PARSEC Supergun Project
- Thread starter low_budget
- Start date
low_budget
Student
I have to admit, I am having some burnout on building Parsec superguns by hand. In the last 2 years I have built over 100 of them.
I will be building one last batch of 6 available late December / early January. Then I need to take a break for a while, then look at my options for building larger quantities and possibly another redesign.
I will continue to offer the bare PCBs with parts list for the v2.2 Parsec.
I will be building one last batch of 6 available late December / early January. Then I need to take a break for a while, then look at my options for building larger quantities and possibly another redesign.
I will continue to offer the bare PCBs with parts list for the v2.2 Parsec.
It does look pretty time consuming to build those. I bought one 8 months ago and finally fired it up the other day. Works great, good work man!
I ripped a bunch of ATX power supplies out of junk PCs, but I can't seem to find a nice form factor. Any recommendations? I'd prefer not to buy one, but I'm not finding what I want. I'd like to build this into a case of some sort and have it all self contained.
I ripped a bunch of ATX power supplies out of junk PCs, but I can't seem to find a nice form factor. Any recommendations? I'd prefer not to buy one, but I'm not finding what I want. I'd like to build this into a case of some sort and have it all self contained.
obitus1990
Enthusiast
rcboosted
Enthusiast
Totally understandable on the burn out. I hope I'm not out of line for suggesting this. Perhaps you can offer a kit with PCB and all the parts that is needed? The user can then assemble it together themselves. I tallied up the price of all the parts that is needed via the BOM list and it was over $200 including the PCB before any shipping.
So I received my Parsec and it won't sync with my Sony PVM 20L2MD. Can anyone suggest some troubleshooting steps i can take to try and correct this issue. Also, should I be looking at TTL sync as a possible correction for this issue?
Ok, so at someones suggestion on discord I went ahead and soldered the jumper for TTL sync this morning to see if it makes a difference and it doesn't appear to make any difference. I'm attaching some video to show what I am getting along with showing how i connected the TTL circuit. I am connecting this item via 8pin din to SCART to BNC direct into my PVM 20L2MD.
Also, I attempted hooking the parsec up through my OSSC, but also got no sync.
View: https://www.youtube.com/watch?v=U9PuxrbxV8s
(TTL)
View: https://www.youtube.com/watch?v=XuqGah_LVvY
(TTL OFF)
Also, I attempted hooking the parsec up through my OSSC, but also got no sync.
View: https://www.youtube.com/watch?v=U9PuxrbxV8s
(TTL)
(TTL OFF)
Attachments
Last edited:
low_budget
Student
It's been a long time since my latest update.
I have decided I will not be building or selling any Parsec superguns anymore. I am finding myself too busy with other projects like the NESessity NES motherboard replacement.
I am willing to share the PCB design files and other documentation with anyone interested. I don't have gerber files; the PCB files are in .pad2pad format that requires the free CAD software from www.pad2pad.com
There still seems to be much demand for the Parsec, and I don't want my work up until now to go to waste. Now anyone can make their own Parsec superguns and even make improvements if they want to put the work into it.
I have decided I will not be building or selling any Parsec superguns anymore. I am finding myself too busy with other projects like the NESessity NES motherboard replacement.
I am willing to share the PCB design files and other documentation with anyone interested. I don't have gerber files; the PCB files are in .pad2pad format that requires the free CAD software from www.pad2pad.com
There still seems to be much demand for the Parsec, and I don't want my work up until now to go to waste. Now anyone can make their own Parsec superguns and even make improvements if they want to put the work into it.
hatmoose
Enlightened
This is an amazing and generous thing that you do!
There are thousands, probably tens of thousands of hours of work that have gone into this - to offer to share that is absolutely amazing.
Totally hear you on the burnout and time taken - I’ve been assembling Miniguns recently, probably 4-5 hours per board, and Parsec has even more work in it. Truly a labor of love!
Thank you very much for this
low_budget
Student
I added the files to google drive; here's the link: https://drive.google.com/drive/folders/1TuKNDK7kQ6Lk38z5NDo_T8bR-494Cp9B?usp=sharing
There are 2 available PCB files, one is complete and ready to order (v2.2 final) while the other doesn't have the ground plane connected and is suitable for making changes.
enjoy!
There are 2 available PCB files, one is complete and ready to order (v2.2 final) while the other doesn't have the ground plane connected and is suitable for making changes.
enjoy!
obitus1990
Enthusiast
Awesome! I love the Parsec, and have built two of them from PCBs I bought from you. It's too bad that the free CAD program can't export the board files to gerbers or any other usable format, because it locks you in to using that particular board house, and they are expensive, with a minimum quantity of 50 boards @ $10 each.
Last edited:
hatmoose
Enlightened
#freethenipple - uh, I mean #freetheparsec, this injustice cannot stand!
I see here that for the low, low price of $100 they will convert from their proprietary format to gerber. Do we jointly believe that this is good value? Will they actually honor the deal?
https://www.pad2pad.com/convert/
obitus1990
Enthusiast
Halfsies?
low_budget
Student
Since I'm making the Parsec open source, I would rather the files be accessible and usable to as many people as possible.
I will contribute $50 toward converting the Parsec v2.2 file to gerber. On the Pad2Pad website is says "Your credit card will be authorized at the time of order placement, but will only be charged if a Gerber file can be successfully generated." So it may not be possible to generate a file? I guess it's worth a try?
Unfortunately I learned how to design PCBs using the free Pad2Pad software, and all my designs are in their proprietary format. I've tried other software and found it was taking me too long to learn and many things were done differently than I was used to.
While Pad2Pad is more expensive than other PCB manufacturers, they are based in the US and I've only had 2 defects out of 1,000s of boards I've ordered from them. Except for the pricing, I've had good experiences with them.
Yes I regret not making all my various PCB designs in gerber format from the beginning but I guess the offer of FREE PCB DESIGN SOFTWARE suckered me in.
Lookin forward to the files being made availabe i have the v2.0 of your board and would love to put together the v2.2 
Also I just got a CPS2 game will the PARASEC work right off with the CPS2 or do I need to adjust anything for it to work it currently is being used with my neogeo mvs boards
Also I just got a CPS2 game will the PARASEC work right off with the CPS2 or do I need to adjust anything for it to work it currently is being used with my neogeo mvs boards
The order finally completed. Strangely, I was only charged $50. Here are the files, enjoy!
If you use them and can afford it, please consider making a contribution to share the cost. You can do so via Paypal (link). I will remove the donation link if I receive a total of $40 (the $50 I paid minus $10 that I will contribute myself).
Important note about the gerbers: Some of the through-hole pads use the board-outline layer instead of the drill layer. This is not an artifact of the conversion process; it was done that way in the original files too. This causes the holes not to be rendered when previewing the gerbers, and means that there is a potential for the boards to not come out correctly if the manufacturer and their software doesn't notice. I contacted JLCPCB and they said there would be no issue ordering through them, but that orders should contain the note "There are slots and holes in the outline layer, please check it".
Another edit: The pick-and-place (CPL) file is missing. I'm going to have to call Pad2Pad tomorrow and determine why. I might have to make the file by hand (god forbid). Also, some of the footprints in the actual file are not configured as footprints, so I might have to re-do the conversion anyways to fix it. I kinda wish I had the schematic so I could just remake the thing in EasyEDA.
If you use them and can afford it, please consider making a contribution to share the cost. You can do so via Paypal (link). I will remove the donation link if I receive a total of $40 (the $50 I paid minus $10 that I will contribute myself).
Important note about the gerbers: Some of the through-hole pads use the board-outline layer instead of the drill layer. This is not an artifact of the conversion process; it was done that way in the original files too. This causes the holes not to be rendered when previewing the gerbers, and means that there is a potential for the boards to not come out correctly if the manufacturer and their software doesn't notice. I contacted JLCPCB and they said there would be no issue ordering through them, but that orders should contain the note "There are slots and holes in the outline layer, please check it".
Another edit: The pick-and-place (CPL) file is missing. I'm going to have to call Pad2Pad tomorrow and determine why. I might have to make the file by hand (god forbid). Also, some of the footprints in the actual file are not configured as footprints, so I might have to re-do the conversion anyways to fix it. I kinda wish I had the schematic so I could just remake the thing in EasyEDA.
Last edited:
low_budget
Student
Thanks Javaguru for converting the design files. I'd like to help out in any way I can to get the converted gerber file complete and correct.
Attached is all the schematics I have. The Parsec design has changed a few times, and I never did make a complete schematic of the entire supergun.
The circuits for the BH7236AF, EL1883, and THS7316 are all based off their datasheet example circuit. The RGB input from the JAMMA connector goes into the attached hand drawn circuit. RGB output from the circuit goes to the THS7316 and BH7236AF RGB inputs. The BH7236AF is fed a 3.579545 MHz clock from an oscillator and configured for NTSC luma trap. The sync signal from the circuit passes through the EL1883 sync separator and goes to the composite sync output and sync input of the BH7236AF. A LM1881 sync separator can be substituted if the EL1883 isn't available.
Component video isn't generated on the PARSEC; a separate PCB is required available from Viletim as a component add-on for the NESRGB. The PARSEC only has the connectors and RGB signals tapped off the BH7236AF pins 21-23. This PCB is available here: https://etim.net.au/shop/shop.php?crn=203&rn=549&action=show_detail
The circuit for the -5v regulator is based off the PTN78000AAH datasheet as well. A DPDT switch controls power to the regulator and selects the -5v source connected to the JAMMA connector.
I realize some things were missing or done differently in my PCB design. There is no modifiable drill layer in the Pad2Pad software I'm aware of, so rectangular holes were created with the board outline layer. I tried to make all the components as footprints with part #s in the BOM but I didn't get all the connectors and switches. There's no pick and place layer because I ordered the PCBs and assembled them by hand.
One possible improvement I was thinking of but never got around to implementing was swapping the THS7316 with a THS7374.
Attached is all the schematics I have. The Parsec design has changed a few times, and I never did make a complete schematic of the entire supergun.
The circuits for the BH7236AF, EL1883, and THS7316 are all based off their datasheet example circuit. The RGB input from the JAMMA connector goes into the attached hand drawn circuit. RGB output from the circuit goes to the THS7316 and BH7236AF RGB inputs. The BH7236AF is fed a 3.579545 MHz clock from an oscillator and configured for NTSC luma trap. The sync signal from the circuit passes through the EL1883 sync separator and goes to the composite sync output and sync input of the BH7236AF. A LM1881 sync separator can be substituted if the EL1883 isn't available.
Component video isn't generated on the PARSEC; a separate PCB is required available from Viletim as a component add-on for the NESRGB. The PARSEC only has the connectors and RGB signals tapped off the BH7236AF pins 21-23. This PCB is available here: https://etim.net.au/shop/shop.php?crn=203&rn=549&action=show_detail
The circuit for the -5v regulator is based off the PTN78000AAH datasheet as well. A DPDT switch controls power to the regulator and selects the -5v source connected to the JAMMA connector.
I realize some things were missing or done differently in my PCB design. There is no modifiable drill layer in the Pad2Pad software I'm aware of, so rectangular holes were created with the board outline layer. I tried to make all the components as footprints with part #s in the BOM but I didn't get all the connectors and switches. There's no pick and place layer because I ordered the PCBs and assembled them by hand.
One possible improvement I was thinking of but never got around to implementing was swapping the THS7316 with a THS7374.
Attachments
Thanks for posting all of that. What chip does the Viletim board use? I would prefer to integrate it in the redesign. I'm going to get started on the schematic in EasyEDA.
Edit: Found it (link). It's a SGM8052 op-amp. Here's a hasty attempt at reverse-engineering it based on the photos since it seems to be out of stock at the moment. This might have mistakes since I can't see under the chip or the board.
Edit 2: I think I might have a basic idea how the board works now. It seems to be doing some analog math. According to WolframAlpha's simplification of the derivation of Pb and Pr from RGB using Rec. 601 constants:
Cb = 0.5B - 0.331264G - 0.168736R
Cr = 0.5R - 0.0813124B - 0.418688G
So I believe it is probably doing something similar to this math using the op-amp and resistors. How such a mess of resistors was solved for though, I don't have a clue. Interestingly, it seems to be taking a bit of a "shortcut" by not subtracting the Blue component from Cr (I guess since it has a small coefficient) which probably results in bluish pixels being slightly redder than they should be. This is assuming the schematic is accurate, though. If someone has one of these boards, please post a picture of the back. It would help me a great deal.
Edit 3: Big thanks to samplehunters who reached out and gave me some very helpful information about the board. I've replaced the schematic with a revised one which I made with his help.
Edit: Found it (link). It's a SGM8052 op-amp. Here's a hasty attempt at reverse-engineering it based on the photos since it seems to be out of stock at the moment. This might have mistakes since I can't see under the chip or the board.
Edit 2: I think I might have a basic idea how the board works now. It seems to be doing some analog math. According to WolframAlpha's simplification of the derivation of Pb and Pr from RGB using Rec. 601 constants:
Cb = 0.5B - 0.331264G - 0.168736R
Cr = 0.5R - 0.0813124B - 0.418688G
So I believe it is probably doing something similar to this math using the op-amp and resistors. How such a mess of resistors was solved for though, I don't have a clue. Interestingly, it seems to be taking a bit of a "shortcut" by not subtracting the Blue component from Cr (I guess since it has a small coefficient) which probably results in bluish pixels being slightly redder than they should be. This is assuming the schematic is accurate, though. If someone has one of these boards, please post a picture of the back. It would help me a great deal.
Edit 3: Big thanks to samplehunters who reached out and gave me some very helpful information about the board. I've replaced the schematic with a revised one which I made with his help.
Attachments
Last edited:
I'm getting close to putting the finishing touches on a redesign. How would people feel about me bundling a multi-rail Mean Well PSU with a plug on it (a-la the Cbox Mini) and raising the price by ~$40? I was planning on charging somewhere around $100-120 for the fully assembled board so that'd put the final unit price somewhere around $150. $40 is about what a half decent ATX power supply costs so the total cost of ownership wouldn't be much higher. These numbers are preliminary but definitely won't go any higher.
Thoughts?
Thoughts?