Overclock.net banner
181 - 200 of 203 Posts
The actual value of the Vref-Refin circuit is 15k ohms, & the value of R52 is currently 33k ohms.

I'm not sure why the resistance of R52 is all over the place.

The card is currently not functional. I think I accidentally shorted the FB pin to the Comp pin with the flux, working on un shorting it
The flux is not able to short things.

But it is what will make the solder pads to un-short.
As you did in the previous picture, put a lot of flux, and try to just poke at the shorted pad delicately, with the iron tip.
Until these separate one to each other, you don't want to have solder on the iron tip, or just a microscopic amount.
Just to be able to poke the 2 pins covered with flux, reshaping themselves and getting them straight, without linking each other.

Good luck, it is not an easy task either, the VRM controller is quite small.
Patience and flux are key things when doing soldering job. (y)
Good job for finding out the previous issue.(y)
 
Discussion starter · #182 ·
The flux is not able to short things.
My tub of tinning flux would beg to differ!

So I was able to un-short the FB pin to the comp pin, and so the Feedback circuit is no longer preventing the card from starting.

However, as I suspected, having the Over Current Protection Divider (R52) double in resistance from 18.5k ohms to ~35k ohms is in fact a problem. When I attempted to boot the card in the system, I got four red beeps, which indicates power supply overdraw... I'm not going to speculate as to how much power this card just attempted to pull, but after taking the card out the system works fine.

Now I believe the only issue preventing the card from functioning is R52's increased resistance, so I will dig into the schematic and see what other SMD's may be impacting R52's resistance to see if they are shorted or have come loose.
 
My tub of tinning flux would beg to differ!

So I was able to un-short the FB pin to the comp pin, and so the Feedback circuit is no longer preventing the card from starting.

However, as I suspected, having the Over Current Protection Divider (R52) double in resistance from 18.5k ohms to ~35k ohms is in fact a problem. When I attempted to boot the card in the system, I got four red beeps, which indicates power supply overdraw... I'm not going to speculate as to how much power this card just attempted to pull, but after taking the card out the system works fine.

Now I believe the only issue preventing the card from functioning is R52's increased resistance, so I will dig into the schematic and see what other SMD's may be impacting R52's resistance to see if they are shorted or have come loose.
As i said, i'm not able to follow your steps and what you are doing anymore, unfortunately.
The flux is what allow you to remove and un-short pin tied together, which is almost impossible to do without flux, on such small packages pins.
By soldering flux, i mean this.
Material property Font Liquid Tints and shades Rectangle

And this is the usual tin lead-free based solder wire.
Cylinder Tin Gas Nickel Auto part


Which motherboard are you using, so i can check the error codes messages?
By the way, which multimeter you got, to replace the analogue one?

To be honest, i started looking for a cheap 1030Gt on second hand market, to be able to sent it to you.
That you can start from fresh, because i can't figure out what you are doing and why things keep changing constantly, when they shouldn't.

Because if you soldered the mod after the R52, and the schematic corresponds to your card, we shouldn't even have touched or bothered with the FS/OC pin.
Soldering a 500K Ohms resistor after R52, would not produce enough damage to the circuits, since it is a high resistance value, applied to low voltages circuits.

:)
 
Discussion starter · #184 ·
Which motherboard are you using, so i can check the error codes messages?
Font Rectangle Parallel Number


By the way, which multimeter you got, to replace the analogue one?
Klein Tools MM400

To be honest, i started looking for a cheap 1030Gt on second hand market, to be able to sent it to you.
That you can start from fresh, because i can't figure out what you are doing and why things keep changing constantly, when they shouldn't.
That's very kind of you, I appreciate the gesture but I will continue attempting to repair this card. I'm pretty sure I know what is wrong with it: When using the flux to un-short the uP1666Q pins, some of the flux got on surrounding resistors, & one of those resistors actually lifted off the board by itself & then re-soldered itself to the pad next to it.

Unfortunately the resistor is totally fried now, so I'm left trying to figure out what it was. It was R68, which is part of the RefAdj circuit, so it seems that without this resistor, this circuit isn't working properly and causing the card to draw too much power:
Rectangle Slope Font Parallel Handwriting
 
View attachment 2527027


Klein Tools MM400


That's very kind of you, I appreciate the gesture but I will continue attempting to repair this card. I'm pretty sure I know what is wrong with it: When using the flux to un-short the uP1666Q pins, some of the flux got on surrounding resistors, & one of those resistors actually lifted off the board by itself & then re-soldered itself to the pad next to it.

Unfortunately the resistor is totally fried now, so I'm left trying to figure out what it was. It was R68, which is part of the RefAdj circuit, so it seems that without this resistor, this circuit isn't working properly and causing the card to draw too much power:
View attachment 2527028
Gotcha, if you manage to provide the values of the other resistors, i can maybe try to calculate back the missing resistor value.

As posted previously, this is the formula that is used, to calculate the operating Vcore voltage.
Font Screenshot Parallel Number Document


According to the schematics and the datasheet, but not totally sure:
R56=R2
R71=R3
R62=R5
R61=R4
R68=R1
Rectangle Slope Font Line Parallel


I can do the calculation with Vref 0.7v, duty cycle 100%, and the vcore voltage of 1.05v, to isolate R68, which should correspond to R1 in the circuit.
But the schematic have to match at least a bit your card, to be able to be accurate enough.

And to be honest, i would gladly send you another 1030Gt, i can get them for 50/70euro i think second hand, but it would take a while to reach you in the US, from Europe. ;)

The multimeter you got seems to be decent enough, i did not read the manual, but you shouldn't have issues reading correctly the resistance and voltages values, with the auto-range function.

EDIT: Usually it is hard to fry these small resistors, so aside the resistors get stuck to the iron for quite a long time, and burn it down.
You should be able to read it, check which value it was, and possibly resolder it, with some patience.
 
View attachment 2527027


Klein Tools MM400


That's very kind of you, I appreciate the gesture but I will continue attempting to repair this card. I'm pretty sure I know what is wrong with it: When using the flux to un-short the uP1666Q pins, some of the flux got on surrounding resistors, & one of those resistors actually lifted off the board by itself & then re-soldered itself to the pad next to it.

Unfortunately the resistor is totally fried now, so I'm left trying to figure out what it was. It was R68, which is part of the RefAdj circuit, so it seems that without this resistor, this circuit isn't working properly and causing the card to draw too much power:
View attachment 2527028
You aren't using desoldering wick? And what flux do you have? Rosin ?
 
Discussion starter · #187 · (Edited)
R56=R2
R71=R3
R62=R5
R61=R4
R68=R1
All measurements are pad-to-pad
R56 = 15.7k ohms
R71 = 4.1k ohms
R62 = 309 ohms
R61 = 13.2k ohms
R68 was bubbled off the board by the flux, I guess thats why the card is probably not booting haha. It now measures a random number in the range of mega-ohms

You aren't using desoldering wick? And what flux do you have? Rosin ?
I'm not using a deslodering wick because I haven't even soldered anything yet, I will obviously use one if I need to. I'm using lead-free tinning flux for tinning wires/potentiometer leads & I'm using water soluble non-tinning flux for removing excess solder from the poorly manufactured board.

The issue I think here is that I accidentally used the tinning flux on the board and that bubbled off R68 and fried it, though no other components were harmed or moved.
 
Discussion starter · #188 · (Edited)
I've just probed the board, and confirmed that the Circuit you outlined below does indeed operate the same way you outlined:
Rectangle Slope Parallel Font Schematic

Here is an image of the board annotated with the above resistors, as you can see R68 is missing from its pads. BTW is there any way to clean the flux residue from the board? It seems to have coagulated and formed a thin film that isopropyl alcohol will not remove, though it might just be the effect of an overheated pcb from the flux itself.
Circuit component Passive circuit component Hardware programmer Microcontroller Computer hardware


Edit: See below for a cleaner pcb, I guess I just wasn't trying hard enough with the isopropyl alcohol:
White Black Style Font Electronic component
 
I've just probed the board, and confirmed that the Circuit you outlined below does indeed operate the same way you outlined:
View attachment 2527161
Here is an image of the board annotated with the above resistors, as you can see R68 is missing from its pads. BTW is there any way to clean the flux residue from the board? It seems to have coagulated and formed a thin film that isopropyl alcohol will not remove, though it might just be the effect of an overheated pcb from the flux itself.
View attachment 2527163

Edit: See below for a cleaner pcb, I guess I just wasn't trying hard enough with the isopropyl alcohol:
View attachment 2527170
So i have gone the rabbit hole, comparing roughly the galax, gigabyte and msi 1030gt pcb.
I found here a quick review, the galax and gigabyte dual-phase share semblable pcb, the msi is a bit more different.

So, supposing that the output VRM under load is roughly the same, between 1.0v to 1.05v, the controller resistor network's circuit values should be roughly the same.

Using the output VRM voltage formula (that i couldn't make a sense of how to solve it properlyo_O) and the boot, standby mode voltage formula, i was able to roughly get back the R1 value.

Supposing all these things, that the schematic wiring is the same, and you read the resistor network values correctly, R1 = 5680 Ohms.
Which is close enough, compared to the 6190 Ohms, coming from the galax schematics.

Your card:
R56 = 15.7k ohms
R71 = 4.1k ohms
R62 = 309 ohms
R61 = 13.2k ohms
R68 = 5.68k ohms

The galax schematics:
R56=20.5k ohms
R71=4.32k ohms
R62=309 ohms
R61=16.5k ohms
R68=6.19k ohms

If you have some spare electronics, you can probe the resistors on the pcb with the multimeter, until you find one with roughly the same value, same size.
Or you can get yourself a 10K Ohms variable resistor, replacing the R1 resistor, functioning as the fix and potentially as volt-mod.

Hope the card will boot again!
Good luck! (y)
 
Discussion starter · #190 · (Edited)
I was able to replace R68 with a 10k potentiometer set for 5.6k ohms:
Passive circuit component Circuit component Hardware programmer Microcontroller Electronic engineering

I then tested the two wires with a multimeter and it read ~4.17k ohms, which is suspiciously close to the value of R71.

I tried booting the card, & again was prompted with a power supply overload warning (I tested in two computers, not a PSU issue) I believe that the card is trying to draw more than the 75W limit from PCI-E, which is triggering the warning.

Since I was suspicious of the potentiometer measuring 4.17K ohms after I 100% set it to 5.6k, I decided to increase the pot until it read 5.6k again. Unfortunately the card still failed to boot, with the computer giving a PSU overload warning which I assume is to protect the MOBO from PCI-E overdraw.

I think I did fix one issue by replacing R68 with the 10k potentiometer, & I do think that this setup could be used to mod voltage once the card will boot, but I don't think R68 was the reason the card cannot boot. I believe the issue is that uP1666Q's current limit has been increased to an unsafe level somehow.

As I've noted previously, when the card was working, R52 (The voltage divider between FS/OC & Vref) measured 18.5k ohms, however for some reason R52 started to read as ~35k ohms & ever since the card triggers PSU overload warnings.

Here is how uP1666Q sets the OCP limit:
Font Parallel Number Document Rectangle

I'm almost certain that R52 is the R2 resistor in Figure 4 above, and it would make sense that increasing this resistor would change the current limit, in this case to a unsafe value which triggers OCP.

Now, I don't actually think R52 is a 18.5k resistor, or even a 35k resistor. I think it’s true value is probably closer to 65k ohms like the MSI & Galax models. I highly doubt that it is just varying on its own, as it has no physical damage and is firmly attached to both pads. I think some other component is pulling down/pulling up its resistance, perhaps a resistor or capacitor, but I cannot figure out what that is from looking into schematics/datasheets. I could be wrong.
 
I was able to replace R68 with a 10k potentiometer set for 5.6k ohms:
View attachment 2527285
I then tested the two wires with a multimeter and it read ~4.17k ohms, which is suspiciously close to the value of R71.

I tried booting the card, & again was prompted with a power supply overload warning (I tested in two computers, not a PSU issue) I believe that the card is trying to draw more than the 75W limit from PCI-E, which is triggering the warning.

Since I was suspicious of the potentiometer measuring 4.17K ohms after I 100% set it to 5.6k, I decided to increase the pot until it read 5.6k again. Unfortunately the card still failed to boot, with the computer giving a PSU overload warning which I assume is to protect the MOBO from PCI-E overdraw.

I think I did fix one issue by replacing R68 with the 10k potentiometer, & I do think that this setup could be used to mod voltage once the card will boot, but I don't think R68 was the reason the card cannot boot. I believe the issue is that uP1666Q's current limit has been increased to an unsafe level somehow.

As I've noted previously, when the card was working, R52 (The voltage divider between FS/OC & Vref) measured 18.5k ohms, however for some reason R52 started to read as ~35k ohms & ever since the card triggers PSU overload warnings.

Here is how uP1666Q sets the OCP limit:
View attachment 2527286
I'm almost certain that R52 is the R2 resistor in Figure 4 above, and it would make sense that increasing this resistor would change the current limit, in this case to a unsafe value which triggers OCP.

Now, I don't actually think R52 is a 18.5k resistor, or even a 35k resistor. I think it’s true value is probably closer to 65k ohms like the MSI & Galax models. I highly doubt that it is just varying on its own, as it has no physical damage and is firmly attached to both pads. I think some other component is pulling down/pulling up its resistance, perhaps a resistor or capacitor, but I cannot figure out what that is from looking into schematics/datasheets. I could be wrong.
Nice job fixing R68.

As you said, the card should have booted even without R68, which correspond to R1 in the schematics.
@buildzoid made a video, where he explains how to VREF/REFIN volt-mod the 3060Ti gpu core and memory.

He refers R1 as the REFADJust resistor, which he partially disconnected in order to get full gpu core voltage control.
Without having the Nvidia drivers, messing with the voltage regulation, if it deems the voltage is too high.
So it should have worked and indeed, R68 is a nice way to volt-mod Nvidia gpu, keeping intact the default voltage regulation.
To get full voltage regulation, one would just disconnect the REFADJust R1 resistor.

Here is the video.

Now, the card not booting because something is shorted, or there is an issue with the OCP protection, i dunno.
I tried to understand the FS/OC pin, the controller probe the voltage on the pin the first time for the switching frequency, the second time for the OCP.

There is a voltage divider that goes from VREF to ground, connection FS/OC, R1 = R46, R2 = R52, VREF = 2.0v, you should get around 1.0v at the FS/OC pin when it probe the OCP.
There is another voltage divider that goes from +12PEX to FS/OC and use the same R46 and R51, which is around 1.0v too.

Dunno if it safe to run the card, but if the cards is not shorted somewhere, i would read the VREF, REFIN voltage to check i find VREF = 2.0v, REFIN = 0.6v-1.05v.
You can also check the PGOOD pin of the controller, you should find 2.0v is everything is ok, 0v if the controller does not start up due to issues.
 
Discussion starter · #192 ·
So it should have worked and indeed, R68 is a nice way to volt-mod Nvidia gpu, keeping intact the default voltage regulation.
To get full voltage regulation, one would just disconnect the REFADJust R1 resistor.
Thats good to know. This mod would likely give much more precise voltage control as well it seems?

I have cross-referenced the original pictures I took of the front & back of the PCB, & nothing has fallen off or been partially disconnected. All the IC pins are correctly attached & nothing appears to be shorted.

Dunno if it safe to run the card, but if the cards is not shorted somewhere, i would read the VREF, REFIN voltage to check i find VREF = 2.0v, REFIN = 0.6v-1.05v.
You can also check the PGOOD pin of the controller, you should find 2.0v is everything is ok, 0v if the controller does not start up due to issues.
It appears safe to run the card, however as I mentioned, it seems that since its triggering PSU overload the computer is not supplying it any power. I am also not seeing anything above a few mV when probing the board while plugged in, & often those values drop back off to 0 mV. The PGOOD pin is also not showing any voltage.

I have no idea what could have gone wrong here. Prior to the card failing for the first time, I had never used a soldering iron on the board. When I tried the FB mod, I didn't even solder the wires to the board, just affixed them with glue, but even then the card worked fine for a while after trying that mod.

The only thing that has changed on the whole board to my knowledge, is the pad-to-pad resistance reading of R52. As I mentioned earlier, R52 (+C62) is in parallel with R46, so whatever I'm reading for pad-to-pad resistance isn't even its resistance of R52, its just the resistance value of the circuit between FS/OC & Ground:
Rectangle Product Slope Font Parallel


Note that both R46 & R52 both measure the same exact pad-to-pad resistance of ~35k ohms. I will keep checking for shorts/disconnected & making sure that the SMD's are all wired up the way the schematic specifies, but I have no leads as to why this card is causing PSU overload.
 
Thats good to know. This mod would likely give much more precise voltage control as well it seems?

I have cross-referenced the original pictures I took of the front & back of the PCB, & nothing has fallen off or been partially disconnected. All the IC pins are correctly attached & nothing appears to be shorted.


It appears safe to run the card, however as I mentioned, it seems that since its triggering PSU overload the computer is not supplying it any power. I am also not seeing anything above a few mV when probing the board while plugged in, & often those values drop back off to 0 mV. The PGOOD pin is also not showing any voltage.

I have no idea what could have gone wrong here. Prior to the card failing for the first time, I had never used a soldering iron on the board. When I tried the FB mod, I didn't even solder the wires to the board, just affixed them with glue, but even then the card worked fine for a while after trying that mod.

The only thing that has changed on the whole board to my knowledge, is the pad-to-pad resistance reading of R52. As I mentioned earlier, R52 (+C62) is in parallel with R46, so whatever I'm reading for pad-to-pad resistance isn't even its resistance of R52, its just the resistance value of the circuit between FS/OC & Ground:
View attachment 2527474

Note that both R46 & R52 both measure the same exact pad-to-pad resistance of ~35k ohms. I will keep checking for shorts/disconnected & making sure that the SMD's are all wired up the way the schematic specifies, but I have no leads as to why this card is causing PSU overload.
Usually, when the motherboard or power supply complain about something drawing too much current, it means there could be a short or issues somewhere.

This card usually draw little current, so if when plugged into the pci-e slot, the motherboard complain about overcurrent protection, chances are something is shorted.

Not sure that a messed up VRM controller, would be enough to trigger the motherboard, usually the card simply does not boot, alongside a black screen.

So i would check the 3.3v and 12v coming from the pci-e slot, that these are not shorted.
Also, as said above, there are two voltages dividers reaching the FS/OC pin, one coming from the 12v rail with R46 and R51 = 1.0215v and the other from VREF with R46 and R52 = 0.9165v.
The voltage at C48 capacitor should be around 1.0v, if i believe the schematics, so i would check these components, the voltage at C48 and if there is any short there.
 
Discussion starter · #194 ·
Usually, when the motherboard or power supply complain about something drawing too much current, it means there could be a short or issues somewhere.

This card usually draw little current, so if when plugged into the pci-e slot, the motherboard complain about overcurrent protection, chances are something is shorted.
I've checked the board a few times, and it doesn't seem like anything is shorted, here are some photos if you would like to check for yourself:
Light Black Circuit component Hardware programmer Font

Circuit component Passive circuit component Hardware programmer Electronic engineering Electronic component

Circuit component Black Hardware programmer Font Electronic engineering

Circuit component Font Electronic engineering Hardware programmer Electronic component


So i would check the 3.3v and 12v coming from the pci-e slot, that these are not shorted.
Also, as said above, there are two voltages dividers reaching the FS/OC pin, one coming from the 12v rail with R46 and R51 = 1.0215v and the other from VREF with R46 and R52 = 0.9165v.
The voltage at C48 capacitor should be around 1.0v, if i believe the schematics, so i would check these components, the voltage at C48 and if there is any short there.
3.3v & 12v lines do not appear shorted. No decoupling capacitors are shorted, nor are any of the SMD's near uP1666Q.

R51 is unused pads on my board & always has been (referencing old pictures) C48 is also unused pads and always has been.
 
I've checked the board a few times, and it doesn't seem like anything is shorted, here are some photos if you would like to check for yourself:
View attachment 2527632
View attachment 2527633
View attachment 2527634
View attachment 2527635


3.3v & 12v lines do not appear shorted. No decoupling capacitors are shorted, nor are any of the SMD's near uP1666Q.

R51 is unused pads on my board & always has been (referencing old pictures) C48 is also unused pads and always has been.
Since you have the schematics, i would check also the other power delivery chips, checking that these are also short free.
As usual, you want to check the resistance of all the power rails at the input, and at the output of the power rail.

If everything is fine, at this point i would agree with you, the VRM controller is so messed up, that it bring down the whole gpu.
You can check this video from Louis Rossmann, he runs a MacBook repair shop, i linked you the moment showing how to rework and reflow the controller's pad.


A lot of flux, a sharp small iron tip with a bit of solder on top, then do as he shows, rework each controller side.
To remove any shorts and recreate shiny and well-defined single pads.

By the way, what are these marks?
Black Font Hardware programmer Technology Electronic device


Can you look closer at this part of the pcb, was it like this when you got the card?
Circuit component Passive circuit component Hardware programmer Font Engineering
 
Discussion starter · #196 ·
Since you have the schematics, i would check also the other power delivery chips, checking that these are also short free.
As usual, you want to check the resistance of all the power rails at the input, and at the output of the power rail.
I have gone through the board a few more times, & there are no shorts. I just specifically went through the IC's again & they are not shorted.

If everything is fine, at this point i would agree with you, the VRM controller is so messed up, that it bring down the whole gpu.
You can check this video from Louis Rossmann, he runs a MacBook repair shop, i linked you the moment showing how to rework and reflow the controller's pad.
I've actually done this a few times already because I figured this must have been the issue since this is the only component I've even come near with a soldering iron. To make sure, I reflowed the pins a few more times & tested it a few more times & the card still triggers PSU overload and prevents the computer from booting.

It seems that every time I come near the board with a soldering iron, there is a chance that the resistance value of R52 will change, as now I'm reading it as 87k ohms, whereas it was 18.5k originally & then 35k for a while.

All the pins of uP1666Q are properly attached to their pads & not shorted to one another, so at this point I don't think that's the issue, but I could be wrong.

Those marks you called out are old bits of thermal paste. The part of the PCB you called out near the lower choke has always looked like that, its just a very strangely shaped PCB trace, there is no warping of the board.

The last time the value of R52 changed, this is what happened:
I reflowed the pins of uP1666Q to make sure they weren't shorted for the 2nd or 3rd time.
After doing so, I measured the resistance of R52, & it initially measured zero, but then slowly began to count upwards.
I was hopeful that R52 would stop at 18.5k ohms, but it kept counting, this time all the way to 87k ohms.

This behavior of the resistor "resetting" before slowly counting up to a new value leads me to believe that something is wrong with a capacitor that is in parallel with R52, however I have not found any capacitor that is in line with R52 on my PCB. What else could cause the resistance reading of a recently reflowed pin to steadily increase besides some messed up capacitor? I wonder if while doing the FB mod, I damaged a capacitor enough to make it fail after the next few boots without the FB mod applied?
 
I have gone through the board a few more times, & there are no shorts. I just specifically went through the IC's again & they are not shorted.


I've actually done this a few times already because I figured this must have been the issue since this is the only component I've even come near with a soldering iron. To make sure, I reflowed the pins a few more times & tested it a few more times & the card still triggers PSU overload and prevents the computer from booting.

It seems that every time I come near the board with a soldering iron, there is a chance that the resistance value of R52 will change, as now I'm reading it as 87k ohms, whereas it was 18.5k originally & then 35k for a while.

All the pins of uP1666Q are properly attached to their pads & not shorted to one another, so at this point I don't think that's the issue, but I could be wrong.

Those marks you called out are old bits of thermal paste. The part of the PCB you called out near the lower choke has always looked like that, its just a very strangely shaped PCB trace, there is no warping of the board.

The last time the value of R52 changed, this is what happened:
I reflowed the pins of uP1666Q to make sure they weren't shorted for the 2nd or 3rd time.
After doing so, I measured the resistance of R52, & it initially measured zero, but then slowly began to count upwards.
I was hopeful that R52 would stop at 18.5k ohms, but it kept counting, this time all the way to 87k ohms.

This behavior of the resistor "resetting" before slowly counting up to a new value leads me to believe that something is wrong with a capacitor that is in parallel with R52, however I have not found any capacitor that is in line with R52 on my PCB. What else could cause the resistance reading of a recently reflowed pin to steadily increase besides some messed up capacitor? I wonder if while doing the FB mod, I damaged a capacitor enough to make it fail after the next few boots without the FB mod applied?
At this point, as said above, i'm unable to provide knowledgeable insights.
The part of the pcb you have been working on, have been butchered too much, which is normal when one start doing electronics.
And without having another card or the right schematics, to check the real resistance, it is really hard for me, to think where the issue could lie.

As i said above, i would rather start with a fresh card, doing just the mod without touching anything else.
And then use the same good card, to troubleshoot the 1st.

But as it stand, i'm sorry, i have no other ideas, aside sending you a new gpu.
Or opening another thread asking for people with the same GT1030, to provide some components values readings.
😕
 
Discussion starter · #198 ·
At this point, as said above, i'm unable to provide knowledgeable insights.
The part of the pcb you have been working on, have been butchered too much, which is normal when one start doing electronics.
And without having another card or the right schematics, to check the real resistance, it is really hard for me, to think where the issue could lie.
Understandable, it is unfortunate that the PCB's seem to vary.

As i said above, i would rather start with a fresh card, doing just the mod without touching anything else.
And then use the same good card, to troubleshoot the 1st.
It unfortunately seems like this model is quite difficult to come by. For reference, I bought this card the week or two after the GT 1030 came out, so this is quite an early PCB design in the products life cycle, given that it is still in production today, albeit on a newer PCB revision.

But as it stand, i'm sorry, i have no other ideas, aside sending you a new gpu.
Or opening another thread asking for people with the same GT1030, to provide some components values readings.
😕
That's fine, I think I will do a more thorough probe of the PCB to compare each circuit vs the MSI schematics to make sure things are at least connected. I didn't really do a comprehensive probe of the PCB the first couple times, so perhaps a more full scan will provide some insight.
 
181 - 200 of 203 Posts