Expired DMD Led Color DMDMK66 128x32, 192x64, 128x16 manufactured by Pinballsp

[pinballsp]

.

Stern/Sega WhiteStar identify, where connect to get 12V DC power supply for last "DMDST32 12VDC Ready Edition". This information, will be also useful to connect DMDMK64 to 12V DC.

 

[pinballsp]

.


Stern SAM identify, where connect to get 12V DC power supply for last "DMDST32 12VDC Ready Edition". This information, will be also useful to connect DMDMK64 to 12V DC.

 

[pinballsp]

.

CAPCOM identification, where connect to get 18V DC supply for last "DMDST32 12VDC Ready Edition". This information, will be also useful to connect DMDMK64 to 12V DC.
DMDST32 and also DMDMK64 accept up to 36VDC. With CAPCOM, connect to 18VDC instead of to 12VDC, no problem about it, work perfect.

 

[pinballsp]

.



All pictures, by pinball manufacturer, to connect DMDST32 to 12V DC, available in a PDF, to download from this MEGA link.
https://mega.nz/#!Ac8WkR4A!Mq8hY8eDN25KiyXRSGFsyEc8UY12xwDrx-ghhf0nAcI



.

 

[pinballsp]

.


Today I received the new PCB, for last version I will do about DMDST32, this is the "12V DC Ready Edition", to connect direclty to 12V DC, so do not need external power supply.

Also I have add some improvements to design. Add 2 x buffers to may connect a Topper with P5 led panels in mirror mode, add a push button fo enter in update mode, now the led panel power cables are very short and directly soldered under PCB (for better look), flat cable with custom correct length (no more original flat cable Folded and flanged), Switched DC-DC centered and with horizontal inductor.

And this new PCB, install a new LED panel design, according with manufacturer " Refresh rate more good and more better remove shadow effect ", so may be solve the BLEEDING problem. As soon as I assemble, today or tomorrow, I will test to confirm.


Tomorrow I hope start to ship the first batch or "DMDST32 12V DC Ready Edition". Thank you to everybody for patience to wait, unfortunately the led panel manufacturer changed design without warn about it, so I had to change PCB design as fast as possible, send to factory and wait. At any rate I think this is the best DMST32 design, I will post pictures about product assembled as soon as possible.

 

[ducks]

Great as ever that you post details about everything you do so openly.

One question ... I assume you plan to include support for SMARTDMD / PIN2DMD colouring files that have already been done, on the MK64/66 platform?

Also, will your own software colouring solution use the same file format etc so it's also backportable to them, where the colouring does not exceed their feature set?

Edit: Oh, and what were the new and old refresh rates on the LED panels?

 

[pinballsp]

Great as ever that you post details about everything you do so openly.
One question ... I assume you plan to include support for SMARTDMD / PIN2DMD colouring files that have already been done, on the MK64/66 platform?

I must check yet if may I import pin2dmd format files. At any rate for pin2dmd there are currently only a few colored files and all (except one) are for SAM. The big problem to colouring for pin2dmd is that their editor is very simple, not optimal colouring by hand pixel by pixel, people need months to do a game. My idea about software to apply advanced color is very different.

Also, will your own software colouring solution use the same file format etc so it's also backportable to them, where the colouring does not exceed their feature set?
Edit: Oh, and what were the new and old refresh rates on the LED panels?

Mine will not use same format, and files will be encrypted, only will work with DMDMK64. If I may understand current format of pin2dmd color files, I will prepare a software to import it (DMDST32 to DMDMK64), but DMDMK64 to DMDST32 will not be possible.

About led panels, problem is the bleeding, new panels seem improve a features that may be solve bleeding, but I must test yet. Between today and tomorrow I will assemble first 12V DC Edition DMDST32 with new led panels, then will confirm if solve or not bleeding.

 

[Pick Holder]

The main issue I find with your design Luis (and all the others) is it would improve the panel greatly if you had two 5v outputs on each PCB to feed the P2.5 RGB LED panels.

I am sure the actual 'data' part of the design is very low current draw, as the panels take most of the current.

How do you pick up the 5v to panels on this design?

 

[ducks]

I must check yet if may I import pin2dmd format files. At any rate for pin2dmd there are currently only a few colored files and all (except one) are for SAM. The big problem to colouring for pin2dmd is that their editor is very simple, not optimal colouring by hand pixel by pixel, people need months to do a game. My idea about software to apply advanced color is very different.

I'm pretty sure at least half the files for PIN2DMD were actually done for SMARTDMD originally (like AC/DC and Metallica that PinballMikeD did). I don't know if the 2 platforms share the same file format, or if the colouring work for SMARTDMD on Pinball Browser was just exported to the PIN2DMD format. The only full recolouring (still not pixel by pixel) I have seen specifically for PIN2DMD is TSPP which the Russian guy recently did.

 

[Pick Holder]

I'm pretty sure at least half the files for PIN2DMD were actually done for SMARTDMD originally (like AC/DC and Metallica that PinballMikeD did). I don't know if the 2 platforms share the same file format, or if the colouring work for SMARTDMD on Pinball Browser was just exported to the PIN2DMD format. The only full recolouring (still not pixel by pixel) I have seen specifically for PIN2DMD is TSPP which the Russian guy recently did.

I have read that its not easy to get these files into a Pin2DMD for use on a real pinball.

My family guy is coloured nicely with one of Luis Pin2DMD panels out of the box - but the Visual pinball version looks even better but I can not work out how to get the new ROM file into the game as I have tried with pinball browser but to no avail. I need another 'file' that needs to come from the author.

Anyhow - you will be happy with one of these @ducks - I have used 5 up to now.

 

[stoptap]

Did the WPT colourisation ever get finished and released ?

 

[ducks]

I'm sure I will be. Just inclined to wait for the DMDMK66 if as Luis says it will have more advanced colouring solutions.

My only concern is that some of the old files are usable. I think the AC/DC and Metallica ones could definitely be done better on a more advanced platform than SMARTDMD as they were originally done for, but it would be a good stop gap until someone does something better with the MK hardware and software set.

 

[pinballsp]

The main issue I find with your design Luis (and all the others) is it would improve the panel greatly if you had two 5v outputs on each PCB to feed the P2.5 RGB LED panels.

I am sure the actual 'data' part of the design is very low current draw, as the panels take most of the current.

How do you pick up the 5v to panels on this design?

No, its not the problem. Time ago, in one of the DMDST32 versions, I added 2 x 74LS45 buffers between the microcontroller STM32 and the led panels, and this did not solved anything.
And about 5v power supply, in new 12V DC version, this work with a 5 Amper DC-DC switched, and this connect directly to both led panels.

There is no problem about 5v power supply or data signals from MCU to led panels data. All was tested time ago. Bleeding is a problem of firmware and led panel design.

 

[Pick Holder]

Good news about the power connectors. I hope in your new version (MK66) it will be easy to put colourised files into a normal pinball. Using your ST board it works perfect in visual pinball but not a stern SAM game. This is not your fault, its down to the way pin2dmd does things, and pinball browser.

I look forward to seeing the new STM32 boards completed. How far away is MK66 to being released. I think you should concentrate on that! I know plenty of people that would buy these displays as they find pin2dmd too much of a headache. Most people want a plug in and play option.

 

[pinballsp]

Using your ST board it works perfect in visual pinball but not a stern SAM game. This is not your fault, its down to the way pin2dmd does things, and pinball browser.

What is exactly the problem with Stern SAM ??.

How far away is MK66 to being released. I think you should concentrate on that! I know plenty of people that would buy these displays as they find pin2dmd too much of a headache. Most people want a plug in and play option.

Yes, I know. My only doubt now is whether to follow the design with Kinetis microcontrollers or use Raspberry pi 3. I've thought about preparing a prototype with Raspberry Pi3 and test it, its a very powerful option, better than any microcontroller. Now is the moment to choose.

And also I have thought about do development step by step and post each detail like I'm doing now with the Creature Holo Mod, so may I see how people is interested, suggestions and comments, I work faster than try to do all and then publish a product ready to work.

 

[BigIan]

Interested in CFTBL Mod, keep us informed

 

[pinballsp]

Interested in CFTBL Mod, keep us informed

OK, I have open a post in Pinside with all information about current development status.
https://pinside.com/pinball/forum/t...-creature-from-pinballsp-is-under-development

 

[ducks]

If the Pi3 offers potentially better features and colouring solutions, then do it.

Added benefit that many more people are going to be instantly familiar with the hardware and able to jump in and help with software or additional mods or features.

What is exactly the problem with Stern SAM ??.




Yes, I know. My only doubt now is whether to follow the design with Kinetis microcontrollers or use Raspberry pi 3. I've thought about preparing a prototype with Raspberry Pi3 and test it, its a very powerful option, better than any microcontroller. Now is the moment to choose.

And also I have thought about do development step by step and post each detail like I'm doing now with the Creature Holo Mod, so may I see how people is interested, suggestions and comments, I work faster than try to do all and then publish a product ready to work.

 

[Pick Holder]

Stern colourisation uses more than one palette. It will switch on Vp but you need an extra file from the original author to do a real pinball.

I put colour rom into my family guy as well as the palette and it looks wrong.

Again. Not your problem Luis.


On the Pi 3 thing - why use it when you can use something simpler.

Your running Linux usually and I think it's a bit unstable.

I think anything made with a pi is home made and not pro.

Carry on the way you do things.

I'm sure soon he will have it wifi ed up to adjust stuff from a pc or phone

 

[ducks]

One other potential advantage of the Pi solution is that if it develops a fault, it's on a separate board and you can just whip it out and replace it.

An all in one board like the ST32 or MK66 solution? Not so much ....

 

[Pick Holder]

I disagree on this one. A pi3 will cost about £35. Yes it is cheaper but if a pi was used it would have to have a daughter board added on, as well as power supplies.
When one part fails then you have to find what has gone! Easy for a tech but the new boards are plug and play.

I have had two pi sbc fail on me in my projects. They are cheap and nastily made.

These boards are great quality as are components.

As a qualified electronics engineer I would recommend these over a pi driven solution.

Looking forward to a hopefully better firmware that does not require licensing and repeative updates.

 

[Paul]

I'm on the pi bandwagon. Much more powerful, wifi and Bluetooth naitive, lots of 3rd party support with the community behind it... have seen some astonishing work done on pi's
I run several for small automation and have never had any board issues.

(From another fully qualified electronic engineer!)

 

[pinballsp]

.

I'm really confused about if do DMD with Rpi3 or Kinetis microcontroller. Each one has advantages and disadvantages.

Rpi3 is a SOC with a Multitask OS (Linux), so its not dedicated to only one task and this may be a problem with software depending from timings, also do not has many direct I/O ports, but its so powerful that may be may compensate these disadvantages. Rpi3 also has a GPU, so may play video in High Resolution (1080) or control any TFT to display animations.

A custom device with a dedicated microcontroller, do not has dependencies from third party hardware, but do not has so much memory to store firmware and RAM. And this is a hardware that run ONLY the firmware, no more task at same time. So though not so powerful like Rpi3, run only one task is a dedicated device and finally may be even powerfull than Rpi3 to run same firmware. Also a microcontroller has many I/O ports, MK66 is a 144 pin TQFP, so more than 100 I/O ports available.

Both may use SD card to store config files, color palettes, and coloured animations, both access to SD by parallel SD port 4 bit, so its a very fast access. My MK66 also may install a flash memory 512Mb to store color animations for very fast access, but may be its enough SD 4 bit parallel access, it's cheaper and easy to do hardware.

 

[ducks]

.

I'm really confused about if do DMD with Rpi3 or Kinetis microcontroller. Each one has advantages and disadvantages.

Rpi3 is a SOC with a Multitask OS (Linux), so its not dedicated to only one task and this may be a problem with software depending from timings, also do not has many direct I/O ports, but its so powerful that may be may compensate these disadvantages. Rpi3 also has a GPU, so may play video in High Resolution (1080) or control any TFT to display animations.

A custom device with a dedicated microcontroller, do not has dependencies from third party hardware, but do not has so much memory to store firmware and RAM. And this is a hardware that run ONLY the firmware, no more task at same time. So though not so powerful like Rpi3, run only one task is a dedicated device and finally may be even powerfull than Rpi3 to run same firmware. Also a microcontroller has many I/O ports, MK66 is a 144 pin TQFP, so more than 100 I/O ports available.

Both may use SD card to store config files, color palettes, and coloured animations, both access to SD by parallel SD port 4 bit, so its a very fast access. My MK66 also may install a flash memory 512Mb to store color animations for very fast access, but may be its enough SD 4 bit parallel access, it's cheaper and easy to do hardware.

What's the SD card bus width on the pro cameras that supporter faster or fastest SD cards? If the Pi / MK66 can support faster SD cards, I'd have thought that would be more than enough. Though I guess initial latency is more important than bandwidth, for this data?

 

[Pick Holder]

The SD card holds the serial key and palette on pin2dmd. A few k only.

A pin2dmd takes 2 seconds to boot and start working. A pi takes longer. Especially if it has to load an os.

I do not understand why people are looking for another solution when this one is fine.

Is there a need to spend more time and effort?

The led panels cost the same whatever you use.

If you use a pi you'll still need a psu of 5amp, the new board has psu inbuilt. Plus you will need a daughter board with the pi to connect the dmd data and send data to the panels.

The cost will probably be more than this solution. You will spend weeks working on it to get same result.

You will end up with something like the dmdextender from Dr Pinball that drives vga monitor(s). That retails for £60 plus cost of pi and cost of screen panel.

 

[ducks]

The SD card holds the serial key and palette on pin2dmd. A few k only.

A pin2dmd takes 2 seconds to boot and start working. A pi takes longer. Especially if it has to load an os.

I do not understand why people are looking for another solution when this one is fine.

Is there a need to spend more time and effort?

The led panels cost the same whatever you use.

If you use a pi you'll still need a psu of 5amp, the new board has psu inbuilt. Plus you will need a daughter board with the pi to connect the dmd data and send data to the panels.

The cost will probably be more than this solution. You will spend weeks working on it to get same result.

You will end up with something like the dmdextender from Dr Pinball that drives vga monitor(s). That retails for £60 plus cost of pi and cost of screen panel.

I think part of the attraction is the possibility of incorporating it into an ecosystem of mods which Luis may or may not intend to build - like his CftBL holo mod. A Pi surely makes this rather more achievable.

Also, let's say in 2 years time much denser LED panels become affordable, and we can have 256*64 instead of 128*32 for the same or similar price. Presumably, aside from upscaling the data to double the horizontal and vertical resolution, a Pi based solution would require little if any modification. MK66 or other ... maybe a redesign?

 

[Pick Holder]

I think part of the attraction is the possibility of incorporating it into an ecosystem of mods which Luis may or may not intend to build - like his CftBL holo mod. A Pi surely makes this rather more achievable.

Also, let's say in 2 years time much denser LED panels become affordable, and we can have 256*64 instead of 128*32 for the same or similar price. Presumably, aside from upscaling the data to double the horizontal and vertical resolution, a Pi based solution would require little if any modification. MK66 or other ... maybe a redesign?

The resolution of the signal coming in will not be improved.

Are you talking about what colordmd do with an LCD monitor and the MM remake? That work needs a lot of work programming.

These displays look great as is. I would post a video but my phone doesn't do it justice.

 

[ducks]

The resolution of the signal coming in will not be improved.

Are you talking about what colordmd do with an LCD monitor and the MM remake? That work needs a lot of work programming.

These displays look great as is. I would post a video but my phone doesn't do it justice.

1) I know.

2) Not really. What they do involves adaptive scaling, smoothing / anti-aliasing etc. It's rather different than raw upscaling.

Correct me if I'm wrong, but doubling the X and doubling the Y axis resolution (4 x the pixels) should be very simple to upscale?

A denser grid of LEDs at the moment are likely unfeasible due to a) availability b) price c) probably lower reliability. Those issues will be overcome with time, and smaller LEDs should mean less bleed / glow, and better responsiveness (less ghosting), as well as the obviously higher resolution.

 

[Pick Holder]

I've not noticed ghosting on these displays. Each dot currently is probably a little smaller but the light that is given off it you don't really notice. It is a different light to a normal plasma dmd.

Most people who are into pinball love the rgb led panels.

Have you purchased one of these @ducks as I think I can get one of these going with your bop2.

 

[ducks]

I've not noticed ghosting on these displays. Each dot currently is probably a little smaller but the light that is given off it you don't really notice. It is a different light to a normal plasma dmd.

Most people who are into pinball love the rgb led panels.

Have you purchased one of these @ducks as I think I can get one of these going with your bop2.

There's a little bit. Not too bad.

I definitely prefer them to the LCD solution that ColorDMD use and similar ones SMARTDMD used in the past. I have no quarrel with it at all. But this isn't the pinnacle of the technology ... I'm saying there'll be much better available in future, and probably sooner than we think (LED wise). A Pi based solution may be more future proof.

No. Still waiting until Luis comes out with either his Pi or MK66 based solution.

Also still haven't got round to installing the BoP 2.0 kit ... or doing the playfield swap

Preparation for and then the actual move to the Netherlands have proved a little time consuming .....