• Please review our updated Terms and Rules here

Debugging an TU56 on a TD8E in a PDP-8/e

I've seen this kind of thing before... factory mis-assembly or mis-applied revision. Good catch in either case.
 
Hi All;
Just for Comparison, Could/would You show a picture of the Correct placement of the cap, and the one that You found that is/was not correct ?? Sometimes, Comparison pictures, Really help the rest of us, when looking for problems, not necessary the same problem, but how easy it can be to have something in the Wrong hole, even from the Factory..

THANK YOU Marty
 
Does_not_work.jpg
The image shows the SN7400 E38 on the TD8E board. DEC installed an 0.01 uF capacitor between +5V and the CC R/W (1) H signal from the R/W Flip-Flop E30. The left leg of the cap is in the correct via, the right leg should go in empty the via to the left of the cap.

Does_work.jpg
This is where the cap is supposed to go.
 
Hi All;
M_Thompson, Thank You.. It makes me wonder if the Assembler was dyslexic.. Great photos..
It should also be easier for you to do the fix now..

THANK YOU Marty
 
Marty said:
All I can say is "WOW !"
Click to expand...

The PDP-9 has been the biggest challenge so far. It is like a PDP-8/S on steroids. It is 18-bit, has no integrated circuits, runs at 1 MHz, and is microcoded. The transformer based microcode store was difficult to get working because of bad solder joints. The TC59 tape controller is just like a PDP-8's TC58 tape controller, but has a 18-bit interface. The TU20 tape drive was used on many other DEC systems. We are looking for a TC02 DECtape controller. It is the same as a PDP-8's TC01, but with an 18-bit interface.
 
The TD8E and TU56 DECtape on the PDP-8/E is now working, mostly.

This morning it failed diagnostics when they it to write and then read back the Data Register. It took me a few hours to find a 7474 flip-flop IC, E03 for the UTS signal, that had both the 0 and the 1 outputs high. This was sending extra CLK signals to the 8271 ICs that make up the Data Register and corrupting the data. We have found lots of bad 7474 ICs, so I replaced the chip but it didn't change anything. I then found that both the SET and PRESET signals on the 7474 were low making both the 0 and the 1 outputs high. Eventually I found that the WTM switch was in the ON position, and should not be for normal use. That was a dumb mistake.

Even after setting the WTM switch to the correct position, the TD8E still fails diags when it is waiting for a quad-line flag. I need to investigate that further.

I was able to use David Gesswein's resttd8e program to make a bootable OS/8 DECtape from an image on my laptop. There is something not quite right about the DECtape that I made because when I boot OS/8 from the DECtape it halts when I try to do a directory listing. Listing, reading, and writing the DECtape from OS/8 booted from Kyle Owen's serial disk emulator works fine. Maybe the OS/8 image on the DECtape doesn't match the hardware configuration that I have? I have several TU56 DECtape images, so I will try the others.

I still need to look for the possible AC clearing glitch when executing the SDLD instruction.
 
Hi All;
I wish I could be there in the Thick of things, Just observing, not that I could be of any use or of any Help..
But, that is Just toooo far away..
Keep up the Good and Great work..

THANK YOU Marty
 
Marty,
Moral support is always welcome. Sometimes debugging is a little frustrating, so it is nice to have another enthusiast to discuss progress with.

Maybe someday I will find an RK8E board set and cable so I can add an RK05.
If not, I will just emulate the controller and drive with an FPGA.
 
Hi All;
M_Thompson, "" Moral support is always welcome. Sometimes debugging is a little frustrating, so it is nice to have another enthusiast to discuss progress with. ""
Boy don't I know this, I wish I would have had someone around when working on the 11/40 and 11/45..
Thank You for the Kind words..

THANK YOU Marty
 
The TD8E ran a complete pass of the MAINDEC-08-DHTDA today. It ran for 20 minutes without a single error.

I tried to format a DECtape using DEC-8E-EUZC, but it quickly died with a timing error.

I made another OS/8 floppy using restrx01. That works fine.

I used one of the batch files on an RK05 diag pack to make an RX01 diag diskette. That was an interesting process to watch because it setup the diskette in 8-bit format instead of 12-bit format.
 
It may be too late for this thread, but I did finally replicating the DEC drawings for the TD8E. The current stuff is online at:

http://svn.so-much-stuff.com/svn/trunk/Eagle/projects/DEC/Mxxx/M868/

If you don't have Eagle, "M868Hsch.pdf" is my version, and "M868.pdf" is the crappy low-res scan of the DEC schematic.
The file "diffs" lists the errors I found in my previous attempt, if that matters.

Vince
 
The TD8E has a simple 74123 Retriggerable Monostable Multivibrator to create the clock signal to write the timing track on a new DECtape. It was running at a 4.5 μs interval and should be 8.2 μs. There is a trimpot on the board to adjust the timing. Setting it to the correct frequency fixed the problem where the it would not format a DECtape. I was able to format four never used 1979 vintage DECtapes. It actually takes 6 passes to format and check the tape.

I looked at the SDLD glitch issue. One input to the 8881 bus driver for C0 is delayed by 20 ns as it goes through the 8251. There is a sizable glitch on the C0 bus line (see below) until the SDLD instruction is decoded by the 8251 and deactivates the bus driver. I ran a little SDLD test program for 20 minutes without a single error, so maybe my little 20 slot 8/e doesn't see the SDLD glitch where a longer (up to 80 slots) system might have a problem.

We need to look at the processor schematics to see how the C0 signal clears (or not) the AC register. Maybe it is tolerant of a C0 glitch at the beginning of the IOT instruction cycle.

SDLD_C0_Glitch.jpg
The upper trace is CC AND 09 H, the lower trace is C0 on the Omnibus backplane.
 
Last edited:
Hi All;
M Thompson, I would try it on a Breadboard or in the air, and take a scope picture and compare it to what You already have try using a 7400 or 7404 and see whether one or the other makes any difference..

THANK YOU Marty
 
We made a DECtape on the PDP-8/e containing OS/8 using Dave Gesswein's dumprest program. It actually booted and ran from the tape! It is painfully slow to run programs because it needs to spin a lot of tape to get to most programs.

So now we have one DECtape, one 8" diskette, two serial ports, and 32k of RAM working on this system. We need to determine if the diskette problem is in the analog board in the drive subsystem, or in the disk drive. The easiest way to do that is to swap the two drives and see if the fault follows the drive. The open connection to one of the timing track coils in the TU56 tape head will be more of a challenge. I can measure the capacitance at the connector contact and see if I can determine if a wire is attached to the connector contact. If not, I can use a Dremel tool to dig out the epoxy in the back of the connector and see if I can repair the broken connection. If the broken connection is in the tape head it will be a lot more challenging to fix.

Once we get both diskette drives and both DECtapes working it will be time to look for an RK8E disk controller and attach a removable disk drive. Debugging the RK8E is supposed to be a big challenge. The RK8E also uses Data-Break (DMA) so it will exercise logic in the processor that has not been used yet.
 
We removed the magnetic tape head from the right side of the TU56 DECtape on the PDP-8/e. The connection to the outer Mark Track was open. The tracks are redundant, so we could probably get it to work with just one Mark Track, but that would be inviting problems in the future.

We took the cover off the bottom of the tape head and found that the wires from the Amphenol connector go into a pool of cured epoxy in the bottom of the head. The wires to the connector tested OK, so the problem is in the tape head. This is probably not repairable.

Anyone have a TU56 tape head that they would sell?
 
m_thompson said:
We removed the magnetic tape head from the right side of the TU56 DECtape on the PDP-8/e. The connection to the outer Mark Track was open. The tracks are redundant, so we could probably get it to work with just one Mark Track, but that would be inviting problems in the future.

We took the cover off the bottom of the tape head and found that the wires from the Amphenol connector go into a pool of cured epoxy in the bottom of the head. The wires to the connector tested OK, so the problem is in the tape head. This is probably not repairable.

Anyone have a TU56 tape head that they would sell?
Click to expand...

If you can't come up with a new head, and if you can conclude you have nothing to lose, I've had very good luck with products from www.dynaloy.com. They specialize in solvents specifically designed to dissolve specific epoxies: http://www.dynaloy.com/products/potting-removers-and-encapsulation.
My guess is that the head is wound with some type of MIL-SPEC polyurethane based insulated magnet wire, while the potting epoxy is likely amine based. You might give them a call and ask which solvent will attack the potting and leave the wire insulation alone. If successfully de-potted, it may be a simple wire break... or you could rewind the head if you had to...not for the faint of heart, but I don't know how rare the heads are. Just sayin :)
 
You must log in or register to reply here.
Back
Top