biggestsonicfan
Professional
My whole draw to specific Superguns was the ATX PSU compatibility. I've only used one Mean Well and it generates too much noise for my board. Just my personal two cents.
- Thread starter low_budget
- Start date
If you don't mind, can you give me some more details on what specific board and PSU you used, and what type of noise you mean? I would like to be aware of any pitfalls that exist. Mean Wells are specified with fairly low ripple, but if the specific supply is not suited for the load maybe that could be the cause of issues.
Regarding your concern about ATX vs Mean Well though, I'm planning on the solution of using a single high-quality 12V supply and implementing two high frequency (MHz) switching converters on the board to generate the 5V and -5V rails with less than 20mV of ripple and all of the protection circuitry you'd expect from a high quality PSU. They will be rigorously tested before final production to make sure everything is correct and suitable but I wouldn't anticipate any noise issues.
If ATX is a major feature for you that you don't want to see go, I'd take that into consideration.
Edit: There was a glut of low-quality fake Mean Wells for a few years. You should be able to check the serial number to be sure on their website: https://www.meanwell.com/serviceReport.aspx
Edit 2: After discussing with multiple people and given the lack of response I'm going to make the working assumption that user error or some other anomaly was involved in biggestsonicfan's issues and move forward with replacing ATX with a Mean Well supply. ATX supplies are much more notorious for having low quality filtering and poor quality control, and I have heard stories now from people who themselves ran into issues with good power supplies by misusing them due to inexperience.
Regarding your concern about ATX vs Mean Well though, I'm planning on the solution of using a single high-quality 12V supply and implementing two high frequency (MHz) switching converters on the board to generate the 5V and -5V rails with less than 20mV of ripple and all of the protection circuitry you'd expect from a high quality PSU. They will be rigorously tested before final production to make sure everything is correct and suitable but I wouldn't anticipate any noise issues.
If ATX is a major feature for you that you don't want to see go, I'd take that into consideration.
Edit: There was a glut of low-quality fake Mean Wells for a few years. You should be able to check the serial number to be sure on their website: https://www.meanwell.com/serviceReport.aspx
Edit 2: After discussing with multiple people and given the lack of response I'm going to make the working assumption that user error or some other anomaly was involved in biggestsonicfan's issues and move forward with replacing ATX with a Mean Well supply. ATX supplies are much more notorious for having low quality filtering and poor quality control, and I have heard stories now from people who themselves ran into issues with good power supplies by misusing them due to inexperience.
Last edited:
Somewhat redoing the design again because I'm migrating to a new CAD program and I am moving to a single 12V supply with onboard regulators. I plan on selling Mean Well GST60A12-P1IR as well as a good quality lower-cost Chinese PSU from Alibaba as options. I am interested in offering the lower-cost PSU because based on my knowledge, the regulation of the 12V rail for JAMMA is less-critical and I'll be doing my own high-quality regulation of +5V and -5V. So for some people, the added cost of the Mean Well ($15 vs $5, so $10 extra) may not be justified.
Open to feedback, as always.
Open to feedback, as always.
biggestsonicfan
Professional
I did not see this reply initially. I pulled out the supply to check and there is no serial. Guess it's fake.
obitus1990
Enthusiast
I’d leave the ATX connector on it. If you don’t want to use an ATX PSU, a high quality PicoPSU from
Mini-box can be used. I’ve used it on my parsec for several years without issue.
Mini-box can be used. I’ve used it on my parsec for several years without issue.
For the record, do you remember where that power supply came from? Maybe add a picture too. Others may unwittingly be using similar fakes.
I've already decided to go forward with the single-rail supply, FWIW. It will reduce total purchasing cost and eliminate the variable of poor-quality ATX PSUs. If the end result is unsatisfactory in testing I'll go back to ATX, but I am optimistic.
obitus1990
Enthusiast
@low_budget
I have a v2.1 Parsec that the power switch has ceased to function. It stays ON, and the only way to power it down is to unplug the PSU. Testing the switch revealed it to be stuck always on, so I replaced the switch with a new one, but, the problem persists. Measuring voltage across where the pads where the switch reads at roughly 2V. Is there some other component that is in control of powering it on and off that may have failed?
Thanks!
I have a v2.1 Parsec that the power switch has ceased to function. It stays ON, and the only way to power it down is to unplug the PSU. Testing the switch revealed it to be stuck always on, so I replaced the switch with a new one, but, the problem persists. Measuring voltage across where the pads where the switch reads at roughly 2V. Is there some other component that is in control of powering it on and off that may have failed?
Thanks!
B0ingB0ingMissSaturn
Student
Heya! Curious if this was still an on-going project. It's been a while since the last update. I took that zip and sent an RFQ through to JLCPCB and included a note about the note.
I plan on sharing the costs for a one-off, qty 5 and qty 10 to see where they are at on pricing (at least at this time, it's subject to change of course lol).
Also got curious if OSHPark's auto builder would work with these files but I got a critical error of "Cannot find board outline". Sent a manual RFQ to them along with the same notes about the slots and holes in the outline layer and sent the files. So we'll see what they say back. If it's something they are able to run, I'd have to imagine they will be a lot more of an affordable option. Plus you can share a purchase link (at least used to many years ago, prolly still an option!) so others can easily buy it.
Reason I want to document all of this process starting at this point, is for others looking for a supergun that can do Composite/S-Video as well as support everything else it does. Can not appreciate enough the work of the person who designed the board and those involved in converting/updating the files. I need a few myself, so hope I can kick something back. Throwing this out there in the meantime.
Though will mention, I would 100% be on board to buy your redesign from you. Which is why I asked if this was still going (I guess in terms of design time, it's not been THAT long lol)
Last edited:
I've talked to the project owner and he told me he is retired for good from this project, that's why he made it open source. He also told me that maybe someone else around these forums could help us on getting this thing done, but I don't think there's someone here working on it either. I was trying to put my hands on one of these but got no answers on this forum yet.
Hi, it's stilll ongoing! Some IRL stuff happened that I had to drop everything to deal with, but I'm getting the ball rolling again. I've been thinking about the design a bit, and I'm going to add a few things to make the design even more robust and failsafe. Updates to come!
Some updates. I am getting custom step-down audio isolation transformers made to lower the BOM cost (the transformers were by far the most expensive thing on the whole board) and also simplify the design, removing the need for the attenuation resistors. I am also starting to complete orders for the components for a first-revision board run, but I still need to finalize the transformers and then route the board.
I am also looking for thoughts from people about whether the audio amps on arcade PCBs really cope well with the high input impedance of these isolation setups. Specifically, I am wondering whether anyone has noticed any reduction in sound quality or coloring of the sound when connected to a line-level converter in a supergun vs directly to an 8-ohm speaker. Perhaps it's not really a problem and the sound quality is fine, but I am considering supporting a dummy-load daughterboard so that electrically the PARSEC looks more like an arcade speaker, and thus the audio may sound more authentic (as silly of a concept as that may be when the game's native environment hardly allowed for much audio fidelity!)
Let me know what you think.
Edit: I posted a dedicated topic about the audio question over in the main arcade forum section.
I am also looking for thoughts from people about whether the audio amps on arcade PCBs really cope well with the high input impedance of these isolation setups. Specifically, I am wondering whether anyone has noticed any reduction in sound quality or coloring of the sound when connected to a line-level converter in a supergun vs directly to an 8-ohm speaker. Perhaps it's not really a problem and the sound quality is fine, but I am considering supporting a dummy-load daughterboard so that electrically the PARSEC looks more like an arcade speaker, and thus the audio may sound more authentic (as silly of a concept as that may be when the game's native environment hardly allowed for much audio fidelity!)
Let me know what you think.
Edit: I posted a dedicated topic about the audio question over in the main arcade forum section.
Last edited:
Some updates for today.
I am replacing the overcomplicated TPS63710-based -5V supply with the LM2776 inverting charge-pump as used in the Jammafier. I was concerned that the 0.2A that the LM2776 can supply would be insufficient, but after seeing a lack of reports of that and confirming that JAMMA PCBs don't do anything strange with -5V like some pre-JAMMA PCBs I am no longer worried. Theoretically the inverter configuration of the LM2776 may provide better symmetry between +5V and -5V as well, though voltage drop probably makes this moot.
Samples of the audio isolation transformers should be manufactured within the next day or so. Once they arrive here I will test them in-circuit. A friend of mine who has access to a lab-grade audio analyzer will help me evaluate their performance compared to the Triad Magnetics transformers. I will be testing 9:1 step-down transformers as well as 1:1 models with attenuation resistors as in the previous design. Due to the high DC resistance on the long winding of the 9:1, it's actually likely that the 1:1 design will be what I end up going with. I will test and confirm for certain which option is best.
I will be setting up my test-bench to test power-supplies soon, to see if there are any Chinese PSUs out of the samples I have from various manufacturers that are worth using. If they don't offer comparable performance to the Mean Well out of the gate, I might not bother with them. Even still, I am including enough safety features in the design that a failed PSU should pose no threat to the supergun or attached PCB.
On the note of safety features, I am adding ESD protection diodes to all inputs and outputs. It's a cheap way to get a bit more peace of mind, and perhaps it will protect at least one PCB from a nasty static shock. This is not a common feature even on high-end superguns, so I hope it will be another good selling point
I am replacing the overcomplicated TPS63710-based -5V supply with the LM2776 inverting charge-pump as used in the Jammafier. I was concerned that the 0.2A that the LM2776 can supply would be insufficient, but after seeing a lack of reports of that and confirming that JAMMA PCBs don't do anything strange with -5V like some pre-JAMMA PCBs I am no longer worried. Theoretically the inverter configuration of the LM2776 may provide better symmetry between +5V and -5V as well, though voltage drop probably makes this moot.
Samples of the audio isolation transformers should be manufactured within the next day or so. Once they arrive here I will test them in-circuit. A friend of mine who has access to a lab-grade audio analyzer will help me evaluate their performance compared to the Triad Magnetics transformers. I will be testing 9:1 step-down transformers as well as 1:1 models with attenuation resistors as in the previous design. Due to the high DC resistance on the long winding of the 9:1, it's actually likely that the 1:1 design will be what I end up going with. I will test and confirm for certain which option is best.
I will be setting up my test-bench to test power-supplies soon, to see if there are any Chinese PSUs out of the samples I have from various manufacturers that are worth using. If they don't offer comparable performance to the Mean Well out of the gate, I might not bother with them. Even still, I am including enough safety features in the design that a failed PSU should pose no threat to the supergun or attached PCB.
On the note of safety features, I am adding ESD protection diodes to all inputs and outputs. It's a cheap way to get a bit more peace of mind, and perhaps it will protect at least one PCB from a nasty static shock. This is not a common feature even on high-end superguns, so I hope it will be another good selling point
Hi all,
Some more updates. I'm waiting on a few more component samples from various suppliers to reach my OEM and then I'm going to get a big shipment of them to the USA for me to evaluate and develop with. Just in AC power cables alone, I have samples from about 10 different companies... I will select the best from among them and pair them with the Mean Well supplies which do not include them.
The design has evolved a bit more. In addition to a on-device configuration menu directly inspired by Frank_fjs's Minigun, I am including a toggleable BTLE 5.3 wireless configuration to quickly change all of the PARSEC's settings from the comfort of a smartphone app. By using BTLE 5.3 which includes the new auto-TX power negotiation algorithm, it should be possible to alter settings externally even from within a metal candy cab, without requiring the use of high TX power the rest of the time. The bluetooth is definitely scope creep... but I will be releasing this feature as a firmware release once everything else is done and shipped.
I'm moving the power-supply watchdogs and other critical hardware monitoring functions into a second small microcontroller (less than 25 cents of additional cost) that will be unencumbered by the GUI and bluetooth code, to ensure that it is responsive to transient faults for the full protection of attached PCBs. I agonized at length over it before realizing that it actually simplified the design, and took all critical board functions out of the BT SoC's responsibilities meaning that everything will still work if it crashes or contains bugs (much more likely since the code will have much more complexity).
I am also now the proud owner of an Altium license. God help my soul. And by soul I mean wallet. I might be forced to make a Kickstarter to pay for the final production run... but that would not happen before the product is done. Overall, I think there's probably about 6 months to go until this thing is out the door and I may have stable employment again by that point that would allow me to front the production costs. We'll see.
Some more updates. I'm waiting on a few more component samples from various suppliers to reach my OEM and then I'm going to get a big shipment of them to the USA for me to evaluate and develop with. Just in AC power cables alone, I have samples from about 10 different companies... I will select the best from among them and pair them with the Mean Well supplies which do not include them.
The design has evolved a bit more. In addition to a on-device configuration menu directly inspired by Frank_fjs's Minigun, I am including a toggleable BTLE 5.3 wireless configuration to quickly change all of the PARSEC's settings from the comfort of a smartphone app. By using BTLE 5.3 which includes the new auto-TX power negotiation algorithm, it should be possible to alter settings externally even from within a metal candy cab, without requiring the use of high TX power the rest of the time. The bluetooth is definitely scope creep... but I will be releasing this feature as a firmware release once everything else is done and shipped.
I'm moving the power-supply watchdogs and other critical hardware monitoring functions into a second small microcontroller (less than 25 cents of additional cost) that will be unencumbered by the GUI and bluetooth code, to ensure that it is responsive to transient faults for the full protection of attached PCBs. I agonized at length over it before realizing that it actually simplified the design, and took all critical board functions out of the BT SoC's responsibilities meaning that everything will still work if it crashes or contains bugs (much more likely since the code will have much more complexity).
I am also now the proud owner of an Altium license. God help my soul. And by soul I mean wallet. I might be forced to make a Kickstarter to pay for the final production run... but that would not happen before the product is done. Overall, I think there's probably about 6 months to go until this thing is out the door and I may have stable employment again by that point that would allow me to front the production costs. We'll see.
Last edited:
