RIPJAY Hardmod: Breaking NVIDIA's Rules w/ Shunt Resistor Piggyback, Ft. HiCookie
video description
Date: 2020-05-06
Related videos
Comments and reviews: 10
slaur42
Electrical engineering / physics explanation for a couple things here. Many metals have a positive temperature coefficient for their resistivity. I couldn't find one for the liquid metal alloy, but all of its components (gallium, indium, tin) are positive. By the way, so is copper. For copper it is around 0. 4% per degree C. This means, as the shunt resistor heats up, its resistance increases, and in the end, this causes a lower current. As your graphs show. This should also happen, if you just solder a copper wire on top of the shunt resistor, so it may be mechanically better than using liquid metal, but not electrically. Now, shunt resistors built for that purpose are made from metal alloys, where that temperature coefficient is as small as possible across the whole defined operating temperature range. This results in a smooth constant current delivery, if all other variables are also held constant. Even the cheapest 8 milliohm shunt resistor on Digikey that I looked at was +- 100 ppm/degree C. Copper is 400 times that, and galinstan (liquid metal) is probably in the same range as copper.
reply
Electrical engineering / physics explanation for a couple things here. Many metals have a positive temperature coefficient for their resistivity. I couldn't find one for the liquid metal alloy, but all of its components (gallium, indium, tin) are positive. By the way, so is copper. For copper it is around 0. 4% per degree C. This means, as the shunt resistor heats up, its resistance increases, and in the end, this causes a lower current. As your graphs show. This should also happen, if you just solder a copper wire on top of the shunt resistor, so it may be mechanically better than using liquid metal, but not electrically. Now, shunt resistors built for that purpose are made from metal alloys, where that temperature coefficient is as small as possible across the whole defined operating temperature range. This results in a smooth constant current delivery, if all other variables are also held constant. Even the cheapest 8 milliohm shunt resistor on Digikey that I looked at was +- 100 ppm/degree C. Copper is 400 times that, and galinstan (liquid metal) is probably in the same range as copper.
reply
Brian
What the shunt does: All current for each pci-e cable goes through the shunt(s. There s a chip that monitors the voltage across the shunt. V=I R, so it can infer the current by the voltage drop across the shunt. 10A through a 5mOhm shunt will have a Vdrop of 50mV. If the card is set to limit at say, 50A per PCIE, then the Vdrop will be 250mV. Adding a 8mOhm shunt in parallel makes the shunt effectively 3mOhm. I = V/R, so at 250mV max Vdrop, you can run 83. 33A now for each PCIE. That s the theory behind it. You technically could just hardwire the shunt, cut the traces and feed a set voltage to the power controller and have effectively unlimited power. I m a Professional Electronics Repair Engineer for a large US Company you ve definitely heard of and would love to help with any component level mods, repairs, or advice you need. DM me! Love Guanghua too, I try to go through there and Syntrend next door when I m in Taipei.
reply
What the shunt does: All current for each pci-e cable goes through the shunt(s. There s a chip that monitors the voltage across the shunt. V=I R, so it can infer the current by the voltage drop across the shunt. 10A through a 5mOhm shunt will have a Vdrop of 50mV. If the card is set to limit at say, 50A per PCIE, then the Vdrop will be 250mV. Adding a 8mOhm shunt in parallel makes the shunt effectively 3mOhm. I = V/R, so at 250mV max Vdrop, you can run 83. 33A now for each PCIE. That s the theory behind it. You technically could just hardwire the shunt, cut the traces and feed a set voltage to the power controller and have effectively unlimited power. I m a Professional Electronics Repair Engineer for a large US Company you ve definitely heard of and would love to help with any component level mods, repairs, or advice you need. DM me! Love Guanghua too, I try to go through there and Syntrend next door when I m in Taipei.
reply
Travis
im kinda new to some of this, as im sure ur goin to see by this question, but why cant we just edit the firmware to do away with the power limits or at least set them to something we can work with? i imagine once u find it in one firmware it would b easier to find in others is it locked down or somethin in some way i dont know about all u should need is a simple hex editor right? if u knew what resistor value to use u could have a subtle on board ammeter n voltmeter it would actually b pretty cool im also a geek n i love #s stats n sensors so that might jus b me idk: p but the problem with the card limiting the power is in the firmware right? n that should b able to fixed semi easily right? jus possibly time consuming i guess
reply
im kinda new to some of this, as im sure ur goin to see by this question, but why cant we just edit the firmware to do away with the power limits or at least set them to something we can work with? i imagine once u find it in one firmware it would b easier to find in others is it locked down or somethin in some way i dont know about all u should need is a simple hex editor right? if u knew what resistor value to use u could have a subtle on board ammeter n voltmeter it would actually b pretty cool im also a geek n i love #s stats n sensors so that might jus b me idk: p but the problem with the card limiting the power is in the firmware right? n that should b able to fixed semi easily right? jus possibly time consuming i guess
reply
ajm8127
Watching him hold the resistors like that at 7: 00 is cringe worthy. He should take a seat to make himself more stable and grip the part by it's sides. For something like this I would first apply some flux then position the resistor to line it up with the original and then apply slight downward pressure with the tweezers while heating the surface mount pad with the iron. Once one side is soldered just hit the other side with the iron. You might need to reapply the flux before soldering the other pad. Add solder sparingly. I am 100% sure that PCB is RoHS compliant so it uses lead free solder, but leaded solder melts at a lower temperature so it might make it easier to work with.
reply
Watching him hold the resistors like that at 7: 00 is cringe worthy. He should take a seat to make himself more stable and grip the part by it's sides. For something like this I would first apply some flux then position the resistor to line it up with the original and then apply slight downward pressure with the tweezers while heating the surface mount pad with the iron. Once one side is soldered just hit the other side with the iron. You might need to reapply the flux before soldering the other pad. Add solder sparingly. I am 100% sure that PCB is RoHS compliant so it uses lead free solder, but leaded solder melts at a lower temperature so it might make it easier to work with.
reply
Patrik
I have tried this on my 1070ti asus strix, i tried piggy back and changed shunt resistors (i used math +25 percent power. This card has two power limits, i have done first, ,high stage, 12V current measurement. In msi afterburner power limit was on 70-80 percent, but it still triggers power limit. it helped little with OC, i lightly tried to mod, low stage, current measurement. Changed some resistors on mosfets, have done nothing, i dont wanted to make smoke machine. I also filled empty capacitors spaces with low esr caps, that was better +150mhz mem clock: ) i hope gamers nexus read this: )
reply
I have tried this on my 1070ti asus strix, i tried piggy back and changed shunt resistors (i used math +25 percent power. This card has two power limits, i have done first, ,high stage, 12V current measurement. In msi afterburner power limit was on 70-80 percent, but it still triggers power limit. it helped little with OC, i lightly tried to mod, low stage, current measurement. Changed some resistors on mosfets, have done nothing, i dont wanted to make smoke machine. I also filled empty capacitors spaces with low esr caps, that was better +150mhz mem clock: ) i hope gamers nexus read this: )
reply
Ste
Am gonna spam this to you andJays2cents and LTT and the rest because you have more contacts that i could even get hold of and very off topic for this vid but please read. At this time of shite for all of us i would like to see konami / capcom / namco and a few others put together an old skool gaming pack and release it for free on ps4 crapbox and pc and would be nice if you guys could all reach out to them. I will try my best my emailing them to but i am nothing to them so need everyone to reach out. we need an old games pac with pacman rainbow island super puzzle fighter and many more
reply
Am gonna spam this to you andJays2cents and LTT and the rest because you have more contacts that i could even get hold of and very off topic for this vid but please read. At this time of shite for all of us i would like to see konami / capcom / namco and a few others put together an old skool gaming pack and release it for free on ps4 crapbox and pc and would be nice if you guys could all reach out to them. I will try my best my emailing them to but i am nothing to them so need everyone to reach out. we need an old games pac with pacman rainbow island super puzzle fighter and many more
reply
Lars
soldering on a heavy ground or power plane is not fun. It is not recommended to go op to 450 degrees or more on the soldering iron, the flux will burn away, so keep the solder on the tip fresh or add extra flux to make a good connection. Also do not reheat multiple times to fix a solder without adding flux or start over by removing old burned solder. Use a as big a tip as possible to have as much energy stored in the tip as possible to transfer to the object, only then go up in heat to bruteforce more energy into the solder tip if needed.
reply
soldering on a heavy ground or power plane is not fun. It is not recommended to go op to 450 degrees or more on the soldering iron, the flux will burn away, so keep the solder on the tip fresh or add extra flux to make a good connection. Also do not reheat multiple times to fix a solder without adding flux or start over by removing old burned solder. Use a as big a tip as possible to have as much energy stored in the tip as possible to transfer to the object, only then go up in heat to bruteforce more energy into the solder tip if needed.
reply
Kappi1997
8: 50an accurate description for those interested. The Power controller checks the power usage by measuring the voltage over the shunt resistor(the higher the current flowing, the higher the power, the higher the voltage over the resistor. By lowering the resistance the voltage per Amp over the resistor will sink but the controller isn't that smart it just sees a okey i'm pulling just a few amount of watts but in reallity he is pulling much more just not knowing the resistance changed.
reply
8: 50an accurate description for those interested. The Power controller checks the power usage by measuring the voltage over the shunt resistor(the higher the current flowing, the higher the power, the higher the voltage over the resistor. By lowering the resistance the voltage per Amp over the resistor will sink but the controller isn't that smart it just sees a okey i'm pulling just a few amount of watts but in reallity he is pulling much more just not knowing the resistance changed.
reply
rayzimmermin
the guy using the iron doesn't know what he doing and you can tell he keep applying solder to his iron and not the board and component when using an iron you apply heat to the board and component then touch the solder to the board and component not the iron and the solder will flow to the heat so if you apply solder to the iron it will never flow to the board and component soldering the component to the board because the board and component are always cooler than the iron
reply
the guy using the iron doesn't know what he doing and you can tell he keep applying solder to his iron and not the board and component when using an iron you apply heat to the board and component then touch the solder to the board and component not the iron and the solder will flow to the heat so if you apply solder to the iron it will never flow to the board and component soldering the component to the board because the board and component are always cooler than the iron
reply
Mick
Does this shunt Mod apply to the 2080ti as well or is this only possible on the titan? Also what is the best Bios to get the highest power draw out of a 2080 ti FE or Reference Card? Choose your favorite term for that one lol. This question is for Steve if you see this post and have time and feel like answering this question and also the GN community if anyone knows I could use the info if you know the answer to either or both of these questions.
reply
Does this shunt Mod apply to the 2080ti as well or is this only possible on the titan? Also what is the best Bios to get the highest power draw out of a 2080 ti FE or Reference Card? Choose your favorite term for that one lol. This question is for Steve if you see this post and have time and feel like answering this question and also the GN community if anyone knows I could use the info if you know the answer to either or both of these questions.
reply
Add a review, comment
Other channel videos
