(Topic ID: 289750)

Williams System 7 Power Supply Question

By ATAX

7 months ago


Topic Heartbeat

Topic Stats

  • 20 posts
  • 6 Pinsiders participating
  • Latest reply 86 days ago by Leaftail
  • No one calls this topic a favorite

You

Linked Games

Topic Gallery

View topic image gallery

9E521392-1730-4AE0-9789-2B4C9614DDD8 (resized).jpeg
Bad_Regulator (resized).jpg
Untitled (resized).png
Proposed Plugins on New Board For The 4 Rogue Connectors (resized).jpg
New Correct Power Supply Board (resized).jpg
Current Incorrect Power Supply (resized).jpg
4 Rogue Connectors (resized).jpg

#1 7 months ago

Hello,

I'm working on my first restoration and looking for some help. I have a Williams Cosmic Gunfight (1982) System 7 machine.

The question I have is regarding the connector cables that attach to the power supply board.

When I bought this machine it apparently had the wrong power supply installed in it (looks like a system 6 power supply). There are currently 4 connectors that are not attached to any board in the system. I purchased the correct power supply board from Marco and I think I can match up the 4 connectors that aren't currently connected, to headers on the new board.

Two of the connectors look pretty obvious on where they should plug in but I'm stumped on the other two. It almost looks like 2 of them should be 1 connector but its split into 2 pieces. If I connect the two connectors to the header on the new board they fit on the single header. One of the two connectors is keyed and still seems to fit correctly onto the header and the 2nd connector fits right next to the first but they are sharing the same header which seems odd.

So I'm wondering if this is correct and if I have everything plugged into the right spots before I flip the power. The last thing I want to do is fry my new board or worse start a fire or something.

Below are some pictures of the old incorrect board which is still in the system, the 4 connectors that are not currently attached to anything and also pictures of the new correct power supply board where I am assuming the 4 rogue connections should be plugged in.

Also, I plan on installing new connectors before I do the final install. If the two connectors in question indeed both connect to the single header, should I purchase 2 replacement connectors or consolidate both of the old connectors into a single new connector?

Could you guys please let me know what you think? Any help would be much appreciated.

Thanks!

4 Rogue Connectors (resized).jpgCurrent Incorrect Power Supply (resized).jpgNew Correct Power Supply Board (resized).jpgProposed Plugins on New Board For The 4 Rogue Connectors (resized).jpg
#2 7 months ago

I think you have the connector J8 right. First four pins are one side of GI, white with striped colors. The last four is the other side of GI and solid colors. It should not be that critical the order as long as they are grouped up properly. The manual should have a wiring diagram page with colors.

Isn't that board missing the F7 GI fuse? I thought they had a 20a GI fuse on the power board in system 7. They did jumper it out in later system 11 when they fused each individual GI string after it leaves the power board. I would make sure you have a fuse(s) somewhere in the GI circuit. Since that board does not have a F7 GI fuse on the power board each string should get one 5a fuse after it leaves j8.
Untitled (resized).png

#3 7 months ago

On my Barracora those 2 connectors used to be one but burned and broke in half looks like someone just used 2 connectors or the factory was lazy. Let me know if you want a pic? Not home but have glass off for work

#4 7 months ago

Bad_Regulator (resized).jpg
#5 7 months ago
Quoted from barakandl:

I think you have the connector J8 right. First four pins are one side of GI, white with striped colors. The last four is the other side of GI and solid colors. It should not be that critical the order as long as they are grouped up properly. The manual should have a wiring diagram page with colors.
Isn't that board missing the F7 GI fuse? I thought they had a 20a GI fuse on the power board in system 7. They did jumper it out in later system 11 when they fused each individual GI string after it leaves the power board. I would make sure you have a fuse(s) somewhere in the GI circuit. Since that board does not have a F7 GI fuse on the power board each string should get one 5a fuse after it leaves j8.
[quoted image]

Thanks for the info! The board is supposed to be for a System 7 - 11 machine - https://www.marcospecialties.com/pinball-parts/D-8345.

I'll check for the fuses. Thanks!

#6 7 months ago
Quoted from KJL:

On my Barracora those 2 connectors used to be one but burned and broke in half looks like someone just used 2 connectors or the factory was lazy. Let me know if you want a pic? Not home but have glass off for work

Yeah sure, if you have a pic that would be great. It almost looks like these two might have been one and broke. One of them does look a bit burned/melty on the end but the other one doesn't which makes me think maybe they were separate. Would be interested to see what yours looks like.

Thanks!

#7 7 months ago
Quoted from G-P-E:

[quoted image]

What's the deal with the LM337t regulator? I searched and a few people saying they suck but if you use LED displays it shouldn't really matter? I'm new at this so not sure what to make of it. Could you please elaborate?

#8 7 months ago

If you only use LED displays then the entire HV supply isn't needed so pull the fuse.
But these boards are also intended to supply power to plasma displays for the people that have not converted to LEDs.

Using a low voltage regulator in a high voltage supply is asking for trouble.
LM337T regulator is specified to have an absolute maximum input to output differential of 40V.

Can't remember DC input voltage of system 7 regulator off the top of my head so I will use 110VDC for an example.
If the input is 110VDC and the output voltage is set to 100VDC then you have a 10V differential. Theoretically this is fine and will work properly.
BUT there are two times where this absolute maximum 40V differential can be a problem.
1 -- If you throw a momentary short on the output voltage (it happens) then the input to output differential spikes quickly to 110VDC, these regulators tend to die quick and rather excitingly due to this.
2 -- At startup. There is a momentary delay between Vin going high to 110V and the output coming up. During this period, there is a 110V differential. This doesn't kill the part instantly like a short does but still causes a slow death by a thousand cuts type death. One of the websites (Clay's?) had a photo of a blown regulator with a description that stated something like "Burned regulator due to kid turning the power switch off and on repeatedly".

The TL783 on the positive has a far higher maximum input to output differential (125V) and tends to survive much better than the LM338s that the RD board was originally using.

I thought that with RD exiting the business that we were finally rid of these cheap short cut designs. Apparently not.
I'll just keep up my supply of replacement regulators...I sell lots of these.

#9 7 months ago
Quoted from G-P-E:

If you only use LED displays then the entire HV supply isn't needed so pull the fuse.
But these boards are also intended to supply power to plasma displays for the people that have not converted to LEDs.
Using a low voltage regulator in a high voltage supply is asking for trouble.
LM337T regulator is specified to have an absolute maximum input to output differential of 40V.
Can't remember DC input voltage of system 7 regulator off the top of my head so I will use 110VDC for an example.
If the input is 110VDC and the output voltage is set to 100VDC then you have a 10V differential. Theoretically this is fine and will work properly.
BUT there are two times where this absolute maximum 40V differential can be a problem.
1 -- If you throw a momentary short on the output voltage (it happens) then the input to output differential spikes quickly to 110VDC, these regulators tend to die quick and rather excitingly due to this.
2 -- At startup. There is a momentary delay between Vin going high to 110V and the output coming up. During this period, there is a 110V differential. This doesn't kill the part instantly like a short does but still causes a slow death by a thousand cuts type death. One of the websites (Clay's?) had a photo of a blown regulator with a description that stated something like "Burned regulator due to kid turning the power switch off and on repeatedly".
The TL783 on the positive has a far higher maximum input to output differential (125V) and tends to survive much better than the LM338s that the RD board was originally using.
I thought that with RD exiting the business that we were finally rid of these cheap short cut designs. Apparently not.
I'll just keep up my supply of replacement regulators...I sell lots of these.

Thanks I appreciate the additional information here. I've got LED displays but haven't installed them yet so will take your advice and pull the fuse after I do. I've got kids and they like to flip switches so this probably saved me a few bucks

#10 7 months ago
Quoted from barakandl:

I think you have the connector J8 right. First four pins are one side of GI, white with striped colors. The last four is the other side of GI and solid colors. It should not be that critical the order as long as they are grouped up properly. The manual should have a wiring diagram page with colors.
Isn't that board missing the F7 GI fuse? I thought they had a 20a GI fuse on the power board in system 7. They did jumper it out in later system 11 when they fused each individual GI string after it leaves the power board. I would make sure you have a fuse(s) somewhere in the GI circuit. Since that board does not have a F7 GI fuse on the power board each string should get one 5a fuse after it leaves j8.
[quoted image]

I think you are right. I've been looking at a few other boards and they have the F7 GI fuse. Not sure what the deal with this one is but as GPE also pointed out, it has the crappy LM337T regulator. I'm probably going to return this thing and purchase this one from XPin. Let me know what you guys think.

https://xpinpinball.com/product/xp-wms8345/

Thanks guys for pointing me down the right path.

#11 7 months ago

That 2nd board appears to use surface mount versions of the same regulators.
You would be better off with the first one which at least used real heat sinks and easier to replace thru-hole components.

I think you would actually be best off with a no frills board like this:
https://ksarcade.net/wms-8345-power-supply-for-williams-and-data-east.html

This third board uses the standard 'series pass transistor' design for the regulator but has components sized for high voltage.
It costs more to make this type of regulator but is worth the extra expense to do it right.
The original Williams boards also used this series pass regulator design which has lasted for decades.
This third board looks best but too bad they followed Williams design too much and didn't add current limiting. Heck, even a pair of solid state fuses on the outputs would do the trick. Just make sure to use proper size fuses (you should always use proper fuses, regardless of board design).
Third board also appears to use the Texas Instruments 'Simple Switcher' for the 5V regulator (right side of board). Excellent regulator - very reliable with very high efficiency (95%+) so it stays quite cool even under heavy load. Usually the heat sink they show on these provides more physical support than heat dissipation.

This kind of stuff is why we need somebody to do *honest* and in-depth "technical" reviews of various boards and then boil it down to results that a non-techie could understand.

#12 7 months ago

Don't think that barakandl makes a Williams system 7 power supply, but if he did, it would be the better product with upgraded features.

#13 7 months ago
Quoted from G-P-E:

That 2nd board appears to use surface mount versions of the same regulators.
You would be better off with the first one which at least used real heat sinks and easier to replace thru-hole components.
I think you would actually be best off with a no frills board like this:
https://ksarcade.net/wms-8345-power-supply-for-williams-and-data-east.html
This third board uses the standard 'series pass transistor' design for the regulator but has components sized for high voltage.
It costs more to make this type of regulator but is worth the extra expense to do it right.
The original Williams boards also used this series pass regulator design which has lasted for decades.
This third board looks best but too bad they followed Williams design too much and didn't add current limiting. Heck, even a pair of solid state fuses on the outputs would do the trick. Just make sure to use proper size fuses (you should always use proper fuses, regardless of board design).
Third board also appears to use the Texas Instruments 'Simple Switcher' for the 5V regulator (right side of board). Excellent regulator - very reliable with very high efficiency (95%+) so it stays quite cool even under heavy load. Usually the heat sink they show on these provides more physical support than heat dissipation.
This kind of stuff is why we need somebody to do *honest* and in-depth "technical" reviews of various boards and then boil it down to results that a non-techie could understand.

Cool thanks. The only other question I have is why doesn't this board have the F7 GI fuse that the XPin board has? If it is missing this, do I need to install 5A fuses for each string that leaves J8? The one I bought is also missing this which barakandl pointed out.

Thanks again for the help!

#14 7 months ago

Can't answer that one but good catch for barakandl.

I have noticed that there is at least one board manufacturer out there that tends to copy circuitry from other people's designs (even keeping 'easter eggs' that are deliberately put into other designs). This could account for why two of them are both missing the fuse.

#15 7 months ago
Quoted from ATAX:

Cool thanks. The only other question I have is why doesn't this board have the F7 GI fuse that the XPin board has? If it is missing this, do I need to install 5A fuses for each string that leaves J8? The one I bought is also missing this which barakandl pointed out.
Thanks again for the help!

If there is no fuse in the GI anywhere you must add them. Best bet would be to fuse each string with 5a after it leaves the power board in your case. WMS moved this fuse around from before the power board, on the power board, then after power board. But for at least some games I am pretty sure the GI fuse goes at F7 the power board. They say transformers never fail, well, a short at a lamp socket and no fuse is a way to smoke one.

LM2679 is great voltage regulator part and my go to. Just too bad they cost like $6. The datasheet even has a cook book circuit.

I will probably not do this power supply. I actually canceled my WMS 3-6 power supply because I did not want to price compete with others and just did not seem worth the time. Right now focusing on stuff I can easily do with automated surface mount assembly.

#16 7 months ago
Quoted from barakandl:

If there is no fuse in the GI anywhere you must add them. Best bet would be to fuse each string with 5a after it leaves the power board in your case. WMS moved this fuse around from before the power board, on the power board, then after power board. But for at least some games I am pretty sure the GI fuse goes at F7 the power board. They say transformers never fail, well, a short at a lamp socket and no fuse is a way to smoke one.
LM2679 is great voltage regulator part and my go to. Just too bad they cost like $6. The datasheet even has a cook book circuit.
I will probably not do this power supply. I actually canceled my WMS 3-6 power supply because I did not want to price compete with others and just did not seem worth the time. Right now focusing on stuff I can easily do with automated surface mount assembly.

Thanks again for the feedback and advice!

#17 7 months ago

ATAX sorry lost this thread here is a pic for posterity

9E521392-1730-4AE0-9789-2B4C9614DDD8 (resized).jpeg
#18 6 months ago
Quoted from KJL:

ATAX sorry lost this thread here is a pic for posterity
[quoted image]

Your picture I think confirms cosmic gunfight GI fuse should be on the power board at F7. That board should be marketed as system 11 compatible only (maybe 9?) until the fuse is added. Someone is going to melt a transformer.

#19 6 months ago
Quoted from barakandl:

until the fuse is added. Someone is going to melt a transformer.

I agree. IIRC The 2pin with the yellow wires comes from the transformer and one line goes directly to 4 pins in the 9 pin Molex next to it and directly to the GI strings.

The other line goes through the fuse next to it then to 4 pins of that same connector which is pretty flimsy compared to system 6

I had to replace all the headers and plugs and the square disconnect to the Playfield

Not sure how a replacement board could work without the fuse but it’s needed if a GI string shorts

4 months later
#20 86 days ago

I recently picked up a project Black Knight, and found the 3J6 power connector is only pushing 1.57v for the #7-10 +5v regulated connectors (with the ribbon removed). The #6 +12v unregulated connector puts out 14.96v.  What is causing these four regulated connectors to run so low?  Is it a capacitor issue? As I look at the power supply board, I do not see any bulged or burned components, but understand that doesn't necessarily mean anything.  Of note, all GI lighting works great, and I do not have any blown fuses.

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