Modding GX4000 Component Collection Begins

[Badstarr]

Hi all!


As I'm sure some or maybe most of you will know I'm embarking on a project to turn a GX4000 into a fully fledged CPC+. I intend to report back here every now and then on my progress.


I have not started soldering or anything exciting yet, I'm currently looking to find the parts i will need. I have targeted the more hard to find ones first and hopefully ebay has provided a compatible Floppy Disc Controller D765AC-2 765AC NEC D/C 8727 IC NEW and two DRAM chips at decent prices RAM KM41464AP-12 SAMSUNG 18-Pin DIP 41464P. I include the links in case anyone happens to be looking for these parts.


Judging by the DATA sheet as far as I can see the FDC chip is compatible and the DRAM is identical to that installed on some (most?) plus machines! A good start by the looks of it! 


EDIT Also looks like I got the DATA Separator as well based on info in an earlier thread SED9420 on ebay still some available too!

[Badstarr]

Right... I seem to have sourced all the disk interface chips that I need all from ebay after a little lateral thinking! The last thing I need to get is the Resistor Network labeled NR101 on the 6128+ PCB. In the service documentation this is labeled as EXB-F7E681J I have searched high and low and can't find it's spec anywhere, I could open up my Plus machine but I'd rather not as it's spending more time in pieces than I would like! 


Anyone know the spec for this component? 


EDIT I have looked at the schematics for the 6128+ on the wiki (mine were pretty hard to read due too poor resolution) the component in question says 6*680 I presume this means 6 banks of 680 Ohm Resistors, correct?

[Bryce]

Yes NR101 is a 6way (Star) resistor network. It might be difficult to find a 680Ohm part though, but you could always just use six normal 680ohm resistors soldered together if you don't mind it looking a bit messier.

Bryce.

[Badstarr]

Hi Bryce, thanks for confirming! I was going to build the Resistor Network  myself looks simple enough, but to limit the number of potential teething problems with the project I decided to get the item from a hong kong supplier after looking around for ages.


Now...tell me this... does anyone else want one of these? It's just I will have at least 50 of them lol! 

[Bryce]

Looks like you'll be stuck with 49 of them then They are being used as pull-up resistors, but it's quite a strange value to choose. I would have chosen a 1K pull up in that situation.

Bryce.

[Badstarr]

Maybe they had a surplus of 680 ones too? 

[redbox]

Very nice.


I take it you're going to connect the FDC directly to the ASIC rather than going via the expansion port like the previous project did (they just attached a DDI-1 to it)...?

[Bryce]

Well technically, that's exactly the same thing, they just chose to connect the hardware at a point where all the required signals are available in one place (ie: the expansion port) Which isn't all that bad an idea. And the hardware implementation is almost identical.

Bryce.

[Badstarr]

I am going the direct route, it is technically the same thing, however as DDI-1's are a little light on the ground and I no longer have one after I accidentally destroyed it 12 years ago    I am going to build the components onto a "daughter card", if this works well I see no reason why, with a little modification perhaps, it could be attached to an expansion port of a 464 for example, obviously with AMSDOS ROM installed.


All I have to do now is sit back and wait for the components to roll in, it's gunna be like Christmas here! Woohoo!


I decided to source the parts separately rather than getting a 6128 and gutting it for parts as it's hard to predict what variants of data separator etc are going to be installed and I thought it would be quite a cool experiment to see how easily the parts could be found. Assuming they all play nice together it was easier than I thought to get the parts just a bit of lateral thinking was needed when searching ebay etc! 

[Bryce]

Judging by the info on the ebay page (ie: the picture) there should be no reason they shouldn't work. They are the correct parts and I'm not aware of any 765 / SED combinations that don't work together, or is there one I'm not aware of?

Bryce.

[Badstarr]

I think you will be right I can't see any reason why it wouldn't work. I suppose we will get a definitive answer when I solder it all together

I think the basic functions required are all there in the SED I ordered, I checked out the data sheet and I can't see anything untoward, I suspect if it were possible to dissect enough CPC's we'd probably find some machines probably have the same chip fitted. The data sheet doesn't seem to attribute any significance to the extra "A" and "C". I'm not great at deciphering IC numbers but perhaps they are different production generations or something like that? Or maybe the A and C are implied and not present on earlier packages? 

[Bryce]

There are several suffix systems used depending on the manufacturer. But they always signify non-functional characteristics such as package type (DIL/SMD/PLCC), package material (Plastic/Ceramic), operating temperatures, military grade or even delivery packaging (Tray / roll / tube), they never signify any difference in the technical workings of the IC. So you can safely ignore the C.

With transistors it's slightly different, where there is a slight difference between say BC547A and BC547C (but usually only the maximum operating voltage).

Bryce.

[Badstarr]

Thanks for the info! I suspected that the suffixes wouldn't make a difference in function but couldn't be sure. On that basis I ordered a Schmidt Trigger, it arrived just 10 mins ago and it's a SMD part, about the size of a lentil! Doh that's going to require a steady hand to solder but at least it's strictly speaking, the right part. The NAND ICs have also arrived so the parts are starting to roll in yay!

[Bryce]

You ordered a Schmitt Trigger?? What exactly is the device? A schmitt trigger is a type of circuit, not a device. Do you mean a schmitt triggered gate of some sort (NOR / NAND etc) ?

Bryce.

[Badstarr]

Oops I think I was a little inaccurate there, it's IC no 107 in the 6128 plus schematic. Its full title is MC74CH14AN Hex Schmitt Trigger Inverter, so I guess its not a trigger in itself?

[Bryce]

Ok, so it's a six way Hex Inverter with Schmitt triggered inputs. Schmitt triggering is a method used to "straighten" the edges of a signal as it rises and falls. Using it on an input improves the noise tolerance of the IC, using it on an output improves the reaction time of the gate. But why did you go for the Lentil sized SMD version?

Bryce.

[Badstarr]

I was wondering what exactly it was used for to be honest, I checked to see what a Schmitt Trigger did and found a lot about electric sewing machine foot peddles, the idea being that it clipped a sine wave into a square wave so there was more off a definite "on off" to prevent overheating or something to that effect.


The reason I went for the lentil sized version was that it was easier to get hold of, all I really did was check that there was no difference in function on the data sheet then I went for the SMD not really thinking far enough ahead that it may be a lot smaller! I guess you live and learn, but hey at least I got the part even if it's going to be a bit of a challenge to solder at that scale! 

[Bryce]

Well I could go into a long and boring description about hysteresis methods, positive feedback loops and transistor based comparators, but you'd probably loose interest or even fall asleep after the first paragraph Let's just say, the output of the device is partially fed back to the input, so if the input starts rising at a rate of X, then the actual input rises at X + the feedback voltage, making the curve steeper (logarithmic instead of linear) and reaching the threshold point (the voltage at which the bit toggles from 0 to 1 or vice versa) much sooner. The steeper the "edge" the higher the frequency you can switch at.

Bryce.