Looking for Deckel FP2-NC Dialog4/NPP90 EPROMs

I recently got another broken milling machine that needs some love, it's a super cheap Deckel FP2-NC this time.

It randomly gets a BA00 30 BA16 error during bootup. I since have checked all batteries and the SRAMs on the NSP55 board. All power supply voltages are good, too. Changed some dubious caps on the NSV just for good measure. If the machine decides to run, I also get an error 28 sometimes, which points to bad EPROM in the NPP90.

Due to the randomness of the fault, I now believe some EPROMs (there are a total of 33 EPROMs inside that machine, total madness) have gone bad (charge leaked out of the floating gate).

I had that kind of problem in the PLC of my now fixed Maho 700P/TNC135, problem was gone as soon as I put newly programed EPROMs in it. Only a few of memory locations need to change and the machine develops a life on it's own. Also, depending on temperature and operating voltage, a faulty EPROM might read differently, because the reading circuitry is still analog at it's heart (voltage levels gets sensed differently). This is the reason why the error occurs at random.

My software versions
NPP90: 351.04 (4x TMS-2564)
4x NZP59: 352.10 (je 1x 27C128)
NRP53/54: 353.81 (4x 27C128)
NPP54: 353.42/2.19 (8x 27C128)
NEP52: 353.32/2.16 (10x 27C512)
NPP55: 353.21/2.14 (2x 27C512)
NCR53: 352.00 (1x27C256)

All EPROMs should be fairly easy to read, the TMS-2564 can be read with common programmers, if the pins are permuted according to this document:
www.dasarodesigns.com/wp-content/uploads/EPROMUniversalAdapter.pdf
All pins connected 1:1, except these: (P: programmer socket side, C: chip socket side/TMS-2564)
2P=>22C; 20P=>23C; 22P=>27C; 23P=>2C; 27P=>20C

So, if anyone has already dumped their ERPOMs, could you please send me the binaries? It's also quite smart to backup the data inside the ERPOMs while your machine is still running. Any cheap TL866 will work for that, only costs $40 on ebay. Would be especially nice, if I could somehow get newer firmware revision, 2.33 or 3.07 .


TL;DR: Please someone send me the EPROM binaries for/from a Deckel Dialog 4. V2.33/3.07 preferably
Mail: hinzz1 (at) hotmail (.) com
Thanks

Would first try reloading the software...See this thread, post #7
http://www.practicalmachinist.com/v...l-fp3nc-dialog-4-power-start-up-issue-340205/

Cheers Ross

AlfaGTA said:

Would first try reloading the software...See this thread, post #7
http://www.practicalmachinist.com/v...l-fp3nc-dialog-4-power-start-up-issue-340205/

Cheers Ross

Click to expand...

Tried that, didn't help. Also put all SRAM chips on NSP55 onto sockets, so I can run it in minimal 128kB memory configuration. I also connected that 3cell NiCd battery on NSP55 by wire with connector in between, so I can reset the data on NSP55 easily by disconnecting the battery. The weird thing is that the machine was working well again after I socketed the SRAMs on NSP55. Was playing with the machine that evening, everything was working fine, then the next morning (had bad weather that night), back to BA00:30 BA16 error without doing anything to the machine.

Thereby I'm quite certain that is is either a bad connection/broken bond wire/bad chip socket or an EPROM that is currently on the edge of loosing it's data.

I also tapped all socketed chips on the computer boards (NEP/NPP54/NPP55/NSP55) with the back of a screwdriver lightly, in the hope of eliminating any bad contacts in the sockets. Of course, that didn't help either.

In case anyone cares, I got that FP2 running again. Wasn't an EPROM issue this time. BD139 transistor, T8 on NPP90 had visibly blown up, which I noticed as I was changing EPROMs. Replaced it with a 2SD794 from parts bin and machine started fine.

Schematics in orange book aren't exactly readable for NPP90, so I couldn't completely figure out what it's function is, but as long as the replacement works, I'm happy. The "BA00:30 BA16" fault is reproducible by pulling NPP90 before powering the machine, so it's a good indication that there is something wrong on NPP90.

Also got EPROM versions 2.16-2.19 and 2.32 (currently working, paid some money to get the dump). So in case anyone else needs Dialog4 EPROMs, send me an email (and give me a small donation, if you can ;-).

Still looking for V3.07, which apparently has DNC added (not that I need it, but it's always nice to have)

Here are some pics of my little baby:
https://www.ricardo.ch/kaufen/model...zeugfraesmaschine-deckel-fp2nc/v/an935098816/

DeadMahoDude said:

Here are some pics of my little baby:
https://www.ricardo.ch/kaufen/model...zeugfraesmaschine-deckel-fp2nc/v/an935098816/

Click to expand...

Did you really get that for ~500 Euros? Wow!

Yes, was a bit surprised, too ;-) With transport and parts for various repairs, I'm now at around 1k€.

Unfortunately, it had a harder life than my Maho 700 (has better control, so it likely was used daily for the last 30years). Vertical spindle needs regrinding sometimes in the future (bell shape, lots of small dings from tool changing and a small fretting ring, presumably someone crashed it). Bearings are fine, though.

Ball screws likely have some wear too, but I'm not exactly sure, yet. Servos have that high pitched sound and if I use the handwheel, it sometimes has problems reaching a given point accurately. If that happens, I can't unselect the axis on handwheel and have to move a couple of mm more until the handwheel position matches the readout again, mostly X axis that is affected, so I suspect it's wear.

Anyone an idea on how to check ballscrews for wear? Overall, I can't really complain at that price, though. That nice universal table alone is probably worth more than what I spend on the machine so far.

I don't think the high-pitched servo whine reflects a problem. My FP2NC D4 flip-head has the noise. I think it is just the nature of DC brushed servos from the mid-1980s. Regarding ballscrew wear, the machine's glass scale feedback can compensate for a lot of sins in that department. Try milling a large circular profile and sweep it with an indicator. Is it round? Are there lumps at the X and Y extremes, where the screws have to reverse? If not, don't worry about it unless the screws sound like rock crushers and are shedding metal. You can also do the test with the horizontal spindle in the XZ plane, and theoretically in the YZ plane with the vertical head tilted over 90 degrees, though I've never tried that. My 1986 machine still makes nice round profiles that are acceptable as bearing bores in plane vanilla machinery applications. I use light finish cuts and moderate feeds so that the machine can keep up with the commands and to minimize tool deflection.

Look up "circle-diamond-square" test. That is a more rigorous test of a machine's performance than simply milling a circular profile, and it's relatively easy to perform using normal shop measuring tools.

Nice looking machine! Nice score on the table!
Some areas to evaluate.....Rumbling sound raising the "Z" likely the thrust bearing for the screw....If machine has had a life doing any drilling that screw and the thrust bearing will likely need some TLC...
Ball screw nut on the "Z" does not nave any provisions to have pre-load...the nut is one piece. Screws are high quality and for lots of machines replacing the balls will restore "like new" tolerances.

Dialog Control does not have any back lash comp...so excessive wear on the "X" and "Y" will result in less than optimal part geometry...Rich's suggestion is a good one. Less critical on the "Z" as gravity
helps remove any slack from wear.
Have re-balled a number of Deckel screws with success, but its (for me) a matter of trial and error trying different size balls til you get the fit correct.. Generally not something you can just measure and fit
in one go unless you are extremely lucky...changes of size from too slack to "OK" can be effected by a size change of 10-20 millionths.....My technique is to by a range of balls and try them starting small and working up.
I use Bal-Tek in LA, but suppose you have sources closer to you...I use grade 20 balls (accurate in size and roundness at 20 millionths)

Baltec - Home

Thrust bearing on the "X" axis needs to be lubed (grease) and adjusted to have slight stretch on the screw....

Whine from the servo motors is normal...My belief is that its mostly a function of the switching of the regulator board....while the motor "hunts" to hold position.
Sometimes tuning the servo and cleaning the tach will reduce the whine....but pretty much it is always present.

As to the software....3.07 indeed supports DNC...but it requires a hardware change and has some other unpleasant side effects that the earlier software does not...such as rounding errors
As to the hardware: You need i believe the NEP 52 A board in place of the NEP 52...and no you won't find a schematic of that board in the Orange book.


Use lots of lube cycles (M7) ...These machines suffer from wear/gauling issues on the vertical ways (inside face opposite the operators station) Also the bottom face of the "X" axis slide.
Move the vertical slide to the top of its travel and look under the slide on the inside face of the box ways opposite the operators side....
Inspect the bottom of the "X" slide by moving the axis fully right and left....

Use caution when moving the machine from vertical to horizontal and back.....There is a gear shift lever and cam that disengages the vertical drive when going horizontal....The slide has a bit of mass, and
when moving forward to go back to vertical its not difficult to damage the cam plate....Seen at least 3 of the Flip head machines that had broken of damaged cams.....Once the cam is damaged, sometimes
difficult to get the upper slide to move and that makes the repair hard....

Would invest in a full set of way wipers.....For sure the top wiper on the "Z" slide...gets lots of abuse and should be changed regularly...
Would also change the oil in the spindle gearbox....AW 46 for lube, same as you use in the hydraulic sump.

Cheers Ross

DeadMahoDude said:

Y
Ball screws likely have some wear too, but I'm not exactly sure, yet. Servos have that high pitched sound and if I use the handwheel, it sometimes has problems reaching a given point accurately. If that happens, I can't unselect the axis on handwheel and have to move a couple of mm more until the handwheel position matches the readout again, mostly X axis that is affected, so I suspect it's wear.

Click to expand...

My FP7NC with Dialog 4 did that too on the X axis. Deckel Doctor adjusted up the ballscrew support bearing to take as much end play out as he could, and that improved it dramatically. It seems that mechanical backlash allow the servo position to get unaligned slightly with the glass scale, and that conflict prevents you from unselecting the axis on the handwheel. Reversing the handwheel one click will relax the ballscrew without moving the readout position on the screen and allow you to unslect the axis. A new bearing for mine is $2000 so I'm living with it for now, it happens about one out of 15 times. Thrust bearings wear faster than ballscrews, look at them before you condemn your ballscrews.

Finally got that thing back together and running as smooth as my other Dialog11 FP4-NC!

Things that got fixed:

-fixed bad transistor on NPP90

-milled and scraped the galled up lower box ways on X, bronze lining on the inside of the slider box, rebored ballscrew mount at about 0.5mm offset, shimmed gib with 0.5mm Phenolic Pertinax sheet. Photos:
Cutting an inside dovetail oil groove on a manual Deckel FP2

-rebuild Y ballscrew after balls got puked out of a failed ball return piece, found them in the cover below. Milled new ball return thingies from Al-bronze, fitted them with Loctite 270. Reballed screws with new 3.175mm G10 balls that I thankfully got from Bruce/ballen. Adusted ballscrew preload a couple of times (by turning top ballscrew piece on that splined plastic ring by one notch)






-replaced rotten wires in the hydraulic pump. Fully disassembled the pump. Ripped out the electrical wires. Soldered new wires from motor windings to a new piece of circit board that I attached by twisted, plastic coated florist wire to the side of the motor, to relief all mechanical stress from the motor wires. Then soldered new wire leads to that PCB which were then connected back to the electrical panel at the back of the machine (no photos of that, unfortunately).

-Fixed that annoying handwheel positioning error and DC Servo jitter / oscillation problem decribed above. Since it happens on all axis, it likely had nothing to do with tachometers, servo amplifier, worn ways or bad ballscrews.

Because Dialog4 uses 1 DAC that is multiplexed over all 3/4 axis, the most probable cause for problems on all axis is likely found there (after I measured all voltages and ripple on NSV56 and these looked good). So I looked on NRP53 and fiddled on R1 potentiometer while moving an axis left to right with handwheel. (no idea what it does exactly, but since it is directly connected to the DAC, I guessed it would be a good canditate to fiddle with). After a bit of fiddling, it got much better and I now got it so good that I can reliably move an axis in 1/1000mm increments with the handwheel, like the other D11 FP4NC that I have.


I have a few spare bronze and plastic ball return thingies left, if anyone want them, send me a mail. Something to return to the comunity after I got free balls from Bruce .

Bronze ones might require a bit of fitting with a file and Dremel, before they run totally smooth.

Current situation, machine works, things to do:

-regrind vertical spindle

-move machine trough a 1m wide door

-put that gorgeous universal table back on



Scraped X-boxways, mostly scraped for flatness, might not be Richard King quality, but plenty of good enough for my shop: