Searching \ for '[EE] Logic Analyser Recomendations' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=logic+analyser+recomendations
Search entire site for: 'Logic Analyser Recomendations'.

Exact match. Not showing close matches.
PICList Thread
'[EE] Logic Analyser Recomendations'
2008\04\02@012002 by Xiaofan Chen

face picon face
I am tasked to source for  a Logic Analyser which shall meet the
future needs to debug more advanced MCU/FPGA related things.
Currently the other group has an older Modular Tek 700 series
(running Windows 2000) with quite some accessories. The main
problem is that module only has 256k memory depth and it is also
a bit slow now.

I was looking at Agilent 16900 series (16901A main frame +
16950B modules + accessories) but the total budget for a
68Ch/4M memory system is much higher than our budget
of US$30k.

The Tek guy recommends me to look at their TLA5000
series (not modularized).

What is your recommendation? Our group is now more on
the analog side and use 8051 and ARM7 MCUs for processing.
The other  group is more on the digital side with 68k/ColdFire
as the main processor. Both group will need to move to faster
ARM9/ColdFire class MCU which is capable of running bigger
RTOS and FPGA (to replace ASIC) in future work.
Thanks.

Xiaofan

2008\04\02@115735 by Vasile Surducan

face picon face
On 4/1/08, Xiaofan Chen <spam_OUTxiaofancTakeThisOuTspamgmail.com> wrote:
> I am tasked to source for  a Logic Analyser

Are they paying you for this job ?
$30k is plenty enough for any serious 32 bit logic analyser. For an
ARM9, 1Gsps is more than enough. For the faster FPGA on the market
must be a better one.
There is one logic analyser with 18bit&1Gsps via USB2.0 at about $500.
A good engineer could solve a lot of problems with that cheap one.

which shall meet the
{Quote hidden}

> -

2008\04\02@122040 by Lloyd Sargent

flavicon
face
On Apr 2, 2008, at 10:57 AM, Vasile Surducan wrote:
> On 4/1/08, Xiaofan Chen <.....xiaofancKILLspamspam@spam@gmail.com> wrote:
>> I am tasked to source for  a Logic Analyser
>
> Are they paying you for this job ?
> $30k is plenty enough for any serious 32 bit logic analyser. For an
> ARM9, 1Gsps is more than enough. For the faster FPGA on the market
> must be a better one.
> There is one logic analyser with 18bit&1Gsps via USB2.0 at about $500.
> A good engineer could solve a lot of problems with that cheap one.

Who makes this? For $500 (not $5000?) I might look into buying one for  
myself!

I've worked with everything from Tek to PC analyzers (LA5000).  
Personally, I think if you are doing hardcore debugging, you are  
better off with Tek - nothing worse than a multiple condition problem  
that you analyzer can't do. Frankly, this is where Tek logic analyzer  
shine.

The LA5000 (which I have on my desk) is not too shabby (although the  
interface is a little rough).   http://www.linkinstruments.com
IIRC it is about 500 Msps... Cheap at $3500. We did some nice analysis  
on an IDE i/f we had problems with (it was talking to an FPGA).

But the GUI interface... leaves a LOT to be desired... And you don't  
get the nice multiple condition like you do with Tek - you get  
SEQUENCES but that is not exactly the same thing...

Cheers,

Lloyd

2008\04\02@123834 by Funny NYPD

picon face
agree. Do you have a link to the cheap $500 product?

Funny N.
Au Group Electronics, New Bedford, MA, http://www.AuElectronics.selfip.com

{Original Message removed}

2008\04\02@144103 by Vasile Surducan

face picon face
On 4/2/08, Lloyd Sargent <lloydspamKILLspamcannasoftware.com> wrote:
> On Apr 2, 2008, at 10:57 AM, Vasile Surducan wrote:
> > On 4/1/08, Xiaofan Chen <.....xiaofancKILLspamspam.....gmail.com> wrote:
> >> I am tasked to source for  a Logic Analyser
> >
> > Are they paying you for this job ?
> > $30k is plenty enough for any serious 32 bit logic analyser. For an
> > ARM9, 1Gsps is more than enough. For the faster FPGA on the market
> > must be a better one.
> > There is one logic analyser with 18bit&1Gsps via USB2.0 at about $500.
> > A good engineer could solve a lot of problems with that cheap one.
>
> Who makes this? For $500 (not $5000?) I might look into buying one for
> myself!

I forgot the link, sorry. AFIR is that one:
http://www.rockylogic.com/products/ant18e.html

It's still a toy, but looks a good one. I don't need it because I
already have a few analyzers around... to many for me and my spare
time.

Never mix your job with your hobby, sooner or later you'll regret...
:(





{Quote hidden}

> -

2008\04\02@152821 by Josh Koffman

face picon face
On Wed, Apr 2, 2008 at 11:40 AM, Vasile Surducan <EraseMEpiclist9spam_OUTspamTakeThisOuTgmail.com> wrote:
>  I forgot the link, sorry. AFIR is that one:
>  http://www.rockylogic.com/products/ant18e.html

IIRC when I last looked at the Ant it didn't do any decoding - ie you
couldn't tell it these pins were a parallel or serial port and have it
decode the output. That's why I went with this guy:
http://www.pctestinstruments.com/ I like it a fair bit but I'm not
doing anything like what Xiaofan does!

Josh
--
A common mistake that people make when trying to design something
completely foolproof is to underestimate the ingenuity of complete
fools.
-Douglas Adams

2008\04\02@172250 by William \Chops\ Westfield

face picon face

>> Currently the other group has an older Modular Tek 700 series
>> (running Windows 2000) with quite some accessories. The main
>> problem is that module only has 256k memory depth and it is also
>> a bit slow now.

Without knowing much about logic analyzers in general, it seems to me  
that if your company already has some experience with Tektronix  
analyzers, and the only thing they object to is memory depth, then  
your best bet is another Tek (with more memory.)  One shouldn't  
discount the "hidden" costs of having to train up N engineers on a  
different platform...

BillW

2008\04\02@193758 by Xiaofan Chen

face picon face
On Wed, Apr 2, 2008 at 11:57 PM, Vasile Surducan <piclist9spamspam_OUTgmail.com> wrote:
> On 4/1/08, Xiaofan Chen <@spam@xiaofancKILLspamspamgmail.com> wrote:
> > I am tasked to source for  a Logic Analyser
>
> Are they paying you for this job ?

This is part of my job: to help set up the lab.

> $30k is plenty enough for any serious 32 bit logic analyser. For an
> ARM9, 1Gsps is more than enough. For the faster FPGA on the market
> must be a better one.

It is not so easy to be future proof. The FPGA side is a concern.

> There is one logic analyser with 18bit&1Gsps via USB2.0 at about $500.
> A good engineer could solve a lot of problems with that cheap one.

Ture but the cost of tweaking with a cheap instrument may not be justified
in the long run. Within the budget I will tend to buy the best performed
instrument. It is relative lower cost compared to the total project cost
due to long project cycle here.

Regards,
Xiaofan

2008\04\02@194027 by Xiaofan Chen

face picon face
On Thu, Apr 3, 2008 at 12:16 AM, Lloyd Sargent <KILLspamlloydKILLspamspamcannasoftware.com> wrote:
> I've worked with everything from Tek to PC analyzers (LA5000).
> Personally, I think if you are doing hardcore debugging, you are
> better off with Tek - nothing worse than a multiple condition problem
> that you analyzer can't do. Frankly, this is where Tek logic analyzer
> shine.

Just wondering which Tek are you using now?

Xiaofan

2008\04\02@194941 by Xiaofan Chen

face picon face
On Thu, Apr 3, 2008 at 5:22 AM, William Chops Westfield <RemoveMEwestfwTakeThisOuTspammac.com> wrote:
>
> Without knowing much about logic analyzers in general, it seems to me
> that if your company already has some experience with Tektronix
> analyzers, and the only thing they object to is memory depth, then
> your best bet is another Tek (with more memory.)  One shouldn't
> discount the "hidden" costs of having to train up N engineers on a
> different platform...
>

Our group has two 1GHz Agilent 8000 series oscilloscopes which can
interface with Agilent LA easily. So that is an advantage for
Agilent. We also have 4 low-end 200MHz 4-ch Tek 3000 series
Scopes that not many people like so that is an disadvantage for
Tek. Tek lost the deal for the 1GHz scope since its price was
about 30% more than the Agilent last time.

But now it comes to price/performance ratio. I got a good
quotation for Tek TLA5000. So I am waiting for a counter offer
from Agilent.


Xiaofan

2008\04\03@005320 by Vasile Surducan

face picon face
On 4/2/08, Xiaofan Chen <spamBeGonexiaofancspamBeGonespamgmail.com> wrote:
> On Thu, Apr 3, 2008 at 12:16 AM, Lloyd Sargent <TakeThisOuTlloydEraseMEspamspam_OUTcannasoftware.com> wrote:
> > I've worked with everything from Tek to PC analyzers (LA5000).
> > Personally, I think if you are doing hardcore debugging, you are
> > better off with Tek - nothing worse than a multiple condition problem
> > that you analyzer can't do. Frankly, this is where Tek logic analyzer
> > shine.
>
> Just wondering which Tek are you using now?

In our lab a MSO4104, but trully I'm not very happy with it (on the
analogic side).
www.conres.com/test-equipment/mso4000.html

2008\04\03@005746 by Vasile Surducan

face picon face
On 4/2/08, Josh Koffman <RemoveMEjoshybearspamTakeThisOuTgmail.com> wrote:
> On Wed, Apr 2, 2008 at 11:40 AM, Vasile Surducan <piclist9EraseMEspam.....gmail.com> wrote:
> >  I forgot the link, sorry. AFIR is that one:
> >  http://www.rockylogic.com/products/ant18e.html
>
> IIRC when I last looked at the Ant it didn't do any decoding - ie you
> couldn't tell it these pins were a parallel or serial port and have it
> decode the output. That's why I went with this guy:
> http://www.pctestinstruments.com/ I like it a fair bit but I'm not
> doing anything like what Xiaofan does!

This looks good, but if it's about Agilent, that means he want to
inspect the whole memory and find easily a trigger point amount the 32
bits of data, control and address busses. Like I said, you can do
almost the same thing with a $500 or a &20k tool.
Depends on your nerves and boss theory about live.

2008\04\03@015000 by Xiaofan Chen

face picon face
On 4/3/08, Vasile Surducan <EraseMEpiclist9spamgmail.com> wrote:
> On 4/2/08, Xiaofan Chen <RemoveMExiaofancEraseMEspamEraseMEgmail.com> wrote:
> > On Thu, Apr 3, 2008 at 12:16 AM, Lloyd Sargent <RemoveMElloydspam_OUTspamKILLspamcannasoftware.com> wrote:
> > > I've worked with everything from Tek to PC analyzers (LA5000).
> > > Personally, I think if you are doing hardcore debugging, you are
> > > better off with Tek - nothing worse than a multiple condition problem
> > > that you analyzer can't do. Frankly, this is where Tek logic analyzer
> > > shine.
> >
> > Just wondering which Tek are you using now?
>
> In our lab a MSO4104, but trully I'm not very happy with it (on the
> analogic side).
> http://www.conres.com/test-equipment/mso4000.html

I was talking about Logic Analyzer.
As for scope, I am not so happy with the Agilent 8000 series
either but it is usable. Lecroy seems to be the best in terms
of dealing with analog signal with a digital scope. There are
dedicated high end analog scope as well but I have not used
them.

Xiaofan

2008\04\03@015151 by Xiaofan Chen

face picon face
On 4/3/08, Vasile Surducan <RemoveMEpiclist9TakeThisOuTspamspamgmail.com> wrote:
> This looks good, but if it's about Agilent, that means he want to
> inspect the whole memory and find easily a trigger point amount the 32
> bits of data, control and address busses. Like I said, you can do
> almost the same thing with a $500 or a &20k tool.
> Depends on your nerves and boss theory about live.
>
Actually Agilent's LA can do much more than this
and some of the features can be useful in the future.
For example, what about a debug trace soft core for
an Xilinx FPGA?

Xiaofan

2008\04\03@015629 by Xiaofan Chen

face picon face
On 4/3/08, Xiaofan Chen <EraseMExiaofancspamspamspamBeGonegmail.com> wrote:
> On 4/3/08, Vasile Surducan <RemoveMEpiclist9KILLspamspamgmail.com> wrote:
> > This looks good, but if it's about Agilent, that means he want to
> > inspect the whole memory and find easily a trigger point amount
> > the 32 bits of data, control and address busses. Like I said,
> > you can do almost the same thing with a $500 or a &20k tool.
> > Depends on your nerves and boss theory about live.
> >
> Actually Agilent's LA can do much more than this
> and some of the features can be useful in the future.

But I admit I personally have not done anything high
end in the digital side (not doing anything related to
high speed digital or FPGA. What we do in the
analog side it can be said to be pretty high performance
(but they are niche market product). In fact we have
already people from ADI, Maxim and TI visiting us in
Singapore

Xiaofan

2008\04\03@063002 by Alan B. Pearce

face picon face
>> $30k is plenty enough for any serious 32 bit logic analyser. For an
>> ARM9, 1Gsps is more than enough. For the faster FPGA on the market
>> must be a better one.
>
>It is not so easy to be future proof. The FPGA side is a concern.

My experience is that you cannot future proof your instruments. I have seen
it so many times with PROM programmers - "this instrument is software
upgradable to future proof against upcoming products, the only programmer
you will ever need"  -oh yeah, within a couple of years you cannot get
upgrades because things have gone ways they didn't expect, or new
technologies have come out the they cannot handle ...

The same thing happens with logic analyzers and micro development systems.
Anyone still using an HP 64000 system? They were designed to be the alpha
and omega of development systems, but I don't know that they got past 8086
and 68k systems.

The same seem to go with all other instruments. Even though you are spending
$30k, it needs to be written off over about 3 years, as by then it will no
longer be state of the art, just ordinary performance for an instrument in
its class.

All you can do is purchase against current testing requirements

2008\04\03@125130 by Lloyd Sargent

flavicon
face
On Apr 2, 2008, at 11:57 PM, Vasile Surducan wrote:
> Like I said, you can do
> almost the same thing with a $500 or a &20k tool.
> Depends on your nerves and boss theory about live.

Were that true.

I just looked at that $500 toy and the difference between it and a  
"real" logic analyzer  - 8K memory depth is laughable. I saw another  
one that was about the same price and it only had 2K memory depth.

When you blow your memory memory budget too quickly after the trigger,  
the difference between a $500 tool and a $20K tool is indeed real. It  
means the difference between having information and having nothing.

We were trying to do some debugging on an IDE to FPGA communications  
project and someone tried to push a toy they found on EBAY with 1K of  
memory depth. That went nowhere (all it took was a little math to show  
them why it wouldn't work).

Cheers,

Lloyd

2008\04\03@173325 by Matt Pobursky

flavicon
face
I was going to post the same basic sentiment earlier. I have a lab full of
"future proof" instruments that are now obsolete. Not worthless, but also
no longer supported by the manufacturer.

The sales and marketing guys will always play up the "future proof" aspect
of their toys but the reality is they have no clue whether the company will
actually support or extend that product. Past performance of a company
isn't all that great a predictor anymore either.

I always tell those who ask my advice about technology based instruments
(test equipment, computers, etc.) to just buy the very best you can afford
right now that meets all your current needs.

Matt Pobursky
Maximum Performance Systems

On Thu, 3 Apr 2008 11:28:33 +0100, Alan B. Pearce wrote:
{Quote hidden}

2008\04\03@192248 by Xiaofan Chen

face picon face
On Thu, Apr 3, 2008 at 6:28 PM, Alan B. Pearce <A.B.PearceSTOPspamspamspam_OUTrl.ac.uk> wrote:
> >> $30k is plenty enough for any serious 32 bit logic analyser. For an
> >> ARM9, 1Gsps is more than enough. For the faster FPGA on the market
> >> must be a better one.
> >
> >It is not so easy to be future proof. The FPGA side is a concern.
>
> My experience is that you cannot future proof your instruments. I have seen
> it so many times with PROM programmers - "this instrument is software
> upgradable to future proof against upcoming products, the only programmer
> you will ever need"  -oh yeah, within a couple of years you cannot get
> upgrades because things have gone ways they didn't expect, or new
> technologies have come out the they cannot handle ...
>

If you can afford a BP or DataIO, it is quite future proof...

>
> The same seem to go with all other instruments. Even though you are spending
> $30k, it needs to be written off over about 3 years, as by then it will no
> longer be state of the art, just ordinary performance for an instrument in
> its class.

I agree with you. But some of the things does not change that
fast. For example, we need to do a lot of DC/DC converter and
high precision ADC/DAC stuff. Things move very slowly in this
kind of analog design front. So a good scope can last 10 years.
A DC calibarator can last 10 years as well. And our product
can sell more than 10 years.

> All you can do is purchase against current testing requirements
>
Currently our group does not need an LA but the next project may need.
This is a fast changing organization and we need to ramp up the capability
fast to support the business growth.

But I get what you are saying.

Xiaofan

2008\04\04@101656 by Lloyd Sargent

flavicon
face
> If you can afford a BP or DataIO, it is quite future proof...

No, even those are not future proof. The DataIO programmers I was  
using in the 80's would not work with todays parts. Sometimes you HAVE  
to buy new equipment.

>> The same seem to go with all other instruments. Even though you are  
>> spending
>> $30k, it needs to be written off over about 3 years, as by then it  
>> will no
>> longer be state of the art, just ordinary performance for an  
>> instrument in
>> its class.

Here is how I explain it to the powers that be:

Let us assume that you pay your engineer $60K / year. Well, with  
bennies, etc. that works out to around a REAL $120K / year or about  
$57/hour (cheap). One month of his time is about $9000. Let's say  
using that cheap piece of stuff takes him twice as long to complete a  
task - so instead of one month, he spends two. All of a sudden, you  
just spent $18K.

And that is just for ONE task.

I once had an engineer tell me he could debug a hard drive controller  
with just an LED (this was when I was trying to get a logic analyzer  
to find a very nasty race condition - HIS race condition). He held up  
the purchase so I used a 6800 (back a LONG time ago) and sucked in  
data points as fast as possible. I then printed them out and spread  
them down the length of the factory floor. After multiple repetitions  
I managed to catch it (by chance). This was the only proof I needed to  
get the logic analyzer which showed it in minutes (as opposed to my  
days).

NEVER underestimate the power of a tool. Hell, I'm of an era when a  
100 Mhz scope was deemed "future proof"!!!

Cheers,

Lloyd

More... (looser matching)
- Last day of these posts
- In 2008 , 2009 only
- Today
- New search...