Post by thread:[PIC] Quickest way to get an ICD-2 to program a PI

Mauricio Giovagnini wrote: > Artem Zezyulinskiy escribió: > >> In MPLAB IDE you can _Import_ a hex., cof, or cod. files from File Menu. >> And directly programm the PIC with ICD-2 >> > That's true but the user will also have to select the device > in the Configure menu, under the Select Device option. > > An HEX file has no information about the device to which the > code belongs. > You can load MPLAB with a workspace from the command line (that is, a properly configured Windows shortcut). The workspace contains the information about the device. You can also automate the whole process (loading MPLAB correctly, importing the hex file and starting the programming) with AutoIt3. Gerhard

On Thu, Jul 31, 2008 at 5:45 AM, Gerhard Fiedler <spam_OUTgelistsTakeThisOuTgmail.com> wrote: > Mauricio Giovagnini wrote: >> Artem Zezyulinskiy escribió: >> >>> In MPLAB IDE you can _Import_ a hex., cof, or cod. files from File Menu. >>> And directly programm the PIC with ICD-2 >>> >> That's true but the user will also have to select the device >> in the Configure menu, under the Select Device option. >> >> An HEX file has no information about the device to which the >> code belongs. >> > You can load MPLAB with a workspace from the command line (that is, a > properly configured Windows shortcut). The workspace contains the > information about the device. > > You can also automate the whole process (loading MPLAB correctly, > importing the hex file and starting the programming) with AutoIt3. A simpler way is to dump ICD2 and buy the US$35 PICKit 2. It has a console version pk2cmd which does what you want. As a programmer, PICkit 2 is way better than ICD2. Firstly it can control target Vdd. Secondly, you can get good support from the PICKit 2 developer from Microchip Forum. Thirdly, it supports non-MCU device like serial EEPROMs and other Microchip device. http://www.microchip.com/pickit2 http://forum.microchip.com/tt.aspx?forumid=15 Xiaofan

On Thu, Jul 31, 2008 at 9:10 AM, Xiaofan Chen <.....xiaofancKILLspam@spam@gmail.com> wrote: > A simpler way is to dump ICD2 and buy the US$35 PICKit 2. It > has a console version pk2cmd which does what you want. As > a programmer, PICkit 2 is way better than ICD2. Firstly it can > control target Vdd. Secondly, you can get good support from > the PICKit 2 developer from Microchip Forum. Thirdly, it > supports non-MCU device like serial EEPROMs and other > Microchip device. > > http://www.microchip.com/pickit2 > http://forum.microchip.com/tt.aspx?forumid=15 > Some more user feedback about PICKit 2 better than ICD2, even for debugging. http://forum.microchip.com/tm.aspx?m=355232 http://forum.microchip.com/tm.aspx?m=282776 Xiaofan

This is very interesting. We may take a look at this as & when production throughput gets a bit busier. Many thanks Matt Xiaofan Chen wrote: {Quote hidden}> On Thu, Jul 31, 2008 at 9:10 AM, Xiaofan Chen <xiaofancKILLspamgmail.com> wrote: > >> A simpler way is to dump ICD2 and buy the US$35 PICKit 2. It >> has a console version pk2cmd which does what you want. As >> a programmer, PICkit 2 is way better than ICD2. Firstly it can >> control target Vdd. Secondly, you can get good support from >> the PICKit 2 developer from Microchip Forum. Thirdly, it >> supports non-MCU device like serial EEPROMs and other >> Microchip device. >> >> http://www.microchip.com/pickit2 >> http://forum.microchip.com/tt.aspx?forumid=15 >> >> > > Some more user feedback about PICKit 2 better than ICD2, > even for debugging. > http://forum.microchip.com/tm.aspx?m=355232 > http://forum.microchip.com/tm.aspx?m=282776 > > Xiaofan

Matthew Rhys-Roberts wrote: > Xiaofan Chen wrote: >> A simpler way is to dump ICD2 and buy the US$35 PICKit 2. It >> has a console version pk2cmd which does what you want. As >> a programmer, PICkit 2 is way better than ICD2. Firstly it can >> control target Vdd. Secondly, you can get good support from >> the PICKit 2 developer from Microchip Forum. Thirdly, it >> supports non-MCU device like serial EEPROMs and other >> Microchip device. > > This is very interesting. We may take a look at this as & when > production throughput gets a bit busier. I know this sounds self serving, but the PicKit2 is really not appropriate for production use. It's a good quick and dirty programmer and serves the low cost niche very well, but for production use you need something that fully adheres to the specs. I know Xiaofan thinks the PicKit2 is the greatest invention since the wheel, is the solution to all electronic problems now and in the future, and cures cancer, but recommending it for production use is just irresponsible. For production get a real production programmer. My USBProg (http://www.embedinc.com/products) is one, but there are others too. And yes, it is run from command line programs that can be scripted or launched from within other programs, such as LabView. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000.

Well, every design has its own pros and cons. The most important thing is: the design must work reliably. It seems the PICkit 2 design does very well on the job it is designed and it is robust. Comparing with the higher cost ICD2, PICkit2 is much more robust on hardware and software architecture. Not even consider many advanced features, such as the "Programmer-to-go" feature, end-user can even set up the production line for ICSP programming without a PC. And all the operators have to do is pushing a button. It couldn't be easier. This is not a "one-kill-all" invention, but it does have many working features and capabilities. So, if you own one, enjoy it. Funny N. Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com {Original Message removed}

I imagine Olin was targeting production programmer features recommended by Microchip, such as flash verification at voltage margins. This takes extra production time, though, and I'm certain many products are made without following proper 'production programmer' guidelines. I'd expect that unless the failure rate was high then the tradeoff would be acceptable, and given Microchip's quality, I expect the failure rate is very low indeed. Still, there are specific recomendations that Microchip makes concerning programming for production vs programming for develeopment and prototyping, and one should always mention that when one discusses production programming. Suggesting a programmer might be suitable for production when it doesn't follow manufacturer's guidelines for a production programmer is shortsighted at best. -Adam On 8/4/08, Funny NYPD <.....funnynypdKILLspam.....yahoo.com> wrote: {Quote hidden}> Well, every design has its own pros and cons. The most important thing is: the design must work reliably. > It seems the PICkit 2 design does very well on the job it is designed and it is robust. > > Comparing with the higher cost ICD2, PICkit2 is much more robust on hardware and software architecture. > > Not even consider many advanced features, such as the "Programmer-to-go" feature, end-user can even set up the production line for ICSP programming without a PC. And all the operators have to do is pushing a button. It couldn't be easier. > > This is not a "one-kill-all" invention, but it does have many working features and capabilities. So, if you own one, enjoy it. > > Funny N. > Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com > > > > {Original Message removed}

M. Adam Davis wrote: > I imagine Olin was targeting production programmer features > recommended by Microchip, such as flash verification at voltage > margins. That's one issue, but the PicKit2 also doesn't adhere to the PIC programming specs over the full range of legal USB voltages. Again, the PicKit2 is great as a low cost hobbyist level programmer, but you really don't want to use it where time and failures cost real money, loss of face, pissed of customers, etc. Note that Microchip itself calls the PicKit2 a "development" programmer. Their only current production programmer is the PM3. For the hobbyist that just wants to program and sometimes debug his hand built one-off PIC projects, the PicKit2 is a great value. Cutting corners for a few 10s of $$ on a programmer doesn't make sense in a professional situation, and certainly doesn't make sense in a production situation. > This takes extra production time, though, and I'm certain many > products are made without following proper 'production programmer' > guidelines. I'd expect that unless the failure rate was high then the > tradeoff would be acceptable, and given Microchip's quality, I expect > the failure rate is very low indeed. The failure rate is very low as far as I can tell, but the cost of field failures is very high, and not just in $$. I don't know of any production environment where it's not worth $45 to make sure the programming specs are followed and the units are tested with a little extra margin. > Suggesting a programmer might be suitable for production when it > doesn't follow manufacturer's guidelines for a production programmer > is shortsighted at best. Exactly. And if failure might result in legal liability problems, think of what the plaintiff's attorney would make out of you saving well under $100 and not getting a production programmer for a production job, even if that had absolutely nothing to do with the failure. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000.

Per Microchip's definition of 'production programmer' guidelines, it is still doable with PICkit 2 hardware. It is not the design cannot have the function, it is the development team don't want to add this function to challenge its own more profitable toolset. The good news is: the PICKit2 itself is open to the public from hardware schematic to software source code. And there are people still working hard to get the linux and Mac versions released. When there is such a issue or high demand for this feature, it will be developed. This can be done by Microchip or the third party. The hardware has been well designed, it is just a matter of time to pop more software/features to take the advantage of this reliable platform. For the moment, I don't see any issue if people use it in ICSP programming. If the programmed chip malfunction, it will be caught right away. The open CMD functions might be a good start of adjusting the output voltage for test purpose. Funny N. Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com {Original Message removed}

On Mon, Aug 4, 2008 at 3:42 PM, Funny NYPD <EraseMEfunnynypdspam_OUTTakeThisOuTyahoo.com> wrote: > Per Microchip's definition of 'production programmer' guidelines, it is still doable with PICkit 2 hardware. > It is not the design cannot have the function, it is the development team don't want to add this function to challenge its own more profitable toolset. > The problem is, Microchip's definition of "production programmer" is not at all clear. Regards, Mark markrages@gmail -- Mark Rages, Engineer Midwest Telecine LLC markragesspam_OUTmidwesttelecine.com

Funny NYPD wrote: > Per Microchip's definition of 'production programmer' guidelines, it > is still doable with PICkit 2 hardware. > It is not the design cannot have the function, it is the development > team don't want to add this function to challenge its own more > profitable toolset. But that still leaves you without the capbility, at least for now. And this doesn't address the fact that the USB voltage is used as the 5V reference voltage, which means that it can't even guarantee the proper bulk erase voltage over the full range of legal USB voltages. Again, this is a quick and dirty hobbyist programmer. It's design is appropriate for that, and I think succeeds well towards that goal. I recommend it to hobbyists over my more expensive but robust programmers, but professional and production applications have different tradeoffs. > The good news is: the PICKit2 itself is open to the public from > hardware schematic to software source code. So are other designs, like my USBProg (http://www.embedinc.com/products). The hardware has fully controllable Vdd from 0 to 6V and Vpp from 0 to over 13V. And these voltages are relative to a fixed on-board 5V reference. > The hardware has been well designed, For low cost hobby use, yes. I disagree that is is "well designed" for other purposes, like professional use or production. Microchip is the first to admit it is not intended for production. > it is just a matter of time to > pop more software/features to take the advantage of this reliable > platform. While it certainly works most of the time, it was designed for low cost, not high reliability. Calling it a "reliable platform" is a bit misleading. > For the moment, I don't see any issue if people use it in ICSP > programming. If the programmed chip malfunction, it will be caught > right away. In a one-off hobby situation, sure. In a production situation you can't afford that. You don't know, for example, what the real flash data lifetime is when it wasn't programmed or erased within the voltages according to the programming spec. > The open CMD functions might be a good start of adjusting the output > voltage for test purpose. If you're in a professional situation, how much time do you want to spend on this to save $50 or so on a programmer that was designed for robustness? ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000.

Mark Rages wrote: > The problem is, Microchip's definition of "production programmer" is > not at all clear. It used to be that production programmers verified the programmed data at the Vdd limits of the chip. Verifying with some margin is good practise in any case, but there is more to the full spirit of a "production" programmer than this simple requirement. In production you have to expect parts accross the full spec sheet range of the various parameters. You need something that is careful about adhering to the specs, and something that verifies the result and itself to a good extent. Field failures are expensive, but the cost of a PIC programmer is very little per unit accross all the production lots. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000.

If you don't mind spend a little more money (Actually, this is not true.), our CB0703/BB0703 design includes a dedicated power regulator (and connector etc.) which can be easily used for "programmer to go" feature and it is also capable of compensating the "USB voltage drift" if it ever been an issue. (So far, I haven't heard anybody crying for the "+5V voltage drift" issue on any PICkit 2, at least not from Microchip Forum.) (Actually, our BB0703 has the same price as the Microchip PICkit 2 even with more hardware functions and features. And thanks to the SMD technology, the size our BB0703/CB0703 design is still slightly smaller than Microchips design even with more hardware on board) Again, thanks to all the open source design of PICkit 2 on both hardware and software, so people can fulfill their own need with some performance enhancement. Many friends on this PIClist who either got the Microchip PICkit2 or Au Group Electronics' BB0703, or the CB0703 DIY kits can share your experience and opinion. (I also got friends who bought five to six sets of CB0703 DIY kits from different countries, you might tell more on what you have found when you build your own system from the scratch.) More detail on BB0703 and CB0703 are available at these web link: www.auelectronics.com/System-PICkit2.htm http://www.auelectronics.com/Hardware-CB0703.htm On the bottom of the BB0703 web, there are two youtube links illustrating the "Programmer-to-Go" feature. Funny N. Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com {Original Message removed}

>And this doesn't address the fact that the USB voltage is used as the 5V reference >voltage, which means that it can't even guarantee the proper bulk erase >voltage over the full range of legal USB voltages. I think, PICKit 2 also addressed this issue on its software design with voltage calibration and low voltage programming capability. The following link illustrated a step by step tutorial on how to calibrate your PICkit 2 to prevent the possible issue from happening: http://www.auelectronics.com/Q4.htm In the mean time, PICKit 2 design support both bulk erase and low-voltage erase/programming. So, it has been covered and it is covered in a "smart" way. Funny N. Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com {Original Message removed}

>Field failures are expensive, but the cost of a PIC programmer is very little per unit accross >all the production lots. I agree that Microchip is not a good silicon designer or a software developer. Nor Microchip's demo kits are perfect designed. However if its silicon are so bad, I bet it had been close the door long time ago. Many of the Microchip chips has a silicon errata more than 5 pages. Rule of thumb is: the newer the chip, the more silicon erratas it has. However, once you figured out know how to make it work, It works like a charm. It doesn't matter you use a ICD2 or a PICKit2, or a PM2, or a PM3, or a third part toolset to program it. If it works, then it works. This is one of the reason why we are still in love with Microchip. Funny N. Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com

>While it certainly works most of the time, it was designed for low cost, not >high reliability. Calling it a "reliable platform" is a bit misleading. I remember someone on the Microchip forum tells everyone else, "the ICD2 is very reliable". I think that's a real misleading. (Just kidding.) I remember xiaofan says: "ICD2 is everything but reliable". This is the same experience I got. After I switched from ICD2 to PICKit2 since sept 2007. I haven't broke a PICKit 2 yet. And I don't see how PICKit 2 is unreliable from design to practice. (I used to broke an ICD2 in about 2 month and then get a replaced "one" from Microchip by using it as my daily programming/debugging toolset. Now I use PICKit2, this reliability nightmare went away.) Now I think I have at least saved Microchip couple hundred $ for not sending broken ICD2(s) back anymore. Funny N. Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com {Original Message removed}

Another good point for PICKit 2 is the PICKit 2 design team. They are quite knowledgeable and honest. Besides the good service from local FAEs in Boston area, the PICKit 2 design team is another good example of excellent service from Microchip. On customer support point of view, Microchip is way better than Freescale or Atmel. PICkit 2 is open on its schematic, however Microchip still keep its own PCB layout and component selection private, it seems the team had pushed some component to its limit too much, like you have to "pump the Vpp" at certain situation. The design team found the issue and published the info and solution. So all the Microchip PICkit 2 user can realize the situation and either upgrade the hardware or prevent it from happening. Our BB0703/CB0703 PCB and component design team seems have more experience and a bit more conservative on the components we choose, so we don't have any of the "pumping Vpp" issue from component design level. Also our PCB board are smaller than Microchip's design though we have more components and most of our components are larger in size to fulfill ease of DIY purpose. Overall, the PICKit 2 is a good open-source design as the Microchip C30 compiler. Both of them will have a bright successful future. Funny N. Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com

On Mon, Aug 4, 2008 at 10:57 PM, Olin Lathrop <@spam@olin_piclistKILLspamembedinc.com> wrote: > Matthew Rhys-Roberts wrote: >> Xiaofan Chen wrote: >>> A simpler way is to dump ICD2 and buy the US$35 PICKit 2. It >>> has a console version pk2cmd which does what you want. As >>> a programmer, PICkit 2 is way better than ICD2. Firstly it can >>> control target Vdd. Secondly, you can get good support from >>> the PICKit 2 developer from Microchip Forum. Thirdly, it >>> supports non-MCU device like serial EEPROMs and other >>> Microchip device. >> >> This is very interesting. We may take a look at this as & when >> production throughput gets a bit busier. Hmm, I do not know this is for production. For production, I will not recommend PICkit 2 or ICD2. At least one should go for Microchip PM3. Take note I only subscribe to [PIC] and [EE] now. I do not even know that Matthew is the OP. The first message under [PIC] is from Gerhard (replying to Mauricio Giovagnini). > I know this sounds self serving, but the PicKit2 is really not appropriate > for production use. As I said, if I knew this is for production, I would not recommend PICKit 2. My take on USB PIC programmers. http://mcuee.blogspot.com/2007/12/usb-programmers-for-pics.html Even ProProg/PM3 are only good for low volume mass production. For higher quantity programming (higher throughput), people should consider off-line Gang programmer or vendor provided pre-programming service. Xiaofan

Funny NYPD wrote: > I agree that Microchip is not a good silicon designer Whoa, I never said that, and I don't agree. I think PICs on the whole are well designed, well documented, and superbly supported. > It doesn't matter you use a ICD2 or a PICKit2, or a PM2, or a PM3, or > a third part toolset to program it. For hobby use, yes. I totally disagree if you mean this for professional and especially for production use. > If it works, then it works. Absolutely not! Anyone who has been thru volume manufacturing of a real product will tell you how untrue this is. Working once under controlled conditions is very different from working reliably in the field for a long time. Stuff happens. No matter what you do, some field failures will happen. The more you guarantee margin and test at the margin limits, the more you weed out the weak units and decrease field failures. Field failures are way more expensive than having to rework or toss the occasional outlier unit detected during manufacturing test. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000.

Funny NYPD wrote: > Another good point for PICKit 2 is the PICKit 2 design team. Are you going to flood the list with another dozen responses to one message? Saying it more times doesn't make the point any better, only more annoying. As for the design team, I don't know them personally, but "good" compared to whom? For this to be a meaningful statement you'd have to compare them to other programmer design teams, which will be nearly impossible to do. I've designed four different commercial PIC programmers by now. I doubt anyone at Microchip has that experience. Are you trying to beat out Xiaofan for the title of lead PicKit2 apologist? I agree the thing serves a particular niche and does that well. Can we leave it at that? Of course, like any product, it is not the best solution for other niches. Nor does it cure cancer. ******************************************************************** Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products (978) 742-9014. Gold level PIC consultants since 2000.

Greeting, Olin, It seems you have never suffered with other Microchip tools or silicon designs. I and my team weren't that lucky. I am sure you will be more disappointed if you have been caught by a Microchip SPI errata or an ECAN errata when you are designing a product based on PICs. And when those design fault throw a user in the middle of nowhere, I am sure you will understand why a good design from Microchip worth appreciation. The Microchip dsPIC30F document cannot be claimed "well documented". This I am pretty sure. I won't feel strange when people complaining about Microchip tools or silicon erratas or documents. That's pretty much we have to deal with everyday. This doesn't means I will abandon Microchip tomorrow, it is the way it be, and if you are using PICs, that's what you got. Sad, isn't it? Not really, once you are used to it. Cheer up, and have a nice day. Thanks, Funny N. Au Group Electronics, New Bedford, MA, http://www.AuElectronics.com {Original Message removed}

Olin Lathrop wrote: > > I know this sounds self serving, but the PicKit2 is really not appropriate > for production use. It's a good quick and dirty programmer and serves the > low cost niche very well, but for production use you need something that > fully adheres to the specs. I know Xiaofan thinks the PicKit2 is the > greatest invention since the wheel, is the solution to all electronic > problems now and in the future, and cures cancer, but recommending it for > production use is just irresponsible. LOL!! :) though I share his enthusiasm about PicKit2 -- Ciao, Dario -- ADPM Synthesis sas -- http://www.adpm.tk