Jump to content



Photo
- - - - -

Pinscape Labs update - Plunger Sensors: A New Hope


  • Please log in to reply
70 replies to this topic

#41 JeanDrEaD

JeanDrEaD

    Hobbyist

  • Members
  • PipPip
  • 12 posts

  • Flag: Brazil

  • Favorite Pinball: Attack from mars, Monster of black lagoon

Posted 17 September 2019 - 03:08 AM

I've thank ou @mjr again! Loved the AEDR solution like all in Pinscape... :otvclap: :otvclap:

 

Used chinese one... :juggle:

 

IMG-20191029-083031.jpg


Edited by JeanDrEaD, 29 October 2019 - 11:32 AM.


#42 JeanDrEaD

JeanDrEaD

    Hobbyist

  • Members
  • PipPip
  • 12 posts

  • Flag: Brazil

  • Favorite Pinball: Attack from mars, Monster of black lagoon

Posted 31 October 2019 - 03:31 PM

I've made a batch from chinese model... And waiting more to arrive... Cheap and best solution. Here in Brazil mainly.

 

IMG-20191029-132419.jpg



#43 wrd1972

wrd1972

    Authoring Padawan

  • Platinum Supporter
  • 2,230 posts
  • Location:Central KY. USA

  • Flag: United States of America

  • Favorite Pinball: Funhouse

Posted 31 October 2019 - 04:34 PM

I'll recommend using the 10k linear potentiometer solution instead of the AEDR8300.

 

I did the AEDR on my first build and it worked great for a good long time.  I think 14-16 months.  But it ended up being at my place of work, so it was getting used heavily most days.  What ended up happening eventually is as the 3d printed parts settled, and things loosened,  the things started rubbing and 3d printer plastic started turning to dust and eventually the mechanics just wore things sideways and it was jamming and hitching and giddyuping and skipping and such.  It's a difficult mounting challenge to get it so this doesn't happen.    

 

With a better more solid construction it might have lasted longer.   But for sure the 10k pot is easier to mount more solidly.  I've done three of those now and they don't look like any failure is imminent.  It's a much more robust physical connection between plunger and pinscape.  Here is the mouser.com part number I used.  688-RSA0N11S9A0K   

MJR and I are working on new "printed" parts to allow the Mouser 100mm 10K sliding pot to be used very integrated into the Pinscape design. The design is being finalized now. But I suspect it will be robust enough to go the distance and does not require a lot of material for the printing process. Sure works flawlessly on my cab, and skill shots have never been more genuinely accurately to the real thing. Stay tuned.


My VP Pincab /MAME Arcade  Specs: Dell T3400 workstation with Core2 Quad core 3.0GHZ (Q9650) CPU - 8GB of RAM - Nvidia  GTX 970

40" PF Sony gaming LED TV, Dual 21" Dell monitors in the backbox - Pinscape dual boards - Full DOF - Full MAME arcade support.


#44 KrakenByte

KrakenByte

    Hobbyist

  • Members
  • PipPip
  • 18 posts

  • Flag: Spain

  • Favorite Pinball: Twilight Zone

Posted 18 December 2019 - 06:57 PM

I've thank ou @mjr again! Loved the AEDR solution like all in Pinscape... :otvclap: :otvclap:

 

Used chinese one... :juggle:

 

IMG-20191029-083031.jpg

 

Amazing solution @JeanDrEaD !!!

I see you've add some new resistors and made the new 3.3V conection. I don't know if the channels are still in their original places... Can you explain the new modifications made to the bought solution, please? I'm fealing that I will follow your steps.

Now I’m working on my own dream-project and also going via AEDR-8300 plunger sensor Pinscape Edition (thanks again @mjr for the great and really impresive developement). But I’m in trouble with it. This is the second time I tried to build this item.

First I failed welding the AEDR-8300 because I used so much soldering paste and it broke.

Now, in the second try, I review the solderings and they look like to be OK. But when I conect the 5.5V wire to the standalone KL25Z then it switchs off.

Does anybody know what could be? Is ther any hiden soldering problem? Have somebody discovered any usefull technique to do it right?

Help me, please, I’m frustrated with this one!

Thank you!


KrakenByte Kraken


#45 mjr

mjr

    Pinball Wizard

  • Members
  • PipPipPipPipPip
  • 3,068 posts

  • Flag: United States of America

  • Favorite Pinball: Theatre of Magic

Posted 18 December 2019 - 07:35 PM

Now, in the second try, I review the solderings and they look like to be OK. But when I conect the 5.5V wire to the standalone KL25Z then it switchs off.

 

If the KL25Z resets when you connect a wire like that, it usually means that you've shorted out one of the power supply pins to ground.  So in this case my first guess would be that you have a short to ground on the 5V connection.  If you have a multimeter, try measuring the resistance between the 5V and GND pins on the sensor board.  If you're using a board based on my EAGLE plans, I think it should look like an open circuit (infinite or at least very high resistance).

 

You might also want to check the resistance between the 3.3V pin and GND to make sure that also looks like a high resistance, and likewise between the 3.3V pin and the 5V pin.

 

Sorry it's been so difficult!  It's definitely a tricky little chip to solder.  It's actually a tricky chip just to handle, given how tiny it is.



#46 KrakenByte

KrakenByte

    Hobbyist

  • Members
  • PipPip
  • 18 posts

  • Flag: Spain

  • Favorite Pinball: Twilight Zone

Posted 22 December 2019 - 08:30 AM

Thank you again for your tips, @mjr and, of course, for all your excellent developement and contribution for the virtual pinball scene: AWESOME WORK!

I was soldering it again but it doesn't work. Perhaps the AERD-8300 was damaged in the process or maybe the moisture finished it, I don't know.

But hope is the last to be lost: I'll try it again for third time with the last of my PCB I bought of your EAGLE plans. I feel that I'm going to be lucky this time.

 

Thanks!


KrakenByte Kraken


#47 JeanDrEaD

JeanDrEaD

    Hobbyist

  • Members
  • PipPip
  • 12 posts

  • Flag: Brazil

  • Favorite Pinball: Attack from mars, Monster of black lagoon

Posted 22 December 2019 - 01:43 PM

 

I've thank ou @mjr again! Loved the AEDR solution like all in Pinscape... :otvclap: :otvclap:

 

Used chinese one... :juggle:

 

IMG-20191029-083031.jpg

 

Amazing solution @JeanDrEaD !!!

I see you've add some new resistors and made the new 3.3V conection. I don't know if the channels are still in their original places... Can you explain the new modifications made to the bought solution, please? I'm fealing that I will follow your steps.

Now I’m working on my own dream-project and also going via AEDR-8300 plunger sensor Pinscape Edition (thanks again @mjr for the great and really impresive developement). But I’m in trouble with it. This is the second time I tried to build this item.

First I failed welding the AEDR-8300 because I used so much soldering paste and it broke.

Now, in the second try, I review the solderings and they look like to be OK. But when I conect the 5.5V wire to the standalone KL25Z then it switchs off.

Does anybody know what could be? Is ther any hiden soldering problem? Have somebody discovered any usefull technique to do it right?

Help me, please, I’m frustrated with this one!

Thank you!

 

I've used EAGLE plans to hack... All lines are the same as original, added resistors like Schematic... 4k7 paralel to 3.3v and Channel A/B...

All work awesome great!

New batch has send for brotherhoods...

IMG-20191107-105734.jpg

IMG-20191207-163827.jpg
 



#48 KrakenByte

KrakenByte

    Hobbyist

  • Members
  • PipPip
  • 18 posts

  • Flag: Spain

  • Favorite Pinball: Twilight Zone

Posted 25 December 2019 - 07:38 PM

 

I've used EAGLE plans to hack... All lines are the same as original, added resistors like Schematic... 4k7 paralel to 3.3v and Channel A/B...

All work awesome great!

New batch has send for brotherhoods...

IMG-20191107-105734.jpg

IMG-20191207-163827.jpg
 

 

 

Oh, my god: So impressive new batch of plunger sensors @JeanDrEaD.

I have allready my 3D printed items in plastic and the mirror plate with the scale black and white, only the circuit resists to me.

Do you sell this fully operational sensors? I'll be interested to buy you one of the electronic circuit with wires ready to put in my own plastics.

Or, maybe do you sell these with all parts? Please tell me, I'm tired of wasting my time soldering and welding... if you could help me!

 

Muito obrigado!


KrakenByte Kraken


#49 JeanDrEaD

JeanDrEaD

    Hobbyist

  • Members
  • PipPip
  • 12 posts

  • Flag: Brazil

  • Favorite Pinball: Attack from mars, Monster of black lagoon

Posted 26 December 2019 - 01:42 PM

 

 

I've used EAGLE plans to hack... All lines are the same as original, added resistors like Schematic... 4k7 paralel to 3.3v and Channel A/B...

All work awesome great!

New batch has send for brotherhoods...

IMG-20191107-105734.jpg

IMG-20191207-163827.jpg
 

 

 

Oh, my god: So impressive new batch of plunger sensors @JeanDrEaD.

I have allready my 3D printed items in plastic and the mirror plate with the scale black and white, only the circuit resists to me.

Do you sell this fully operational sensors? I'll be interested to buy you one of the electronic circuit with wires ready to put in my own plastics.

Or, maybe do you sell these with all parts? Please tell me, I'm tired of wasting my time soldering and welding... if you could help me!

 

Muito obrigado!

 

 

I can't shipping out from Brazil, too much expensive... Our currency is worth nothing, sorry :(



#50 KrakenByte

KrakenByte

    Hobbyist

  • Members
  • PipPip
  • 18 posts

  • Flag: Spain

  • Favorite Pinball: Twilight Zone

Posted 05 January 2020 - 08:15 AM

I did it! :db:

 

After the third try, it was a sucessfull! :otvclap:

This time I didn't want any issues and because I didn't sure how was the moisture of the chip, I was baking the AEDR-8300 for 48 hours at 60 °C (140 °F).

Another thing I changed was that I used a little bit less of soldering paste than the last two tries and removed a bit excess of melting soldering metal with a wire.

 

Now it works! It's alive! It's amazing! :dblthumb:

 

Everybody: Thankyou very much! :love39:


I can't shipping out from Brazil, too much expensive... Our currency is worth nothing, sorry :(

 

Don't worry about that, but I keep in mind your excellent hack idea for the future. It could be an excellent spare part. ;)

 

Thanks! Obrigado!


Edited by Popote, 05 January 2020 - 08:15 AM.

KrakenByte Kraken


#51 mjr

mjr

    Pinball Wizard

  • Members
  • PipPipPipPipPip
  • 3,068 posts

  • Flag: United States of America

  • Favorite Pinball: Theatre of Magic

Posted 06 January 2020 - 06:27 PM

Popote - that's great!  The baking was a really good thing to try.  I didn't take all of the humidity warnings about this thing very seriously at first, but apparently it's a big deal for chips like this, because of the type of plastic housing they use for the optical parts.  The heat from soldering can make any trapped water vapor inside the plastic expand and cause tiny cracks that can disable the chip.  I'm glad you were able to get a working sensor!



#52 Gizmooo

Gizmooo

    Hobbyist

  • Members
  • PipPip
  • 10 posts

  • Flag: Germany

  • Favorite Pinball: Funhouse

Posted 05 February 2020 - 04:43 PM

I would like mjr and everyone else to have a fresh look at the CJMCU-83 boards. Why?  For 3$ shipped you only need to add two 2,7k pullup resistors and you can use it with Pinscape. There are no modifications needed to the board itself. Let me elaborate:

 

mjr previously speculated in this thread that the CJMCU-83 board is difficult to use, because it may need level shifters or modifications to the board etc. I was curious and ordered one (aliexpress ) and made some interesting discoveries.

 

The board has 1 resistor (120 Ohm) and 2 capacitors. The data sheet for the 75 LPI sensor AEDR-8300-1K2 recommends to use 220 Ohm at 5V to limit LED current. So why does the little board use such a low (120 Ohm) resistor instead? I found the datasheet of the AEDR-8300-1Wx encoder. Which is a higher resolution (212 LPI) variant of the sensor mjr used in his design.The datasheet is mostly identical, however it specifically mentions use at 3.3V now! For example it mentions to use a 110Ω resistor at 3.3.V or 220Ω at 5V to limit the current of the LED on page 5. So the low resistor value indicates that the CJMCU-83 board is actually designed for 3.3V (even tough all the sellers I found mentioned 5V).

From my first tests it works just fine at 3.3V. I measured that the builtin LED draws 12 mA at 3.3V.

It could read the printed code strip from the Pinscape build guide reliable when quickly swiping the codestrip over the sensor infront of a small mirror. 

 

Advantage of using CJMCU-83:

- no complicated SMD soldering, no risk of damaging the sensor

- cheap (3$)

- only 4 wires needed to the sensor (no additional 5V line)

- no extra sensor board

Disadvantage:

- 3D print needs to be adapted (cable needs to attach to backside of the board, along with the 2 pull-up resistors) 

 

luY4rF2.jpg?1

Above image is just to illustrate the simplicity of the connections (the codestrip needs more room to be moved over the sesnor from from left to right and back ). I

 

Summary of parts and connections between Pinscape (Standalone) and CJMCU-83 sensor baord:

Pinscape: PTD0 --> Sensor Board: CA

Pinscape: PTD5 --> Sensor Board: CB

Pinscape: GND --> Sensor Board: GND

Pinscape: 3.3V --> Sensor Board: VCC

2,7K pullup resistor between VCC and CA

2,7K pullup resistor between VCC and CB

One can directly solder the 2 pull-up resistors to the back of the CJMCU-83 board or somewhere at the Pinscape side of the cable. The front of the board needs to stay free so the code strip can get close to it.

 

Disclaimer: I have not mounted my sensor to the plunger so far. These are my prototype results.

Thanks to mjr for the wonderful playground he named Pinscape.



#53 mjr

mjr

    Pinball Wizard

  • Members
  • PipPipPipPipPip
  • 3,068 posts

  • Flag: United States of America

  • Favorite Pinball: Theatre of Magic

Posted 07 February 2020 - 08:51 PM

Gizmooo - that's a great discovery!  It definitely simplifies things enormously to be able to use a pre-built board like that.  It sounds like it would just need some changes to the geometry of my 3D-printer plans for the part that holds the sensor board to make this totally plug-and-play.  I'll have to order one of the boards and see if I can tweak the designs.  (Or, if you or anyone else gets to it first, let me know what you come up with.)



#54 dbknightx

dbknightx

    Hobbyist

  • Members
  • PipPip
  • 26 posts

  • Flag: United States of America

  • Favorite Pinball: Lord of the Rings

Posted 14 February 2020 - 10:02 PM

First of all, thanks so much to mjr for his fantastic guides, and in particular for the AEDR-8300 option.  I successfully built one last year for my 3-screen cabinet, and opted to use the same approach for a VR PinSim that I’ve just completed.

 

This time it is not quite working properly.  The behavior is odd and I was wondering if somebody might help me troubleshoot.  The symptom is that it works as expected when I pull the plunger, but when I release the plunger it acts as if I am pulling it further.  In other words, rather than the values decreasing as it retracts, they actually keep on increasing.

 

Any idea where I should start looking?  Is this what would happen if one of the two channels were not connected properly?

 

Thanks very much in advance.



#55 mjr

mjr

    Pinball Wizard

  • Members
  • PipPipPipPipPip
  • 3,068 posts

  • Flag: United States of America

  • Favorite Pinball: Theatre of Magic

Posted 14 February 2020 - 10:35 PM

dbknightx - I think if you had only one channel connected, the reading would just go up and down by one "tick" each time the optical sensor crossed a bar in either direction, so it wouldn't really move at all, just sort of vibrate slightly at one position.  I'm not sure what would make it get stuck in a single direction like that.  If you move the physical plunger at a slow, uniform rate back and forth, does the on-screen reading consistently move at the same speed, or is it different going at different speed when the physical plunger is going back (pull) vs forward (push)?

 

Given that it was working before, I think you're right to suspect the wiring.  I'd definitely check that carefully.  Maybe also check the optics to make sure there isn't any dust or grease stuck to the lens.



#56 dbknightx

dbknightx

    Hobbyist

  • Members
  • PipPip
  • 26 posts

  • Flag: United States of America

  • Favorite Pinball: Lord of the Rings

Posted 15 February 2020 - 05:52 PM

Funny you say that - I checked it again today and now it is doing exactly what you described - just moving up and down by one.  It was definitely behaving the way I described above yesterday.  Seems like I probably have some loose wiring or a bad solder joint.

 

By the way this is a completely new build - the one I used for my previous cab was sold with the cab.  My biggest fear was that there was something wrong with the AEDR-8300 or my heat gun soldering.  I guess that is still a possibility but I’ll try the simple fixes first.

 

Thanks for your super prompt response.


Edited by dbknightx, 15 February 2020 - 05:53 PM.


#57 dbknightx

dbknightx

    Hobbyist

  • Members
  • PipPip
  • 26 posts

  • Flag: United States of America

  • Favorite Pinball: Lord of the Rings

Posted 16 February 2020 - 05:56 PM

Well, another quick update.  I had some issues with the pin housing I was using to connect Ch A and Ch B to the Pinscape board, which was causing intermittent connections.  I removed the housing and just connected the pins directly to the header pins to make a secure connection.  However now I'm just back to my original problem, which is that no matter what direction I move the plunger, the numbers count up in the plunger setup screen.

 

Some additional details:

  1. The numbers seem to increase at the same rate no matter what direction I am moving.  They are tracking really well, just in the wrong direction half the time.
  2. If I disconnect one of Ch A or Ch B, I get the behavior you expected, which is the numbers just repeat going up/down by 1.  This is the same behavior no matter which of the two channels I disconnect.
  3. I've checked continuity between Ch A and Ch B - being concerned that there may be a short.  However I'm not able to detect anything with my meter.
  4. I've tried different sets of pins on the Pinscape to connect Ch A and Ch B, with no change in behavior.

At this point I'm not sure what else to try apart from unsoldering and resoldering all of the connectors and resistors on the board.  Do you think something wrong with the resistors (either soldering or otherwise) could cause such behavior?

 

p.s I've ordered a couple of CJMCU-83 boards from China to have on hand in case I can't get this working.  It sounds like that may end up being a preferred and less error-prone solution in the future.  In the meantime I'm just glad I also have a launch button on my pinsim.



#58 mjr

mjr

    Pinball Wizard

  • Members
  • PipPipPipPipPip
  • 3,068 posts

  • Flag: United States of America

  • Favorite Pinball: Theatre of Magic

Posted 17 February 2020 - 06:46 PM

dbknightx - it sure sounds like the channel A/B connections are good, given that saw the +/- 1 tick behavior when you deliberately disconnected one channel.  I'm pretty baffled about what could be causing the one-way behavior - it seems like it has to be a software issue.  Are you on the latest version of the firmware?  If so, it might be worth trying the prior release (from 3-5-2019, linked from the firmware download page, http://mjrnet.org/pi.../swversions.php - scroll down to the "Archived Release Builds" section), just in case I broke something in the latest build.  I made a small change to the AEDR-8300-related code in the last build, so it's possible that I broke something in the process.



#59 dbknightx

dbknightx

    Hobbyist

  • Members
  • PipPip
  • 26 posts

  • Flag: United States of America

  • Favorite Pinball: Lord of the Rings

Posted 18 February 2020 - 01:54 AM

mjr - I tried your suggestion and rolled back my firmware to the 3-5-2019 version, using the corresponding version of the PinscapeConfigTool.  Unfortunately it behaves exactly the same.  At this point I'm not sure what else to try other than redoing all my soldering.

 

There wouldn't happen to be any kind of debug mode in the config tool to allow me to output the raw values coming from Ch A and B?  I understand there is a 4 step A/B pattern that would be used to indicate if it is moving forwards or back - I'd love to be able to dump out the raw values as they change.

 

Thanks again for your help.



#60 mjr

mjr

    Pinball Wizard

  • Members
  • PipPipPipPipPip
  • 3,068 posts

  • Flag: United States of America

  • Favorite Pinball: Theatre of Magic

Posted 18 February 2020 - 03:05 AM

mjr - I tried your suggestion and rolled back my firmware to the 3-5-2019 version, using the corresponding version of the PinscapeConfigTool.  Unfortunately it behaves exactly the same.

 

Thanks for trying - that at least would seem to rule out the notion that it was a bug in the new version.

 

I'm starting to think the only remaining explanation might be a defective sensor chip.  Or maybe the wrong chip?  Is there any chance you might have gotten one of the other sensors in the same line?  This series has a bunch of sensors that work at different line densities - I could believe that this might be what you'd see if you used the 150-line-per-inch sensor with a 75 LPI scale, for example.  I believe the different AEDR-8300 variations are identified by color-coding of the plastic case.  The 75 LPI version has a clear plastic case, and the others are tinted yellow, blue, etc.

 

 

There wouldn't happen to be any kind of debug mode in the config tool to allow me to output the raw values coming from Ch A and B?  I understand there is a 4 step A/B pattern that would be used to indicate if it is moving forwards or back - I'd love to be able to dump out the raw values as they change.

 

Unfortunately there's not any way to do that right now.  But it's a great idea for a debugging tool for this situation.

 

Let me see... I could probably jury-rig a quick test version.  If you can hold off on the re-soldering project for a couple of days, let me see what I can come up with.

 

You're exactly right about the A/B pattern.  All quadrature sensors work on the same principle of two sensor signals that are 90 degrees out of phase with each other.  Digital sensors like this represent the channel positions as binary on/off signals, so you get a two-bit pattern, which inherently can take on four values (00, 01, 10, 11).  Each channel flips its bit value when its half of the sensor moves past the edge of one of the optical bars.  Since the two sensors are slightly offset side-to-side, one of the channels always leads the other by half a bar, and which one leads depends on which direction you're moving.  So if you're moving to the "left", say, you'll see A lead B, so you'll get bit sequences where A flips first, then B flips, then A flips, etc - 00 10 11 01 00 repeat.  If you're moving the other way, "B" will flip first, so you'll get 00 01 11 10 00 repeat.

 

Do you think you could parse the results visually if I had, say, the red LED channel represent "A", and the blue LED channel represent "B"?  So a flash pattern like OFF-RED-PURPLE-BLUE-OFF would represent the 00-10-11-01-00 direction.  You'd have to move the scale very slowly to be able to see the individual stops, but it sounds like what you're seeing is consistent at all speeds, right?


Edited by mjr, 18 February 2020 - 03:06 AM.