Boot hangs with "Detecting IDE Drives" on CDRW

[serafis]

I've just built a new PC based on Soltek SL-75FRN2-L MB and AMD XP 2200+ CPU. It hangs on boot at "Detecting IDE Drives" whenever my Cyberdrive CW058D CDRW is connected, as master or slave (jumpers set correctly), on either IDE channel, even if it's the only IDE device connected. The CDRW works fine in other computers but appears dead in this one (no lights, door won't open) if the IDE cable is connected. When removed, the PC boots just fine. If power is applied without the IDE cable connected, the door will open etc.

I contacked SOltek tech. support in Taiwan. They say that Cyberdrive have told them it's a known problem with the reset voltage on the mainboard IDE Pin 1 (3.3V) being to close to the quiescent voltage at pin 1 on the CDRW (3.24V), which causes it either not to reset, or to reset continuously (I had trouble interpreting the Taiwanese-English here). The CDRW is apparently expecting 3.8V on pin 1, and Cyberdrive recommended changing a resistance (without being specific as to what or where, or indeed how).

All very helpful, and I can understand how this could cause the problem, but it doesn't tell me how to fix it and I'm reluctant to buy a new drive as this one is less than a year old.

Now, someone has told me that if you cut Pin 1 (or wire 1 - reset) on the IDE cable (or pins 1&2 on an 80-wire cable), it is likely to solve the problem. I can see how this might work but, before I take this drastic step, I thought it would be a good idea to ask for some advice in this forum.

Anyone had a similar experience who might be able to help?

Thank you

[Graham L]

More "improvements" on standards. :_|

The IDE standard is for "TTL" levels. That is, "nominal" 5V and 0V levels, with specified levels between those at which the logic is guaranteed to switch, and a range in the middle where it won't toggle. New CPUs and their associated circuits are not TTL, they have much lower swings. That obviously works "most of the time" when connected to TTL equipment. But it's a matter of luck. Conservative design it's not. It's probably cheap. :-(

Disconnecting the reset line from the CDRW will allow its reset input to float. Since the thing isn't working now (;-)) that implies that it doesn't have an actual pull-up resistor to give it a specified "high" level. (That is the resistor they are referring to ... a 4.7 kilohm (or even as high as 10 - 20 k) resistor from that pin to 5V would fix it. I'd be inclined to do that if at all possible ... the reset signal is provided for a reason. It's a Good Idea for all the peripherals to be properly reset with a centrally generated signal.

[serafis]

Thank you very much for your comprehensive reply. I'll be getting the soldering iron out tonight!

Cheers :-)

Jon.

[serafis]

I put in the pull-up resistor as suggested, but no difference - the PC still hangs on POST at "Detecting IDE Drives".

Currently I have the HDD on IDE#1 Master, a CDROM on IDE#1 Slave and the faulty CDRW on IDE#2 Master

So I got an old 40-wire IDE cable and snipped wire 1 (the red wire - reset signal) just where it enters the Master connector, and tried again (leaving the pull-up in place).

Success!! The CD-RW resets when power is first applied (presumably because the pull-up resistor raises pin1 to 5V), the PC boots correctly, detects the CD-RW drive properly and it all works fine.

Without wire 1, the CD-RW doesn't reset if you press the "Reset" button but that doesn't seem to cause a problem, and I can't find any other problems with it set up this way. So I've done the same on the 80-wire Ultra DMA cable (leaving the slave connector's pin1 still connected in case I want to add a new IDE #2 Slave device in the future).

Thanks for the help.