http://www.mouser.com/Search/Refine.aspx?Keyword=2n3584

I have a spare green 7 digit X-pin that I could lend out for testing. PM me your info if you want it.

Unofficial, totally amateurish temp test on the heatsink. Using a meat thermometer suspended between the inner fins around the LM323K.

Space Invaders. Double glass with chaser lights in a widebody cabinet, door closed. 5 LED displays. One hour on attract mode reached a max of 135F. Started a four player game. 5 minutes into it the 5A fuse blew. Temp was 140.

EBD, standard cab with fake plastic back glass and five gas displays. Max temp at any time 120F.

I doubt it blew over temps since its well away in the lower cab. The fuse is new as is the rectifier board. Yikes.

Unless a drop in 5V cause some kind of solenoid driver glitch to lock on a coil.

Bally's reset circuit monitors the 11.9VDC, not the 5VDC. So a dip in the 5V could cause a software issue while not causing a full reset.

Resolved. My "new" problem had nothing to do with the new score displays. Took awhile to track down. SMDH...

IMG_6568_(resized).JPG

Put it on attract mode for an hour followed by non-stop four player games for an hour. Never topped 135F. Not one glitch. Not one blip. Displays look fantastic.

Part II - Received the EZSBC module today. A 5A replacement for the LM78HO5. Went ahead and reinstalled the gas displays to get a baseline. Believe it or else, 138F. So 140F with the LED displays seems not so bad. Hoping the 5A PSU7 will bring it down a little and help it all last a little longer. So I removed the 7805 from the solenoid board and replaced it with the PSU7. Replaced all gas displays with all LED displays. Pulled the fuse for the high voltage circuit. And fired it up. Blows F3 every single time within 3 seconds. A little disappointed and unsure of where to go from here.

Melted R50. WTF?
IMG_6584_(resized).JPG

You now have a switching power supply going in there instead of an old linear supply. The ground current can be quite high for normal operation of a switching supply. You do NOT want R50 in there for a switcher, replace it with a stout wire.

OK. Just reviewed the thread and saw where replacing R50 with a jumper is recommended. Yikes.

I can't tell from the photos on their website but are there any traces on the underside (besides a ground plane)? The PSU5 I have doesn't but it is also much more forgiving around the mounting holes than the PSU7 appears to be. So Ace may be right if your mounting hardware has either a larger washer or a star washer that may be touching any of the surrounding traces.

They really need to include a warning sheet for anyone that plans to use these on the Bally Regulator boards to install a jumper in R50.

Plus, with R50 burnt open, there is no path to ground for the regulator and there is the possible potential for feeding 11.9VDC to everything on 5V depending on how the switching circuit acts when ungrounded.

Post incident, I replaced the resistor and removed the EZSBC to swap back to the original component. Machine fired right up, so no harm, no foul. (One crispy resistor, a burned spot and one lifted trace. Nothing here I have not seen before.) Tomorrow I will replace R-50 with a stout jumper and try, try again.

Nail on the head here. Let's go to the tape!
IMG_6589_(resized).JPG

Then it looks like fiber or nylon washers are in order as well as the directive to remove and resistors in line with ground (R50 on Bally) before installing the 5 amp version.

Aren't you glad you were our ginnea pig?

That's how it appeared to me. The star washer bit right into the neighboring area and created a mess. Hopefully it still works. I'll be looking at perhaps a small lockwasher or something similar. The footprint of the star washer is too dang big and a connection is needed to reach ground. Here's a back side shot for ya.
IMG_6590_(resized).JPG

I don't mind being a guinea pig as long as others learn from my mistakes. We shall see...

I would guess that the short you had completely bypassed all the PSU7 circuitry so it probably survived.
That is a VERY small clearance surrounding the pads. You cannot use an insulator there as you need the electrical interconnect. A small diameter washer such as a NAS620 is probably what you need there.

Hope so!

Noted.

IMG_6598_(resized).JPG
IMG_6600_(resized).JPG

A flat washer (like Ed suggested) in combination with star washer would be my personal choice. With a star washer you damage the circuitboard; cuts through the copper.

Installed and humming right along. Literally. The sound feedback is so bad I can barely make out the sounds of the game. Any suggestions?

I'll put out a temp report after an hour on attract made.

Nope. this has actually been one of my quietest machines. Only my 6M$M was louder. On the other hand,

100F !!!!! Holy crap! That's a 40 degree drop for cryin' out loud! Sold.

I've read through the updates on this thread and I understand the need for clearance on the pads. I'll improve it soon. I don't understand the audio interference since the switching frequency is 800kHz, I think. It's certainly above 500kHz. This design is a lot noisier than the PSU5 because of the higher switching frequency, high peak currents and inductor shape but since it is so far above the audio range I didn't expect problems.

It costs a bunch of money to bring a design that like this into production. For a simple design its in the $400-$500 range for setup costs in the PCB factory, a solder paste stencil and wasted material when setting up the pick and place machine and the reflow oven. I have two options here, keep the design and fix the clearance issues only and live with the noisy design. Or go into another round of development to try to fix the noise issue with no guarantee of success. All feedback and suggestions appreciated.

I can also make a 4A version that is as quite as the PSU5. It can be used where the PSU5 doesn't supply enough current but the noise is an issue.

May I use these photos on my site for an installation guide?

Can someone post the part numbers for those screws and where I can buy some for sizing of the mounting holes.

Should I sell properly sized mounting screws and washers on my site or do you re-use the existing ones?

I'm learning a lot here.
Daniel

The pads are good, it's an application problem. Dummy end-users like me are going to use whatever is available or closest at hand without a second thought. In this case I used the existing hardware and you can see what happened. I would maybe have a note regarding the need to ground it at the pad with respect to washer size. IE- "washer O.D. should be no more than x mm in diameter." Or buy a box of 100 at a local hardware store and send them along with it. Two sets nut/bolt/washer might come out around $.05 per order in the end with no appreciable weight to add to shipping costs. I get the expense of refining a PCB run. The design is fine, just gotta educate the monkeys out here.

As for the noise issue, I wouldn't sweat it right away. I'm doing a lot of repairs/upgrades to this machine and it just happened that after installing the PSU7 the feedback took a big jump in volume. If I get a chance later today, I have another Bally out in the garage set up and I will swap in the SDB with your module and see what happens there. It's a common problem with these early Bally S/S machines. I only mentioned it because it was so noticeable. A cursory review of the schematic shows this particular sound card runs off 5v all together. I'll take some readings at the test points and see what we got. Results to be posted later.

Thanks for looking in on this. I'll be at TPF setting up Thursday afternoon and you are welcome to come and check out the machine for yourself. The way I see it, once I get some LED GI bulbs in the backbox I may get it down to near room temperature. Nice.

Shawn

And feel free to use my pictures as you need.

At work, companies always include new mounting hardware with replacement boards, because too long of screws may ground the board to a non-reference ground.

They know that in the field where many techs may have serviced equipment, there could be an assortment of screws installed as replacements.

Same with these Bally regulators. Who knows what different sized lock washers were lost and replaced, or even substituted by Bally themselves over the years?

Much safer to include the proper size washers, and note their importance in the installation instructions.

Best solution imho would be, to use smt washers soldered on the board by the manufacturer, so the end user only needs 2 srews.

http://www.smtnet.com/news/index.cfm?fuseaction=view_news&news_id=4248

OK, so all the test points on the sound card tested well within standards.

Next, I tested the board in my EBD and found NO appreciable difference in the amount of feedback between a normally configured SDB and one with the PSU7 installed. The EBD has normal gas displays.

Next I retested the Space Invaders using a the normal board first, then the one with the PSU7. You can hear the difference for yourself. I changed nothing between the videos except for the SBD board.

Without PSU7 -

With PSU7 -

Looking at the schematics, the Eight Ball Deluxe sound board has its own LM323 +5V regulator, whereas the Space Invaders sound board is using +5V from the solenoid driver board (PSU7 in this case).

Presumably some will use the PSU7 for other (non Pinball) applications and they will have different length screw requirements.
I'd prefer a larger screw face area on the next batch of PSU7 to eliminate screw head issues altogether.

Daniel, I presume it's OK from the pic above but are any of the vias on the bottom of the PSU7 live? i.e. can one bolt the PSU7 directly onto a heatsink without fear of live vias inadvertently shorting to common/ground? Some will probably remove the heatsink but I'm one that would prefer to leave the heatsink there so anyone in future could easily revert to the original regulator for whatever reason.

Can you simply upgrade the PSU5 to 4A or does it need the expense of a new design?

Hopefully you and Mk1Mod0 can catch up at TPF and do a little analysis of the humming issue.

I still have the heatsink under it but I don't believe it to be necessary.

It appears that his thread is becoming one about PSU7 regulators. My question is if it makes sense to install the lock washer underneath the board, between it and the heat sink? It does not serve a locking function as much as a bite and it has a thru trace.

Do you mean between the back of the regulator and the top of the heat sink? Or between the bottom of the heat sink and the top of the SDB?

While that washer might bite into the aluminum heat sink thereby grounding it, it may be difficult to insure a good contact with the regulator's ground at the surface of the heat sink (black ones are anodized) and the back of the regulator module has solder mask covering it. And it almost would seem like overkill to have to use a star washer between the regulator and heat sink AND the heat sink and SDB PCB pad.

I would instead use a star lock washer on the underside of the PCB where it can bite into the ground pad unless you are using a "kep nut" (the kind where the star washer is part of the nut).

A simple split lock washer the size of the screw could be used on the top side at the PSU7 pad but I have mixed feelings on this. By using a washer, you introduce a second surface where we have to insure good electrical contact but the cut of the split washer may do a good job cutting into the pad and screw head. Hard to know without doing actual tests under heavy current loading.

It might be just as good to leave it out and to just make sure to completely clean the underside of the screw head (wire brush) or use brand new screws to insure good contact with the regulator's pad. But I would definitely use one on the underside of the SDB PCB.

The noise would have to be on the 12v line to come through the speakers... no? The amplifiers are powered by 12v.... unless the sound generator is effected and it comes from the analogue output there.

Could the 12000uF filter cap be failing on the board making excessive hum? Can you check the ripple on the 12v line?

Board was recapped and all ground mods completed two years ago.

No clue how to check the ripple on the 12v line. Not even sure what that means. If it can be done with a dmm, I can do it. No o-scope in this house.

That will all depend on how your DVM reacts to AC on a DC voltage. What you do is set your meter to AC. Lock it in to a low setting if you can (2 or 20 volts if available). An auto range is pretty useless unless it has the hold feature.

You should see less than 1 volt of AC across the 11.9vdc across the filter cap.

If your meter won't tell you the presence of AC, then you might see the normal DC reading fluctuate when the buzzing increases (high score displayed).

Some really cheap meters may just tell you that there is 2X the 11.9v there because of how they read AC. If you get a reading higher than a few volts, disregard it. Your game would not be functional if there were like 5-6 volts AC on the unregulated 11.9.

Fluke 115. Set to AC MIN it read 0.00 with ground on SDB ground and pos on either side of the 12,000 cap at any point in the high score / last game score cycle.

Posting these pics for reference and because I said I would when they arrived.
These are two LT1003CK 5V linear regulators which were designed as a 5A replacement for the LM323 with some better characteristics.
Both purchased over ebay from China. The old one date stamped 9306 was sold as used. The newer one date stamped 1048 was was sold as new. Based on previous pictures they don't look fake. I've even soldered the top cover back down on the old unit and am currently running it in my machine for test (amusement

LT1003CK_Decapped.jpg

LT1003CK.jpg

Look like the real deal to me with the multiple wiskers.

How much did they cost?

Old one:
www.ebay.com/itm/151896786969
The two legs were normal length so had never been cut. Maybe they were just old stock.

New one:
www.ebay.com/itm/201024477998

Erm, I just noticed they have none left however what I got came with a business card from these guys so this is the original source of the newer ones and they still have stock:
www.ebay.com/itm/261235641206

Thanks for posting the links. I just ordered 3 off the first link since there were still 17 left (now 14). I figured it was worth it just to make sure my 78H05K drawer wasn't empty!

Thanks again for being the guinea pig and posting the results!

My pleasure, it was thanks to your idea of cutting the head off to look inside and prove those cheap LM78H05 were fake.
BTW, both of these came well packaged, unlike the fake LM78H05 which came with legs sticking out from an envelope!

I installed the psu7 tonight and here is how it went. Didn't read this thread and just pulled the old regulator and heatsink and installed the psu7. Reinstalled and no boot. Pulled out board, checked work, asked mk1mod what he did. I used smaller hardware, pulled r49, and pulled r50 and installed a jumper. Machine boots but the noise is unbearable. I have the same displays as mk1mod and the same noise issues. Machine is a Nitro Ground Shaker. Any ideas on noise reduction?

Would love to figure it out. Problem is that with regular displays and PSU7 there is no problem. With led displays and LM78H05 there is no problem. However, with led displays AND the PSU7, the noise is horrendous. No one noticed it over the din of TPF but in the house it is unbearable. Some kind of interaction going on there.