975 replies to this topic

[mjr]

> What's astonishing with your picture that the accelerometer chip U7 is missing at the top of the board

 

Ah, well, that would be one way to deal with the out-of-production accelerometer chip - just don't install one!

[rickh]

> What's astonishing with your picture that the accelerometer chip U7 is missing at the top of the board

 

Ah, well, that would be one way to deal with the out-of-production accelerometer chip - just don't install one!

Dang it!  I missed it and I have new glasses.  I want a refund! :-)  

[mjr]

 

> What's astonishing with your picture that the accelerometer chip U7 is missing at the top of the board

 

Ah, well, that would be one way to deal with the out-of-production accelerometer chip - just don't install one!

Dang it!  I missed it and I have new glasses.  I want a refund! :-)  

 

Good thing the CPU is still in production.  They'd ship the board with just a bunch of resistors installed otherwise.

[fhjui]

I would try to return this board to the distributor, with a note that an important chip is missing.

The distributor certainly doesn't know this and it's important information for them that they could clarify with nxp.
This doesn't help you now, but I would choose the new pinscape pico project from mjr instead

[mjr]

> This doesn't help you now, but I would choose the new pinscape pico project from mjr instead

 

That's definitely an option, and probably the better option for the long term, although the trade-off is that the Pico project is still very immature.  The KL25Z project is stable and reliable, so if you can find a working original KL25Zs, that'll likely be a faster route to a working build.  But the Pico project runs on hardware that widely available now, at a very low price point, and will likely be available for years to come.  And it does have some nice improvements from the KL25Z project.

[soup001]

A bit strange, the whole thing...

• What's astonishing with your picture that the accelerometer chip U7 is missing at the top of the board, perhaps a similar chip is mounted on the bottom. It would be helpful if you took a photo of the bottom side. The manufacturer's informations all refer to the old revision, no current information.
• Your photo is a bit blurry, you can't read the name of the KL controller properly. It looks like "MKL25Z129VLK4", but I couldn't find such an IC. It must be called "MKL25Z128VLK4".
• Otherwise the board is populated with the same ICs, except for the USB-C ports.
• Please check the content of the file "LASTSTAT.TXT" when the controller is started in bootmode. It must be logged "ready".
• The two files "SERCDC89.." are installation files for virtual com driver. You don't need them here. They are visible when the usb connector is connected to the programmer port (not switched to bootloader mode)
• When connected to the programmer port the RGB LED flashes green/yellowish every 2 seconds four times, then 2 seconds pause.
• The bootloader version of newer boards (~ 2018) is 1.12, but thats ok. Installing the pinscape firmware works well on these boards (boards with round edges and accelerator chip populated on top)

A picture from KL-board from 2018

 

Have uploaded top/back photos hopfully clearer. In regards to the points made:

- I only just noticed the labelling 'ACCEL' and can see the solder points but no chip, also seeming no chip on the other side

- The KL controller indeed is marked MKL25Z128VLK and what appears to end with a 4.

- I didnt even think the old board not having USB-c, should of been a given how long its been in production.

- Correct, does state "Ready."

- Ta.

- When connected no lights are displayed, but is recognised by Windows11 as a "Pinscape Controller"

- Ta.

[soup001]

I would try to return this board to the distributor, with a note that an important chip is missing.

The distributor certainly doesn't know this and it's important information for them that they could clarify with nxp.
This doesn't help you now, but I would choose the new pinscape pico project from mjr instead

Yep, have contacted them requesting a refund to which they have been apologetic and are following up on trying rectify the situation but it does appear with they listing 1,638 globally availble that the manufacturer having instead gone with the new version (minus the accelerometer chip). I'll start looking into the pinscape pico project. Ta.

[soup001]

> This doesn't help you now, but I would choose the new pinscape pico project from mjr instead

 

That's definitely an option, and probably the better option for the long term, although the trade-off is that the Pico project is still very immature.  The KL25Z project is stable and reliable, so if you can find a working original KL25Zs, that'll likely be a faster route to a working build.  But the Pico project runs on hardware that widely available now, at a very low price point, and will likely be available for years to come.  And it does have some nice improvements from the KL25Z project.

I've only just collected all the parts, havent even gone into cabinet design so could be a while before i'm even close to being completed 

The only availble KL25Z I could find locally was $100+ AUD incl shipping, and just over half of that from Ali. If the intent is for the Pico to do everything the KL25Z can do, and hopefully more, then I'm good for that as the option. I'm just building a 1080p system with exciters/shaker/digital plunger and nothing more (will move to a 4K with all the bells and whistles later if desired).

 

Appreicate everyones quick responses (including from the man himself MJR!), was a relief that I hadnt missed something simple and good to have ppl point out the accelerometer chip missing (I assume this being for the 'tilt function') as a strong case supporting a refund given they advertised it having come with one (which it would of should the correct model having shipped!). Thanks again!

Edited by soup001, 27 February 2025 - 07:07 AM.

[fhjui]

Ok, as expected, there is no acc. chip at the bottom of the board. That means NXP sells the board without this chip and without any hint or warning for that not populated IC.

I have no idea why the board doesn't work with mjr's firmware but I don't know how his software handles a board without acc. chip.

The chip is for nudging, the board should also work without this function.

Since I would like to know what is going on at nxp, I wrote to the developer of the KL25z directly. Maybe I will get an answer.

Edited by fhjui, 27 February 2025 - 12:08 PM.

[Suikazz]

Ok, as expected, there is no acc. chip at the bottom of the board. That means NXP sells the board without this chip and without any hint or warning for that not populated IC.

I have no idea why the board doesn't work with mjr's firmware but I don't know how his software handles a board without acc. chip.

The chip is for nudging, the board should also work without this function.

Since I would like to know what is going on at nxp, I wrote to the developer of the KL25z directly. Maybe I will get an answer.

 

if the IC is not present it can't be initialized, therefore the program will probably crash at a certain point.

the accelerometer ic MMA8451 is EOL for over a year now and as long as stock lasted NXP shipped the old design including the acc.

Now it seems out of stock and the redesigned the board for whatever reason and left the acc. footpreint empty.

[fhjui]

 

Ok, as expected, there is no acc. chip at the bottom of the board. That means NXP sells the board without this chip and without any hint or warning for that not populated IC.

I have no idea why the board doesn't work with mjr's firmware but I don't know how his software handles a board without acc. chip.

The chip is for nudging, the board should also work without this function.

Since I would like to know what is going on at nxp, I wrote to the developer of the KL25z directly. Maybe I will get an answer.

 

if the IC is not present it can't be initialized, therefore the program will probably crash at a certain point.

 

This can be answered by mjr

the accelerometer ic MMA8451 is EOL for over a year now and as long as stock lasted NXP shipped the old design including the acc.

Now it seems out of stock and the redesigned the board for whatever reason and left the acc. footpreint empty.

 

This is not new information for me.
I just didn't want to write detailed novels here in the forum that are uninteresting for the average reader.
What annoys me is that nxp sells a product that doesn't meet the specifications without any comment, basically cheating the customer.

In addition, nxp/freescale is itself the manufacturer of the Acc.chip.
That's why I want to know what they were thinking.
 

[mjr]

> if the IC is not present it can't be initialized, therefore the program will probably crash at a certain point

 

That seems unlikely from the reported symptom, in that it doesn’t appear that program is in fact crashing.  It’s also unlikely from the nature of I2C, where it doesn’t really cause any problems to read/write chips that aren’t present on the bus.  The symptom looks like it’s just having trouble with the USB connection, which could be something unrelated to the hardware change (something on the host side, even), or it could be due to some change they made in the USB adapter.  If they changed the shape of the board and removed the accelerometer, who knows what else they changed.  The MKLZ25 SOC contains the USB adapter, so it’s not like they subbed out some other part there, but maybe they’ve tweaked the chip itself in some subtly incompatible way.

[mkotek]

I have also purchased my board from RS Components and they have not changed their description nor put the notice that is now visible on NXP side about Rev. I not to include the accelerometer. Given the popularity of the board, I would think maybe the cause of non-function could be determined (I have the board but intend to return it soon, yet still can run the tests if needed)?

[mjr]

> ( Given the popularity of the board, I would think maybe the cause of non-function could be

> determined (I have the board but intend to return it soon, yet still can run the tests if needed)?

 

Thanks for the offer, but there's probably not much I can do to look into it remotely.  If someone wanted to send me one of the new ones to test, I could probably poke at it a little bit and see if there's a straightforward fix.  But I don't think it's worth a lot of effort, unfortunately.  The "popularity of the board", as you say, doesn't really apply to THIS board - the ORIGINAL board is the one that's popular, in large part because of the integrated (and quite good) accelerometer.  Without the accelerometer, it loses a lot of the appeal, especially now that there are some good alternatives that run on the Raspberry Pi Pico.

Edited by mjr, 16 May 2025 - 04:06 PM.

[mkotek]

What I mean is, there will be a lot of people buying the new KL25Z thinking it will work unless you publish a very clear information in the Build Guide. Myself did not even think my replacement KL25Z will be unusable until I tried to program it today.

[mjr]

> What I mean is, there will be a lot of people buying the new KL25Z thinking

> it will work unless you publish a very clear information in the Build Guide.

 

I did recently add a warning about this - see:

 

http://mjrnet.org/pi...sid=pscOverview

But I know there's a lot of material there to sift through.  I don't think it would be practical to add a warning to every single page there, but if there are any other especially prominent entrypoints where an additional warning would be helpful, let me know and I'll add something there too.