speed strips ...do they work?

csimon

http://www.plycon.com/hw/speedstrip.htm
looks interesting ...my eyes can't handle the other methods therefore I haven't touched the multipliers yet.

"Please take a moment before deciding to purchase and install this product to make sure that it will work with your AMD Athlon™ XP processor. Locate the label on the chip, as shown above, and be certain that the model number starts with “AXDA”. If it does - you're all set! If it does not - do not buy this product! "

seversphere

what are you supposed to do with that - cut a tiny peice/shape to go on the pins you need?

Geeky1

Seversphere, no... it just goes over the pins to unlock the chip, not to change the multiplier.

And Csimon, it should work... for $15, it's worth a shot, and Plycon is a reputeable (if expensive) company...

WuGgaRoO

yeah buy it...im cursious to see if it works

csimon

I'll let you know in a few weeks when I order my new processor ...prices are dropping again next week ya know.

shwaip

Have you tried the wire trick on the bottom of the processor yet? That's all that this is doing, and it is only connecting two adjacent pins. Another easy way to do it is to use conductive ink on the back of the mobo instead of the pins.

csimon

shwaip had this to say
Have you tried the wire trick on the bottom of the processor yet? That's all that this is doing, and it is only connecting two adjacent pins. Another easy way to do it is to use conductive ink on the back of the mobo instead of the pins.

I may try the wire trick however I don't think I'll be doing the ink thing ...don't want anthing permanent!

I won't be ordering the speedstrip this week as they are out of stock but perhaps next.

AFA the wire trick ...where do I go to see how it's done?

shwaip

http://www.beachlink.com/candjac/Workarounds.htm

HOW TO EXECUTE, DETAILS OF MODIFICATIONS.
See Pics to visualize Multiplier block diagram, bridges, and to locate all sockets/pins listed below.
Changing an 8X Bit Value HI to a LO, to access 12.5X and lower from default Multipliers = 13X and higher.
Tbred's rightmost 5th L3 bridge will be Open. Palomino's L10 bridge pair will be Open/Closed L to R.
General solution is to ground that signal circuit. Options:-
1:- Tbred, close the open 5th L3 bridge with a conductive trace around its open pit, no need to insulate the pit, restorable.
2:- Palomino, change L10 bridge pair from Left/Right Open/Closed to Closed/Open.
3:- Tbred and Palomino, connect AJ27 and nearby AH28 Vss/ground socket holes with fine wire u-loop, replace CPU.
Or connect the "bases" of socket pins AJ27 and AH28 with conductive ink.
4:- Tbred and Palomino, solder wire to socket AJ27, ground it thru 100 ohm resistor, (remotely switch/toggle).

Changing an 8X Bit Value LO to a HI, to access 13X and higher from default Multipliers = 12.5X and lower.
Tbred's rightmost 5th L3 bridge will be Closed. Palomino's L10 bridge pair will be Closed/Open L to R.
General solution is to reset that signal circuit HI. Options:-
1:- Tbred, open the closed 5th L3 bridge.
2:- Palomino, change L10 bridge pair from Left/Right Closed/Open to Open/Closed.
3:- Tbred and Palomino, connect AJ27 and nearby AH26 Vcc/Vcore socket holes with fine wire u-loop, replace CPU.
Or connect the "bases" of socket pins AJ27 and AH26 with conductive ink. Note risk of accidental Vcore short circuit since no current limiting resistor as in 4:- below.
4:- Tbred and Palomino, solder wire to socket AJ27, connect to Vcore thru 100 ohm resistor, (remotely switch/toggle).

I guess you'll need to mod the processor itself, so you may or may not want to do it

csimon

hmmm ...any idea where to get the wire?
I can't use the speed strip afterall my Xp2000+ isn't a tbred. I ran wcpuid and it said AthlonXP (model 6) ...and I'm pretty dang sure that's the palamino core ...so.

muddocktor

The wire out of an ide cable will work just fine. Just use 1 strand of wire per socket hole.

I did this trick on my P3S to raise the vcore to 1.625 for booting reliability. It's was a real pita to get the fine little wire right in the socket holes.:banghead:

csimon

it seems so simple ...am I missing something here?

ok I have a kr7a-r and xp2000+ pally ...don't I just thread one end of a short pc of copper wire into each slot of the green mark and then do the same to for the red mark? and then when I seat the processor back it should keep it in place right?
whats the odds of really screwing up my system if I do it this way?

csimon

omg ...I took a strand from a serial cable which I assume is the same as an ide cable when it comes to the wire. It is silver in color (aluminum maybe) and is about the thickness of a hair ...is that the right wire ir should I get it from an ide cable instead? I was thinking it would have been copper.

edit:/ I placed a hair next to it and it's pretty dang close ...just seems too small.

shwaip

3:- Tbred and Palomino, connect AJ27 and nearby AH28 Vss/ground socket holes with fine wire u-loop, replace CPU.
Or connect the "bases" of socket pins AJ27 and AH28 with conductive ink.

All you need to do is the red wire!!!

Straight_Man

Essentially, the wires in IDE cables are aluminum, yes. Consistently.

One thing to note, the UDMA wires may be too tiny, you want wire from a 40 conductor cable (true IDE as opposed to the 80 conductor EIDE\UDMA 80 conductor cable for heat stability reasons as the EIDE\UDMA wires could possibly melt if you get the core voltage a bit too high). What is actually best is a wire wrap tool and the finest wire wrap wire you can get-- you need the insulation off the contact part that goes on the CPU pins-- 2 wraps on each pin, not a lot or the CPU will not seat right, and you will get field effect from a LOT of turns. To reverse, you unwrap the wire on the pins. If you wrap right the end of the wire will be wrapped around the pin, and the wraps will move to the base of the pin against the small brass (possibly gold plated) contact ring (where pin is "spread out looking" in ceramic outer CPU coating layer. Since the conductors can be seen in the picture of the strip, emulate those with wire wrap as to what pins to use and make sure the insultaion is ON areas that go past any other posts. This method lets you CAREFULLY remove any misconnected stands of wire wrap if you have a problem with that.

Make sure ends are WRAPPED around pins\posts so nothing sticks beyond ring\plate at base of pin, and make sure no wire wrap strands corss each other. Welcome to the breadboard\prototyping wire wrap method.

SIDE note: You CAN use a fine tip non-permanent marker and carefully mark only tip of pins to be wrapped, the socket grabs long part of each pin and not the tip so tip does NOT have to be conductive if you use a very small amount of ink from the fine-tip marker and work one pair at a time from a printout of the strip picture. IF you use aluminum, be very careful not to overtighten the wrap as you will break the aluminum strands and CAN break the copper wire wrap wire also.



HINT set: Practice on pin connectors that are on a scrap motherboard will teach you how to torque just right. you do NOT need powered wire wrap tools to do this, though pros used them when they did a LOT of wire wrapping. AFAIK, JameCo has wire wrap tools for hand use still-- 28-36 guage wire will work given you are doing a single wrap or two, bigger the wire, smaller the number and fewer wraps used. DO test by putting CPU in, and then pushing lever down on socket with CPU seated, and then get CPU off by releasing levr and removing and turning over and inspecting with a 3-7X magnifier so you cna make sure you have no extra ends before finalizing seating and powering up.

Mouser Electronics has the finest conductive ink pens I know of if you prefer that route.

John-- who started using computers when boards WERE often wire-wrapped to prototype and has wire wrapped wire.

F@H score note: One thing, the site appears to have problems connecting to the F@H site right now, my real F@H points are above 3K and I am at position 237 overall and gaining 3 positions average a day. Will email admin with this info and a time stamp also and what is not connecting.

csimon

from what I've read so far it seems by unlocking bridges I will only get another 50mhz out of this chip. this doesnt seem to be the overclocker at all.

muddocktor

OK everyone, he has a pally proc which the L1 bridges are lasered, which means the proc is definitely multi locked. He will have to fill the pits and connect all the L1 bridges to unlock the cpu unless he just want to arbitrarily choose a higher multi and go with it.