I have a couple of Epson PF-10 portable 3.5″ floppy drive units. One is in very nice cosmetic condition, the other is pretty battered. When I initially received them, the battered unit seemed to hold a charge and would spin a disk but never successfully read or write to one. The other unit worked for a short while, albeit intermittently before completely dying on me.
The PF-10 is an odd beast, taking standard DSDD 3.5″ disks it uses the same 40 track format as the TF-20 5.25″ drive resulting in a formatted disk capacity of 278k. I don’t know how common battery operated portable floppy drives were but the PF-10 is not unique, a quick search turned up this HP Drive and this drive for the Tandy TRS-80 and I’m sure I once read about a unit for the Cambridge Z88 though I can’t find reference to it.
The PF-10 has a replaceable main NiCad battery which can also be substituted for four dry C cells. In order to prevent the drive dying in the middle of operation when the battery runs low there’s an additonal sub-battery hardwired inside the unit, a single 4.8v 100 mAh cell. A battery light on the front of the unit lights when the main battery is low to warn you that the sub battery has kicked in and will subsequently flash when the sub battery itself is running low.
I read sometime ago in the manual that when using dry cells in the unit it may still not work if the sub battery is flat and this is what I found leading me to conclude that the sub battery was dead as I’d made numerous attempts to charge it, no surprise really after 25 years or so. Putting a meter on the battery confirmed this.
I had envisaged trying to replace the sub battery but my soldering skills are poor and these drives are pretty rare, I don’t want to be responsible for totally wrecking one. I dug out the battered unit and took it apart thinking I might be able to clean and align the heads on it. However putting a meter on the sub battery on this unit showed some life in the old cell.
I therefore decided to take both units apart and swap the boards so that the board with the working sub battery would be combined with the drive that had worked in the past. I then swapped out the original Epson main battery and replaced it with a modern NiMh pack. On connecting up the drive to the PX-8 things initially looked promising as I was able to read the directory of a disk, however the sub-battery light then immediately came on and shortly after began flashing, then the drive died.
I knew the NiMh pack had a good charge and a meter appeared to show the sub-battery also holding a reasonable charge. When I plug in the AC adapter the unit works fine which it didn’t do before but at the moment I can’t understand why the unit won’t run off of its batteries and the manual does advise that you shouldn’t run the drive with the adapter attached.
So some success, I can use the drive albeit with the adapter attached.
Lately I’ve been experimenting with new batteries for my Epson PX-8’s (I have three PX-8’s now, it’s a worry I know.) Of the three’s original batteries, one still holds a reasonable charge, the other two however require the AC adapters to be permanently connected. This is not a huge issue but with high capacity NiMh batteries pretty cheap these days I decided to look into replacing the original NiCad’s.
The original NiCad’s are 4.8 V 1100 mAh packs, ideally I wanted replacements that were of equal or greater capacity. Having hunted around and looked at a myriad of options it became clear the most limiting factor was the size of the battery compartment on the PX-8. Initially I was tempted by a 4 pack of 1.2 V AA cells at 2700 mAh, however there was no way of squeezing these into the compartment even with a homemade pack. I also looked at new NiCad’s but really wanted to go with NiMh.
In the end I tried two options, a ready made four pack of Sanyo Eneloop AAA cells at 800 mAh and some loose Duracell AAA’s at 1200 mAh secured in a simple 4 cell battery holder. Both options are working well and lasting for what seems like forever.
I’m no expert on battery charging so don’t fully understand what the consequences will be of using the supplied Epson adapter and the PX-8’s charging circuit on the modern NiMh’s. From what I do understand the PX-8 will charge for 8 hours and then drop into trickle charge with its standard NiCad. If you disconnect the AC adapter and plug it in again the charge cycle will start again possibly leading to overcharging. The AC adapter is an unregulated 6 V 600 mAh model. (Epson HOOAAU-A)
As far as I can tell, the NiMh’s are taking somewhere in the region of 2 hours to fully charge which seems to me to be what’s considered fast charging.
With July rapidly approaching my thoughts are turning this year’s Retrochallenge. I have a number of possible avenues to explore. Continuing with the battery theme I really would like to try and revive my PF-10’s. I’m pretty sure the problem lies with dead sub batteries inside the units. The main batteries are easy to swap out, however the sub batteries are soldered to the logic boards and it’s been a long time since I’ve been near a soldering iron.
I also have a Commodore 128D which I’d like to get fully acquainted with including using it with CP/M, in fact it would be nice to try and get it talking to a PX-8. On that theme I’ve been experimenting with the PX-8 RS232 ports and quite fancy writing some sort of two player game with communication via this route. Decisions to be made.
I recently stripped down my two PF-10’s and managed to reassemble one working drive from the components. The drive is working quite well, the only real problem is the battery is only lasting about 30-40 minutes before needing a lengthy recharge. However this at least gives me enough time to perform a few tasks.
Today I formatted some disks using Copydisk, I found some old 1.0 MB double sided double density disks which seem to work well, the high density ones that I tried did not work. I then daisy chained the TF-20 and the PF-10 to the PX-8 by running the cable from the TF-20 into the back of the PF-10. This required opening up the PF-10 battery compartment in order to flick a dip switch to ensure there were no drive letter conflicts.
This set up gives you the PF-10 as drive F and the TF-20 dual drives as D and E. The PF-10 came with some old disks and I took the opportunity to copy their contents using PIP onto some 5.25″ disks just in case the PF-10 dies again. This included some original copies of Microsoft Basic and Compiler. That proved to be enough work for the main battery and the back-up battery light began to flash. It’s a shame the unit won’t run off of the AC adapter however I’m just thrilled it’s now working at all.
I shall use Filink to further back up the files I transferred today onto my bridge machine and I shall see if there’s any chance of building another working PF-10 from the left over components.