What's new
What's new

German- "fixgra checksum" ?

Milacron

Administrator
Staff member
Joined
Dec 15, 2000
Location
SC, USA
On Dialog 11 control, the result of "Selbsttest Laeuft" is "NGE fatal error: Fixgra checksum"

NGE is one of the Grundig modules.

I'm hoping this is the result of missing parameters but can't tell from the German...Martin ?
 
Hi Don,

It's common practice to protect programs loaded into Read-Only Memory (ROM) with a checksum, which is formed by simply adding the contents of all the memory locations and storing the resulting value. A test is performed by doing the same calculation and comparing with the stored result (or other similar algorithms).

The first suspect would be a socketed ROM. If there are any on the board, pull them out and plug them back in to clean the contacts, and run the test again.
 
The first suspect would be a socketed ROM. If there are any on the board, pull them out and plug them back in to clean the contacts, and run the test again.
There are, and I will try that...but, I've done that many times in the past on other machine controls and it *never* was the problem. So, I'm curious, have you, or anyone you know, ever reseated IC's and had that actually cure the problem ? Just wondering if this whole reseating business is really an urban myth !

Having said that, I did have one instance of where I cleaned the contacts with eraser and reseated some PC boards that did seem to help...but never an IC.

So, what's the real likelihood of that actually doing any good...5 percent?....001 percent ?
 
Hi Don,

Yes, I have had IC lead problems that were corrected by reseating. Usually on equipment that was operating in poor environmental conditions (humidity, fumes, etc.). But I've seen an awful lot of equipment. Been doing microprocessor hardware and software design since the 8080 was introduced in 1974.

If I had to put a number on it, I'd say probably 1% to 2% success rate. But that's an infinite improvement over not trying it in the first place :D

And the pencil eraser trick on board edge connectors has saved many a pound of bacon :D
 
Usually on equipment that was operating in poor environmental conditions (humidity, fumes, etc.).
This particular control is pristine...like new inside. Sounds like I'm SOL :(

I think "selfsttest laeuft" is " self service in process" but still can't figure out what "fixgra" means.
 
It is probably the German equivalent of an electrical engineering term that is not well known in the popular venacular. Diode might be a good example. I'm guessing it is something totally different in German. 95% of the native English speakers (probably more) wouldn't know what a diode is either, however.

As for the reseating... Yes, I have had that make a difference. Not so much on EPROMs but other chips. The failure mode isn't specific to one type or the other, however, so it is not a bad idea.

The previous comment about the formulation of the checksum is right though. So I'd center in on the EPROMS or PROMS. Sometimes (often) the chips have stickers on the top of them which show the checksum values. If that is the case, you can zero in on the one causing the problem (while you have them out of the socket
). Assuming you can find the faulty one quickly, I wouldn't imagine that you will have that many problems getting a replacement. DonS or Arno can probably help there.

Technically an EPROM has a stated life of about 12 years, as I recall (one of the reasons I make copies of all of my EPROM based machinery). From a realistic perspective, it should be much greater than that. Given these thoughts, I'm surprised that your 'late' model control would be having memory problems.

--Alan
 
Alan, et all...I don't know if I spoke about this already, but until last night, I've been "blind" on the FP2A because the monitor doesn't work. Also when one hits the control "on" button nothing happens. I've been assuming all this time that nothing happens because of the machine sitting for 3 years disconnected and the Saft NiCad battery voltage go too low and the control lost parameters.

Recently, I figured out a way to connect a standard PC monitor to the monitor 25 pin cable*

To my great annoyance, once my blindness veil was finally lifted, instead of the expected screen where I could finally check to see what parameters were lost and such...instead I get a message that is basically saying the Grundig won't boot up past a certain point due to a "fatal error" I presume there is no way to check the parameter situation until this error is resolved.


And guess what module, out of all the modules, the NGE is ? Why, the MONITOR module of course !!! :mad: :rolleyes:

So now I don't know what to think. It just seems too coincidental that out of all those boards, the one board showing a "fatal error" just happens to be the same board that has something to do with the monitor. Makes me wonder if the signal to my machine monitor is screwing up that monitor, but is "good enough" to run a PC monitor somehow ??

Or is the monitor screwed up of it's own accord (btw, the monitor failing is supposedly common on these as the whole CRT and associated electronics are totally enclosed and with no fan) AND the monitor card screwed up of it's on accord (and yet not "too" screw up to run a PC monitor and have it's red "not ready" LED off and it's green "on" LED glowing pretty as you please) , with no relationship between the two events ? Just seems impossible cooincidence.

I even tried booting up with monitor cable dissconected at the NGE and then reconnecting it. Instantly the same "fatal error" appears on the screen of the PC monitor.

So, the bottom line is..out of all the 'fatal errors' I could possibly get, this is the worst possible one, as it's the most inexplicable !

Aaaaarrrrgh !! :mad: :rolleyes:


dekk40.jpg


*Now, aren't you curious how, with the Dialog 11 monitor only showing one thin line on the screen, no wire color coding, and no schematics, did I figure out which wires were R,G,B,V and H ? ;)

dekk10.jpg


Btw, the strange looking color gradients you see on the PC CRT was a temporary thing that cleared up after the photo was taken and now looks perfect...yet another mystery ! :confused:
 
Maybe it requires a proprietary monitor to work properly. Some disk and tape drives have special firmware to stop them working in an environment other than that intended by the company selling them. If you wanted to upgrade an older HP workstation it HAD to have the special software in the driver.

It's so you don't buy cheaper from somewhere else.

Chris P
 
bcs, I thought about that. But I asked Volker (the German guru of older CNC Deckels at DMG)last week if the monitor had to work for the machine to boot up and he said "no". He also figured the machine had simply lost parameters.

But I presume until the control boots up there is no way to install the parameters...I could be wrong about that, but to upload via RS232 you have to hit various keys on the keyboard, and I see no evidence that any of the keys are "connected" to anything until the control boots up. Plus I was hoping to avoid the whole RS232 thing and just enter any missing parameters manually since I now have a printout of them.
 
Hi Don,

Your machine is probably dead until it passes self-test. If any failure is detected the machine will just lock up with the error message displayed. This is common practice with any microprocessor-based equipment. It would certainly be true of a machine tool where there are safety considerations.

The built-in monitor has lost vertical deflection. You should have a raster, even if it's blank. Most likely a blown power transistor.

The color gradient can be caused by an external magnetic field. The built-in monitor may have some magnetic shielding to protect it, but the PC monitor certainly does not.
 
Sinlge line on the machine's monitor, I see... Is it possible that the vertical sync lead is broken and that the monitor is 'painting' everything in one line? If you have a scope, you should see a sawtooth waveform on the monitor board's HSync and VSync. Be aware that all CRTs have flyback transformers in them to generate the voltage for the electron gun. They can give you a very good zap! So be careful in there.

A monitor *should* be input only, of course. So Vogler's comment makes perfect sense. Baring some noise related or impedance issues, I can't see how the faulty monitor could be affecting your start-up sequence. Of course, a designer can do *anything* to trip you up if they want too! You've been respectably tenacious in trying to figure out a way to find a suitable monitor though... I applaud you on that front.

Can you shoot a picture of the NGE board? Many monitor cards will have a character generator, so maybe that is where the checksum error is coming from (character generators are often nothing more than specialized EPROMs or PROMs, especially in stuff that old).

--Alan
 
The color gradient can be caused by an external magnetic field.
That's interesting as the color was fine on the PC monitor the first time it came on and only got "wierd" on me when I moved the montior stand to another position relative to the machine. Just another Red Herring thing I guess.

So, I guess it's 99 percent that the Dialog 11 monitor is blown. But how on earth could the monitor card, out of all those cards, just happen to be the one with the fatal error....and yet does seem to "work" perfectly fine ?

Gawd....I have some "unique" problems don't I ? :rolleyes: ;)
 
Don,
what do the Grundig modules do ?

I reread the post and I don't understand how you are drawing the connection that the video driver is bad.

The screenshot shows me that the video driver seems to be working fairly well.


Years back I had a simple raster display that only gave me one horizontal line. The fault was in the display driver not amplifying the signal (thus no deflection in the gun). Sure doesn't appear to be the problem here.

The checksum could be in any of the control boards. I agree it will probably have a sticker on a DIP style IC. "Fixgra" doesn't mean anything outside of the Dialog11 context, so hopefully somebody will make it out.
 
I reread the post and I don't understand how you are drawing the connection that the video driver is bad.
I'm not drawing the connection that the video driver is bad, the "Fatal Error NGE" is drawing that conclusion, and I'm as mystified as you as to why.

However, just got off the phone with Volker and now it seems that sometimes when a monitor goes bad it in turn causes the NGE board to go bad via the communications with the other modules. It seems it's not just just a graphics board but also a communication board. There is even the possiblity of lightning/voltage spike, harming both at once. Cute, huh ? :rolleyes:
 
If that is the case, (get ready to smack me) but is there any feedback from the monitor?

If the monitor itself died like in the flatline pic, how would the Dialog11 know? There needs to be something special about the monitor.

When you spliced in the PC monitor, was there more to it than RGB&synch? Any idea what else is going on in that 25 pin cable?
 
"If that is the case, (get ready to smack me) but is there any feedback from the monitor?"

I assume not based on converstations with Volker. As already stated, he said monitor does not have to work for machine to boot up.

"If the monitor itself died like in the flatline pic, how would the Dialog11 know?"


How do you know it does know or needs to know this ?


"When you spliced in the PC monitor, was there more to it than RGB&synch? Any idea what else is going on in that 25 pin cable?"

Nope, just 5 wires (and 5 ground wires...one for each function). Why they used a 25 pin connecter is beyond me...probably just to throw off people like me from doing what I did !

The only other monitor related wires are one (3 wire) cable coming off the Dialog 11 monitor for the remote brightness control pot, and one (2 wire) cable that is a mystery. I get a reading of 12 VDC from it's terminals on the monitor board, but can't see where it goes...seems to go to the keyboard or keyboard interface board. But, whatever it does, there is voltage getting to it (~from~ the monitor, not to the monitor) even with the monitor not connected to signal, only connected to 220 volt A/C input.
 
monitor goes bad it in turn causes the NGE board to go bad via the communications with the other modules.
It sounded like there is some method for a signal to get back to the brains?

From a normal electronics point of view, you can run the broken monitor for 100 years and it won't damage the board that is driving it.


Guessing that the other two wires may be for a 'display off' switch of some type? Just a swing in the dark tho.
 
Hi Don,

The two wires to the interface board may be for synchronizing a joystick or similar device. If the display supports a cursor, its position information has to be recorded whenever you click a mouse button or perform a similar select function.

The interconnect is probably a signal from the interface card to the video card to latch that position information into a port that can be read by the processor.
 








 
Back
Top