• Please review our updated Terms and Rules here

Dual 8 inch QUMETRAK 842 rack mount floppy drive restoration

Do these QUMETRAK 842 drives Spin continuously when powered on?

I made a straight through cable. Edge connector on one side. I went over the first drive cleaning everything and repairing a couple minor wire damage.
 
Ok. Im in a bit over my head and could use some guidance. Im using the cable I made (from a standard straight 50 pin scsi cable with a cardedge connector on the end) to the first drive. Im using the D-Bit FDADAP board as if its uneeded, it shouldnt hurt either right? And its a convenient way to adapt the 50 pin to 34 pin.

I have a kryoflux and a greaseweazle.. I just want to try testing this drive to see if it is working a little. Can someone provide some settings and things I should run. Also what should the "DS" Jumper be set to when making these tests? DS1 through DS4? They are both currently set to DS3 for the Cromemco I assume.

I may need new belts. The motor doesnt seem to have enough pull to turn the disk when I have an 8" cleaning disk in the drive. Whats a good source for 8" drive belts?


IMG_20240417_104459.jpgIMG_20240417_104227.jpgIMG_20240417_104235.jpgIMG_20240417_104251.jpgIMG_20240417_104258.jpg


EDIT: Just something to note. Both drives were set to DS3 and both drives have both Terminators in place.
 
Last edited:
I don't use the FADAP thing, as the Qume drive doesn't need TG43. In point of fact, I don't use any adapter board.

I go from a 50 conductor flat cable from edge connector to a female DD-50 connector on the external box. Drives are set to DS1 and DS0. From the DD-50 I use a round cable (I'd have to open it up to see, but I don't believe that it's much more that 34 wires to a DC-37 male. This gives me the option to use a 50 pin cable should I need one.

On another box, It's 50 conductor flat cable all the way to a DC-37 male solder-cup cable. No adapter board.
I do have adapters from Micro Solutions, which are similar to the French board I linked to in another thread. I don't use them, but others seem to find the adapter boards convenient.

All of my external diskette drive boxes, regardless of size, from 3.0" to 8" end in DC-37 connectors.

Unless you really need TG43 or the eye candy, the FDADAP is nice, but not necessary.

As far as cable hookup, just match the signals with their 5.25" versions, the exception being HEAD LOAD, which goes to the MOTOR ON signal of the 5.25".

And yes, the motors on AC-powered drives run continuously.

As far as belts, I can't tell you--the 842 uses a rubberized fabric flat belt (you can't use round one, as the drive pulley is a crowned one.) Maybe others have some idea.
 
Ill just use the fadap as I already own it regardless of whether it does anything. Im sure a cable would work, but for what I have on hand its fine.
 
VERAULT said:
Ill just use the fadap as I already own it regardless of whether it does anything. Im sure a cable would work, but for what I have on hand its fine.
Click to expand...

For proper operation, the old school QumeTrak 842 does require a "TG43" input signal. In the Qume manual the "TG43" signal is actually referred to as the "Low Current" input signal. Older drives like the QumeTrak 842 lacked the "smarts" built into later model 8" drives that kept count internally of the track the head was positioned on, eliminating the need for an external TG43/Low Current signal.
For basic testing purposes the QumeTrak 842 will certainly work without the TG43/Low Current signal. However if you use the drive to save files (particularly in DD mode) on the higher tracks you will see many errors reading back data if the TG43/Low Current signal is not implemented - an issue I have tested myself using this very same drive.
 
Never noticed that, but then the 842 in my shop is used mostly for reading (after all, that is my business--nobody wants copies of the old media). The Siemens FDD-200 drives are my daily drivers.

I'll add that RWC is the old way of handling the inner cylinder bit-shifting; most modern floppy controllers allow for a different pro-active precompensation technique, shifting the timing on the bits without reducing the write current.

If you're writing your disks using one of the flux-transition thingummies (Catweasel, Fluxengine, etc.) you need to figure in write precomp. The NEC 765 applies it at track 42 automatically.
 
Chuck(G) said:
If you're writing your disks using one of the flux-transition thingummies (Catweasel, Fluxengine, etc.) you need to figure in write precomp. The NEC 765 applies it at track 42 automatically.
Click to expand...
Could very well be my problem. My settings with the kryoflux and greaseweazle are all wrong.
 
So I decided to go direct method and attach the 8" drive to a 486. I setup the bios for Drive A: 1.2MB HD

at first I got drive not found errors but I remembered I needed to change the DS jumper. The bios "drive not found" error messages went away.. But whenever I try to do a task with the drive I get "Not Ready". The Stepper motor never moves the head at all.

So unless Im still having something configured wrong Id say this first drive has a fault.
 
Last edited:
You need to jumper the RDY line. Pin 34 on a PC is not connected internally so your Ready for the 8" is probably floating ?
Chuck would know. I think he told me about it years ago. Can't remember exactly how that goes without some more digging
but it seems like just make the drive always ready.

Larry G
 
retrogear said:
You need to jumper the RDY line. Pin 34 on a PC is not connected internally so your Ready for the 8" is probably floating ?
Chuck would know. I think he told me about it years ago. Can't remember exactly how that goes without some more digging
but it seems like just make the drive always ready.

Larry G
Click to expand...
Doesnt the Dbit fadap do that for me? This video by Texelec has it connected as a 1.2MB floppy direct to a pc.


I think before I can find faults I need to determine if I am connecting this up correctly.
 
Last edited:
May I assume that your PC "MOTOR A" signal is activating the "HEAD LOAD" pin on the 842? You should hear a "clunk" when the drive is accessed by the PC. Also, check your "Drive Select" jumpers.

The 842 has a plethora of configuration options; since you don't know where this one came from, you need to check those against the "Factory Defaults" in the OEM manual on bitsavers.
 
Chuck(G) said:
May I assume that your PC "MOTOR A" signal is activating the "HEAD LOAD" pin on the 842? You should hear a "clunk" when the drive is accessed by the PC. Also, check your "Drive Select" jumpers.

The 842 has a plethora of configuration options; since you don't know where this one came from, you need to check those against the "Factory Defaults" in the OEM manual on bitsavers.
Click to expand...
The only document i have come across is the MAINTENANCE manual. Would that cover default configurations?
 
Depends on what you're using for a host. A "blind' drive that doesn't handle sectored data will simply pulse the INDEX line every time it hits a hole. An index hole can be determined because it occurs halfway between two sector holes. But using a generic soft-sector (e.g. WD 17xx or NEC 765) FDC? Nope.
 
Chuck(G) said:
Depends on what you're using for a host. A "blind' drive that doesn't handle sectored data will simply pulse the INDEX line every time it hits a hole. An index hole can be determined because it occurs halfway between two sector holes. But using a generic soft-sector (e.g. WD 17xx or NEC 765) FDC? Nope.
Click to expand...
So Id need a more versatile FDC? Can you recommend one. Are any of these new Quad Floppy boards that were all the news last year up to the task?

Reason I bring it up.. I went through my 8" disks today. The vast majority are DS/DD 32 sector 1" floppies. Id just like to know if Im screwed with these Qumetrak drives.
 
The old-style hard-sector controllers for 8" were pretty much "anything goes". That is, there was no standard encoding. You can't read, for example, a Future Data disk in a Wang 2200. There may have been some hard-sector LSI controllers, but mostly, they were pieced together from SSI/MSI logic.

Your 842 qualifies as a "dumb drive"; that is, it's meant for soft-sector media. The real gotcha is that the READY signal is obtained by timing the interval between index pulses. Obviously, a hard-sectored disk will mess this up a bit. Is the Kyroflux smart enough to work with hard-sectored media? I don't know.

In contrast, the Siemens FDD-200 is a relatively "smart" drive that can deal with hard sector media and separate the index pulses from the sector holes. So it can use hard-sectored media as if it were soft-sectored.
 
In case anyone is interested, I have a Power One PSU that I traced out the schematic for one day. It's a 28V version though, and I'll have to check the model number later. But perhaps the schematic is similar to the one in this thread.

The two AC sources are different taps on the transformer. Mine uses 2N3055 transistors and some big Motorola TO-3 diodes as a rectifier.
 

Attachments

  • Schematic_2023-04-28.pdf
    350.8 KB · Views: 0
Bentendo64 said:
Mine uses 2N3055 transistors
Click to expand...
I suspect (as I've said before) that pretty much any commodity TO-3 NPN with appropriate ratings would work with the standard reference design. I suggested 2N5907 and 2N3055. The OP's uses "house numbered mystery" TO-3s.
The only exceptions to this that come readily to mind is where a complementary PNP-NPN pair might be used or a Darlington device, but a linear PSU application typically doesn't need a high-gain configuration. I've also seen some PSUs with TO-3 50W Zeners, but those tend to be the higher voltage units (e.g. 200V). Since those are 2-terminal devices, it would be easy to spot them.
All in all, linear PSUs are delightfully simple, unlike their switching cousins.
 
You must log in or register to reply here.
Back
Top