Successfully saved a nofix GSCARTSW. A few questions

The place for all discussion on gaming hardware
Post Reply
halfspec
Posts: 9
Joined: Wed Dec 08, 2021 3:26 pm

Successfully saved a nofix GSCARTSW. A few questions

Post by halfspec »

This is a year too late for the gscartsw support thread, but I’m hoping I might lure @superg back (or someone else smarter than me) by offering my eternal thanks and being as concise as possible to save time:

TL:DR

I bought a cheap broken GSCARTSW lite 1.5 off of ebay that turned out to be a nofix from a prominent member of r/consolefix. The root cause of the issue was a short from blue/green multiplexors U1/U2, but I also had to deal with the mess the previous owner made of the power stage(s) while trying to debug. I’m happy to say the unit is working 100% now, but I’m hoping I may get some insight / IDs of the following two components:

#1 – TVS Diode D1 – superg refers to this part in this thread as SMAJ5.0A-13-F. However, there is a problem with this part that I’ve encountered.
With the stock 9VDC PS, the TVS clamps the input to 7.2V and sends it on to the 5V buck circuit but it is sinking enough current to make the SMAJ5.0A-13-F boiling hot to clamp @ 7.2v. The unit operates like this and the fuse never opens, but I can’t imagine this is correct. Comparing pictures (see below) of what I’ve got (SMAJ5.0A-13-F) and other peoples’ gscartsw, it appears the SMAJ5.0A-13-F is comically small for the PCB footprint, therefore I think I’m dealing with a misunderstanding / mis-identification of the part. Any ideas for what the specs for the original TVS are?

#2 - L3, L4, L5, L7, L8, L9 inductors / chokes. I don’t have a LCR meter, and while these are all working, the PO was really rough on them and they look like they’ve been cooked 10x over. I’d like to replace them for longevity reasons but also for aesthetics. Anyone have good information on the specs of the OG inductors?


Long version:

I bought a cheap broken GSCARTSW lite 1.5 off of ebay hoping for a simplish multiplexor replacement. The seller ended up being a prominent member on r/consolerepair/ and I was disappointed to find they picked apart a large portion of the power stage(s), failed to fix the issue, and hastily put it back together [incorrectly] for the sale. I suspect they never got to the multiplexors due to not being familiar with the supergun spec violations that kill them, or wouldn’t wait for parts availability.
I’m happy to say, the root cause of the issue was a short in blue / green multiplexors U1/U2. I also ended up replacing all of the SP6661 charge pump vregs because I found that the previous tinkerer incorrectly reworked one 180 degrees out of alignment and I didn’t trust any of them after that.

The initial symptoms of the switch when it arrived post purchase was no LED, ~1v on the 5v rail, and a blistering hot U17 charge pump vreg.
U17 was found to be 180 degrees out of alignment so that was replaced along with U18 and U19.
With the vregs fixed, I started removing the chokes to determine what was shorting the 5V rail to ground. I determined that the multiplexors U1-U4 behind L3 & L7 were responsible. After waiting ~2 months for mouser to restock, I received my replacement 78-DG4051EEQ-T1-GE3 multiplexors and started reworking. ~100ohm short was still present after removing blue multiplexor U1, but gone after removing green multiplexor U2. U1 may have also been shorted, but U2 definitely was because after removing these two components, I got the 5V rail back with L3 & L7 installed + the return of the LED!
Replacing U1 and U2 netted a fully working unit with one caveat…

I noted the current TVS diode D1 was getting hot enough to melt plastic which in this case was the top of my NES frontloader the GSCARTSW was sitting on while testing! No sweat though. It was my fault and all of my consoles are just beaters, so no worries. Anyway. D1 was one of the components the previous ‘fixer’ replaced and they also included a few spares with the sale. Unfortunately I don’t have the OG TVS, hence this line of questioning. It appears correct for the application since sugerg mentioned the same part number this as SMAJ5.0A-13-F when concerning another gscartsw lite. It’s also oriented correctly with the cathode connected to the fuse output and the anode connected to ground, which correctly configures it as a ~7.2V clamp that will sink current for voltages higher, with the extreme being PS > 12V which in theory should sink enough current to open the fuse and save the downstream components… But I digress..
I never got a temp off of the TVS because I stopped running it for appreciable amounts of time and eventually moved to a 7V PS (since the GSCARTSW Lite 1.5 specs its acceptable voltage input as 7-12VDC) and removed the TVS entirely since @ 7V there shouldn’t be any reason for clamping or transient protection (aside from like lightning). However, it was getting hot enough to smoke flux. Additionally, while still installed, I measured that it was clamping the 9VDC input to 7.2V before sending it on to the downstream 5V buck converter (U15 I suspect, but my eyes aren’t good enough to read anything off of that tiny mofo!).
The biggest standouts to me as problems are:
#1 - The size of SMAJ5.0A-13-F. It matches the datasheet’s size specs for the part being ~2mmx4mm so I don’t think this is a mistake and the previous fixer truly sent me SMAJ5.0A-13-F diodes vs a mixup from his/her bench but comparing it to other people’s GSCARTSW lite D1’s it looks comically small:

Mine:
Image
[please don’t judge me too harshly. I’m not done cleaning up the flux, and the diode is simply sitting on top of the pads for the picture / not soldered]

Others:
Image

This makes me think the SMAJ5.0A-13-F is actually for another variant of GSCARTSW. Perhaps the newer version that uses a 5v / microusb input for power. I suspect this because my experience with discrete components like this is that larger parts are large to dissipate more heat. If this TVS were 2-3x larger, it might not get much hotter than ambient. Similarly, I’ve always used TVS diodes as clamps for big transients… Not a full time clamp like this appears to do. That leads me to the other standout.

#3 – Clamp voltage. The SMAJ5.0A-13-F’s max output clamped voltage is 7.2V and its max input voltage is 9.2V. SuperG includes a 9VDC PS with the GSCARTSW, but the PCB says it can take up to 12VDC. 12V would exceed the SMAJ5.0A-13-F’s max voltage input and seeing that it is roasting with the 9VDC input, I’m sure it would trip the fuse and/or kill the SMAJ5.0A-13-F with 12V input.
Additionally, TVS’s aren’t meant for full time clamping duties. They’re supposed to clamp voltages the power input stage is never supposed to see. I.E., If the GSCART is rated for an input of 7-12V, I’d expect the TVS to clamp voltages over 13V… Not voltages over 7V. That just means, anytime you apply power with anything over 7V (actually 6.4V if you want to get nitty gritty), the TVS is sinking current to clamp the voltage. That’s what makes me think the SMAJ5.0A-13-F is actually intended for a version of GSCARTSW that normally only sees a 5VDC input like superg’s version that utilize USB power input. That would make sense for the SMAJ5.0A-13-F since it would clamp voltages over 6.4V which it should never see when powered by USB.

Like I mentioned earlier, currently, the suspect TVS is removed and I’m running off a 7VDC PS just fine, but I’d like to return the power input circuit to its original form + utilize the OG 9VDC power supply in the future, which is why I wanted to ask these questions.
Thoughts?

Finally. It’s not as important, but I was wondering if anyone has the specs / part numbers for the - L3, L4, L5, L7, L8, L9 inductors / chokes. I don’t have a LCR meter, and while these are all working, the PO was really rough on them and they look like they’ve been cooked 10x over.

Any help is greatly appreciated.

Lane
User avatar
superg
Posts: 1597
Joined: Sun Jul 06, 2014 4:31 pm

Re: Successfully saved a nofix GSCARTSW. A few questions

Post by superg »

#1 D1 stock part: http://www.digikey.com/product-detail/e ... ND/3042727
#2 L2-L10 stock part: http://www.digikey.com/product-detail/e ... -ND/806759

#2 are ferrite beads, they are applied just to be able to easier troubleshoot multiplexer issues e.g. you can even short these pads if needed.
#1 advice: No way to tell what was done with the switch by the previous repairman but in general if you see a big voltage drop and it gets hot, replace all the multiplexers that you haven't replaced. When excessive voltage breaks through some multiplexer it travels through the power rail and damages the other up to some extent, I speak out of past experience repairing these. It might look normal and even operate somewhat normal and still be damaged.

EDIT: sorry switches availability suck at the moment, there is still some parts shortage and there is an ongoing attempt to switch to a different FPGA.
halfspec
Posts: 9
Joined: Wed Dec 08, 2021 3:26 pm

Re: Successfully saved a nofix GSCARTSW. A few questions

Post by halfspec »

superg wrote:#1 D1 stock part: http://www.digikey.com/product-detail/e ... ND/3042727
#2 L2-L10 stock part: http://www.digikey.com/product-detail/e ... -ND/806759

#2 are ferrite beads, they are applied just to be able to easier troubleshoot multiplexer issues e.g. you can even short these pads if needed.
#1 advice: No way to tell what was done with the switch by the previous repairman but in general if you see a big voltage drop and it gets hot, replace all the multiplexers that you haven't replaced. When excessive voltage breaks through some multiplexer it travels through the power rail and damages the other up to some extent, I speak out of past experience repairing these. It might look normal and even operate somewhat normal and still be damaged.

EDIT: sorry switches availability suck at the moment, there is still some parts shortage and there is an ongoing attempt to switch to a different FPGA.

WOW! I got a reply! Thank you so much! I posted not expecting anything but this is a fantastic help!

The SMCJ12CA makes soooo much more sense with a 19.9V Clamp! I know from an electrical standpoint this will completely solve the issue I was having with the mistaken replacement using the SMAJ5.0A-13-F.
Nothing else on the board is even warm and every input / output works, so I don't suspect anymore damage. I didn't mention it earlier, but I did replace U3 and U4 as well. Audio works perfectly, so I didn't mess with U5 and U6, but I'm not the kind to turn down good advice. I'll replace them as well.

Thank you again for taking the time to reply. I really appreciate it and I'm a huge fan of your work.

Lane
halfspec
Posts: 9
Joined: Wed Dec 08, 2021 3:26 pm

Re: Successfully saved a nofix GSCARTSW. A few questions

Post by halfspec »

Just wanted to follow up and offer my thanks again.

I replaced the TVS, ferite beads, and U5 and U6 to net a much healthier gscartsw.

Put a pin in it because it's done :D

Image

Image

Lane
Post Reply