(Topic ID: 290458)

Bally AS-2518-22 Voltage regulator problem with high voltage

By Inkochnito

6 months ago


Topic Heartbeat

Topic Stats

  • 78 posts
  • 14 Pinsiders participating
  • Latest reply 5 months ago by Inkochnito
  • Topic is favorited by 4 Pinsiders

You

Linked Games

No games have been linked to this topic.

    Topic Gallery

    View topic image gallery

    PinCompares (resized).jpg
    IMG_0023b.jpg
    pasted_image (resized).png
    20210417_214727 (resized).jpg
    Stern_High_Voltage_Regulator_schematic (resized).jpg
    IMG_0017a.jpg
    340623_4334341271808_10133355_o (resized).jpg
    switching (resized).png
    Screenshot_20210415-201326 (resized).png
    Heat_Sink_Wakefield_Vette (resized).jpg
    AS-2518-22_front_good (resized).jpg
    AS-2518-22_front_bad (resized).jpg
    AS-2518-22_back_good (resized).jpg
    AS-2518-22_back_bad (resized).jpg
    SDB_HV1.png
    Bally_AS-2518-22_High_Voltage_Regulator_schematic_measurements (resized).jpg

    There are 78 posts in this topic. You are on page 2 of 2.
    #51 6 months ago
    Quoted from G-P-E:

    ... and this is why when one transistor fails, they all do.

    Using quality parts helps, but yes, should either of the drivers short (in a particular way) that would place excess voltage on the other driver. I add it's never happened to boards I have repaired in 30 years using quality parts.

    If you really want to be super safe, use a TIP48 rated at 300V but I have never found it necessary. Sometimes overkill isn't always needed to achieve a reliable result.

    When rebuilding this particular circuit a repair person should replace ALL active devices regardless, so your point is rather moot.

    I try and use suitable parts that can EASILY be obtained at SENSIBLE prices.

    #52 6 months ago

    The question is why do we see the HV circuit blow out more than we should?

    The only time I've had to re-repair a board was when a friend disconnected/connected a display live and it killed the HV section.

    I guess another reason is when the 100k 1/4W resistors in the digit drive circuit on the displays go up in flames.

    From Dolly Parton onwards, Bally revised the high voltage F2 fuse on the rectifier board from a slow blow to a fast blow (same 750mA rating).

    #53 6 months ago
    Quoted from wayout440:

    So you bought 40 transistors to fix one game? Geez, why not leave some for other folks trying to fix their games?

    Nope, I fix boards for a lot of other people.
    This will allow me to fix 20 boards.
    They'll be gone in a year or so, maybe two.

    #54 6 months ago
    Quoted from Inkochnito:

    Nope, I fix boards for a lot of other people.
    This will allow me to fix 20 boards.
    They'll be gone in a year or so, maybe two.

    I kind of figured that might be the case. Though I do suppose some might hoard parts, as long as you're making use of them it's good.

    #55 6 months ago
    Quoted from Inkochnito:barakandl Would this heat sink work for the TIP50G?
    https://nl.mouser.com/ProductDetail/Wakefield-Vette/291-H36AB?qs=VVKQmw408U%252Bz99MOvnriAQ%3D%3D
    No screws to get lose, not too heavy.
    Giving just a bit more cooling....
    [quoted image]

    yes that should be fine. i measured the temp of MJE3439 used in this circuit just using a large area of PCB foil to heat sink and neither one is over 40c so just about any heatsink is fine and even naked does not come close to maximum temp.

    #56 6 months ago

    Just to summerize the possible replacement parts....

    Q22 and Q23 are 2N3440 and is getting harder to find.
    Replacement can be a TIP50, but be aware of the correct orientation (as with all transistors).
    The use of a small heatsink is advised for Q23 (like Wakefield-Vette/291-H36AB).

    Q21 is a 2N3284 (very expensive and hard to find).
    A good replacement is a BUX85 mounted on the original heatsink (very easy to do, cut the middle leg).

    The zener diode VR1 is a 1N5275 and is getting harder to find.
    A good replacement is a Z4KE140A zener diode.

    The adjustment pot RT1 (25K) is an open type and cause bad adjustment.
    This can/should be replaced with a Bourns 3306P-1-203 (20K).

    #57 6 months ago

    That's why the aftermarket supplies are a no brainer. Highly efficient linear designs that generate less heat, short circuit protection, 5 amp 5V rails, LED indicators, mosfet drivers for faster coil response and more. It's one thing to save an expensive MPU board with a few dollars in parts, but replacing a power supply or solenoid driver board with an aftermarket board for years of reliable operation often makes more sense.

    #58 6 months ago
    Quoted from wayout440:

    That's why the aftermarket supplies are a no brainer. Highly efficient linear designs that generate less heat, short circuit protection, 5 amp 5V rails, LED indicators, mosfet drivers for faster coil response and more. It's one thing to save an expensive MPU board with a few dollars in parts, but replacing a power supply or solenoid driver board with an aftermarket board for years of reliable operation often makes more sense.

    I think you mean:

    "Highly efficient SWITCHING designs"

    Linear is what the originals are......

    #59 6 months ago
    Quoted from pins4u:

    I think you mean:
    "Highly efficient SWITCHING designs"
    Linear is what the originals are......

    No. Linear, as stated below. Anyways, really not a debate, but an option. The prices are so affordable on a lot of the aftermarket stuff, if you want to save time and get one up and running quick and reliable.

    Screenshot_20210415-201326 (resized).png
    #60 6 months ago

    Actually it IS a debate, they have 100% for certain made an error in that wording.

    This is their 5V section and it is, without any doubt, a switching supply as can be seen by the chip used and the toroidal choke.

    Anyone can make a mistake and this is mistake on their part for sure.

    switching (resized).png
    #61 6 months ago

    Well, really they could have simply said "modern" designed power supply and that would suffice. The choke input will give much better voltage regulation over a varying load vs. capacitive input. My personal preference is if a supply needs a minor repair such as a filter cap I just make the minor repair and test it, but if it needs a lot of work, just replace it with a new supply and and move on. That's the first thing I overhaul when restoring a game - start with clean, reliable power. It's the heart of the game.

    #62 6 months ago

    Not worded correctly:
    "This board uses modern linear power supplies which means it offers high efficiency and less heat with no more big heat sinks. This board will keep the backbox cooler!"
    Sorry but all linear regulators operate by dissipating excess power as heat. Given the same input voltage and output voltages then it will dissipate the *exact same amount of heat* ... regardless of actual linear regulator method used.

    I believe they meant to say "This board uses modern switching regulator for +5V supply which means it offers high efficiency...."
    Definitely a big difference and the 5V regulator will run much cooler.

    The inductor (choke) is not part of the input but is actually part of the switching regulator's output current reservoir. The inductor's collapsing magnetic field provides power to circuit while switching regulator's internal FET is turned off (about 60K times per second depending on specific part).
    Strange but I do not see the bulk capacitance on the input of that board's low voltage supply. These regulators specify an input capacitor and output capacitor based on the fact that the voltage they are getting is already filtered DC voltage. This was one of several major RD design flaws.

    #63 6 months ago
    Quoted from Quench:

    The question is why do we see the HV circuit blow out more than we should?
    The only time I've had to re-repair a board was when a friend disconnected/connected a display live and it killed the HV section.
    I guess another reason is when the 100k 1/4W resistors in the digit drive circuit on the displays go up in flames.
    From Dolly Parton onwards, Bally revised the high voltage F2 fuse on the rectifier board from a slow blow to a fast blow (same 750mA rating).

    The transistors seem to give up the ghost for no good explainable reason at times. I have had at least one go short Vin = Vout while running in my home. The only thing wrong was transistors and a couple resistors burn presumably after the fact. I am sure you can have human error accidents too. The display PCBs will get covered in dust, ash, and whatever. At some point maybe the HV can arc over. I have not had the HV circuit get blown out in any of my replacement boards that I know of. Small sample size and short length of time.

    This can happen too. High voltage is a bitch.
    340623_4334341271808_10133355_o (resized).jpg

    #64 6 months ago

    G-P-E In your professional opinion is a new supply, even with this design flaw, still a better option than significant repairs to a 20 or 30 year old supply? Another thing to keep in mind is not all aftermarket supplies are the same, with all the different games and aftermarket companies out there. I've only had a few Xpin and Rottendog supplies and they worked very well.

    #65 6 months ago
    Quoted from barakandl:

    I am sure you can have human error accidents too.

    Yes, I also had someone call today who accidentally swapped a display with the power on and blew the SDB's high voltage circuit.

    Quoted from barakandl:

    This can happen too. High voltage is a bitch.

    Not as bad as yours but those 100k resistors can go out in a blaze of glory.

    IMG_0017a.jpg

    #66 6 months ago

    That burnt display i pictured came in a dead, sitting in a warehouse for probably a long time, game. The SDB was missing but I assume the HV section got smoked too.

    I have seen the MPU board 1K2 series resistors burnt to a crisp too. I guess when the digit driver goes short C to E it can put high voltage through the connector that goes back to the PIA chip with that 1K2 resistor in series.

    When plugging in the display live HV is at Pin 1 and ground is at 13. Some interesting situations happen when the ground is still floating?

    #67 6 months ago
    Quoted from wayout440:

    G-P-E In your professional opinion is a new supply, even with this design flaw, still a better option than significant repairs to a 20 or 30 year old supply? Another thing to keep in mind is not all aftermarket supplies are the same, with all the different games and aftermarket companies out there. I've only had a few Xpin and Rottendog supplies and they worked very well.

    First of all - age is not a deterrence to functionality for solid state electronics. Sure you may need to recap boards and replace connectors. But overall, they were sound designs for decades. Finding replacement parts for often damaged components, on the other hand, may become an issue for 30 year old boards (and some of the new boards). Some parts are NLA and getting hard to find. Always a work around but it still makes it harder to maintain (e.g. 3081 transistor arrays). A replacement is usually better than a board that needs a significant amount of repairs but there are some boards out there that are better than others.

    If you look at any board long enough - you can always spot something and think "I can do better". Does it make the original design or the replacement design worse? Not necessarily, mostly just 'different'.

    Regarding reliability.
    The original boards were solid designs for 20-30 years. But heat and typical 'pinball machine abuse' of electronics takes a toll. There is at least one board maker out there that does not take the abuse very well. For XPin and Alltek boards - I have never had anybody asking for replacement parts. The other board maker you listed is a far different story.

    I have seen a minor problem and two major problems with replacement power supply boards:

    The minor issue: low voltage supply not having bulk filtering capacitance in addition to the added regulator's input cap (low-ESR, high ripple current). The regulator's are expecting rectified and filtered DC power at the input of their circuitry. This means bulk capacitance in addition to the regulator's input capacitor. Most designers just like to see a separation of bulk filtering and the input cap. However, depending on what specific caps they use for the regulator's input capacitor, this one is easily overcome. Also, paralleled caps tends to work better than a single cap (lower series resistance and higher ripple current capability) - but this is a typical manufacturer suggestion, not an issue.

    Major issue 1: wrong type or undersized catch diode in the low voltage supply. This must be a high speed, high current, low forward drop diode to properly supply current through the output inductor. There are many choices but one maker was using 1N5401s in Gottlieb supplies for awhile - until they started burning up. If you have a Gottlieb supply with burnt catch diode, proper replacement would be an 80SQ045N.

    Major issue 2: one piece voltage regulators for the high voltage supply. None of the one piece regulators are designed for a 170-190V supply. Some could barely do +125V and others aren't even rated high enough to do the 60V for Gottlieb supplies. I have sold MANY replacement regulators for one of the two supplies you mentioned - and can no longer get the regulators they used (obsolete).
    The designers of the replacement boards see the spec "Vin to Vout maximum voltage" and believe they can stay within that. However if somebody shorts the output - I have seen the one piece regulators literally explode. And then there are the turn-on surges where input voltage is up to full voltage and outputs are not yet pulled up (this was the one that the National Semi FAEs warned us about that kill the regulators over time - sort of a death by a thousand cuts issue). Some fast protection has been added by a couple suppliers - if fast enough, it does protect from blowing the regulator but the power-up surges will eventually kill them. Just depends on how often you power up and down the machines.

    The only place I can see the proper use of a one piece HV regulator is a Texas Instruments TL783 in a Gottlieb system 1 or 80. For the other supplies, series pass transistor regulators is the only way to go for these supplies. Bally's HV regulator is a series pass transistor and they almost had it right but they used parts that were very close to the absolute max ratings. A bit of beefing up of transistors (such as TIP50 and BUX85 mentioned above), good heat sinking AND current limiting -- these could have been indestructible.

    #68 6 months ago

    Quench -- second photo, did it actually burn a hole through a 'ceramic' capacitor?

    #69 6 months ago

    My replacement checks those boxes as long as you count a fuse as current limiting the HV output =D.

    FWIW stern's later MPU200 game with the HV shutdown circuit does not seem to work, at least not in all cases, as the sdb board HV section will get damaged on overcurrent. Fuse probably should be added back in for those games.

    #70 6 months ago
    Quoted from barakandl:

    My replacement checks those boxes as long as you count a fuse as current limiting the HV output =D.
    FWIW stern's later MPU200 game with the HV shutdown circuit does not seem to work, at least not in all cases, as the sdb board HV section will get damaged on overcurrent. Fuse probably should be added back in for those games.

    Do you mean the 4 transistor version like this schematic?

    Stern_High_Voltage_Regulator_schematic (resized).jpg
    #71 6 months ago

    Q24 was added to provide what is called "Foldback Current Limiting".
    All output current goes through resistor R73. If the supply is shorted then the entire output voltage will be dropped across R73. Q24 will be activated by the higher base voltage than the emitter voltage (base side would be close to 185V but the emitter side would be approaching zero volts). This will immediately cause Q24 to turn on which in turn pulls Q22 low which will then turn Q21 nearly off (still slightly on). Foldback current limiting does work well in a simple series pass regulator circuit but when combined with high voltage and the voltage adjustment circuitry, it doesn't work quite as well.

    The idea behind this was good but it was a poor part choice for Q24. In a fault condition, his voltage rating is too low. Q24 can fail which eliminates the current limiting and turns the regulator back up to full strength. High voltage supply really should have been fused.

    #72 6 months ago
    Quoted from Inkochnito:

    Do you mean the 4 transistor version like this schematic?[quoted image]

    Yes. That's it. The shutdown circuit is nice idea but the fuse should not have been deleted.

    Based on all the boards i have amassed seems more typical of the stern sdb without the fuse to be burnt up when the HV section has a problem.

    20210417_214727 (resized).jpg

    #73 6 months ago

    Here is a different looking replacement to the other brands mentioned here listed on Pinball Life site.

    pasted_image (resized).png
    #74 6 months ago
    Quoted from barakandl:

    I have seen the MPU board 1K2 series resistors burnt to a crisp too.

    Wow, haven't seen that.

    Quoted from G-P-E:

    Quench -- second photo, did it actually burn a hole through a 'ceramic' capacitor?

    Nope
    That's the capacitors internal disk that's uncovered.

    Quoted from G-P-E:

    The idea behind this was good but it was a poor part choice for Q24.

    I think about a year ago you mentioned a MPSA42 transistor (same as used on the display boards) was a better choice for Q24 (Vceo of 300V)? apposed to the factory used 2N3904 which is only 60V spec.

    First time I'd seen one of these particular H.V capacitors (red epoxy on the positive side). It measures good with my ESR tester but was causing the 750mA H.V F2 fuse on the rectifier board to blow every time I powered on. One for the memory bank.

    IMG_0023b.jpg

    #75 6 months ago
    Quoted from Quench:

    Wow, haven't seen that.

    I've seen it happen when you put the display connector in reverse on the display board.
    It instantly fries the 1K2 resistor on the MPU.
    Some cheap ass repairman did not put in a lock pin and the connector got twisted around.
    I failed to see the red wire on the wroung side of the connector before it was too late....

    #76 6 months ago
    Quoted from Quench:

    It measures good with my ESR tester but was causing the 750mA H.V F2 fuse on the rectifier board to blow every time I powered on. One for the memory bank.
    [quoted image]

    leaking too much DC if the ESR is good?

    Quoted from Inkochnito:

    I've seen it happen when you put the display connector in reverse on the display board.
    It instantly fries the 1K2 resistor on the MPU.
    Some cheap ass repairman did not put in a lock pin and the connector got twisted around.
    I failed to see the red wire on the wroung side of the connector before it was too late....

    PIA probably not happy about sinking high voltage either. I can't remember if the PIA was bad when I have encountered this, but likely to at least kill one port.

    #77 6 months ago

    Don't mean to go off topic, but I really struggled with some of the (counterfeit) HV driver chips on the display boards (Williams World Cup 1978, System 3). Drove me nuts. I knew the chips were suspect, but original chips are almost impossible to get. So I gambled and lost that bet. Shipping just added insult to injury.

    Long story short, it was simply easier to swap over to Wolffpac's LED's. His kits were only $99 - goodbye fake high voltage driver chip issues and didn't have to worry about any HV issues down the road.

    Granted, some people like to keep everything original as possible.

    #78 5 months ago
    Quoted from mbwalker:

    Granted, some people like to keep everything original as possible.

    That is why I keep repairing these boards.
    Most of the time it is not only the HV section that needs to be repaired.
    The 5V caps needs to be replaced and the upgrade needs to be done.
    Solenoid drivers and CA3180's need to be checked too.

    There are 78 posts in this topic. You are on page 2 of 2.

    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