GridSeed 5-chip USB miner voltage mod

[gbyg]

No problem.  Those parts should be pretty cheap, hopefully that's all that was damaged.  Only way to find out is to replace them with identical spec replacements.  Someone mentioned the possibility of temporarily bypassing those parts to verify functionality of the gridseed - however if it were mine, I'd replace the parts before testing it.

:EDIT:  What thermal paste did you use?  Non-conductive I hope.  You might want to clean that up while you've got it apart.  In my opinion, there's no real point in using thermal paste unless you are positive the heatsink makes full contact with all 5 gc3355 chips.  On all the gridseeds I have there are 2 components which are taller than the gc3355's and prevent full contact.  I milled out part of the heatsink over those two components so the heatsink made full contact.   I posted a pic in this thread somewhere of it.  However, if you do that, then the heatsink will likely make contact with all the capacitors surrounding the gc3355's.  So I covered those with electrical tape to prevent a short.

thermal paste i used was left over non-conductive MX-4. it will be fun to clean this mess!!! I will think about the milling out that small part as you did.

[amix]

Just leave the top off. Add small indivdual heatsinks to the top of the chips. It allows a better airflow leabing it open. The air flow fro  the fan os pretty totally redticted due to the pcb. Leaving the top off also allows you to add a heat sink to the voltage regs. There is only. The top of the chips toucjing the heatsink. Really for mini g scrypt is it needed a smaller one or ones would be better. Remember these where desgined wattage and heat wise for the sha or dual mining mode.

[dyland]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

[wolfey2014]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

I'm going for top hash rate. I don't have to worry about power bills here.
I think 49.9k is the best option overall though. And it's up to 1250/1275 actually. Some actually hash pretty profitably at that speed.
Good luck!

[dyland]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

I'm going for top hash rate. I don't have to worry about power bills here.
I think 49.9k is the best option overall though. And it's up to 1250/1275 actually. Some actually hash pretty profitably at that speed.
Good luck!

Thanks very much Wolfey! I'm at about .09 cents a kw here, which is about .08 cents USD.

Can you PM me your email? I might have a few questions when this is performed.

EDIT: Jesse's guide is definitive on the 49.9k mod, correct?

[wolfey2014]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

I'm going for top hash rate. I don't have to worry about power bills here.
I think 49.9k is the best option overall though. And it's up to 1250/1275 actually. Some actually hash pretty profitably at that speed.
Good luck!

Thanks very much Wolfey! I'm at about .09 cents a kw here, which is about .08 cents USD.

Can you PM me your email? I might have a few questions when this is performed.

EDIT: Jesse's guide is definitive on the 49.9k mod, correct?

Sure, just click one of the links under my account name....
good luck!

[cbutters]

Hey everyone, Just wanted to report back on my 10 miners from gaw,
2/10 did not work out of the box,

1 had clamped the fan wires and shorted everything out when I plugged it in. - Fixed this by repairing the fan wires
1 hashed, but nothing got accepted - fixed this by reflowing the solder with a heat gun... hashing like a champ now.

I modded 4 with the 49.9K resistor, and was running them averaged 1225mhz stable / average 507 khash and found the heat and power consumption excessive - 29ish watts when pods were hot.
I now have modded all to with a 47K resistor and total mhz average after tuning is 1190mhz / 501

47K vs 49.9K Comparison:

                       Hashrate Ave.        Watts            MHz (ave after chip tuning)        hash /Watt
47K                     501.65                20-22          1190                                              22.8
49.9K                   507.6                 28-29          1225                                              17.5


My definitive Thoughts
-Use a 47K resistor - the extra heat / power management is not worth the minimal hashrate increase.
-use sandor111's cpuminer
-No HW errors does not mean your miner is running optimally, some gridseeds will run slowly at clocks that are unstable, but not throw HW errors.
-use autotune for debugging and testing max frequencies of chips, but lock them in manually in your config script.

[amix]

Hey everyone, Just wanted to report back on my 10 miners from gaw,
2/10 did not work out of the box,

1 had clamped the fan wires and shorted everything out when I plugged it in. - Fixed this by repairing the fan wires
1 hashed, but nothing got accepted - fixed this by reflowing the solder with a heat gun... hashing like a champ now.

I modded 4 with the 49.9K resistor, and was running them averaged 1225mhz stable / average 507 khash and found the heat and power consumption excessive - 29ish watts when pods were hot.
I now have modded all to with a 47K resistor and total mhz average after tuning is 1190mhz / 501

47K vs 49.9K Comparison:

                       Hashrate Ave.        Watts            MHz (ave after chip tuning)        hash /Watt
47K                     501.65                20-22          1190                                              22.8
49.9K                   507.6                 28-29          1225                                              17.5


My definitive Thoughts
-Use a 47K resistor - the extra heat / power management is not worth the minimal hashrate increase.
-use sandor111's cpuminer
-No HW errors does not mean your miner is running optimally, some gridseeds will run slowly at clocks that are unstable, but not throw HW errors.
-use autotune for debugging and testing max frequencies of chips, but lock them in manually in your config script.

46.2K draw 1.43A  17.16W /MHZ 1160 /hash rate 490
 5W extra for 11k with the 47K!
watt is effectored by the resistor AND MHZ, so it good to put lower resistor and higher hash rate, as opposed to higher reisitor and lower hash rate.

[wolfey2014]

Hey everyone, Just wanted to report back on my 10 miners from gaw,
2/10 did not work out of the box,

1 had clamped the fan wires and shorted everything out when I plugged it in. - Fixed this by repairing the fan wires
1 hashed, but nothing got accepted - fixed this by reflowing the solder with a heat gun... hashing like a champ now.

I modded 4 with the 49.9K resistor, and was running them averaged 1225mhz stable / average 507 khash and found the heat and power consumption excessive - 29ish watts when pods were hot.
I now have modded all to with a 47K resistor and total mhz average after tuning is 1190mhz / 501

47K vs 49.9K Comparison:

                       Hashrate Ave.        Watts            MHz (ave after chip tuning)        hash /Watt
47K                     501.65                20-22          1190                                              22.8
49.9K                   507.6                 28-29          1225                                              17.5


My definitive Thoughts
-Use a 47K resistor - the extra heat / power management is not worth the minimal hashrate increase.
-use sandor111's cpuminer
-No HW errors does not mean your miner is running optimally, some gridseeds will run slowly at clocks that are unstable, but not throw HW errors.
-use autotune for debugging and testing max frequencies of chips, but lock them in manually in your config script.

Right, but keep in mind:
not all miners are the same hence your comparisons do not apply in general to all.
And, the watt/heat is not excessive w/49.9k. At least not with my pods.
I wonder what Gridseed has changed since the first mods were done on these things?
We know their quality control leaves quite a bit to be desired.
I'll be glad when I don't have to do diddly with my miners anymore. Just buy them, mine them, make them pay off ROI and profit from then on! BIG TIME! 

[geetash]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

i have used 48k resistors (measured and selected from a pack of 5% 47k 1/8W resistors) to mod my pods and they are running at 1200MH/s with almost zero hardware errors with a power consumption of 25w~ per each measured with a kilowatt meter. power supply is a 850 silver.
i don't know who got this readings but its not consuming 30+ watts.

[wolfey2014]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

i have used 48k resistors (measured and selected from a pack of 5% 47k 1/8W resistors) to mod my pods and they are running at 1200MH/s with almost zero hardware errors with a power consumption of 25w~ per each measured with a kilowatt meter. power supply is a 850 silver.
i don't know who got this readings but its not consuming 30+ watts.

Right. 30+W sounds ridiculous if not suspicious....
Yah, 48k 5% tol is close enough I recon as long as that 5% drift doesn't go the wrong way i.e. down below 45.6k which is useless for overclocking above 950 or 50.4k which is a tad too high but not harmful to other than power consumption hence profit.
I might get lazy on this batch and do the same thing though. It's mine (punn) anyway

[cbutters]

When I was modding my gridseeds, I accidentally removed one of the wrong ultra small resistors... I had to resolder it back in place, and it was absolutely ridiculous... I can't even believe I got it back on. An ant literally walked across my table at the same time as I was working on it, and the ant looked enourmous. the resistor was about the size of an ant's leg. I kept losing it over and over again from touching it and it would stick on my finger, or the tweezers, etc...
It is especially impressive I got it back on since usually when I take them off, it gets sucked inside of the ball of solder on the tip of my soldering iron... those resistors are way to small to make dumb mistakes guys I think I need a new soldering iron tip before I do any more.

[nst6563]

When I was modding my gridseeds, I accidentally removed one of the wrong ultra small resistors... I had to resolder it back in place, and it was absolutely ridiculous... I can't even believe I got it back on. An ant literally walked across my table at the same time as I was working on it, and the ant looked enourmous. the resistor was about the size of an ant's leg. I kept losing it over and over again from touching it and it would stick on my finger, or the tweezers, etc...
It is especially impressive I got it back on since usually when I take them off, it gets sucked inside of the ball of solder on the tip of my soldering iron... those resistors are way to small to make dumb mistakes guys I think I need a new soldering iron tip before I do any more.

Use a fine tip...or better yet...a hot-air system is your best friend.

[michelem]

I will start a new thread when ready but in the meantime I would like to share with you the work in progress behind this new project based on sandor cpuminer.

I'd like to test it with a great bunch of pods (20 or more) but I have only 3 now, may be someone of you could test it? It requires a linux controller like a raspberry with a web server (something like a Scripta img should work very fine, just change the document root to the Minera directory).

If there is someone who wanna try it please leave me a PM, I think I will release a public beta-version next week.

Hope you like it:

Quoted for awesomeness, I applaud your efforts.

Thanks sandor
Do you think it's possible to add an option to cpuminer to detect automatically the gridseed devices? Of course you will not be able to set per device freq with this option on but it would be useful to start the command in autotune without writing every single dev in the launcher string. I think looking at the dev ID do the job, Gridseed has always 0483:5740 as ID.

[geetash]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

i have used 48k resistors (measured and selected from a pack of 5% 47k 1/8W resistors) to mod my pods and they are running at 1200MH/s with almost zero hardware errors with a power consumption of 25w~ per each measured with a kilowatt meter. power supply is a 850 silver.
i don't know who got this readings but its not consuming 30+ watts.

Right. 30+W sounds ridiculous if not suspicious....
Yah, 48k 5% tol is close enough I recon as long as that 5% drift doesn't go the wrong way i.e. down below 45.6k which is useless for overclocking above 950 or 50.4k which is a tad too high but not harmful to other than power consumption hence profit.
I might get lazy on this batch and do the same thing though. It's mine (punn) anyway

well you have to measure and get the value you want from a pack of low precision resistors.
what you mean 5% drift?
each 47k 5% resistors will range between 45.x and 51.x when producing them, but its value is not going to be a variable value after that. each resistor will have a fixed value between 45.x and 50.x
A 48kohms 5% resistor measured by a DVM will remain 48k ohms unless influenced by temperature or other reasons. It should not be mistaken that the value varies randomly from 45k to 51k (5% difference)

[happydaze]

How electrically efficient are these gridseed pods when used as a space heater?
I know they are not an "efficient" heater so I am not arguing for that use.  Please assume for this question that all other heat in the house is electric and that heat is required for 9 months of the year.  Also, please ignore purchase cost of the pods versus buying an actual heater.  This info could influence miners who live in cold climates when considering a voltage mod.

My guess is that 75% of the electricity used ends up as heat.  Do you think my guess is high or low or about right?  Probable less waste heat ratio at low wattage (stock LTC) compared to high wattage (dual mode).

[wolfey2014]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

i have used 48k resistors (measured and selected from a pack of 5% 47k 1/8W resistors) to mod my pods and they are running at 1200MH/s with almost zero hardware errors with a power consumption of 25w~ per each measured with a kilowatt meter. power supply is a 850 silver.
i don't know who got this readings but its not consuming 30+ watts.

Right. 30+W sounds ridiculous if not suspicious....
Yah, 48k 5% tol is close enough I recon as long as that 5% drift doesn't go the wrong way i.e. down below 45.6k which is useless for overclocking above 950 or 50.4k which is a tad too high but not harmful to other than power consumption hence profit.
I might get lazy on this batch and do the same thing though. It's mine (punn) anyway

well you have to measure and get the value you want from a pack of low precision resistors.
what you mean 5% drift?
each 47k 5% resistors will range between 45.x and 51.x when producing them, but its value is not going to be a variable value after that. each resistor will have a fixed value between 45.x and 50.x
A 48kohms 5% resistor measured by a DVM will remain 48k ohms unless influenced by temperature or other reasons. It should not be mistaken that the value varies randomly from 45k to 51k (5% difference)

You answered your own question....

[amix]

How electrically efficient are these gridseed pods when used as a space heater?
I know they are not an "efficient" heater so I am not arguing for that use.  Please assume for this question that all other heat in the house is electric and that heat is required for 9 months of the year.  Also, please ignore purchase cost of the pods versus buying an actual heater.  This info could influence miners who live in cold climates when considering a voltage mod.

My guess is that 75% of the electricity used ends up as heat.  Do you think my guess is high or low or about right?  Probable less waste heat ratio at low wattage (stock LTC) compared to high wattage (dual mode).

if its hashing away and using 20w itll be producing 20w heat. hashing produces heat it doesnt turn it into another form of energy. ok there a small loss in the fan and the leds. but its all gone out as heat if not from the gridseed chips themselves then the voltage reg crystals and other pieces.

You could measure it putting it in a metre by metre box giving a 1m cubed volume area, of airt at room temp make sure only the grids are inside disconnect the fan and power it up and time how long its takes to rise the temperature in the box by 1 degree. take an ampage reading of just the grid seed, not the power convertor or from the wall or psu but at the grid. find the energy co efficient of air compare contrast and youll find your answer.
something like that anyway

[wolfey2014]

How electrically efficient are these gridseed pods when used as a space heater?
I know they are not an "efficient" heater so I am not arguing for that use.  Please assume for this question that all other heat in the house is electric and that heat is required for 9 months of the year.  Also, please ignore purchase cost of the pods versus buying an actual heater.  This info could influence miners who live in cold climates when considering a voltage mod.

My guess is that 75% of the electricity used ends up as heat.  Do you think my guess is high or low or about right?  Probable less waste heat ratio at low wattage (stock LTC) compared to high wattage (dual mode).

if its hashing away and using 20w itll be producing 20w heat. hashing produces heat it doesnt turn it into another form of energy. ok there a small loss in the fan and the leds. but its all gone out as heat if not from the gridseed chips themselves then the voltage reg crystals and other pieces.

You could measure it putting it in a metre by metre box giving a 1m cubed volume area, of airt at room temp make sure only the grids are inside disconnect the fan and power it up and time how long its takes to rise the temperature in the box by 1 degree. take an ampage reading of just the grid seed, not the power convertor or from the wall or psu but at the grid. find the energy co efficient of air compare contrast and youll find your answer.
something like that anyway

Although a bit more expensive power bill wise, GPU's would be perfect for your 'space heating' needs
And they are becoming cheaper and cheaper now too, I believe.

[geetash]

will this mod work with hashra controla?

The mod will work for any mining software, you just need to change the chip frequency.

Nervous to attempt, but an extra 100 khsh across 20 miners translates to an extra 2Mhsh. Anyone want to share any best practices aside from OP?

I did the 49.9k mod on all 14 of my gridseeds and I'm glad I did. I'm running CPUminer and all of my chips are stable at 1200mhz and up. The pool and CPUminer and both reporting very close to each other so I'm happy!!

Thanks for the follow up; can you list the tools/hardware you used and your new hash rate? Also if possible, anyone know the page in thus thread where I find the 49.9k mod offhand?

Go back to page 1, search for - wolfey2014 - all my posts will come up, and all shall be revealed to you

Thanks Wolfey,

I'm still a bit confused though:

what's the better mod between these two below? The guy I have doing it is only familiar with 47k mod.

VMOD 3: (47k): Up to freq=1175 (500kH/s poolside average)  30 +/-4 watt
VMOD 3: (49.9k): Up to freq=1200 (510kH/s poolside average)   33 +/-4 watt

i have used 48k resistors (measured and selected from a pack of 5% 47k 1/8W resistors) to mod my pods and they are running at 1200MH/s with almost zero hardware errors with a power consumption of 25w~ per each measured with a kilowatt meter. power supply is a 850 silver.
i don't know who got this readings but its not consuming 30+ watts.

Right. 30+W sounds ridiculous if not suspicious....
Yah, 48k 5% tol is close enough I recon as long as that 5% drift doesn't go the wrong way i.e. down below 45.6k which is useless for overclocking above 950 or 50.4k which is a tad too high but not harmful to other than power consumption hence profit.
I might get lazy on this batch and do the same thing though. It's mine (punn) anyway

well you have to measure and get the value you want from a pack of low precision resistors.
what you mean 5% drift?
each 47k 5% resistors will range between 45.x and 51.x when producing them, but its value is not going to be a variable value after that. each resistor will have a fixed value between 45.x and 50.x
A 48kohms 5% resistor measured by a DVM will remain 48k ohms unless influenced by temperature or other reasons. It should not be mistaken that the value varies randomly from 45k to 51k (5% difference)

You answered your own question....

your reply didn't sound like you knew this before.


edit :

just noticed that you have bold influenced by temperature
so you think that the given tolerance value will save the resistor from heart drift ?
nope tolerance means precision of the resistor value in production process. Variation of resistance with temperature is independent from tolerance.