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 Info Reference (MODERATED) Setting Up FSB Clocking and CPU Clock Ratios Setting Up FSB Clocking and CPU Clock Ratios

Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) duvell Profile Here is a quick table of the resistors you need to add or remove to change the FSB frequency and the CPU clock ratio. There are two sets of resistor locations that makeup these settings. According to the table you will need to either add a 0 Ohm resistor or remove an exisitng resitor. I will use U to designate an unpoulated resistor site and P to designate a resistor site populated with a 0 Ohm resistor. Most of this information was found by others, or I read in one of the specs, or I have otherwise made it up. Use any of this at your own risk your milage and results may vary. And if you don't know what you are doing don't try it. Depending on the processor you are using, you may need to change the processor core voltage. That information is not included here. You should be able to use these with the Winchip and Winchip2 without modifying any of the voltage settings. Setting the Front Side Bus FSB Freq. PCI Freq R331 R330 R229 R208 60 30 (CPU/2) U U U U 66.8 33.4 (CPU/2) U U U P -- Default IO Setting 70 35 (CPU/2) U U P U 75 25 (CPU/3) U U P P 97 32.3 (CPU/3) U P U U 83.3 27.7 (CPU/3) U P U P 95.25 31.75(CPU/3) U P P U 100 33.3 (CPU/3) U P P P 75 37.5 (CPU/2) P U U U 96.2 32.0 (CPU/3) P U U P 83.3 41.7 (CPU/2) P U P U 105 35 (CPU/3) P U P P 110 36.7 (CPU/3) P P U U 115 38.3 (CPU/3) P P U P 120 40 (CPU/3) P P P U 124 41.3 (CPU/3) P P P P Setting the CPU Frequency Ratio uses the same notation. Some chips have different sets of ratios depending on what version they are. For example later versions of the k6-2 have a 6x multiplier that earlier versions did not. Double check the values for your specific chip before making these settings. The settings I provide should be okay but I make no guarantees about how safe or accurate the information is. CPU Clock Ratio Settings CPU Clock Ratios R328 R130 R327 Winchip Winchip2 AMD K6-2 P P P RSVD 4.5x 4.5x P P U 5x 2.33x 5x P U P 4x 4x 4x P U U RSVD 2.66x 5.5x U P P RSVD 2.5x 2.5x U P U 3x 3x 3x -- Default IO Setting U U P 2x 3.33x 2x (6x in newer revs ?) U U U 4x 3.5x 3.5x Thanks to Codeman, William, Solder Boy and others who have contributed to compiling this info. 04-25-2000 07:29:21

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) Hawk-I Thanks Duvell for compiling all that info!!! So, lets say a guy has an I-O with a Winchip/WinChip2 and a stock heat sink/Lasagna Cooler. Bumping the mult. to 3.5 to reach 233 MHz is probably going to be OK for long term use, at least with the lasagna cooler. What about bumping the mult. to 3.5 and the FSB to 70 MHz to reach 245 MHz? Any opinions on short/long term stability? 04-25-2000 14:02:15

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) Hawk-I Nevermind, I read the posts in the technical stuff section. For others who are interested, a WinChip2 with Cooler runs stable (for at least 2 people) at Mult 2.33 * 100 FSB. Booting is possible at 2.5/2.66 mult * 100 FSB, but stability and the need for greater cooling becomes an issue. A lower FSB setting might be used to allow higher mult. (2.5/2.66) to work, but I believe people try to maximize FSB to achieve the highest memory bandwidth. 04-25-2000 19:10:05

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) QuintLeo The AMD K6/2 from the 400 up uses the "Intel" 2x settings for a 6x setting. 350 and down retain it as a 2x setting - I'm not sure about the 380. Down side, K6/2 needs 2.5 or less (speced at 2.2) core voltage.... 04-29-2000 22:40:37

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) hevnsnt What about the stability of the "stock" Winchip.. I would like some more info on that. (ACK have not got my Winchip2 yet!) -hevnsnt 04-30-2000 00:26:34

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) Tinman Profile Duvell: On the FSB Settings, you have R229 listed as one of the places needing a zero ohm resistor. I tried those settings for my K6-3 400 and ended up with an overall CPU speed of 266 Mhz. I was wondering if that was a typo, and you meant R329. Is this the case? If not, I'm back to the drawing board in search of the proper settings. And I would like to thank you for compiling that information, it sure makes things alot easier than reading all the different threads. 05-03-2000 15:32:17

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) duvell Damnit Tinman I refuse to beleive it until you post some pictures that we can all see! Sorry, I couldn't help myself. When I saw the name Tinman it brought to mind all of the grief you were given because you were being helpful and reporting valuable new information. I don't have a board in front of me so I cannot verify the actual R numbers. There is a pad for a 4 position DIP switch over near the clock chip. Three of the 4 resistor positions are actually in the middle of the switch foot print. The other is off to the side. You can measure accross the pads of the switch to verify what you are really set at. One side will all be tied together and grounded (I think this is the side with pins 1,2,3,4). If one of the zero ohm resistors is in place, it will short one of the pins on the other side (8,7,6,5). So if you check the continuity from pin 1-8, pin 2-7, pin 3-6, and pin 4-5 you should beable to verifyeha the FSB is really set to. If I did mislabel my chart with an incorrect R number, the data on the chart is still good. For a 100MHz FSB you should see "1000". Which means pins 2-7 and 3-6 will be grounded. And either 1-8 or 4-5 will be grounded too. I cant remember which was teh MSB and which the LSB. But you will see the pattern "0001" or using the notation of short and open, "OSSS". Sorry I don't have more specific info right now. But it is fairly intuitive. Run a program like Sandra 2000 that reports detailed system info. It will tell you what the FSB is really running at and what the CPU is really running at. It takes some of the guess work out of it. --duvell 05-03-2000 23:37:33

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) Tinman Profile Duvell, I'll never forget that day for a LONG time, one for the books. I showed them picyures of the cut pins and the goo, and they still didn't believe it, or didn't want too. Then they thought they were so important that I would screw up my own board just to screw with them. Heh. Anyway, I thought R229 was labled wrong, it should be R329. Thanks for the info, it helped alot. I have a K6-3 400 here, and the L2 cache makes a world of difference, to bad I don't have any pics :) Only problem is, I can't get it to run any faster than 167 Mhz, yet SI benches 3 times as fast, and it seems to be running atleast twice as fast as my Winchip 2 200 @ 233Mhz. My multiplier is 2.5 and a FSB of 66. For some reason I was unable to produce anything better. I am using the Tennmax cooler, and it's running at about 79 Degrees F according to Hardware Monitor Pro. I'm happy, but not happy enough, I am going to play with it more when I reup on some more patience and sleep. Once again, thanks for posting the info, it helped ALOT. Trail and error sucks, and digging through threads is almost as bad. You saved me alot of work. BTW: didn't you mention you were going to get a K6-3 400? May have been solder boy. Anywho, catch you later on. No Pics, No Lies, No Sh_t. Tinman 05-04-2000 11:37:21

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) duvell Tinman: Yes I did get a K6-3 but havn't done anything with it. I also got some surface mount DIP switched (like the ones near the memory) because they will mount perfectly where the switches for the FSB and CPU multiplier go. I can remove the existing 0 ohm resistors and then try the various combinations with ease. Adding and removing the resistors isn't that difficult but its a pain when you want to experiment with several combinations. My real question for you is what did you do to get your core voltage down to 2.2V, or are you running at the lower limit of 2.5V? Please post info if you have a way to set the core to 2.2V. I wonder if the K6-3 requires too much power for the IO to deliver when you try to run it at the higher frequencies. For CMOS power consumption increases proportionately with the frequency; or it may be worse, with the frequency squared. --duvell 05-04-2000 16:41:30

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) Tinman Profile Duvell My voltage wouldn't go any lower than 2.5. I have overclocked cpu's and ran the voltage at 3 10th's higher with great success, so with that past experience I took the chance with the K6-3 400. It's been running for almost 18 hours at this voltage without a burp or complaint from the IO. If I were you, I'd let the guinea pig [ me :) ] run the test for a week or so, and let me re-post my results here. But I honestly feel the the 3 10th's will not be a problem. Now I have a question for you. I have a power supply here capable of 3.16 amps, if the power requirements are exceeding the IO's capabilities, do you think this power supply could help, or do you think the IO would starve for voltage regardless? I think you've hit the nail on the head. As I said in my previous post, I can't do any better than 166 Mhz [ 2.5 x 66 ], all I get is a black screen. I did get it to post at 200 Mhz once or twice, but as soon as the bios posting was over it just rebooted itself. A place called battery biz has every tip size imaginable for power bricks, and all of them are rated at 3.16 Amps, $45 + shipping seems very fair to me. I had this brick here already, but it isn't mine. The tip fits the IO perfectly. I am going to retry my settings using it. If I have any greater success I will certainly post my findings here. I honestly don't feel the low Mhz yield is a waste, I can see a performance increase, even though the video is slower than a turtle. I wish I could dedicate more of my ram toward video, but I haven't found a way to do so as of yet. That would further improve performance, 2 meg is a joke by todays standards. Tinman 05-05-2000 00:50:02

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) duvell Thanks for the reply Tinman. I figured I would try the 2.5V because I agree with you. Running the voltage up 10% or so should be fine and will probably run faster, although hotter. Solder Boy first posted about the 2.5V limit on the normal technical board. I think the thread name was K6 at 300MHz or something close. Search on his name and you'll find it. He said he used an external source to supply 2.2V for the core of his K62 and was able to runn much faster than the 233MHz wall he had hit. But he soon ran into some bigtime thermal problems. I have asked him for some details on what he did. What I have done many times before (I designed processor boards for workstations for several years) is to isolate the core voltage plane and then redrive it from an external supply. Often we would have a stack of three or four external supplies powering the different power planes and a function generator plugged into the system oscillator socket. It's not difficult at all. It would be nice if we had some schematics though. If you can find the source of the core voltage. I mean the actual outputs from the voltage regulator an isolate those (desolder the pin or sonething similar) then you can attach your external power supply at that point or accross a cap on that voltage plane. There are three pairs of large 0 ohm resistors surrounding the socket7 socket. One pair on each of three sides. In the voltage modifications I have seen, these are changed. I assumed it was changing from a single voltage to a split voltage. I would venture a guess that if all three sets were removed the core voltage would be completely isolated, and not driven. This is speculation on my part, but these are things that shouldn't be too difficult to test with a meter. Good luck and let me know what you find. If you like drop me an email at: duvell at beer dot com --duvell 05-05-2000 06:24:50

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) mr_binary Profile Duvell: I'm finally getting around to looking at doing the CPU mod. Codeman's pictures show that BF0-BF2 are set from the unpopulated SW1 located to the right of the CPU socket just above R130. Your chart says that the clock multiplier is set by populating/unpopulating R130, R327, and R328. Who's right? Has anyone tried setting the multiplier by using SW1? Seems like an easy mod would be to move SW3 since all 4 switches are open ... 06-07-2000 21:07:16

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) duvell mr_binary: The resistor positions I mentioned are connected to SW1. They did it this way so they could remove the resistors and add a micro DIP switch (like SW3) or they could just put zero ohm resistors. I just called out the actual resisitor locations. It was pointed out in another thread that I may have two of the resistor locations reversed, as far as which BE bit they correspond to. It is in another thread but I can't remember where. If you put the switch from SW3, it will be easy to experiment and see. duvell 06-08-2000 05:52:52

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) Jim Hearne Yes, there is a mistake or 2 in the list, the 2.5 settings give something else which i forget. Fitting the switchs is definatly preferable. JIM 06-08-2000 11:34:09

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) mr_binary Profile Thanks for the quick responses. Do I understand correctly then that I can remove R130, then move SW3 to SW1 and set the multiplier by using the dip switches? Or are the resisters still needed? I'm hoping to get an AMD k6-2 to run at 2.2 volts 66MHz FSB x 4.5. Does anyone know whether anyone has gotten this to work without the use of an external power supply? I haven't seen any posts that claim to run 300 MHz without an alternate power source for the CPU. 06-08-2000 15:16:06

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) duvell mr_binary: You are correct about R130. If you are addiing DIP switches, then remove the zero ohm resistors. If you left R130 in place it would short one of the switches. I don't recall anybody running at 66/300. It would be at the upper limit of what the power supply can drive. duvell 06-10-2000 19:07:12

RE:Setting Up FSB Clocking and CPU Clock Ratios (modified 0 times) jasper sorry if that has been asked/answered but how do you make it permenently set the clock settings? i can get mine to 225 (or whatever) but it doesnt stay after rebooting... ? jasper 06-22-2000 12:34:32

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