I-Appliance BBS The Official Source for Internet Appliance Upgrades and Mods Click Here! BBS Main List | Sign In | Sign Up | Search | Help | Linux-Hacker.net Reply to Thread | Printer |

Home / I-Opener Areas / I-Opener Technical Stuff Hacked V3 boost to 350 Mhz, low risk mod. Pretty Easy

Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) BadFlash Profile | Email These changes will take a V3 (Goo’d bios, WINCHIP, no SW4) to 350 MHz with very minor changes. Needed: K6-III+ 350 or faster Lasagna cooler (or eqiv) 1/4 watt 10K ohm resistor. 4’ 24 guage wire Soldering Iron with fine tip. Fine solder. Solderwick Experience soldering Steady hands. Optivisor (unless you have GREAT eyes) General notes: 1. Work quickly so you don’t overheat and damage traces. 2. If the area to be soldered has no solder on It, tin and remove excess with solderwick. Keep in mind note 1. 3. Tin all leads before attempting to solder. 4. Use clips or other tools to hold components in place before soldering. The soldering iron will drag them away from where you are working. Fingers are a poor choice for this. 5. Be sure the solder flows on both the board and the component. Keep in mind note 1. 6. All references are given with the CPU socket located in the Upper Right. Instructions: 1. Remove CPU. 2. Remove R60 and R61 and save. These are located above the CPU socket 3. Move these 0 ohm resistors to R184 and R185 located at the lower right corner of the CPU socket. This provides for split voltage CPU’s. 4. Verify the value of R342. R343 and R343 are located at the top right of the sodimm socket. Should be 10K. R343 should be empty. You will need a resistor equal to the value of R342. This does not need to be a surface mount resistor. You can carefully bend the leads of the resistor and trim the leads so it will fit between the solder pads for R343 and the screw pad for the RF shield. Solder in place. 5. De-solder and remove SW3. Use your solderwick to remove the excess solder and heat all the pins on one side at the same time will gently prying up on that side with a fine blade jewelers screwdriver. You want it to come up on one side, not rip it out by the roots. Repeat on the other side and remove SW3. 6. Remove R130 just below SW1 at the right side of the CPU socket. 7. Solder the dip switch removed from SW3 in place. Set the dip switch so the the middle two switches are ON and the outer two are OFF. This sets the multiplier to 5X. 8. Cut 4 feet of 24 gauge speaker wire. Coil this so that every second turn is in the opposite direction to avoid making It an electromagnet. Solder this in parallel to R302 to boost the available current above 6 amps. The wire is actually a low value resistor. 9. Power up the board and measure the voltage at the lower pin of Q16 and be sure it reads 2V+/-.1V. 10. Drop in your K6-III+, and fan, close it up & test it. For V4 and V5 owners, all you need to do is set the dip switches on SW4 (located above the SODIMM socket) to both off to set Vcore to 2.0V, then do steps 5 through 10. The hardware changes will get you to 333Mhz. You can use SOFTFSB to raise the bus speed from 66 to 70 MHz to get 350MHz. My tests with 23C air temperature gives me 42C CPU temps and 73C board temps near Q16 as read by Motherboard Monitor 5. These temps are safe for continuous duty. 10-29-2000 12:52:13

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) Turbo3 Profile Running your K6-III+ at a vcore of 2.0v and 500MHz stands a very good chance of burning out your power brick. I recommend a 1.6-1.7v core voltage. Although the iopener power regulator can supply 3.5 volts at 7 amps the external power brick can not sustain that level for continuous operation. But at the lower vcores you need better cooling for Q16. I also do not agree that 73c is a save temp level to operate at. The capacitor that is between the heat source (Q16) and the temp sensor is only rated for 80c. If the sensor is at 73c then the capacitor being closer to the heat source is going to be hotter. The capacitor is right on the same copper foil as Q16 whereas the temp sensor is close to but not on the copper foil. However, you can probably run for some time before the capacitor dies. Component life testing is done at elevated temperatures to speed up the affects of time. If you make only these changes then you should stay in the 300-333MHz range. People who want to go higher need to use lower vcore and provide cooling solutions for Q16. If you must run at a vcore of 2.0v then look into a higher wattage 19v power brick. 10-29-2000 15:12:37

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) BadFlash Profile | Email I think that is what I said here. I'm offering something here for the average guy. I don't want anyone going above 350 with these changes, but these are easy to do , low cost changes with little risk of board damage due to the changes. 350 seems to be very stable. Got a reference for 80C? My book gives me 85C for the capacitors. In any event, this is enough margin for me. If you can tell me where the temp readings are comming from that would be a great help. 450 or 500 MHz without adding cooling resitors would be insane. I don't suggest that the average person should try to lift Q16 to add cooling resistors. What I am suggesting here is the maximum bang for the buck with average skills. Getting a K6-III+ running at 333-350 gives you a pretty zippy beast without serious risk. 10-29-2000 19:49:47

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) Turbo3 Profile Why pick such a high vcore for the K6-III+? A value of 1.8 or lower is better. If a V3 owner is going to go to all the trouble to lower vcore they should pick a value that puts the least stress possible on the iopener and 2.0vcore is not it. A vcore of 2.0 only makes sense for a V4/V5 owner with an SW4 and only because they can get that value without making any changes. If you can get the current down with a lower vcore then you might be able to skip adding your 4 foot of wire to R302. But I would need to do some measurements to see if that is possible. I should have said around 80c. Yes, most caps run in a -40c to +85c range. But life time at the extremes can be in the 1000 hour range. The 73c measurement from the board only tells you that the cap is hotter than that value. A higher ambient (room temp) will also increase this value so someone running this in their car on a hot day could see much higher values. I would prefer to see a temp in the high 50c range to give a safe margin. But to each her/his own. 10-29-2000 22:17:16

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) BadFlash Profile | Email I tried it at 1.6V with no bypass on R302 and it kicked out as soon as I plugged it in. I picked 2V as it was really easy to find a 10K resitor. Smaller values will give lower voltages. 2V worked and gave me temps I would deal with. I am not as comfortable as you are at running my K6-III+ at lower than rated voltage. Folks should feel free to do whatever they want as far as that goes. It isn't much trouble to modify following my instructions. My second one took me around 30 minutes, but you are right. Soldering one value is as good as any other if it works. I wasn't going to recommend people do something I hadn't tried. You and others have had great luck running at low voltages. I haven't. Again, my goal here was to give great gain with low cost and minimal risk at messing up your board doing the changes. You could be right about the wire causing some of my gremlins. Could be some sort of inductive spike despite my best efforts. Shorting out R302 didn't work for me, as it appears to work for you, so all things may not be equal here. Reducing core voltage does not have as much an effect on reducing Q16 temps as one would hope. As the voltage drops in the processor, the voltage drop across Q16 increases. The curent may be lower, but due to the increaced voltage drop, it stays just as hot. Sure the CPU may run cooler, but that has never been an issue if you are using a fan. 10-30-2000 09:53:20

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) BadFlash Profile | Email It may be counter intuitive, but lowering Vcore below 2 volts to 1.8V may increase the heat load on Q16. My calculations show you have to go down to 1.5 volts before you end up with the same heat load on Q16. I have no hard data to support this, but if you added cooling resistors, you do. You can measure the voltage drop across the cooling resistors vs Vcore and can see what is happening and calculate the current. I assumed that the current drawn by the processor is linear with respect to voltage (unproven) and acts like a resistor. Q16 adjusts its voltage drop to whatever you tell it. Heat load on Q16 is V*I. If it is 7 amps at 2.2 volts. it would be 6.36 amps at 2 volts and 4.77 amps at 1.5 volts. The voltage drop accross Q16 is 1.5 volts at 2 volts Vcore and 2 volts @ 1.5 volts vcore. Volts times amps is the same. If you want enough current to go to 450Mhz, you need to drop the voltage, but that does not appear to help you with Q16 heating. 10-30-2000 10:33:24

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) Turbo3 Profile I suggest you install a pot to vary vcore with the case closed. I have found that a lower vcore always gets me a cooler Q16. Power is Volts times Amps. With lower vcore the current seems to drop faster than the voltage going up across Q16. What speed were you trying to run at with a vcore of 1.6? 10-30-2000 10:35:10

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) BadFlash Profile | Email 333 initally. 66X5 and then use SOFTFSB to get to 500. All my tests are done with the cover on, system closed up. Been trying to avoid using a pot. I'll bite the bullet with my next one. What is the minimum resistance of R302 you have been able to use? 10-30-2000 14:05:01

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) klotz Profile Are these instructions valid for early V4 (SW3 but no SW4)? 10-30-2000 17:27:03

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) BadFlash Profile | Email Yes, as long as you can figure out what you have and what you don't. If you have a rise266 you already have the split voltage. Just run down the list of what is there and what isn't. If you have a winchip, for all intents, treat it like a V3. The V3,V4 and V5's all have very similar screenings. Use the references I've listed and see if you can locate everything and what is installed and what is just a pad. 10-30-2000 20:43:30

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) klotz Profile Jack and all, Thank you for this concise discussion. I have an early V4 (Goo’d bios, no SW4). It last had a K6-2 in it from a previous owner. After all the reading on the various threads, it seems to me that below <2v the 333 MHz K6-III will have trouble with the HLT instruction and perhaps other problems, and furthermore the additional voltage drop on Q16 may be a problem. So I'd like to go with 2V to run the chip at 333MHz. But I'd like to use a resistor instead of wire because I'm concerned about the inductance and possible RF problems (noise). What value resistor should I use? On another thread I read about using a 0.012 ohm resistor instead R302 details: The one I used was a 0.012 ohm, 1 watt, 1%, 2512 (size), Dale WSL-2512-R012 DigiKey also has some for about $2.69 each by Panasonic DigiKey #P12MCT-ND. Also, is there any advantage to putting a heat sink on Q16 if you have the lasgna fan? Perhaps a piece of copper over to between the fan the the CPU? Or just a finned heat sink on Q16 itself with no other cooling on it. Thank you! 10-31-2000 17:29:35

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) Turbo3 Profile Yes, that is the resistor (R302, 0.012 ohms). Yes, place a heat sink on Q16 if you can. My daughter's system is setup with a K6-III 333MHz running at 333MHz with a heat sink on Q16 and the 0.012 ohm current limit resistor. (It could also be running at 300MHz, I will have to check on that.) R302 is hard to remove, so don't. Just soldier the 0.012 on top of it. In fact if you solder it on top you can use any value from 0.050 to 0.012 ohms as long as it is the correct size and wattage. 10-31-2000 17:38:06

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) leighklotz Profile (I'm the same user as klotz; couldn't get the password to work even though I got e-mailed a new one.) Today with the help of a friend who does SMT work for a living, "I" upgraded my early V4 following these instructions. I used a 0.02 Ohm SMT resistor in parallel with R302 instead of the wire, and used a 10K Ohm SMT one in place of the 1/4 watt suggested. It ran first time. I let it sit in the BIOS for a while and the main board temp got to 50C and I turned it off. This is with the plastic cover but no RF shield, lasagna fan, no heat sink on the voltage regulator. I will go to Radio Shack and get a heat sink and some thermal paste and try to attach it to the lasagna fan (between it and the CPU?). These instructions are for 5*70MHz but I wonder if it might be better performance to do 3x100MHz? Would this have roughly the same power consumption, or would running the bus at 100MHz take more power? Thank you, Leigh. 11-07-2000 16:54:27

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) leighklotz Profile I should add I'm running Linux so I can't use SoftFSB, though I will look for an equivalent program... 11-07-2000 16:55:02

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) leighklotz Profile Here's my results, at 2.06v: 166, 200 MHz boots and works 233 MHz boots if USB unplugged 300 MHz boots up to disk access, then stops 333 MHz stops when Linux says "testing HLT instruction" I haven't tried 300 or 333 w/o USB. 11-07-2000 17:57:32

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) BadFlash Profile | Email What did you do for R302? This is the only reason I can think of that you wouldn't be able to get to 333MHz. I've done this same basic hack with 3 I-O's, 2 V3's and a V4 with no troubles so far. 11-08-2000 06:27:15

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) Tackhead Profile Leighklotz: I can help you on the HLT instruction - just boost your VCore. I had the same problem on a Windoze K6-III installation. The thing was stable at 1.91V, but when I installed any one of three separate "CPU cooler" programs, the IO would lock up hard when the HLT was executed. It confused me for weeks, until I boosted VCore, and then the problem went away. I traded more heat on Q16 during CPU-intensive operation for the ability to cool it off during idle time. If your IO is running fine, and then the screen powers off, you've got an R302 problem. If the IO's running fine and then locks up solid with the display intact, you've got an undervoltage condition that exhibits itself when you execute the HLT from which you never recover. If the video is corrupted when it crashes, I'm not sure what the problem is, but my hunch is that it's an undervoltage condition. 11-08-2000 10:18:51

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) leighklotz Profile Thank you all for the help -- I hope that others can benefit as well. I'm using a 0.02 Ohm resistor (SMT) in parallel with the existing R302. It's labelled 20M. The screen does not blank out -- I just get bad behavior -- HLT really does appear to halt! (At least it doesn't execute the old HCF instruction -- Halt and Catch Fire. :) My vcore is 2.06 according to the BIOS and 2.05 according to my DMM. (It's not a Fluke but it is yellow. :) I went to RS today and bought some 1 Ohm power resistors and a heat sink for Q16. I will try upping the voltage a little. The only SMT resistors I could find bigger than 10K were 33K (at Fry's Palo Alto), so I can get either 33K or 16K by paralleling them. Thanks for the help, Leigh. 11-08-2000 11:49:15

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) BadFlash Profile | Email 1 ohm is way too big! The basic idea is to supply at least 2.3 V to Q16 at 7 amps. V4/V5 supply appears to be 3.3V, V3 and below appears to be 3.5V. Figure you want the cooling resistors to drop 1 volt at 7 amps, This would mean you want a 1/7 ohm resistor. If you are running a lower Vcore, you can adjust the resistance a little higher. The supply to Q16 needs to be at least .3 Volts above Vcore to work properly. I'm using 3 .47 ohm resistors in parallel. 5 watt .47 ohm resistors are available from radioshack. .02 may be too big for R302. 11-08-2000 14:22:38

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) leighklotz Profile I tried another K6-III and have the same problems with my early V4 (No SW4). I still get OK at 200MHz and reboots at 333MHz with the disk plugged in and HLT failures with the disk removed. (It runs at 200MHz but given I already have a 233 MHz Winchip 2 and that's faster for some tasks, I'm reluctant to settle.) The rebooting seems to indicate that the power supply is cutting out AND the HLT shows my Vcore is too low. I've tried shorting out R302, but it gives no improvement. I'm beginning to think that I have a bad connection on R302 and will re-solder it. I have an 0.02 Ohm resistor in paralle with the 0.015 Ohm resistor, supposedly giving me 0.0086 Ohms. Since I also have trouble with the HLT instruction, at 2.07V, so I think I may have to boost it. I get 1.18V at the junction of R342 and R343, and at Pin 2 of U16, and 4.961V at Pin 1 of U16. I have 10K resistors in both (well, they measure at 9K but my DVM reads in Kilohms). Any suggestions on a VCore to aim for for 266MHz or 300MHz and a small heat sink on Q16? I may be able to go for 16K Ohms. 11-12-2000 18:07:08

CPU Question (modified 0 times) scooter360 Profile | Email Here's my question: I have a ver 4 with no goo sw3, and 4 onboard RISE 266, IDE Connectors intact, hacked ver5 bios. What I'm wondering is I have a AMD k62 400 cpu with a core voltage of 2.2v I could care less of running the cpu at 400 or even at 333. I want to run it at either 266 or 300. The sole perpose of this unit is to be a mobile mp3 player running Win98SE. My problem is Win98SE is not loading the audio drivers. I do have the correct yamaha drivers downloaded and installed. from what some guys are telling me the cpu is the problem. What others are telling me that the chip is not a yamaha chip. I'm a newbie in the IO world but not when it comes to computers and electronics. What is the switch settings for sw3 and sw4? please email or just reply thanks. Pooter mineispooter@yahoo.com http://www.adsl5.com 12-20-2000 17:01:36

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) ASPguy Profile | Email sw switch settings are here: http://www.linux-hacker.net/cgi-bin/UltraBoard/UltraBoard.pl?Action=ShowPost&Board=technical&Post=1632&Idle=0&Sort=0&Order=Descend&Page=0&Session=aspguy.9769143702451 as for your sound, open up and look at the sound chip on the motherboard. if you can find a chip the size of a square inch that say yahama then you have yamaha sound. 12-20-2000 18:39:23

RE:Hacked V3 boost to 350 Mhz, low risk mod. (modified 0 times) IO_Dood Profile I have a Problem with the switch seetings. I have perfromed the hack as per Badflash ( ihave a v4 with 128 megs and a k62 333 @ 200Mhz). I have just moved the pos switch to sw1 and it works. But as for the correct multiplier settings it appears like its not what you other guys post. (I searched the multipliers and set them correctly but they do not correspond to what I found on the board. Heres a copy of what I found posted by whyo With the sw1 & sw2 in this position ______________________________________________________ ----SODIMM------ - - - - - - - - - - - - - - - - - - - - - SW2 - - - - - - - - - - - - - - - - - - - - 4 3 2 1 -SW1- 4 3 --- CPU 2 1 ____________________________________________________ CPU Multiplier SW1 3 2 1 --------- X X X ----4.5x X X O ----5 x X O X --- 4x X O O ---5.5x O X X --- 2.5x O X O --- 3x O O X --- 2x O O O --- 3.5x Bus Speed SW2 4 3 2 1 ------------ X X X X -----124 Mhz X X X O -----120 Mhz X X O X -----115 Mhz X X O O -----110 Mhz X O X X -----105 Mhz X O X O ----83.3 Mhz X O O X ----96.2 Mhz X O O O ----- 75 Mhz O X X X ---- 100 Mhz O X X O ------95 Mhz O X O X ----- 83 Mhz O X O O ----- 80 Mhz O O X X ----- 75 Mhz O O X O ----- 70 Mhz O O O X ----- 66 Mhz O O O O ----- 60 Mhz My multipliers are not as they say above when i switch the coresponding switchs. My question is am i missing anything??. I assumed the bus was set to 100Mhz as I had origianlly had a Rise chip in there, and when I did switches for a 2.0v core My K6 2 ran fine at 200Mhz without any other mod. Now I just want to try it at 2.5 multiplier but cant seem to get it. Any help would be appreciated. PS Im running at 300Mhz now but as I do not hat a fan I wanna lower the speed to 250Mhz. Thanks 12-20-2000 19:33:19

Reply to Thread | Printer | All times are PST Powered by UltraBoard v1.62

I-Appliance BBS General - NEWS, Breaking NEWS - linux-hacker.net BBS on CD Questions - User Groups in Your Town - E-Bay, Yahoo Auction LINKS ONLY (MODERATED) - Jokes & Humor - usdtv box I-Opener Areas - I-Opener General Posts - I-Opener Technical Stuff - I-Opener Technical Info Reference (MODERATED) WebSurfer Areas - WebSurfer General Posts - WebSurfer Technical Stuff - WebSurfer Technical Info Reference (MODERATED) Other I-Appliances - Virgin Webplayer - WebPal - MailStation - 3Com Audrey - ThinkNIC / Oracle NIC - Gateway/AOL Connected Touch Pad - Philips/AOL-TV - ACER NT-150 STB - MSN Companion - Compaq IA - Emachines IA - IBM NetVista IA - Intel dot.station - MSNTV2 Network Computers - Fujitsu-Siemens Scovery Linux Net Computers - GallantWeb Server Consoles - Dreamcast - xbox hacking Web Pads - Epods One General Post - Panasonic CF-01 - Fujitsu Stylistic 1200 - Cybiko Thin Clients - Thin Clients DigitalVideoRecorders - DTV Recorders, Replay, TiVo, Etc... - Tivo Hacks Linux - Midori Linux General Board - ITX MotherBoards Single Board Computers - Earth Computers mARMalade Elite Member Area MISC Areas - The Swap Meet WE TAKE NO RESPONSIBILITY FOR ANY TRANSACTION HERE AND NO COMMERCIAL POST PERIOD !!! - Wireless Net - Cameras - Robotics - Routers/Gateways - Mattel JuiceBox