You mean disable Slow refresh?

I can try again when I get home. I haven't looked at EDO vs. non EDO RAM either.

The benchmark differences was between BIOS defaults and the settings on the screen. I didn't go into trying out individual settings as I won't be tweaking it and just leaving it on the default settings.

Nope, I mean 'enable' slow refresh. You had a post showing a screen shot of the "tuned BIOS", whereby "Slow Refresh" was set to "disabled". This isn't so much a tweaking setting, it just means that you don't have to refresh the capacitor in your DRAM cells so often to maintain the memory state. Theoretically, it is suppose to free up time for the system to do other tasks, but I have not noticed any benchmark improvements. I like to think of this setting as reducing wear and tear on your DRAM.

Ah, I see it now 😀

Yes it didn't make a difference. 1 tick difference in Doom -timedemo demo 3.

Also using standard non-EDO ram makes hardly a difference. 1360 realticks with tuned BIOS.

I tried "EDO DRAM Read Speed" and it made no difference. Got the exact same Doom score.

Those were also the conclusions I came to. The only real benefit to FPM (non-EDO) memory is if you wanted to fun faster cache timings with FSB's greater than 33 MHz. Since you are not interested in tweaking, this doesn't matter, but in general, I have had better luck running FPM memory with faster cache timings at 40 and 66 MHz compared to EDO memory on PCI-based 486 motherboards.

I've got an MB-8433UUD-A Ver 2 and the Odin battery kicked the bucket.
The amount of solder on the pins is too little for me to try to desolder it with my crappy tools.

Is there a way to work around an external battery? What is exactly an Odin battery and why does it have so much pins?

Stojke wrote:

I've got an MB-8433UUD-A Ver 2 and the Odin battery kicked the bucket.
The amount of solder on the pins is too little for me to try to desolder it with my crappy tools.

Is there a way to work around an external battery? What is exactly an Odin battery and why does it have so much pins?

Here is a guide for the Dallas/Odin chip modification without desoldering.

http://www.mcamafia.de/mcapage0/dsrework.htm

Usually what I do to remove ICs is to just clip them off with some very fine wire clippers...then you can desolder the stubs one at a time (try to leave enough to pull on with tweezers). If the amount of solder is not enough, just add more.

It's even possible to put in a new socket without using a solder sucker. All you have to do is use a pin or a needle to poke a hole in the solder after you've heated the pad. It probably takes a little longer than using the correct tools, but it works okay and doesn't seem to cause any damage.

I'm not sure how well this would work on an odin chip. I have done this to remove SGS Thompson RTCs, but those had pins that were more exposed on the sides like a regular IC. I suppose if there is enough clearance at the bottom you could use some really fine clippers.

Or if you have two 45+ W irons, you can solder each row of pins together on the back of the board, and then heat each side simultaneousy with the who irons and lightly tap the chip out. You might be able to do it with one iron if it's hot enough and you switch between the two rows quickly.

I do not beleive there is sufficient clearance on this board to clip off the RTC pins.

The ODIN on this board is the same as the Dallas 12887.

Yes, the main problem is that the space around the battery is very confined making it impossible to clip the pins.
Should i attempt the drill and solder hack or solder thin wires (from PATA 80pin cable) from below the board?

Drill? No.

I am not sure if connecting two RTC's in parallel will yield the desired functionality, but you could try.

Adding a battery parallel will increase capacity.
Since its known what values are inside the Odin chip i think simply dragging lines from below and adding a socket on top will solve the problem.

I have not noticed any benchmark improvements. I like to think of this setting as reducing wear and tear on your DRAM.

I wouldn't think of DRAM refreshes from significantly affecting its life. I mean it probably does but DRAM is still cheap and it lasts generally lasts a loong time
but wouldn't reducing the refresh potentially reduce stability?

The ODIN/DALLAS is not just a battery, it contains a real-time clock (RTC) module. So you if you double-up the module, you also double-up the RTC. Even if you are to just add a coin battery to the bottom of the motherboard, I still think you need to cut the lines which connect the battery inside the old RTC. Connecting two batteries of differential potential in parallel is a no-no. The old battery is undoubtably at a very low voltage and the new battery will be at a much higher voltage, so you need to discect the old RTC to disconnect the battery. It is generally acceptable to connect two batteries of any voltage in series (to increase the voltage), or two batteries of the same voltage in parallel (to increase the max current output), but not two batteries of different voltages in parallel.

smeezekitty wrote:

I have not noticed any benchmark improvements. I like to think of this setting as reducing wear and tear on your DRAM.

I wouldn't think of DRAM refreshes from significantly affecting its life. I mean it probably does but DRAM is still cheap and it lasts generally lasts a loong time
but wouldn't reducing the refresh potentially reduce stability?

If the system is refreshing the DRAM less, the system has more time for other tasks. I have not had any issue with 60 ns FPM DRAM not working with "slow refresh". Most chips support this.

Not only is it located in a very inaccessible area but its not socketed.
Some one decided to engineer that area while high.

I am not sure I will desolder it properly. And cutting/drilling the plastic case can look ugly.

Some v3.0 boards have a RTC socket. I suspect all v3.1 boards have the RTC socket. If you are attempting to desolder, use approx 435 C for no more than about 5 sec. at a time and use a solder sucker. Do not try to pry the RTC off as this will pull out the through hole conductor.

I have desoldered ICs before with an electric pump, but this has way too little solder on the pins for it to catch anything properly.
I have access to an hot air station as well as an temperature adjusting soldering iron.

Invest in a de-soldering station with pump. Such a life saver. It made removing the old RTC and putting in a socket very easy.

What is the model of your desoldering tool?

Anonymous Coward wrote:

What is the model of your desoldering tool?

ZD-985