(Topic ID: 218125)

What are your thoughts on the Spike II system?


By LoserKid3

1 year ago



Topic Stats

  • 435 posts
  • 82 Pinsiders participating
  • Latest reply 1 year ago by mbwalker
  • Topic is favorited by 15 Pinsiders

You

Linked Games

Topic Gallery

There have been 18 images uploaded to this topic. (View topic image gallery).

2018-08-03 08-51-03.JPG
Pinball Controller Task Split (resized).PNG
SBC (resized).PNG
lpkf2 (resized).PNG
lpfk (resized).PNG
515-0174-00 (resized).jpg
grommet (resized).png
20180702_192610 (resized).jpg
PP4_06_CPU_800x600 (resized).jpg
WOZ_CPU (resized).PNG
scared (resized).jpg
Microchip Details (resized).jpg
eScope (resized).PNG
resistors-on-match-head (resized).jpg
images1 (resized).jpg
images (resized).jpg

Topic index (key posts)

3 key posts have been marked in this topic

Post #26 A report from an operator and a life cycle expectation. Posted by xTheBlackKnightx (1 year ago)

Post #215 Things to consider from a senior technician’s perspective. Posted by xTheBlackKnightx (1 year ago)

Post #262 Words of expierence and a warning. Posted by xTheBlackKnightx (1 year ago)


Topic indices are generated from key posts and maintained by Pinside Editors. For more information, or to become an editor yourself read this post!

You're currently viewing posts by Pinsider cynric.
Click here to go back to viewing the entire thread.

13
#32 1 year ago

Taking a look at Spike (1) from a technical stand point: The system uses some really good and modern concepts; let me go into some details.
It is using an isolated serial bus for communication. This basically confines faults to the faulty nodes. Even high-voltage-to-switch failures will only fry the switch node, but leave the CPU ok.
A single high power medium voltage for all mechs and all nodes allows a simpler/more efficient power supply. Having a proper high power DC system eliminates common fliptronic problems like bad rectifiers, fluctuating supply voltages and most of the stuff that often make WPCs sluggish under certain circumstances and are a bitch to figure out and fix.
5V for all lighting simplifies node/driver design immensely.
Distributing the computation effort throughout the machine (and having at least a little processing power on the nodes) allows for all the neat things we've come to like, like RGB leds, instant kickbacks, more controlled mechs...
Endless processing power. Even Spike 1 has a 400MHz CPU with 64MB of ram. For something a computationally simple as a pinball game this is plain joyful overkill

But there are downsides too.
SMD can be tough to repair if you're not used to it. I personally like it more than through hole (especially with old boards that tend to loose their traces), but most people will probably have a hard time with SMD. But they are definitely repairable, saying that you only have the option to buy new boards is bogus.
Parts can be harder to obtain. From what I've seen Stern uses stock common parts, but you don't likely get them at your radio shack. Iirc the bus is powered by some automotive transceiver chips, so even that should not be a problem. However we do not know how supply will be in 15 years when there is a new generation.
And the imho most important argument is that Stern uses programmable controllers with embedded flash memories on all the node boards and the cpu. You can not just replace those with a new blank part. While the controllers are not hard to acquire, the software is. Unless Stern releases binaries for the node boards or they can be extracted from either existing boards (which can usually be prevented) or the spike update files, we are pretty much screwed. The node board NXP MCUs do come with a basic bootloader, however I would not count on spike just accepting and reprogramming a blank one. If it does, then this is point is invalid .
The CPU itself does have an embedded ROM that does boot the system from the SD card, so that should be fine. However the CPU is a BGA chip, swapping that is something that required special equipment to do.
On the Spike 1 CPU board there is a second AVR controller (with embedded flash); I guess that is taking care of the RTC and maybe the DMD. This would be prone to the flash problem too. And since flash usually just has a guaranteed data retention time of 10-20 years this might really become a problem somewhere down the road..

#35 1 year ago
Quoted from 85vett:

I think my biggest frustration/concern with Spike and Spike 2 is that the advent of the node boards was supposed to make things easier (not talking cost) but they are not using the same node boards across the games. This causes me concern as the lower production games may have a hard time getting replacements real soon.
If they would just use the same node boards then it would make things much easier, you could have a spare or two on hand for your collection and it would reduce overall cost due to volumes.

True.. Does anyone know if older versions can just be replaced by newer ones of the same type? SW still uses the 4 driver / 16 switch / 16 LED board labeled "4 Coil NODE" that GoT uses too, just with a different part number. So to me it looks like there are only a few main playfield node types (drivers/switched/lights): 4/16/16, 8/8/0, 9/32/8 and a few primitive serial extensions. Newer games seem to use the 9/32/8 boards more often.
At least for the serial boards like through and switch / led drivers swapping should not be a problem, for the others we'll have to see ^^

Promoted items from the Pinside Marketplace
$ 279.95
Lighting - Led
Pin Stadium Pinball LEDs
6,666 (OBO)
Machine - For Sale
Toronto, ON

You're currently viewing posts by Pinsider cynric.
Click here to go back to viewing the entire thread.

Hey there! Got a moment?

Great to see you're enjoying Pinside! Did you know Pinside is able to run thanks to donations from our visitors? Please donate to Pinside, support the site and get anext to your username to show for it! Donate to Pinside