So I was trying the hack to 40 MHz for the Q605 and related models which is described here: http://homepage.mac.com/schrier/q605_40.html and I can not get it to work.
First, I switched the resistors to the 33MHz configuration as instructed and that worked fine. Then I lifted pin 9 of U17 and tied it to 5V through a 4.7K resistor. At that point I could get the start-up bong but no video or hard drive activity. Same battery as works at 33 MHz.
So I replaced the MC88920 with a MC88916DW80 and still had the same problem.
Thinking that the MC88916DW80 has been sitting around for the better part of ten years, I tested it by putting U17 back to stock configuration and the machine once again booted fine at 33 MHz. So that should confirm that the MC88916DW80 is installed and functioning, since the machine couldn't boot without it.
BTW, I have replaced the 68LC040/25 with a full 68040/33. So there should be no issues with the CPU at 40 MHz.
Anyway, I suspect one of three possible problems.
1) This Q605 has a 343S0161 chip instead of a 343S1135 at U17. It could be that this differently numbered chip is actually different in function.
2) The pin number given in the instructions could be wrong. Typically, one counts chip pins by starting at pin 1, which is marked (usually with a dot or indentation) and counting counter-clockwise from there. So the chip at U17 would be numbered like this:
But it is possible that the author of the article counted pins in some other manner, perhaps meaning pin 16 when he wrote pin 9. This is the main reason I'm writing to the list. If someone else has successfully implemented this speedup, then that would eliminate the second possibility.
3) There could just be something wacky about my hardware.
I have three other Q605 boards on hand, so when I'm not so exhausted I may try again, but I'd like to read the collected wisdom of the group first.
i have a Performa 475, which is the same as a Q 605, LC 475, and LC 476
they are all the same sept there is one surface mount resistor that changed the ID of the system from a LC to a Quadra. that will not make any diffrence.
you have to find a faster clocked PLL, something that came from a 40mhz 68k Mac.
i am going to post a few pages for you , the first one has a list of sites that have got these systems to run at 40mhz.
here is another good site for info
info site 2
here is another good link (i had to use web archive) he states he got his to run a 42mhz, you have to replace the PLL with a faster one ( he say's the model number of it as well)
he also say's any faster than 42mhz your serial ports will stop working.
info site 3
just as it say's in the site you linked to above
If you have any problems you will want to replace the PLL (MC88920) with a faster one. The MC88920 is rated at 50MHz (25MHz bus). The MC88916DW70 is rated at 70MHz (35MHz bus), and the MC88916DW80 is rated at 80MHz (40MHz bus). The MC88916DW70 is used in many of the other Centris/Quadra level machines.
all you are doing is replacing the same PLL with the same PLL and so you will get the same results, in order for you to get 40mhz you will need to get the 80mhz (40mhz bus one) MC88916DW80. or unless he Author of that site got the pin out numbers wrong or you somehow misunderstood ( it can happen)
i plan on having this done to my Performa 475.
Full 040 40mhz cpu (still running at the stock 25mhz speed) will move the surface mount resistors to get 33mhz some time. has heatsync installed.
Apple ethernet PDS card
External LaCie SCSI CDRW drive (reads great and burns great with toast delixe 4.x)
Mac OS 7.6.1
I'll check out your links. Thank you. However, regarding your advice about the PLL, read my third paragraph in my original posting. I already replaced the PLL with the 80 MHz version.
ok so you can put the 80mhz PLL in its stock mounting points and it will boot up but only see 33mhz, have you tryed to move the surface mount resistors to 40 mhz with the 80mhz PLL in its stock config on the board. cause i am thinking the method of lifting pin 9 of U17 and tied it to 5V through a 4.7K resistor can be for a diffrent PLL (AKA stock PLL) to get it to work at 40mhz.
i didnt see anywhere in your post about you setting the surface mount resistors to set the system to 40mhz with the 80mhz PLL with all its legs in the stock mounting points. this could be your problem in why its not seeing 40mhz and only 33mhz cause you do have the surface mount resistors set for 33mhz and not 40mhz.
you might need to buy a few proper voltage of resistors (if you dont already have them) before you can set the mobo to 40mhz cause IIRC there is only the 2 which isnt the proper ones you need to set to 40mhz.
just test it out with the 80mhz PLL with all the legs in there proper possition on the mobo and set the system to 40mhz.
the whole set to 33mhz thing was if you also replaced the oscolator clock driver crystal so you could set it to 33mhz and depending on the oscolator you replaced the stock one with would determine the new settings for the 33mhz setting. there is ones you can buy that would make the 33mhz setting into 38mhz and there is ones to make the 33mhz settings to 42mhz, and the 40mhz setting into 50mhz. but if i read right you only replaced the PLL with the 80mhz one with will keep the stock settings for 20mhz, 25mhz, 33mhz, 40mhz, will still be what they are
No, I have not tried that. The instructions I used do not say to move the resistors to the 40 MHz position. The instructions say to put it in the 33 MHz position and then pull pin 9 high.
I am willing to hear that the instructions I used were incomplete. Are you saying that you have information that with the pull-up modification the resistors should be in the 40 MHz position?
It is probably worth trying the resistors in the 40 MHz position. But that is not what the 40 MHz Speed-up instructions say to do.
I have about two hundred 4.64 Kohm resistors on hand. I have no memory of why I bought them, but they should substitute fine for 4.7Kohm resistors.
The PLL has always been in the proper position. There are two chips involved in this modification. U17 is a custom chip of Apple's which performs some clock function. The PLL was the MC88920 and is now a MC88916DW80. Lifting pin 9 and tying it high is done to U17, the Gazelle chip, not to the PLL.
There is a different 40 MHz speed-up in which one uses a separate oscillator and (IIRC) connects it to one of the pads for R93. That modification allows one to use different oscillator speeds. However, the modification I think I'm doing does not involve a separate oscillator.
I may have to try the other modification style if this one won't work, but I'd rather not, because it's rather inelegant adding several wires and an extra board.
However, I found a note at the end of the oscillator based instructions stating that if one gets a good start-up bong but no video, switch the VRAM from 80ns to 70ns. My VRAM here is 80ns, so I need to reconfigure it so that pin 9 is high and try some 70ns VRAM.
yea i also seen the it say about the 70ns vram (but good luck finding any cause i looked for about a month and could not find any)
i think the instructions you have can be flawed. you should be able to set it to 40mhz with the new PLL in place in the bootable config and it should work. some systems will and some wont. allot of the problems with the video at 40mhz was the fact people didn't replace the PLL with a 80mhz PLL and the slower PLL wasn't locking on to the faster speed ( i could be wrong).
if it dont work then you will know, but you will never know until you try, and you will have the experience of it if it does work to lead people to go this route and how to do it.
i seen allot of the 40mhz clock up sites using the clip on oscillator. i seen other sites refering to some clock chipping methods limit the screen rez to only having 640x480, i would think you would want to stay away from the method. the site didn't really say what clock chipping methods caused this to happen.
i do hope that that the type of resistor you have does work (even tho 4.64 ohm is real close to 4.7 it just might be enough to make a difference to not work)
i do hope that just by changing the surface mount resistors to 40mhz does work for you.
if you haven't seen
click on the expansion link, top right
yea me and Dana benchmarked our system's. i have a Performa 475 and she has a Quadra 605. her's is overclocked to 33mhz (a full 040) and mine is still at the stock 25mhz (a full 040 40mhz cpu).
my system is running 7.6.1 and at first the system running Macbench 1.0 was showing some weird results.only in a few video test's was showing below the Performance of a LC II in copybits small and large and in a few other test's. but somehow they fixed them self and the system is showing how it should. it beats me what was wrong.
yea she knows her stuff when it comes to the Q605, LC475, LC476, and P475 (all the same system's but some have a slightly diffrent design of pizzabox case).
when you get it running at 40mhz i want you to run Macbench 1.0, 2.0, 3.0 cause i would love to see how much of a speed up when at 40mhz over 33mhz and the stock 25mhz. i still have mine and dana's MB 1.0 2.0 and 3.0 test results to compair with
any updates on getting it to 40mhz
I haven't had a chance to touch it since slightly before I opened the topic. I must get on it soon, because I'm assembling the machine for my 5-year-old. I will report back when I have new developments.
ok cool, no rush here. glad to see you didnt give up on it. yea i can understand having a family to take care of and all. no rush man. you got to tend to life over computers. i can hardly wait to see the results of this. wish you the best of luck
Have you tried using a real 40mhz 040 chip? Not all chips are made with the same tolerances and some may not stand up to overclocking at all especially after 15 years. It's possible that you may have one that runs only at it's rated speed and no higher. I would try to check that out first if you have access to a 40mhz chip because delving into more complicated issues if your CPU isn't even capable of the desired speed will be a waste of time with a possibility that you may damage something needlessly. I am working on bumping a Q700 to 40mhz now and bought a 40mhz chip off ebay to try out as I don't think the stock 25mhz chip can take such a big bump. Even if you don't have a 40mhz chip, it's worth tracking one down because it will most likely take the extra speed better than an overclocked chip and last longer if you plan to use this machine frequently.
i would love to see if the OP finished this or not. i have been away for a wile (no internet)
my 475 has a 40mhz 040 in it but i have yet tryed to get it to 40mhz yet.
i wanted to see if he got it to 40mhz or not i hope he did and not gave up on it.