Title: HP A2655A CD-ROM drive: a short repair story
Post by: MindWalker on May 31, 2025, 09:55:17 AM
Post by: MindWalker on May 31, 2025, 09:55:17 AM
Hi all.
I've had my HP 712/60 for about 1,5 years now. It runs NeXTStep great and it's a neat stylish pizza-box machine.
Recently I came across a matching HP A2655A external SCSI CD-ROM drive on eBay. Sold as untested and without power supply. It came with an Adaptec PCI SCSI controller and cables for some reason and it was slightly on the expensive side. But it would look nice next to the 712... I put in a slightly lower offer and was happily suprised when it was soon accepted.
Some time later the drive arrived, looking slightly scratched and yellowed (at least compared to the computer) but nothing drastic. Even the two plastic clips that hold the top cover on were still intact.
The drive itself is a Toshiba XM-3401B (double-speed caddy-load) unit. I gave the drive a cleaning and lightly lubricated the optical assembly rails on which it slides. The PCB looked clean and none of the caps were visually leaking so I didn't replace anything for the moment. I tested the drive separately using a modern 5+12V power supply and my Windows NT box that I had at hand with the appropriate SCSI interface. The drive worked just fine, nice!
The case uses a funny style of moulded styrofoam frames that hold everything place. In fact there are no screws in the case at all! The same style of moulded foams are used in the 712 workstation too for the hard drive and floppy drive mounts. The casework is also identical; the top part of the case is released with the two plastic tabs on the rear and then the top slides out and off.
There was a matching tape drive and a hard disk drive (obviously without the door on front) and these are all listed in the 712 series manuals.
The power board sits on the back corner of the case, requiring the lengthy power switch arm.
The PSU connector was labeled as 12V DC on the case but the connector was a 4-pin mini DIN so I had to take the PCB out to figure out the pinout. From google searches I expect the original PSU having been a 12V 1.65A unit.
Probing the DC connector and the board I figured out that 2 left pins were GND and 2 right were 12V in. Later on I found another image of a probably correct PSU that confirmed this on the label. I build a small adapter cable with the mini-DIN plug on one end and a more usual barrel-jack on the other so I can use this with a generic 12V supply.
Looking at the power PCB inside the case, I soon found an issue. The board takes 12V in and passes it thru to the drive and to the small case fan. Then there is a small DC-to-DC converter that regulates it down to give a 5V line (used by the drive). The 5V line was flying high at around 5,6V and a related 5,6V Zenerdiode was getting very hot. Looking in further I found that the DC/DC-converted itself was outputting a way high around 6,5V output, surely not a good sign for a part that should be pretty precise 5V out! Everything else in the power board was passive and appeared good. The zener was doing it's best to keep it down. Don't know it drive had been exposed to the over-voltaged 5V line but at least it had survived and I was happy I checked this before trying to operate it now.
The DC/DC-converter module is an open-case but very small and partially epoxied over the components so I didn't really attempt to probe it further. Probably something had drifted out of spec making the output regulation off. I started looking on Mouser and I was able to find a similarily spec'd modern part, identical in 12V-in-5V-out, max 1.5A output and even a similar three-pin packaging. At about 5 EUR/piece this was an easy way forward and I soon had this in order.
Original DC/DC converter on the left, new one on the right.
With the new DC/DC-converted the voltages measured fine and testing with the drive nothing got even slightly warm. Using my bench power supply I measured the power draw on the 12V around 500mA during disc operations with very brief peaks to around 1A. Idling at about 250mA.
I assembled the drive and it has taken in all discs I've thrown at it with no errors. Solid operation.
Drive with caddy. I've always liked the caddy-operated CD-ROM drives, they always feel like a piece of serious tech! I also like the fact that there is practically no seek noises nor vibration so they operate much UNLIKE the modern cheap and noisy plastic'y drives. With this particular drive I can't hear the drive at all behind the case fan's noise.
This will look nice next to the 712. The case is a bit too large to retrobight, but I might consider it at some point.
;D Thanks for reading!
I've had my HP 712/60 for about 1,5 years now. It runs NeXTStep great and it's a neat stylish pizza-box machine.
Recently I came across a matching HP A2655A external SCSI CD-ROM drive on eBay. Sold as untested and without power supply. It came with an Adaptec PCI SCSI controller and cables for some reason and it was slightly on the expensive side. But it would look nice next to the 712... I put in a slightly lower offer and was happily suprised when it was soon accepted.
Some time later the drive arrived, looking slightly scratched and yellowed (at least compared to the computer) but nothing drastic. Even the two plastic clips that hold the top cover on were still intact.
The drive itself is a Toshiba XM-3401B (double-speed caddy-load) unit. I gave the drive a cleaning and lightly lubricated the optical assembly rails on which it slides. The PCB looked clean and none of the caps were visually leaking so I didn't replace anything for the moment. I tested the drive separately using a modern 5+12V power supply and my Windows NT box that I had at hand with the appropriate SCSI interface. The drive worked just fine, nice!
The case uses a funny style of moulded styrofoam frames that hold everything place. In fact there are no screws in the case at all! The same style of moulded foams are used in the 712 workstation too for the hard drive and floppy drive mounts. The casework is also identical; the top part of the case is released with the two plastic tabs on the rear and then the top slides out and off.
There was a matching tape drive and a hard disk drive (obviously without the door on front) and these are all listed in the 712 series manuals.
The power board sits on the back corner of the case, requiring the lengthy power switch arm.
The PSU connector was labeled as 12V DC on the case but the connector was a 4-pin mini DIN so I had to take the PCB out to figure out the pinout. From google searches I expect the original PSU having been a 12V 1.65A unit.
Probing the DC connector and the board I figured out that 2 left pins were GND and 2 right were 12V in. Later on I found another image of a probably correct PSU that confirmed this on the label. I build a small adapter cable with the mini-DIN plug on one end and a more usual barrel-jack on the other so I can use this with a generic 12V supply.
Looking at the power PCB inside the case, I soon found an issue. The board takes 12V in and passes it thru to the drive and to the small case fan. Then there is a small DC-to-DC converter that regulates it down to give a 5V line (used by the drive). The 5V line was flying high at around 5,6V and a related 5,6V Zenerdiode was getting very hot. Looking in further I found that the DC/DC-converted itself was outputting a way high around 6,5V output, surely not a good sign for a part that should be pretty precise 5V out! Everything else in the power board was passive and appeared good. The zener was doing it's best to keep it down. Don't know it drive had been exposed to the over-voltaged 5V line but at least it had survived and I was happy I checked this before trying to operate it now.
The DC/DC-converter module is an open-case but very small and partially epoxied over the components so I didn't really attempt to probe it further. Probably something had drifted out of spec making the output regulation off. I started looking on Mouser and I was able to find a similarily spec'd modern part, identical in 12V-in-5V-out, max 1.5A output and even a similar three-pin packaging. At about 5 EUR/piece this was an easy way forward and I soon had this in order.
Original DC/DC converter on the left, new one on the right.
With the new DC/DC-converted the voltages measured fine and testing with the drive nothing got even slightly warm. Using my bench power supply I measured the power draw on the 12V around 500mA during disc operations with very brief peaks to around 1A. Idling at about 250mA.
I assembled the drive and it has taken in all discs I've thrown at it with no errors. Solid operation.
Drive with caddy. I've always liked the caddy-operated CD-ROM drives, they always feel like a piece of serious tech! I also like the fact that there is practically no seek noises nor vibration so they operate much UNLIKE the modern cheap and noisy plastic'y drives. With this particular drive I can't hear the drive at all behind the case fan's noise.
This will look nice next to the 712. The case is a bit too large to retrobight, but I might consider it at some point.
;D Thanks for reading!
Title: Re: HP A2655A CD-ROM drive: a short repair story
Post by: Nitro on May 31, 2025, 10:58:08 AM
Post by: Nitro on May 31, 2025, 10:58:08 AM
Your HP is in great condition. I've always liked the styling of them. I didn't know there was a matching CD-ROM drive. It looks right at home with the 712/60. Nice find!
Title: Re: HP A2655A CD-ROM drive: a short repair story
Post by: paolo.bertolo on May 31, 2025, 01:51:57 PM
Post by: paolo.bertolo on May 31, 2025, 01:51:57 PM
I have the very same combo.
Very neat and snappy machine for NextSTEP.
Pity that there's not much ported software around.
My CD drive has the usual loading tray, though. "Funny thing": the unit can only operate when actively connected to a SCSI controller.
Else, you can't even open the tray. Some may get confused and consider the unit defective.
Works with HP and with black hardware, too.
Very neat and snappy machine for NextSTEP.
Pity that there's not much ported software around.
My CD drive has the usual loading tray, though. "Funny thing": the unit can only operate when actively connected to a SCSI controller.
Else, you can't even open the tray. Some may get confused and consider the unit defective.
Works with HP and with black hardware, too.
Title: Re: HP A2655A CD-ROM drive: a short repair story
Post by: MindWalker on May 31, 2025, 03:15:33 PM
Post by: MindWalker on May 31, 2025, 03:15:33 PM
@paolo.bertolo: very interesting that your drive has a tray-loader inside. I guess the same door mechanism should work with that too, if the tray sits at a correct height. The caddy needs to be pushed quite far inside so the tray might is probably much nicer to operate.
Have you checked what drive it is inside? The Toshiba in my drive has jumper to "Allow insertion & removal of CD's by command or eject button"/"Prevent insertion & removal of CD's" (also another jumper for "Normal operation mode"/"Audio disc reproduction mode" - not sure what they thought was the use for these modes!). Perhaps your drive has some similar special mode that's currently on?
Have you checked what drive it is inside? The Toshiba in my drive has jumper to "Allow insertion & removal of CD's by command or eject button"/"Prevent insertion & removal of CD's" (also another jumper for "Normal operation mode"/"Audio disc reproduction mode" - not sure what they thought was the use for these modes!). Perhaps your drive has some similar special mode that's currently on?