5080, I was able to identify the passive components associated with the 6x switching outputs. The measured output voltages for these 6x circuits, plus the MP8715 circuits are shown on the first photo, along with the corresponding board test points. The transistor in the lower left of the photo appears to be used to switch the supply from the battery grip (if one is installed on the camera).

Obviously, I don't know what these voltages are intended to be but the 3.286 V (3.3V), 3.296 V (3.3V), and 1.799 V (1.8V) correspond to standard logic supply voltages. The 0.907 V (0.9V) could also be a core supply for the main Pentax custom Application Specific Integrated Circuit (ASIC). The other supply voltages, 1.361 V, 4.034 V, 5.505 V, and 6.500 V are all fairly close to standard 0.1V supply increments.

For the switcher circuit on the small board underneath, I was able to probe the output with an oscilloscope and see a very low amplitude signal. The second photo shows this signal using a vertical resolution of 50 mV per division and a time base of 100 us per division. I was also able to probe this signal with a Fluke 115 multi-meter and measured only 9.8 mV RMS on the AC mV range (standard AC V range used previously did not have enough resolution). I was not ale to see any signal on the control input to this board when probed with the oscilloscope. Because of the low amplitude, I believe the signal observed on the output is likely just noise, rather than the intended switcher waveform.

I'm still puzzling how this additional circuit is intended to operate. I do not see any components on the top side of the main board that would provide the expected switching control waveform. However, it is possible the PWM controller is mounted on the backside of the main board and currently not visible. It is also possible that this switching circuit is not utilized in live-view mode but this seems unlikely. The only remaining possibility would be if it is somehow related to the use of the electronic shutter in pixel-shift resolution mode? I'm starting to think the next step may involve removing the main control board.