PS3 (Research/Experimental) - NEC/TOKIN Capacitors Replacement - YLOD

Can you please provide the following:
  • /QUOTE]
Sure, it's a CECH-A01 and the YLOD consistently happened within 2 seconds of every bootup. The 6 mOhm capacitors I used are "EEF-LX0E331R".

The the 4.5 mOhms capacitors i'm considering are "EEF-LX0E331R4"

Neither of these caps are tantalums by the way, I figured if Sony wasn't strictly using tantalums then I see no reason as to why I should either.
 
Tantalum may be better because it has lower ESR in the higher frequencies. Here, this graph compares the ESR of your AlPol caps to the tantalum that SONY initially used (2R5TPF470M6L 6mOhm 470uF TaPol) before switching to AlPol.
AlPol vs TaPol.PNG

I'd go with the EEFLX0E331R4 if choosing between the ones you listed. The 330uF allows you to get closer to the ideal 4800uF target at 5280uF for 16x of them...and the ESR would be lower too (4.5 / 16 = 0.281mOhms). They should last longer and be able to handle higher ripple because of that, but any noise or ripple above 12MHz will not perform as well as tantalum. BTW, the NEC/TOKINs are quite good, they have low ESR into the HF also. Even better than TaPol, which is why they are still the best option. Basically, HF noise gets coupled into the circuit from the processor, which has to get past the MLCC Bypass caps first. Once the noise gets to the AlPol Bulk filter caps, it will have a see a higher ESR and heat the cap more than a TaPol would. A-E models still have an array of 36x 0.1uF MLCC bypass caps on the CPU. They were later removed with the switch to 65nm Cell_BE and decrease from 3 switch mode converters to 2. So my guess is that the 65nm Cell had much less noise, pulled significantly less current, and required fewer noisy switching VRM. That allowed them to remove some of the HF bypass caps. All of which reduced cost. The tokins job got easier as the current being drawn across them was smaller, and thus the ripple would be smaller.
 
Last edited:
yesterday I changed the nec/tokin cap with 470uF 6.3v tantalum, at first I am so happy that it gave green light.. but the problem is it stays green with no video/sound.. is this GLOD?
 
yesterday I changed the nec/tokin cap with 470uF 6.3v tantalum, at first I am so happy that it gave green light.. but the problem is it stays green with no video/sound.. is this GLOD?
Please provide the following:
  • PS3 Model
  • Model # or Link to the Capacitors you used and how you installed them. Did you use hot air? Bridge wires? How many of what gauge? And so on. Please be specific.
  • Type of YLOD you initially had:
    • Warm Start (Only boots when warm)
    • Cold-Start (Only boots when cold)
    • Instant (<1s)
    • Non-Instant (1 - 10s)
    • Delayed (10s - 5mins, YLOD occurs after the console warms up or unstable in menu and dies at the drop of a pin. May be followed by a cold start YLOD)
    • Random (can occur anytime under normal stress, like games, movies, streaming apps, but is usually stable in menu)
    • Intense (High Stress, same place in certain demanding games like TLOU or GT6
To answer your question, yes it could be the GLOD. Or the Video might need to be reset. Often times the PS3 will default to the AV out instead of using HDMI. You didn't say how you hooked it up, or to what. Have you tried using an AV cable? When you plug in a PS3 controller via USB, does it sync?
 
I have CECHA first gen model. Yes initially the PS3 had YLOD problem, 5 seconds delay before it beeps and flashing red light. Then I follow the tutorial from NSC youtube to change the caps with a yellow tantalum 470uF 6.3v. I replaced 2 NEC/TOKIN on the RSX with 8 tantalum caps. i didnt use bridge wire since I dont replace all the NEC/TOKIN. After I reassmble the machine, turned it on, it gave green light without video or sound. Even the controller doesnt sync. Tried to reset video several times and used AV or HDMI to no avail.
 
Get to work @RIP-Felix, you think I don't see that board was delivered?!? HURRY UP!
I'm going to call this one PS3#7
EDIT: This was originally found by @squeept. He wrote about it here first. He traded it to me in a Motherboard exchange for PS3 #4, which sadly didn't survive the reball attempt.

Destroyed the case on that inductor (resistance reads okay), crushed C6056 (2.2uF filter for Vo +1.8v RSX_PLL_VDD, and part of one tokin on the top. The tokin on the bottom side is compressed, the plastic is broken a bit but it's in better shape. I took the plastic off the topside tokin and the one in the pic is smashed in that corner, but the resistance +/GND is about 2.3Ohms, so it may still be okay for testing...IDK yet.

The 2-pin main connector is holding on for dear life and bent out of shape! I pressed it back in and while the old clips broke off, at least it's back in the right position. The whole board has a warp to it now. Man they really did a number on it...what a shame! Of course the damage had to be right over the RSX! Lucky number seven..lol!

On a better note, the PCB I ordered from OSH park arrived today too. Good timing! I'll probably test this weekend, so we'll see how lucky #7 is...

Continued in part 2 here...
 
Last edited:
Ow, well good thing Sony built these systems to last. (Remember each console was often sold at a loss) I wonder if they made the boards karate proof too? (Or maybe they only tested that inside the RF shield and the rest or the case)

Hopefully no permanent damage and the tests will still be unaffected, but notice there's also an SMD missing (and probably the one next to it took a beating too?) So I'd probably replace those before testing the board. Good thing is you should be able to quickly borrow whichever components from other boards


So have fun
 
hi
i need some advice according this thread that slowly nec tokin become degraded over time of ps3 use
so tomorrow im going buy used ps3 bc fat still sealed (or just fake seal) so my plan is to change the thermal paste and delid
so if i want make this ps3 like fresh or just wanted to last longer shall i change all the tokin with same new tokin as ps3 have ? since people here doesnt know how to change it to tantalum
 
Lol, and I told you not to re-use a medium flat rate box because they're a bit flimsy to begin with...

I may have inadvertently screwed us out of an insurance claim. Remember that $1 eBay item you bought from me so I didn't get your address messed up? That's now the receipt that values that board at a single dollar... I'll still try to file and see what happens and I'll send you anything they give me.

I have a good feeling it will still work. The impacted BGA chip has nice soft lead solder now, I've scraped off some TOKINs that looked more mangled than that that still tested working, and the stuff that busted off of the coil is just cement encapsulation.

Either way, the pile of CECHA01 keeps getting bigger since I'm slow, so I should be able to find another one soon.
 
...so if i want make this ps3 like fresh or just wanted to last longer shall i change all the tokin with same new tokin as ps3 have?
If there is nothing wrong with them, leave them alone! They can go bad, just like any capacitor, but that doesn't mean they are defective or prone to premature failure. I have not been able to find any evidence to support that idea. If it were true there should be some information about it online, but as of yet all I have seen are echo chambers parroting the the same claims about PS3 and Laptop tokins...anecdote only. If this were a widespread defect, there would be academic and research materials explaining the failure mode. I have been looking, but can't find anything. No lawsuits or arbitration claims...nothing but unsubstantiated claims. I'm leaning towards it being a myth.

A Retrospective Analysis of the NEC/TOKIN Proadlizer (Part 2)
...(continued from page 168)
With continued research I found that NEC/TOKIN rated their proadlizers at 125C for 1000 hrs, but NEC didn't really specify the low temperature threshold (1). Nipon Chemi-Con on the other hand said "when the devices are restored to 20C after the rated voltage is applied for 1,000 hours at 105C" (2). Some PS3's use those caps instead of the NEC/TOKINs and they are similar in performance, endurance and construction. I wouldn't expect them to have wildly different endurance numbers. Using KEMETs calculations for capacitor lifespan (3), I can estimate how long the tokins should last using real PS3 temperatures. That should give us a better estimate.
For Al Electrolytic: L = F x 2^[(T0-Tamb)/10]
For Solid Polymer Al: L = F x 10^[(T0-Tamb)/20]

L is the calculated Lifespan based on the temperature difference expected. F is a multiplication "Factor" I will need to figure out based on the endurance rating in the datasheet. First we need to find F for each the electrolytic and polymer, since proadlizers are actuallu a hybrid between the 2 technologies. That will give us a range. So let's plug in the numbers from Nipon Chemi-Con's proadlizer endurance specification to figure it out:

Flytic = L / 2^[(T0-Tamb)/10]
FPoly = L / 10^[(T0-Tamb)/20]

Flytic = 1000 hours / 2^[(20°C - 105°C) / 10] = 362038.672
FPoly
= 1000 hours / 10^[(20°C - 105°C) / 20] = 17782794.1

Now that we have F for both electrolytic and solid polymer aluminum capacitors, I can plug in more realistic temperatures. Let's say for now that room temperature is 20°C and the hottest the caps will get is 85°C. That's a more realistic average.

Llytic = F x 2^[(T0-Tamb)/10] = 362038.672 x 2^[(20°C -85°C)/10] = 4000 hrs
LPoly
= F x 10^[(T0-Tamb)/20] = 17782794.1 x 10^[(20°C -85°C)/20] = 10,000 hrs
Using these calculations, a proadlizer should last between 4000 and 10,000 hours. Now, the tokin are closer to a solid polymer aluminum capacitor than they are to regular Aluminum electrolytics. Therefore they should last closer towards 10,000 hours than the 4000 hour mark.

I would like to measure the actual temperatures on different models of PS3, so we have an accurate number to use in the calculations above, but since I only have A models that's the best I can offer. You may have noticed that Aluminum electrolytics have 4x the expected lifespan for every 20C decrease in delta T, Polymer have a 10x. I chose a 20°C decrease because...
capacitor-life-vs-temp-png.30338

As far as I can tell, the NEC/TOKINs are meeting specification. Based solely off what others users have reported in this and the SYSCON thread, consoles seem to experience the YLOD in the range of 150-300 days uptime (3600 - 7200 hours). I would actually like to have large dataset of confirmed YLOD and the amount of uptime they had, but the anecdotal reports from a couple of users is all I have to go off of ATM. Interestingly, from videos I've watched on CPU simulations of mean thermal cycles to BGA failure for SAC305 lead free solder balls, 3600 - 7200 thermal cycles doesn't seem unreasonable given SONY's insane fan curves (4). Just an observation.
SONY's engineers modeled all of this math in computer simulations to be sure your console lasted beyond the 1 year warranty, but wasn't going to spend more money building extra reliability in. They even prioritized sound performance over thermals that could have easily extended the lifespan of your console! That's how ruthless they are! That's the price YOU pay when SONY and MS chase ray tracing, and graphical performance instead of prioritizing thermals and reliability. Nintendo has always prioritized gameplay and reliability over these considerations. Perhaps that's because they are primarily a video game company and SONY/MS are not. Nintendo can't survive a bad reputation, whereas SONY and MS can drive their video game divisions into the ground and still make out fine. The PS4 wasn't free from the BLOD, why would you expect the PS5 to be different? At least MS seems to have learned it's lesson after the XB360 billion dollar recall. Who knows, maybe the solution was liquid metal and vapor chambers all along. Time will tell!

The bottom Line:

Assuming a person played 8 hrs/day every day on their PS3, it would take 1.23 - 2.5 years to experience the YLOD. Most hard core gamers don't clock anywhere near that kind of time, let alone the average user. 'In the United States, the average gamer spends 22 hours per week playing video games' (5). At that rate your PS3 should last 3.15 - 6.29 years. How does that stack up to the calculated lifespan of 4000 - 10,000 hours? The proadlizers should perform to specification for 3.5 – 8.7 years, probably closer to 8.7 because they are closer to Solid Polymer Aluminum than they are to electrolytic caps. In other words, the math suggests that we should be seeing tokins go bad now that we are 15 years into the PS3's lifespan. However, all of this is HIGHLY dependent upon the tokins getting to 85°C during use. If it's closer to 65°C then they'll last far longer. Basically you need real measurements for your model PS3 to get a realistic calculation.

Also, using the equation above, you should be able to estimate how long your tokins have left based on the uptime on your console. Just note, this number is based on the specification guarantee, not when you'll actually experience a YLOD. It'll probably last longer.

@squeept you may find this interesting. The equation above for solid polymer says that the tokins should still read within specification after 7.4 days in your drying oven @120°C. After that they are not guaranteeing they'll perform to spec. If you were to put them in for that amount of time and test them on a known working console, they should still work. If you did the same for another set that had been in for 10 days, or 15 days, & 20 days, and so on, we could pinpoint the amount of time needed to cause a YLOD. That number could then be used with the equations above and the consoles uptime to estimate how long the console has until it's tokins fail (theoretically).
 
Last edited:
Just finished installing 32 "EEF-LX0E331R4" caps and 8 bridges in another PS3, this one also had a YLOD after 2 seconds every bootup. Model was CECH-E01 and the GPU and CPU both had flux residue around them, so i'll assume they were either reflowed or reballed in the past. Firmware was on OFW 4.80 so i'll assume it died before late 2016.
 
PS3#7 Update
...continued from part 1 here.

Today I examined the board damage more closely.
BOX Dammage.jpg
BOX Dammage 2.jpg
Tokin & Inductor damage.jpg
L6202 Dammage (Inductor).jpg
L6305 Scrape (Inductor).jpg

Note that these two inductors have scraping damage and the bottom of the box has inductor sized holes where it looks like they punched through as the box was scraping along a metal slide. My guess is that a heavy box was riding on top the USPS medium flat rate box, crushing it, grinding away the cardboard and packing until the inductors themselves were in contact with the slide/chute. That explains the inductor damage. Now, the inductors themselves are actually taller than the RSX, So they may have saved this console's life!

Moving on... Here is the tokin damage on the topside. Notice the circular impression left by whatever hit the board.
C6231 Tokin Damage (1x).jpg
C6231 Tokin Damage (10x).jpg
C6231 Tokin Damage (40x).jpg


For reference here is a closeup of the undamaged upper right hand corner. I flipped it so it would look like a before/after:
C6231 Tokin Undamaged (40x).jpg


Clearly whatever hit here did so HARD AF! You can see in the first picture the smush pattern looks like it came in at an angle an bounced off. It was obviously moving in the direction of the MLCC caps because it pulverized C6056 and took a gash out of C2136. Looks like the Tokins absorbed the biggest impact, but the MLCC caps got some of the fury as well...
C2136 & C6056 Damage (Before).jpg
C2136 Damage (Short).jpg


I spent the morning replacing them from PS3#1's carcass board. Here's the repair...
C2136 & C6056 Damage (Repaired).jpg
Anyway, there's nothing I can do about that tokin. I can remove it and replace it, which I intended to do all along. I guess one silver lining is that the part we knew was bad and I was going to remove anyway, is what got hit. Way to take a grenade NEC/TOKIN!!!
:grenade:
...except that was the whole point of this board. If you remember this was the motherboard that @squeept reballed and had a YLOD while stress testing. It would only occur under load, so we wanted to get good measurements of the tokin noise for the sake of science. Well, I guess the USPS had different plans.

I took the opportunity to fully clean the flux residue under the RSX and surrounding area off. I don't have an ultrasonic cleaner like squeept does, so I made do with CRC QD Electronics Cleaner for the residue under the chip, followed by 99% IPA, a tooth brush, and plenty of paper towels. I can't test until it's had time to fully dry. So I'll save that for tomorrow.

Speaking of cleaning, you know those VIA's near the Tokins?
Thermal VIAs.jpg

Notice the flux residue inside them? Well those are actually thermal vias and they are an important part of the console's thermal design. What they do is isolate heat from the processors from reaching the TOKINs. They are supposed to be clear so that air can help cool the TOKINs. When flux or solder gets inside it reduces the thermal performance of that area, meaning that the tokins or any caps placed there will heat up more than they would have otherwise. So they won't last as long. There's no problem with it electrically though, so for testing I don't mind leaving it there. But if you're going to sell it to a customer, that is one good reason to invest in an ultrasonic motherboard cleaner. I have one at work, but it's too small for this sized board. Luckily there is a poor man's solution. Sewing needles and thread! Thread it through and use some IPA, the thread will soak up the IPA and you can pull it through, back and forth, to clean them up. Shine a light through to see if they are clear.

Continued in part 3 here...
 
Last edited:
@RIP-Felix I've been meaning to do a full probe of all my heater plates to check that everything is even, which because I only have a couple of thermocouples means running 20 or so full cycles and poking around the whole time. That will give me a good opportunity to harvest a crap ton of working TOKIN while I do that. I'll put them all in the oven, which I run at 100C, and start pulling them out at varying times over the next month or two and marking how many hours they were cooking. Next time I have to change any out, I can test them all out and see what we get.

I didn't give it a bath because then I'd have to bake it again, and I figured I didn't want to do anything that could change the current state of the board. Little did I know what the USPS had planned...

Either way, I'd leave the TOKIN as is and give it a shot still. Even if one or two are dead, it should still run. And the CPU was looking just as bad with that sine wave peeking through, so those should still be good for testing if the board is alive.
 
PS3#7 - Initial Testing
(...continued from part 2 here.)

This morning I assembled my test bench and captured a bunch of Oscilloscope images. First, the console now has a 4.5s YLOD. I'm not surprised, considering the damage it received in shipping. At least it's not an instant YLOD!
4.5s YLOD Event (Normal Acq.).png

Here is the startup sequence with normal acquisition mode (showing all the tokin noise):
Startup Seq. (Normal Acq.).png
There are supposed to be discrete steps in the voltage here. You can see that the noise is severe enough to obscure the steps somewhat. The RSX noise (Blue) is much worse than the Cell_BE CPU (yellow), but it's pretty bad too. Here it is with HiRes Acquisition mode enabled, which will average many of the samples and should result in less noise obscuring the average signal:
Startup Seq. (HiRes Acq.).png
Now we can see the discrete voltage steps that are part of the normal power up sequence. The last plateau, where the voltage drops, is the last step before the console makes post, begins loading code, and boots up. I like to trigger on the plateau just before this, because my YLOD consoles that had BGA defects resulting in 0.5 -1.5s YLOD's wouldn't reach this last plateau. Here is same picture from PS3#4, 1.5s YLOD with a BGA defect that I traded to @squeept for PS3#7.
PS3_4_3.png
And here is the same from PS3#3, which is a normal working console:
Working_NEC_RSX&BE.png
You can see there should be a lower voltage plateau, if the console is making it past POST. PS3#7 does! So that's further than my previous YLOD console's have ever gotten. That's a good sign that the console does not have a BGA defect. Now let's zoom in on the second to last voltage plateau. This way I can compare the NEC/TOKIN noise to the other consoles I've measured at the same point in the startup sequence.
RSX Bad Tokin Noise (DC Coupling 10x probe).png
CPU Bad Tokin Noise (DC Coupling & 10x probe).png
RSX Bad Tokin Noise (DC Coupling & 1x probe).png
CPU Bad Tokin Noise (DC Coupling & 1x probe).png
The RSX Noise measures about 120mVpp @450KHz and the CPU is about 50mVpp @480KHz. From what I've speculated before, 50mVpp might not be enough to cause a YLOD. So the CPU might be okay.

It's interesting that the CPU noise looks like a rectified sine wave, whereas the RSX noise is a triangle (saw-tooth) wave. The RSX has definitely switched over to the "BAD" waveform that @squeept identified for the RSX. The CPU is definitely the BAD form too, but it's Vpp isn't nearly as bad yet. Let's look at @squeept's image of the RSX noise before he shipped it to me. It was having the YLOD only under stress testing before. So I want to compare the Vpp before (high stress YLOD) and after (4.5s YLOD):
5Nd9s6e.jpg
siGDTUY.jpg
Before @squeept sent me the board the the RSX tokin noise looked to have been about 30mVpp @ 2MHz and the CPU tokin Noise about 40mVpp @500KHz. After the USPS damage, the RSX Noise measures about 120mVpp @450KHz and the CPU is about 50mVpp @480KHz. The difference seen in the CPU noise is about what I would expect given his settings vs. mine, so I think that hasn't changed much. I don't see any damage to the CPU tokins, so I think they are in the same state as they were when squeept sent the board. The RSX tokins are clearly damaged now, they are absolutely causing a YLOD.

I forgot to add the SYSCON wires when I was repairing the board, so I'll edit back later after I do that...Okay, here it is:
Initial SYSCON.PNG
Notice that the only error is the 1002! Well that is what we've now confirmed to be Bad RSX tokins (RSX VRAM Power Fail). So the scope images confirm what the SYSCON is telling us. Also, notice that the lasterrlog only shows the 80 1002! Well 80 is a power on system state, meaning the error occurred after the powerup sequence, like the o-scope suggested. If the error had occured during the powerup sequence, then the lasterrlog would show the specific step in the POST process it failed. PS3#2 and 4 were getting bittraining errors and showed the full startup sequence up to when the error occurred. PS3#7 is getting past all that, so it doesn't show in the lasterrorlog. That means @squeept did a good job reballing! They are holding tight my man!!!

Alright, we can breathe a big sigh of relief. The console is fine, I just need to replace the bad RSX tokin. I think tomorrow I'm going to remove the one that was struck, then test to see if the console will boot. @squeept hypothesized that a bad toking spoils the bunch and this USPS disaster actually may prove that! This is the only tokin with visible damage. If that's the one that's bad and I remove it, the noise may actually get better. If so, then it proves 1 bad tokin spoils the bunch. If the USPS hadn't pinged that tokin, I wouldn't know which one it was...friggn crazy. Until tomorrow.

Continued in part 4 here...
 
Last edited:
I never captured anything clear enough to see that was a rectified sine wave. Thought it was just a regular sine, neat.

I'd also try booting it up after you remove the damaged one, but before you add new parts. That'll be pretty concrete evidence on the apples theory if it fires up. And my other pet theory on failing spot welds... kind of fits here since this was physical damage to the cap? Eh.

Anyway, at this point, I think anyone that frequently checks in on this thread needs to start VIOLENTLY berating anyone from now about getting the damn adapter and checking the syscon. The oscilloscope angle was never going to be practical for a random person on the street doing one console, but there's just no excuse with the little guy being $5 shipped.
 
PS3#7 Continued...
Alright, we can breathe a big sigh of relief. The console is fine, I just need to replace the bad RSX tokin. I think tomorrow I'm going to remove the one that was struck, then test to see if the console will boot. @squeept hypothesized that a bad toking spoils the bunch and this USPS disaster actually may prove that! This is the only tokin with visable damage. If that's the one that's bad and I remove it, the noise may actually get better. If so, then it proves 1 bad tokin spoils the bunch. If the USPS hadn't pinged that tokin, I wouldn't know which one it was...friggn crazy. Untill tomorrow.

neat.

I'd also try booting it up after you remove the damaged one, but before you add new parts. That'll be pretty concrete evidence on the apples theory if it fires up. And my other pet theory on failing spot welds... kind of fits here since this was physical damage to the cap? Eh.

there's just no excuse with the little guy being $5 shipped.
This is all amazing! Great work, keep at it.

Now quickly let me poke my head out again about the apple thing, and remind you to maybe, just even before taking that mangled tokin off... To try adding the additional tantalum(s) in parallel. See what happens.
Simply because this will be subtle and reversible, whereas the removal will not.
Then you can carry on with the following tests like of course powering on with the mangled one missing etc

Cheers
 
Last edited:
To try adding the additional tantalum(s) in parallel. See what happens. Simply because this will be subtle and reversible...
Well, too late now. I got impatient and removed it already. I'm not done testing yet but the syscon still shows a 1002 and the YLOD occurs faster now (~2s). I'm going to get some O-scope images of the noise shortly. Then I was going to try removing the tokin on the opposite side of the board (underneath the one that got struck), because that one's plastic was cracked. It looks okay otherwise, but maybe not. Not looking so great for the one bad apple spoils the bunch theory, but I haven't looked at the scope yet.
 
80 is a power on system state, meaning the error occurred after the powerup sequence,

I never paid much attention to that portion of the error code (mostly because I never made it that far), but I wonder if we can also make progress on documenting what the steps map to, specifically the prefix. Looking at some of my previous notes, I had a 09-3003/09-3004 when I removed the cell and rsx bridge wires, respectively, but putting them back on resulted on a 40-3034 (which we now believe is bad rsx solder joints). However, i'm pretty sure that people getting 1001/1002 we're getting a different prefix than you are getting with the 80-1001/1002 from your ps3 #7.

The obvious benefit would be to more quickly identify where the failure is happening, or why.

1002 and the YLOD occurs faster now (~2s)
oooh... can you get us the full error code?
 
Last edited by a moderator:
PS3#7 Continued...
(1 TOKIN removed, C6231. The damaged one.)
...continued from part 3 here.
SYSCON, 1 RSX TOKIN removed (C6231, most damaged).PNG
The first time I booted it gave the YLOD in 350ms (Instant YLOD). In the SYSCON that showed up as 10 1002. So it gave out before the startup sequence even had a chance to really get going. However, this is more rare of a YLOD in the current state.
350ms YLOD EVENT (infrrequent).png


Most of the time it makes it 3.8s before the YLOD. That's after the startup sequence. The SYSCON still shows a 80- 1002.
2.8s YLOD EVENT (more often).png
Startup Sequence (2.8s YLOD).png


Here is the RSX Tokin Noise with the damaged tokin removed. So now the RSX only has 3 tokins installed. The RSX noise has increased from 121mVpp to 163mVpp. It has increased by 42mVpp. That suggests to me that the damaged tokin was still helping to remove some noise. If there is a bad apple spoiling the bunch, C6231 wasn't it.
RSX Bad Tokin Noise (DC Coupling & 1x probe).png


And here is another measurement of the CPU noise. I didn't do anything to them, so I wasn't expecting it to change from the previous measurement. I forgot to change the cursor source of CH1 (CPU) so it's measuring CH2 (RSX). The scale of the RSX was set to 50mV/div. The cpu was set to 20mV/division. So BY-AY = 93mV is calculated from the wrong scale. I can multiply the 93mV by 0.4 to get the right number. It's actually 37.2mVpp, which is not significantly different than it was before - as expected. Frequency is about the same too.
CPU Bad Tokin Noise (DC Coupling & 1x probe).png

Continued in part 5 here...
 
Last edited:
Back
Top