[SOLVED] Project to replace CY7C64613 in the ICD2

[metal]

Hi all,

Is there a possibility for a group buy for the PCB, I think this idea sounds good compared to buying a ready ICD 2 from microchip.

Thanks

 

[Beowolf]

metal,
I doubt it would pay off. Whichever is the price you could get, add shipping, and it would be a signifficent ammount.
Why not just make your own PCB, there are hundreds of tutorials on that subject, eqipment is cheap and it is NOT DANGEROUS, unless you drink the chemicals, of course

 

[metal]

Making the PCB would cost me alot more than you are talking about, thanks for the compliment "Drinking Chemicals"...

see ya

 

[togarha]

Hi!!!

If I make a smt version of potyo design, how can I put the ICSP connector for the two pics inside the ICD2 (PIC16F877 & PIC18F4550), must I short all lines in two pics (RB6, RB7, VCC, GND) except the MCLR which must be selected by a jumper (with pull-down resistor in two lines)? or how must I do it?

Something like that?:

Thanks in advance,

Regards,

Added after 39 minutes:

Regarding the design I see that the MCLR lines have 2k2 pull-up resistors, then I think I must to change this resistors to about 100k and change the pull-down resistors of the jumper to 10k resistors, isn't it? other values are welcome...

 

[Beowolf]

metal:

photo-sensitive layered PCB 100x160mm - 2€
enogh for two units

NaOH 10g, for 1l of developer - 0.25€

FeCL3, 0,5l, enogh for 40-60 boards - 5€

transparent foil on which you print layout - 0.10€

CD case in which you press the layou on the PCB - you must haveat least one case without scratches...

1min of noon sunlight,
or 40 min of neon light - priceless

few old ice-cream boxes for the chemicals..

I see no expencive equipment here. Correct me if I'm wrong...

I just saw in hobby project section a thread on home-made PCBs. Look in there, some spannish vide tutorial is there, very similiar to my technique, except that I don't turn the light off when I peel-off the coating(no point, to destroy the film, it needs to be exposed for 40 min with regular light bulb), and I use FeCl3 for etching, and he uses H2O2+HCl, which is more harmful to breathe than FeCl3.

Only thing that can go wrong is the exposure time which depends on the source of UV (sunlight 1-1.5min, regular neon 40min, "modified" street light 4min), distance from it, and on thicness and type of UV sensitive coating. For determing the right ammount, you can use 1/2 of the board that can be chopped in pieces for testing.

Also, i heat up the chemicals to 40°C to spped up the proces. Developing shoud take 2-4 min, etching 5-10 min.

 

[Daemons]

If I make a smt version of potyo design, how can I put the ICSP connector for the two pics inside the ICD2 (PIC16F877 & PIC18F4550), must I short all lines in two pics (RB6, RB7, VCC, GND) except the MCLR which must be selected by a jumper (with pull-down resistor in two lines)? or how must I do it?

I'm not sure I understand what the problem is here or maybe I don't understand your question, but you only need the 877 and 4550 ICSP connectors to program those chips with the bootloaders the very first time. You'll only need to do this once normally, so you do so before closing the case you'll put your ICD2 clone inside.
An SMT version needs all the SAME connections as the non-SMT one. You need the jumpers to isolate the chips while programming them, and you need the same values for the resistors. SMT is identical as far as electronics are concerned. Don't overcomplicate things. This project is complicated enough as it is. :lol:

Btw, I'm also making mine SMT, but I'm using DIP 4550 and 877 (because I have them already), and my PCB is not finished yet.

Daemons

 

[togarha]

Hi!

Daemons:

SMT version does not need the same connections as non-SMT version because pics in DIP you can program outside the ICD-2 (in any programmer), but pics in TQFP you only can program when they are in the PCB, then, it need one ICSP connector to program two pics (only the first time, but it is needed)...

Regards,

 

[Daemons]

SMT version does not need the same connections as non-SMT version because pics in DIP you can program outside the ICD-2 (in any programmer), but pics in TQFP you only can program when they are in the PCB, then, it need one ICSP connector to program two pics (only the first time, but it is needed)...

Isin't that what I just said? :?

you only need the 877 and 4550 ICSP connectors to program those chips with the bootloaders the very first time.

That is what I said... As for the connections, obviously you need to add the ICSP connectors, as NICK7's design incorporates. And NO, you don't need one ICSP connector but two. One for each chip!!! Many design variations with ICSP for both chips are in this topic already.
I don't need them in my SMT version because, as previously stated... I'm using DIP version of the PICs. Everything else will be SMT though.

Daemons

 

[togarha]

ok, we are talking about two different designs (I do not see NICK7 design, I was talking about potyo design, as I write)...

About NICK7 design, it uses 2 ICSP connectors, ok, and a lot of jumpers, but:

It is possible to use the schematic previosly stated by myself to use only one ICSP connector and only one jumper (Without put, it works as normal ICD, put in one point it programs one pic, and put in the other point, it programs the other pic...)...

Thanks in advance...

 

[Beowolf]

Also, you can solder five wires to VPP,VDD,PGD,PGC,VCC, program the chips (in mid-air), and then solder them on PCB. It would last less than the time you need to develop an ICSP connector.
Or just put an jumper on all those pins, so when they are open, connect on them the programmer, wire by wire, and when closed, ICD is operational.

But with all those additional connectors,jumpers,etc.. you will lose all the benefits in size reduction of SMD design.

Also, take care of the case dimensions, I messed up mine.

Does anyone have an schematic in OrCAD of potyo 2 design?

 

[togarha]

Beowolf:

I know that with this connectors we lose some advantages of SMT, but I think it is minor loses than solder cables to the pcb wires...

I have potyo design in OrCAD, but it is not tested, I think it is right, but it is better other view...

I post it, if you find any error on it, tell me please...

 

[Daemons]

About NICK7 design, it uses 2 ICSP connectors, ok, and a lot of jumpers, but: It is possible to use the schematic previosly stated by myself to use only one ICSP connector and only one jumper (Without put, it works as normal ICD, put in one point it programs one pic, and put in the other point, it programs the other pic...)...

Have you tested this before saying it works? I see many post mods here that aren't sound and never tested. :!: I suspect RB6-7 from either chip will interfere with eachother. :| I urge you to test this out first.
If you ever modded an SMT PCB designs before, as I have you'd know it ain't easy. Be sure your SMT design works before you make your board. It'll be too late once the PCB is made, and you find out you can't program the chips.

You know they have SMT mini jumpers that are about the size of a 1206 resistor. I don't think thats too big for an SMT design. If you really don't want jumpers, then just make break links on the PCB, and keep them open to program the chips, and solder them up (jumping them) when you're done. Or use 0Ω 0805 shunts instead of jumpers, adding them only after programming the chips. These can always be removed if you ever have the need to reprogram a chip.

I really don't see the point of making it with only one ICSP. Good luck though.
Did you have a look at the RED SMT ICD2 picture SPARTAN posted in this topic on Dec. 13th. It's pretty small, and still uses 2 ICSPs.

Daemons

 

[Beowolf]

RED SMT ICD2 is 5x9cm, which isn't so smalller than Potyo's 8x10. If you're designing it for home use, no mass production, consider placing components on both sides. On RED all components are on one side as it is for mass production, where is cheaper (signifficantly) to place all components on one side. Maybe you can use two small SMD connectors for ICSP to save space, no jumpers.

P.S. togarha, I haven't looked your design yet, after the hollidays

 

[myster_e]

Rev.C is double sided board with minor modification :
added pot for adjust Vpp(9v..14v)
added rezistor for mclr PIC18F4550
and change capacitor for oscilator with 12pF(C6,C7)

It is possible to obtain typons of this rev C.

This could be really great for all of us !!!

Thx in adavance for your response.


By the way, Happy new year everybody .

 

[Frazor]

Hello and Happy new year,

I'm also interested about a typon for the C version.

Thank you for this great job.

 

[Beowolf]

I have potyo design in OrCAD, but it is not tested, I think it is right, but it is better other view...

I post it, if you find any error on it, tell me please...

You have changed the design, so I can't confirm that it is identical, as it isn't.
I have few remarks, considering decoupling. You should put two capacitors on each PIC, an ceramic one with electrolythic on that LT converter on input, to reduce noise on Vdd.
Also, reduce capacitors on xtal, from 33pF to 22pF.
PICs are connected properly, but you have changed almost all other chips, so I can't help you out on that.

 

[togarha]

Hi!

I will see decoupling capacitors...

About xtal capacitors, y need 33pF (or 27pF) to the xtal due to its load capacitance (value of xtal manufacturer), but I test it when the xtal is on the PCB (minor changes)...

About the other chips, I change the transistors by DG411. I think it is the only change I made, isn't it?

Regards,

 

[Beowolf]

mine was more stable with 22pF.
you changed DC-DC. Originaly it was MC34063.
Either way, It should work...

 

[togarha]

Hi!

In potyo design originaly was the LT, after that it change to inexpensive MC34063...

regards,

 

[jim2098]

Just want to give a thank you everyone that got this project to this point. I have got mine working and I'm very happy with it.
One problem I had was with the 16f877a bootloader. The ones posted did not work until I loaded with the icd661.exe. After that everything seems to work. I even went back and tried them again and it would not connect to the PC.
Any one know what the problem might be?
Again thanks to everone that made this possible.
.