EmbeddedRelated.com
Forums
Memfault Beyond the Launch

MSP430 and Atmel AVR in one JTAG chain?

Started by Grant Edwards October 24, 2007
Does anybody know if either the MSP430 or Atmel ATmega AVR
parts (or associated JTAG interface widgets/SW) have problems
when there are other devices in the JTAG chain?

In the past, I've run into buggy JTAG stuff that couldn't
tolerate having multiple devices in the JTAG chain.  I'm
working on a design with both an AVR and an MSP430, and I'd
like to have a single JTAG header/chain.  Before trying it out,
I thought I'd check to see if anybody already knew it won't
work.

-- 
Grant Edwards                   grante             Yow! I had a lease on an
                                  at               OEDIPUS COMPLEX back in
                               visi.com            '81 ...
Grant Edwards <grante@visi.com> wrote:
> Does anybody know if either the MSP430 or Atmel ATmega AVR > parts (or associated JTAG interface widgets/SW) have problems > when there are other devices in the JTAG chain?
> In the past, I've run into buggy JTAG stuff that couldn't > tolerate having multiple devices in the JTAG chain. I'm > working on a design with both an AVR and an MSP430, and I'd > like to have a single JTAG header/chain. Before trying it out, > I thought I'd check to see if anybody already knew it won't > work.
Read the AVR Datasheet. Older Chip revision had problems and mask out other devices. I am using a chain with a recent AT90CAN128 and CPLD/PROM/FPGA following with success -- Uwe Bonnes bon@elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
On 2007-10-24, Uwe Bonnes <bon@hertz.ikp.physik.tu-darmstadt.de> wrote:
> Grant Edwards <grante@visi.com> wrote: >> Does anybody know if either the MSP430 or Atmel ATmega AVR >> parts (or associated JTAG interface widgets/SW) have problems >> when there are other devices in the JTAG chain? > >> In the past, I've run into buggy JTAG stuff that couldn't >> tolerate having multiple devices in the JTAG chain. I'm >> working on a design with both an AVR and an MSP430, and I'd >> like to have a single JTAG header/chain. Before trying it out, >> I thought I'd check to see if anybody already knew it won't >> work.
I found an MSP430 app note that states the MSP430 violates the IEEE JTAG standard and as a result has to be the first device in the chain.
> Read the AVR Datasheet.
I did (for the 1281V). It didn't say one way or the other. I don't interpret lack of a statement of a feature's absence to be an assurance that a feature is present.
> Older Chip revision had problems and mask out other devices.
So those would be OK if they were the last device in the chain? That along with the MSP430 non-compliance would be a happy coincidence.
> I am using a chain with a recent AT90CAN128 and CPLD/PROM/FPGA > following with success
I found the following note in the JTAG ICE manual: Note: The JTAG ICE does not support several devices placed into a JTAG Chain. For example, the target AVR must be the only device connected to the JTAG ICE. Cripes. JTAG has been around for how many decades and people still can't seem to make tools/parts that implement it correctly? [I also found AVR JTAG interfaces from other vendors that specifically state that multiple device chains are supported.] -- Grant Edwards grante Yow! ! Up ahead! It's a at DONUT HUT!! visi.com
> Note: The JTAG ICE does not support several devices placed > into a JTAG Chain. For example, the target AVR must be > the only device connected to the JTAG ICE. > > Cripes. JTAG has been around for how many decades and people > still can't seem to make tools/parts that implement it > correctly?
I expect there is more of that to come - not just in JTAG. Less and less people even try to look what is below/aside the popdown menus on offer. What you quote relates just to some poorly written tool, though. I cannot think of a part which does not correctly do JTAG bypass (being simply a D-flipflop in the chain); so your plan with the 430 first sounds good (the 430s problems likely appear when it is doing non-bypass things, but that's just a guess). Recently I did choose to connect the 430 JTAG separately on a design with more JTAG-ged parts, though. But I had more excuses to do so, so it was easier to take (free pins on a connector I had anyway, separate chain powerdown capability etc.). Dimiter ------------------------------------------------------ Dimiter Popoff Transgalactic Instruments http://www.tgi-sci.com ------------------------------------------------------ On Oct 24, 7:15 pm, Grant Edwards <gra...@visi.com> wrote:
> On 2007-10-24, Uwe Bonnes <b...@hertz.ikp.physik.tu-darmstadt.de> wrote: > > > Grant Edwards <gra...@visi.com> wrote: > >> Does anybody know if either the MSP430 or Atmel ATmega AVR > >> parts (or associated JTAG interface widgets/SW) have problems > >> when there are other devices in the JTAG chain? > > >> In the past, I've run into buggy JTAG stuff that couldn't > >> tolerate having multiple devices in the JTAG chain. I'm > >> working on a design with both an AVR and an MSP430, and I'd > >> like to have a single JTAG header/chain. Before trying it out, > >> I thought I'd check to see if anybody already knew it won't > >> work. > > I found an MSP430 app note that states the MSP430 violates the > IEEE JTAG standard and as a result has to be the first device > in the chain. > > > Read the AVR Datasheet. > > I did (for the 1281V). It didn't say one way or the other. I > don't interpret lack of a statement of a feature's absence to > be an assurance that a feature is present. > > > Older Chip revision had problems and mask out other devices. > > So those would be OK if they were the last device in the chain? > That along with the MSP430 non-compliance would be a happy > coincidence. > > > I am using a chain with a recent AT90CAN128 and CPLD/PROM/FPGA > > following with success > > I found the following note in the JTAG ICE manual: > > Note: The JTAG ICE does not support several devices placed > into a JTAG Chain. For example, the target AVR must be > the only device connected to the JTAG ICE. > > Cripes. JTAG has been around for how many decades and people > still can't seem to make tools/parts that implement it > correctly? > > [I also found AVR JTAG interfaces from other vendors that > specifically state that multiple device chains are supported.] > > -- > Grant Edwards grante Yow! ! Up ahead! It's a > at DONUT HUT!! > visi.com
On 2007-10-24, Didi <dp@tgi-sci.com> wrote:

>> Cripes. JTAG has been around for how many decades and people >> still can't seem to make tools/parts that implement it >> correctly? > > I expect there is more of that to come - not just in JTAG. > Less and less people even try to look what is below/aside the > popdown menus on offer. > > What you quote relates just to some poorly written tool, > though. I cannot think of a part which does not correctly do > JTAG bypass (being simply a D-flipflop in the chain); so your > plan with the 430 first sounds good (the 430s problems likely > appear when it is doing non-bypass things, but that's just a > guess).
I'll have to read through the app note again, but I'm going to try to wire up a couple eval boards in a JTAG chain.
> Recently I did choose to connect the 430 JTAG separately on a > design with more JTAG-ged parts, though. But I had more > excuses to do so, so it was easier to take (free pins on a > connector I had anyway, separate chain powerdown capability > etc.).
We're going to have no extra space on the board in question, so being able to use a single JTAG header would be a good thing. I suppose I could come up with a way to share most of the connector pins whithout actually chaining the TAP units. -- Grant Edwards grante Yow! I just forgot my whole at philosophy of life!!! visi.com
Didi wrote:
> >> Note: The JTAG ICE does not support several devices placed >> into a JTAG Chain. For example, the target AVR must be >> the only device connected to the JTAG ICE. >> >> Cripes. JTAG has been around for how many decades and people >> still can't seem to make tools/parts that implement it >> correctly? > > I expect there is more of that to come - not just in JTAG. > Less and less people even try to look what is below/aside > the popdown menus on offer. > > What you quote relates just to some poorly written tool, though. > I cannot think of a part which does not correctly do JTAG > bypass (being simply a D-flipflop in the chain); so your > plan with the 430 first sounds good (the 430s problems > likely appear when it is doing non-bypass things, but that's > just a guess). > Recently I did choose to connect the 430 JTAG separately on > a design with more JTAG-ged parts, though. But I had more > excuses to do so, so it was easier to take (free pins on a > connector I had anyway, separate chain powerdown capability > etc.).
While your posts are generally quite readable, they are spoiled by the fact that the complete original message is appended as a quote. Please simply delete that appendage. -- Chuck F (cbfalconer at maineline dot net) Available for consulting/temporary embedded and systems. <http://cbfalconer.home.att.net> -- Posted via a free Usenet account from http://www.teranews.com
>>> Does anybody know if either the MSP430 or Atmel ATmega AVR >>> parts (or associated JTAG interface widgets/SW) have problems >>> when there are other devices in the JTAG chain? >> >>> In the past, I've run into buggy JTAG stuff that couldn't >>> tolerate having multiple devices in the JTAG chain. I'm >>> working on a design with both an AVR and an MSP430, and I'd >>> like to have a single JTAG header/chain. Before trying it out, >>> I thought I'd check to see if anybody already knew it won't >>> work. > > I found an MSP430 app note that states the MSP430 violates the > IEEE JTAG standard and as a result has to be the first device > in the chain. > >> Read the AVR Datasheet. > > I did (for the 1281V). It didn't say one way or the other. I > don't interpret lack of a statement of a feature's absence to > be an assurance that a feature is present. > >> Older Chip revision had problems and mask out other devices. > > So those would be OK if they were the last device in the chain? > That along with the MSP430 non-compliance would be a happy > coincidence. >
No, the early AVRs with problems also needed to be the first in a chain. I think the mega1281 should be fine. Better talk to your local Atmel FAE and ask about a specific device, or send an email to avr at atmel dot com.
> Note: The JTAG ICE does not support several devices placed > into a JTAG Chain. For example, the target AVR must be > the only device connected to the JTAG ICE. >
The JTAG ICE Mk II certainly supports several devices in a chain. There are AVR Studio configurations where you specify where in the chain it is. -- Best Regards, Ulf Samuelsson This is intended to be my personal opinion which may, or may not be shared by my employer Atmel Nordic AB
> Chuck F (cbfalconer at maineline dot net)
I followed your recommendation, while writing this reply but I think it is going to confuse readers of this email. -- Best Regards, Ulf Samuelsson This is intended to be my personal opinion which may, or may not be shared by my employer Atmel Nordic AB
On 2007-10-25, Ulf Samuelsson <ulf@a-t-m-e-l.com> wrote:

>> So those would be OK if they were the last device in the >> chain? That along with the MSP430 non-compliance would be a >> happy coincidence. > > No, the early AVRs with problems also needed to be the first in a chain. > I think the mega1281 should be fine.
It's good to know that some people don't design their parts under the assumption that nothing else exists in the world. :)
> Better talk to your local Atmel FAE and ask about a specific > device, or send an email to avr at atmel dot com.
I'll do that. I'm also going to try to wire up a couple eval boards and make sure it works.
>> Note: The JTAG ICE does not support several devices placed >> into a JTAG Chain. For example, the target AVR must be >> the only device connected to the JTAG ICE. >> > > The JTAG ICE Mk II certainly supports several devices in a chain.
Yup, I just found where the datasheet for the Mk II explicitly says that.
> There are AVR Studio configurations where you specify where in > the chain it is.
I wasn't planning on using AVR Studio, but I can if I have to... -- Grant Edwards grante Yow! I just went below the at poverty line! visi.com
On Wed, 24 Oct 2007 22:50:08 -0000, Grant Edwards <grante@visi.com>
wrote:

>On 2007-10-24, Didi <dp@tgi-sci.com> wrote: > >>> Cripes. JTAG has been around for how many decades and people >>> still can't seem to make tools/parts that implement it >>> correctly? >> >> I expect there is more of that to come - not just in JTAG. >> Less and less people even try to look what is below/aside the >> popdown menus on offer. >> >> What you quote relates just to some poorly written tool, >> though. I cannot think of a part which does not correctly do >> JTAG bypass (being simply a D-flipflop in the chain); so your >> plan with the 430 first sounds good (the 430s problems likely >> appear when it is doing non-bypass things, but that's just a >> guess). > >I'll have to read through the app note again, but I'm going to >try to wire up a couple eval boards in a JTAG chain. > >> Recently I did choose to connect the 430 JTAG separately on a >> design with more JTAG-ged parts, though. But I had more >> excuses to do so, so it was easier to take (free pins on a >> connector I had anyway, separate chain powerdown capability >> etc.). > >We're going to have no extra space on the board in question, so >being able to use a single JTAG header would be a good thing. I >suppose I could come up with a way to share most of the >connector pins whithout actually chaining the TAP units.
Maybe you can use the TI 'LVT8986 device. One can connect 3 secondary JTAG busses to the master JTAG bus. Each of the secondary JTAG busses can be isolated. Regards Anton Erasmus

Memfault Beyond the Launch