• We've upgraded our forums. Please post any issues/requests in this thread.

Voltage Mod Adjustment Circuit

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#1
I thought it might be interesting to assemble a circuit that would allow for voltage increase and decrease on a graphics card (gpu and gmem) while the card is operational. I can't say that this is even possible, there are many factors to consider. Sounds like an interesting challenge.

More to come...

Any ideas?:)
 
Joined
Jun 5, 2005
Messages
4,640 (1.01/day)
Likes
415
Location
Toronto, Canada
System Name old school / new school
Processor 3.0e C0 @ 3.6 / e5200
Motherboard p4p800e-dlx / p5q-DLX
Cooling custom water see sig / air
Memory 2x1g oczPC4000EbPl / 2x2g ocz2rpr1066
Video Card(s) 3850AGP / 4890vaporX
Storage 36g raptor+120g wd / wd 1001fals 1tb
Display(s) BenQ / sharpAQUOS LC-37D64U
Case modded antec plusview / generic
Audio Device(s) audigy 2zs / ASUS Xonar HDAV1.3
Power Supply fan/cable modded powerstream 520 / OCZ 700mxsp
Software Xp pro SP2 / VISTA ultimate OEM
#2
do a serch of volt mods here on TPU ;)
 
Joined
Apr 30, 2006
Messages
1,167 (0.27/day)
Likes
433
Processor 7900
Motherboard Rampage Apex
Cooling H115i
Memory 64GB TridentZ 3200 14-14-14-34-1T
Video Card(s) Fury X
Case Corsair 740
Audio Device(s) 8ch LPCM via HDMI to Yamaha Z7 Receiver
Power Supply Corsair AX860
Mouse G903
Keyboard G810
Software 8.1 x64
#3
 
Joined
Dec 26, 2007
Messages
1,801 (0.49/day)
Likes
307
Location
Toronto, CA
Processor Intel Core 2 Quad Q6600 Kentsfield L724A629 @ 3.8Ghz 423x9 1.47v
Motherboard Abit IP-35 Pro Beta 16.B04 (vDroop)
Cooling 6x Scythe S-Flex SFF21F, EK Multi Option 2 Port, dB-1 Compact, Swiftech MCR80, Swiftech Apogee GT
Memory 2x1 OCZ Flex XLC PC-9200 @ 5-5-5-18 1220 2.3v
Video Card(s) Inno3D Geforce 8800GTS 320
Storage Western Digital Raptor X 150
Display(s) Samsung Syncmaster 226BW 22"
Case Cooler Master RC-690
Audio Device(s) Creative SoundBlaster X-Fi Xtreme Gamer 7.1
Power Supply Thermaltake ToughPower 1000
#4
+1 Frank. Tribe has spoken :)

- Christine
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#5
woohoo

yeah that's the stuff!
I'm thinking tac switches icm's multiple identicle resistors and working in a pci-e psu power connector(realtime voltage display aswell). just ideas for now. Solaris17 and I have been discussing it in another thread so if interested it is on pages 8 and 9 of graphic cards-nvidea-8500/8600/GT/GTS Mod guide:)
 
Joined
Dec 26, 2007
Messages
1,801 (0.49/day)
Likes
307
Location
Toronto, CA
Processor Intel Core 2 Quad Q6600 Kentsfield L724A629 @ 3.8Ghz 423x9 1.47v
Motherboard Abit IP-35 Pro Beta 16.B04 (vDroop)
Cooling 6x Scythe S-Flex SFF21F, EK Multi Option 2 Port, dB-1 Compact, Swiftech MCR80, Swiftech Apogee GT
Memory 2x1 OCZ Flex XLC PC-9200 @ 5-5-5-18 1220 2.3v
Video Card(s) Inno3D Geforce 8800GTS 320
Storage Western Digital Raptor X 150
Display(s) Samsung Syncmaster 226BW 22"
Case Cooler Master RC-690
Audio Device(s) Creative SoundBlaster X-Fi Xtreme Gamer 7.1
Power Supply Thermaltake ToughPower 1000
#6
check XtremeSystems I think someone created a Volt Mod Guide on this exact topic, can't remember who or which card.

- Christine
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#7
...

I'd like to be able to adjust the voltage up and down in very small increments by means of two tac switches and have a realtime display of the voltage:) Small increments with power level adjusted before the card to hopefully avoid any serious reaction to voltage increase while operational
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#8
first thoughts

I don't have much experience with ic's but I have played with the old 555/6 and they have an output of 100m milliamps I'm sure there is an ic with suitable tolerances and a more favorable output 7555/6 is 20 milliamps and will accept 3-18 volts. so wouldn't a slew of 7556's give nice small increments of adjustment and be activated/deactivated by a single tac switch (each)seeing as they would be operating as a flip flop circuits.


haven't done the math yet...so 20 milliamps maybe too mch still?
 
Joined
May 6, 2005
Messages
2,786 (0.60/day)
Likes
435
Location
Tre, Suomi Finland
Processor i7 4770K Haswell, watercooled
Motherboard Asus Z87-C2 Maximus VI Formula
Cooling Fuzion V1, MCW60/R2, DDC1/DDCT-01s top, PA120.3, EK200, 3× D12SL-12, liquid metal TIM
Memory 2× 8GB Crucial Ballistix Tactical LP DDR3-1600
Video Card(s) between GPUs
Storage Samsung 840Pro 256@178GB + 4× WD Red 2TB in RAID10 + LaCie Blade Runner 4TB
Display(s) HP ZR30w 30" 2560×1600 (WQXGA) H2-IPS
Case Lian Li PC-A16B
Audio Device(s) Onboard
Power Supply Corsair AX860i
Mouse Logitech PM MX / Contour RollerMouse Red+
Keyboard Logitech diNovo Edge / Logitech Elite Keyboard from 2006
Software W10 x64
Benchmark Scores yes
#9
Why assemble a complex circuitry when you can achieve the same with 2 trimpots, 4 pieces of wire, DMM and a datasheet for the phase controller?
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#10
lack of knowledge

well thats sounds pretty simple. I've only been dabling in electronics for a couple of months so my knowledge is pretty limited. :)


too funny - wasn't sure was dmm was and anwered the post first before looking it up. I am not familiar with alot of the abreviations used in electronics.
 
Last edited:

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#11
...

I'd have to look for the perfect trimpot but I think it may be a little hard to get the very small adjustments easily. I could build the circuit I mentioned seperate so display and tac switches are usable outside the case and only solder pins to the card to connect wires. Being able to oc with tools such as rivatuner allows the option of not using the oc unless you want or need it. Thus disabling it whenever you want. v-mods with trim pots doesn't seem as freindly as my idea, and can't be disabled quickly. Just my thoughts though
 
Last edited:

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#12
...

I'd like to figure a way to control mutiple identical resistors with a two tac switches (up and down) to add or subtract small amounts at a time, but I'm not sure there is a single ic that is suitable (or even if an ic is suitable)I'm not sure I'm explaining correctly.
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#13
dip switch

a dip switch would work but not quite what I'm looking for. Thanks though
 

W1zzard

Administrator
Staff member
Joined
May 14, 2004
Messages
17,069 (3.44/day)
Likes
17,977
Processor Core i7-4790K
Memory 16 GB
Video Card(s) GTX 1080
Display(s) 30" 2560x1600 + 19" 1280x1024
Software Windows 7
#14


something like this?

you could have total resistance:

R1 + R3
R1 + R4
R2 + R3
R2 + R4

you will probably want to add another resistor before or after the switches which is always present to get a base resistance value.

why not just use a potentiometer?
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#15
Wizzard

First I'll start by saying I'm an electronics virgin(only dabbled for a couple of months). I saw the vmod thread in a google search and it peaked my interest. Decided I want to try it because I've only evr done software oc with graphics cards. I'm not happy with the performance of my card so why not. I've read as much as I could find related to the mods on my card (xfx 8600gt xxx) but I see some large benefits to making the mod a little more complex. The final results I'm looking for are two tac buttons and and realtime voltage display for each mod. The only soldering on the graphics card will be pins for the needed connections making it removable. It will add a small amount of voltage with each increment triggered buy one tac switch and subtract it with the other making it possible to disable the v mod completely. Each increment will have to be small enough to prevent any adverse effects fom the voltage increase. I would like to see if it is possible to avoid shutting down to make the adjustments along with making the adjustments without opening the case. A potentiometer doesn't have the controlled increments I would like. If it does please explain it to me.
thanksfor the diagram I can see how that would function but think there would have to be more switches and resistors to represent the increments I desire. I'll see if I can't put together some diagrams of what I'm thinking. The ic could be used as a flip flop switch with the tac switch. Depending on the ic, the output power along with something to absorb any fluctuations as it's switched could itself function as the increase. The ic's could also be powered by the psu instead of the motherboard to avoid problems on that end.

I haven't done the math or taken measurments of my card so I can't give exact numbers yet. It will also take some testing to figue out the safe size of the increments. I'm going to do more research on ic's to find the best one for the job.
 
Last edited:

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#16
Wizzard

For now well pretend that an icm7555 has the perfect ouputs. It has two states as flip flop. High and low. with a diode and a tac it could function as as on/off switch or without the diode it would have two different output levels. If those levels where set up properly they could represent two increments, lowering the amount of ic's required. Using 7556's would give me four increments with 3 tacs to control it. Of coarse this is all based on the ic having the correct output levels.
Not that I don't like buttons but this would probably be too many, if multiple ic's are required.


If I don't understand this properly please let me know.
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#17
does it exist?

a single ic with many states and the ability to activate/deactivate sequentially thus adding or subtracting states to the complete state of the entire chip is what I have in my head. does it exist?
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#18
...

I quickly put this pic together so have a laugh and get it out and move on. It shows the pencil/conductive ink voltage mod and an idea in the works. The pencil/ink mod seems to work for people but doesn't do the job as I think it should be done. The ic would take power from the psu and add it after the resistor that is being bridged in the manor i previously described. This would avoid taking the extra power from the motherboard the same way a 6pin pci-e power connector would. I can't say if the two power sources are too much for this card setup to regulate,(the 6 pin connector could be regulated at another point) but these cards seem to handle the power via the pencil mod so it may work reliably. The main concern would be suppliing the correct increase. The number of resistors is not accurate and will depend on future calculations, other components to complete the circuit will also be added.



http://forums.techpowerup.com/attachment.php?attachmentid=13319&stc=1&d=1207534202
 

Attachments

Last edited:
Joined
Apr 30, 2006
Messages
1,167 (0.27/day)
Likes
433
Processor 7900
Motherboard Rampage Apex
Cooling H115i
Memory 64GB TridentZ 3200 14-14-14-34-1T
Video Card(s) Fury X
Case Corsair 740
Audio Device(s) 8ch LPCM via HDMI to Yamaha Z7 Receiver
Power Supply Corsair AX860
Mouse G903
Keyboard G810
Software 8.1 x64
#19
heh I just saw your drawing. First you need to understand that the voltage regulator on your video card outputs a certain voltage which is set from the resistance of that pencil mod resistor. Shading it with a pencil lower the resistance and causes the voltage regulator to output a higher voltage. You wouldnt add any current to that resistor as your drawing shows. You just need to lower the resistance across the ends of that resistor. The easiest way to make that kind of circuit is to use a pot + a dmm to monitor the set voltage.



You would need to figure out the min and max resistance needed for the voltages you want. Then use a resistor and pot to set any resistance within the min and max range.

You could integrate everything into a 5.25in Bay slot:

 
Joined
May 6, 2005
Messages
2,786 (0.60/day)
Likes
435
Location
Tre, Suomi Finland
Processor i7 4770K Haswell, watercooled
Motherboard Asus Z87-C2 Maximus VI Formula
Cooling Fuzion V1, MCW60/R2, DDC1/DDCT-01s top, PA120.3, EK200, 3× D12SL-12, liquid metal TIM
Memory 2× 8GB Crucial Ballistix Tactical LP DDR3-1600
Video Card(s) between GPUs
Storage Samsung 840Pro 256@178GB + 4× WD Red 2TB in RAID10 + LaCie Blade Runner 4TB
Display(s) HP ZR30w 30" 2560×1600 (WQXGA) H2-IPS
Case Lian Li PC-A16B
Audio Device(s) Onboard
Power Supply Corsair AX860i
Mouse Logitech PM MX / Contour RollerMouse Red+
Keyboard Logitech diNovo Edge / Logitech Elite Keyboard from 2006
Software W10 x64
Benchmark Scores yes
#20
I'm not sure what you mean by the label "V source motherboard" in that pic? Please elaborate.

I gather your plan would be not to alter the voltage applied on the phase controller's feedback pin but rather to feed the pin your own generated voltage. I assure you, this will not work as a sensible voltage mod.
What it seems you have described here is a monsterously complicated reverse voltage mod, a coupling like this (if it did work, which I doubt) would force the phase controller to actually lower the output voltage of the PWM.

Skitzo said:
v-mods with trim pots doesn't seem as freindly as my idea, and can't be disabled quickly

(...)

A potentiometer doesn't have the controlled increments I would like. If it does please explain it to me.
That's exactly what a potentiometer has - a true stepless control, with a trimmer of optimal rating you can tune the voltage in the scale of millivolts (0.001v). And if you want a way to disable the mod quickly you could just put a dip switch/jumper in series with either leg of the trimmer meaning when you flick the switch/jumper the mod-circuit breaks.
 

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#21
thanks for the input.

Thanks all, I was hoping to get feedback.

EDIT: ok the first part has some erors but I thought it might be usefull to people in seeing where I was going wrong, if not let me know and I'll delete it.

I understand that changing the resistance(pencil mod) increases the voltage being produced by the votage regulator. My understanding is that by shading the resistor current is being added, a path is being created around the resistor (which restricts current) to allow more current to reach the voltage regulator thus increasing the output voltage of the regulator. I also understand that the versions of the particular card I want to apply this too has some diifferneces from it's big brother (mine 8600GT => big bro 8600gts). The 8600gts has faster clock speeds and an added 6 pin pci-e power connector. So that tells me that if my 8600gt will reach 8600gts speeds it will most likely require 8600gts power levels aswell. I realize people are making this mod work by taking the excess power the card requires from the motherboard(pci-e motherboard slot), done by shading the resistor, using conductive ink/paint and adding resistors or trim pots.
Some people are able to add the 6 pin connector to the card (some pcb's have the 6 pins configured so a connector just needs to be added),and just plug it in for the power(so I've read). If adding the 6 pin connector eliminates the need for the other types of volt mods, but isn't an option with a particular card, wouldn't a cross between the mods work aswell. As long as the power I add is complimentary to the power already flowing through the circuit, why would it matter where it comes from as long as the card gets the power it requires. I'm talking about making the same adjustments that everyone else is doing, just looking for another way to do it. Taking the power from the motherboard can be done, it has been done, but I understand it's not the greatest thing to do. I'm just considering options. Yes the circuit I'm talking about is more complex, but it's the end results I'm interested in, the functionality. Do I need it, no. I could go buy another graphics card, that would be the simplist. From what I've read, changing the voltage while the card is in operation hasn't been done yet or is constantly advised against(I'm sure it's been played with though).
Seems like a nice option for someone who wants to oc with voltage increases but doesn't want a perminantly active mod. Trim pots work great for adjusting resistance, and stepless adjustments have their benefits. However, in this situation I think stepped adjusments would work a little better for reducing adverse effects of voltage increase. It seems to me that the smaller the increment the less chance of adverse effects. I know 5 min turning a dial or 5 min pushing a button 100 times... I like buttons better than dials

I need to understand how the voltage regulator works, I think thats where this idea is going wrong. (virgin remember) The feed back pin is where I went wrong. It dosn't feed the volt reg it's power, it sets the level of output. (thanks Frank & Largon)

So adding power isn't wrong, I was just thinking the wrong spot.
Can no one see the benifits of a circuit like this?
How the resistance is adjusted isn't the important part( the tac and ic is just my twist and finding the safe stable method is a given), the main focus would be giving the card the optional extra power from a safe source with the ability to make adjustments (or disable the mod) while the computer is running. Obviously the voltage display is fluff, but hey, it can't hurt right:)
 
Last edited:

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#23
...

Ok, if I skip the added power from psu, what about my idea of switching the resistance levels in steps via the tac switch applied in the same manor as the pencil v-mod. Multiple identicle resistors activated by transistor via flip flop via tac switch. Using the components I know of, this would be a rather larger circuit though. I think it may be possible to use both high and low outputs as increments but it would require a switch to activate/deactivate the ic as well as a trigger and reset switch. Components would have to be added to detect the output and act accordingly. If I could find something like an ic with many states (more than high and low) the detection circuit might be a simpler way to go. I really need to learn more. There may be another component that would do this task the way I want but I havn't learned of it yet.:roll:
 
Joined
Apr 30, 2006
Messages
1,167 (0.27/day)
Likes
433
Processor 7900
Motherboard Rampage Apex
Cooling H115i
Memory 64GB TridentZ 3200 14-14-14-34-1T
Video Card(s) Fury X
Case Corsair 740
Audio Device(s) 8ch LPCM via HDMI to Yamaha Z7 Receiver
Power Supply Corsair AX860
Mouse G903
Keyboard G810
Software 8.1 x64
#24

Skitzo

New Member
Joined
Apr 3, 2008
Messages
720 (0.20/day)
Likes
86
#25
one step adjustment

this circuit should work for a one step adjustment but the voltage increase would most likely be too much for use while the pc is operating. If the resistors used determined the appropriate safe and stable increase, many of these circuits could be joined to provide the adjustments I'm looking for. Each increment would require a tac and the V+ for the ic would need to come from either +5 standby from the psu or a battery (with the ic's I've worked with, when they lose power, the next time it is powered it will be in the high state.

I will have to add a diode to make low = off
and draw a better diagram.:)
 

Attachments