The differences between ICD2 clone designs

[jaydag71]

I have been looking for the true ICD2 clone design schematics and all ascociated documentation, think that I have found it, and the documenation states that the design was originated from ths discussion board and another. It states that the design is a reverse engineered ICD2. I am hoping that someone here knows which design I am referring to, the other link is in german or something, but their design is not one of the simple approaches, rather it uses the cypress USB chip and a bunch of what look like logic and chips. **broken link removed** that is the site, the rest of them use the simple approach with the FTDI usb-rs232 bridge, and few other parts. What is the difference between the approaches I wonder, as a beginner to electronic design I cannot see what they are, so before I go and build one, why shoudl I not go for the easy way out? Does the larger design have more functionality? thanks in advance, jaydag71

 

[William At MyBlueRoom]

buy ftdi232rl

It appears to be accurate, among the best ICD2 clone I've seen. Problem is getting the cypress chip (seems to be in the real ICD2)

Personally I like this the following ICD2 Clone (Serial)

**broken link removed**

I really think the voltage doubler using the RS232 chip is brilliant.

The most basic clones do not have the variable voltage supply. Also if you use a 16F877A this is the firmware to use.

www.myblueroom.com

Bill

 

[jaydag71]

ftdi232rl eeprom

Yeah, so I am just waiting for the rest of the semiconductors to get here, i have to make a few adjustments to the layout to accomodate some alternate packages, but I should be good to go by the weeks end. I have ordered the TI usb controller IC and am going to look into using that instead, I have seen it used in one of the design approaches that I was considering, so the conversion should be minimal, but concerning aquiring the cypress usb IC, that should be no problem at all as I have found dealings with them in the past to be pleasant and prompt, so it could go either way at this point. Thanks for the firmware link, I already had it though.

 

[philba]

chip ftd1232 buy

these seem much more complex than the Stolz ICD2 design. **broken link removed**

Note that his USB design can be further simplified by using the FTDI 232RL which needs no oscillator or external memory.

 

[albert22]

ftdi232rl buy

Anybody knows what is the funcion of Pin 6 (RB3 target of the ICD2) ?
I have found no info on the Microhip docs. In every example that they show this pin is not connected.

Some ICD2 clones implement this signal and others not.
Most of the clones use an OR gate (RB1+RB6) to activate the OE of the output buffer. However in the "original ICD schematics" (posted at www.mcu.cz) the OE is implemented with a NAND gate giving a different function.
Also, the supply for the output buffer of this signal in some cases is Vcpu and in other it is Vtarget (which I think is more correct)

 

[qk]

ftdi 16f876 schematic

Hi jaydag71!

I´m also looking for a "good ehogh" ICD2 clone. I don´t care if is USB or serial.
Have you build your ICD2?

Thanks

Added after 22 minutes:


Note:

I was checking different ICD2 versions ("stolz", "nebadje", "MCUServer",... versions).
All of them are "similar". They use different parts to make the same function: switch voltages (including a "no usual" DG411 analog swich).

I´m getting crazy...
Do anybody knoks a EASY IDC version (only serial, it´s OK), made off "regular" parts, that can debug 16F877A and compatible with MPLAB?

Please, help me!

 

[Kripton2035]

+example +dg411 +circuit -datasheet

you can buy a cheap one fully built here :
**broken link removed**

micro4you is a member of this forum...

 

[philba]

ftdi232rl

Hi jaydag71!

I´m also looking for a "good ehogh" ICD2 clone. I don´t care if is USB or serial.
Have you build your ICD2?

Thanks

Added after 22 minutes:


Note:

I was checking different ICD2 versions ("stolz", "nebadje", "MCUServer",... versions).
All of them are "similar". They use different parts to make the same function: switch voltages (including a "no usual" DG411 analog swich).

I´m getting crazy...
Do anybody knoks a EASY IDC version (only serial, it´s OK), made off "regular" parts, that can debug 16F877A and compatible with MPLAB?

Please, help me!

Actually, the serial version of stolz's design is very simple. If you just want to program PICs, you can go much simpler. But if you want debugging, I think you can not get any better.

 

[William At MyBlueRoom]

icd2+micro4you+usb+to+serial

Anybody knows what is the funcion of Pin 6 (RB3 target of the ICD2) ?
I have found no info on the Microhip docs. In every example that they show this pin is not connected.

It's the low voltage program pin. Not worth using as you lose the PIN after it's programmed (Has to be grounded)

The stolz design does work though. There is a catch to all the serial programmers.

After you program a bootloader...
You MUST on your windows machine (from 95-XP)

USE comm1 or comm2 only
DISABLE your comm port FIFOs
ENABLE hardware handshaking

Added after 3 minutes:

In a previous message I mentioned...

I really think the voltage doubler using the RS232 chip is brilliant.

Built it, the voltage doubler does not work

Ah well, a 78l08 & 7805 combo work fine. 13V

 

[qk]

ftdi232rl buffer disable

Thanks people.

I already have several "easy programmers" (serial and parallel).
Now I need an "simple emulator", I mean, an ICD2.

Stolz´s ICD2 Clone (**broken link removed**) seems to be simple.

Questions:

1) Must I programm the Clone ICD´s PIC with the firmware provided by Microchip ("Bl??????.hex"). Is the hardware of Stolz´s ICD2 compatible with Microchips´ICD2?

2) .hex file is compatible with PIC16F876 and PIC16F877?

3) I found some differences between Stolz´s circuit and some reverse_engineered_ICD2s like http://www.harald-sattler.de/assets/images/USB_ICD2_Schematic.gif (in RA3 and RC1 pins)
Is stolz sch OK?

4) MPLab "sees" USB port like serial COM, does it? So, MPlab doesn´t care if you use USB (with FTD1232) or COM (with MAX232), ok?

5) Must I use MPLab 6 to use this ICD Clone? Can I use MPLab7?

Thank for your help!

 

[philba]

ICD2 revisited??

answers to your questions:
1) yes. that is the boot loader. it loads the program that knows how to talk to the specific target chip. MPLAB does all the work once you have the bootloader burned in.

2) yes - I've used that file in an 877 even though it was built for an 876. there is aslo a bootloader compatible with the A versions floating around. if you need it, I can give you details.

3) stolz's is ok - a lot of people have built one sucessfully. there are many many ways to get the voltage levels for programming. in the case of the sattler design, I really could not see much difference beyond a few resistors to provide more isolation. Not strictly necessary but a good idea

4) yes. MPLABs uses serial I/O. It doesn't care how the actual transport of the bits happens. In fact I've seen a design that uses ethernet as the transport (kind of a stupid idea, though).

5) people have reported success with MPLAB 7 and stolz ICD2s.

Now, if you are going to build a USB ICD2, I recommend you used the FTDI232RL chip. It will take a little thinking to get the schematic right but it's a much simpler chip and you can get rid of a bunch of parts.

 

[qk]

Re: ICD2 revisited??

Thanks Philba!
I will build it and I´ll tell you what happened.

 

[Kripton2035]

ICD2 revisited??

be aware that with serial icd2 or usb-to-serial chip like fdti chips, you get a very slow device for debugging...you must use a usb to parallel chip (cypress) to have a good speed in debug mode...

 

[philba]

Re: ICD2 revisited??

be aware that with serial icd2 or usb-to-serial chip like fdti chips, you get a very slow device for debugging...you must use a usb to parallel chip (cypress) to have a good speed in debug mode...

you sure about that? I have heard of lots of people getting full speed with FTDI. I don't see how a parallel interface will work since the ICD2 firmware only works on 16F87x chips and expects a serial interface. Also, MPLAB only knows how to talk to a com (serial) port. people report it working with baudrates up to 115kb.

could you point us to more data on this point?

 

[manu]

Re: ICD2 revisited??

Here is a great project :
**broken link removed**

 

[philba]

Re: ICD2 revisited??

Here is a great project :
h**p://membres.lycos.fr/silicium31/Electronique/PIC/FreeIcdEnglish.htm

this is an ok design but a) it uses an older USB chip (newer ones like silabs 2103 and FTDI232RL don't need crystals or eeprom and are cheaper) and b) it uses transistors instead of an op amp to drive the ICSP lines. On the positive side, it uses a max 630 to generate Vpp instead of a DC-DC converter.

I think the best design would use the max630, opamp and FTDI232RL for usb.

 

[unhappy]

Re: ICD2 revisited??

I have built the LEPROHON version having parts 7407 max232 16F876A
but am not able to connect to the MPLAB ver6.2 to 7.31
HELP????? any one tuned in

 

[Kripton2035]

Re: ICD2 revisited??

be aware that with serial icd2 or usb-to-serial chip like fdti chips, you get a very slow device for debugging...you must use a usb to parallel chip (cypress) to have a good speed in debug mode...

you sure about that? I have heard of lots of people getting full speed with FTDI. I don't see how a parallel interface will work since the ICD2 firmware only works on 16F87x chips and expects a serial interface. Also, MPLAB only knows how to talk to a com (serial) port. people report it working with baudrates up to 115kb.

could you point us to more data on this point?

yes you can get 115kb with FDTI chips on an icd2 clone
but you get some 1.5Mb ! with cypress usb chips !!!
and so if your icd2 is to program chips, it's ok but to debug program real usb (cypress) chips are far better...
I'm building actually the silicium31's icd2 clone... seems to work but I have problems with mplab and icd2 driver... shoud be ok in some days..

Added after 3 minutes:

Here is a great project :
h**p://membres.lycos.fr/silicium31/Electronique/PIC/FreeIcdEnglish.htm

this is an ok design but a) it uses an older USB chip (newer ones like silabs 2103 and FTDI232RL don't need crystals or eeprom and are cheaper) and b) it uses transistors instead of an op amp to drive the ICSP lines. On the positive side, it uses a max 630 to generate Vpp instead of a DC-DC converter.

I think the best design would use the max630, opamp and FTDI232RL for usb.

ok but is it sooooooo fast compared to any serial or usb-to-serial icd2 clones....
and the chip is not that old..; it is the only real usb chip that is the same as microchip original icd2.
it seems that a new design of icd2 with a 18f4550 is on the way / already shipped by microchip ? anyone has infos on it ?

 

[Kripton2035]

ICD2 revisited??

well after some hours spend in reinstalling usb drivers, mplab, and so on... finally I had some soldier between two pins of the cypress usb chip.. 0.65mm between pins it's not so easy to solder...
and now the icd2 module works sooo fine !

 

[qk]

ICD2 revisited??

Finally, I built my Stolz ICD2 (serial version) but I get a "ICD0286: Unable to communicate with ICD".
All of the I/O data lines are "alive" (they change voltages when I try to connect) .
I´m using a F877 with BL010101.hex.
I´ve already disable FIFO buffers.
I cannot download ICD2 OSystem neither.
Vpp circuit is not already built.
Any idea?