1. Home
  2. Electronics Design
  3. Oscilloscopes
  4. Page 2

Oscilloscopes

As mentioned up above, I use

formatting link
Your mileage may vary.

Some things to consider:

  1. The USB logic analyzers tend to split into two groups based on their capture mode. There is a group with deep sample memory that always captures in real time. This leads to a trade-off between duration and precision, where sampling at 100 MHz into a 1 Mb buffer fills the buffer in only 10 msec.

The other class has shallower buffers (typically a few Kb) but can tag each sample with a timestamp, so sampling at 100 MHz is still possible throughout an event that lasts several seconds. On the other hand, if there's a channel with a lot of activity then the buffer will be filled correspondingly quickly.

  1. Pay attention to the maximum input voltage range and also to the trigger range. There are some inexpensive USB LAs that have a fixed front-end range of 0-5 VDC and a fixed trigger threshold but ones that handle at least +/- 20 VDC with adjustable triggers aren't much more expensive.

  1. Back in the day, one feature of LAs was the ability to load a machine language interpreter and, with the proper assignment of channels to address, data, and control busses, to trace program execution and trigger on events like the change of a memory location. Nowadays, at least in microcontroller land, not so much. Instead, look for the ability to interpret common serial protocols like CANbus, I2C, etc.

--
Rich Webb     Norfolk, VA
Reply to
Rich Webb
Loading thread data ...

I recently purchased a used logic analyzer from Tek for $450. I looked at the USB LAs as well but if you want something nice (lot of depth, at least 1Gs/s, versatile triggering and input specifications specified at the tip of the probe) you'll pay a lot more.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------
Reply to
Nico Coesel

A TLA704 can be bought cheap if you have some patience. This runs plain Windows98 (or Windows2000 if you upgrade the memory) instead of some archaic Unix.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------
Reply to
Nico Coesel

=3DA0I =3D

=3DA0Also, I=3D

=3D

AVR =3D

a =3D

=3D

At that speed the choices are get good stuff or go without. Did that = price=20 include pods and probes?

Reply to
JosephKK

=20

looking=20

at=20

on=20

=20

would=20

nice=20

That is pretty good for an LA at that price.

Reply to
JosephKK

Thanks for the info. Myself not knowing anything about logic analyzers I'm not sure if I should get the one in your link or something like a HP 1630 for around $100 shipped. Much of what I'm interested in would be the serial protocols but more like Allen Bradley Remote I/O and DH+. I have some Allen Bradley PLC's here at home and could set up limited communication, and try to determine the commands. I think the Data Highway + and Remote I/O are electrically RS-485, there are 2 wires and a shield for connections. They only range from 57.6K baud to 230.4K baud, shouldn't require anything too high performance.

Do any of the out dated logic analyzers, such as the HP 1630(A)(D)or(G),

1631 or 16500 show the data being communicated through serial, or is that more of a recent feature?

RogerN

Reply to
RogerN

BTSOOM -- They should all *show* the signals. Whether or not they have data interpreters that can be told to translate "57.6, 8-N-1, inverted" to ASCII characters is left as an exercise for the student. ;-)

I haven't worked with those AB protocols but it may be that a couple of serial ports fed by two MAX488 and one MAX232A are all that's necessary. The MAX488s each have one RX and TX for 422/485 to digital, the MAX232A has two RX and TX digital to 232. Hook those up to a PC with a couple of (possibly USB) serial ports and you've got both sides of the conversation.

--
Rich Webb     Norfolk, VA
Reply to
Rich Webb

Yes. It came with 3 16-channel probes and an extra 4M deep module.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------
Reply to
Nico Coesel

circuits?=20

looking=20

starting at=20

on=20

4=20

would=20

nice=20

discussion

Overdue

After Nico's find i may have to amend that.

Reply to
JosephKK

1241.

top.

looking=20

their

buffer

filled

machine

I'm=20

1630=20

serial=20

Allen=20

try=20

are=20

They=20

too=20

=20

that=20

Actually that was split off and handed to a device called a protocol=20 analyzer. It would basically leach off the serial bus 232, 422, 423,=20

485 or whatever and record and decode the transactions.
Reply to
JosephKK

Warning, I hear the connector from the pod to the rest of the world is unobtainium. In many cases, the pods ran plugged directly into the microprocessor interface adaptor box and the little brown plastic block with the patchwires, and the micrograbber clips, walked away or got stepped on years before they got surplused. Or worse, some gorilla was charged with collecting up anything that looked like a cable and sending it off for copper recycling. Pod and all.

(Yow, it's been 25 years). As I remember, the 1630(D) didn't do serial protocol analysis as a stock feature. You could customize the unit with disassemblers and the like if you had the microcassette drive with the A or D. (The G uses the HPIB diskette drive. I think you could upgrade, but the G came out long after we got the D at where I worked.)

Their speciality was state analysis (memory bus cycles) for microcontrollers. Software machines. The A had less channels (32?) and would max out on an 8 bit micro, the D was good for the 16+ bit machines of the day (48? channels). One neat feature was to display a bit field as analog oscilloscope type display (with the quirk that it used unsigned integers so the negative numbers were on the top half, upside down).

Timing analysis was limited to, I think, only 8 channels. And maybe only 4 channels at top speed (100 MHz?). Not very deep, 2k samples or something like that.

If you had to use one with a serial protocol, it might be better to build a UART/USRT on a project board and try to figure out what you got as parallel data.

I think that Tek or Sony/Tek was bigger with serial protocol analysers.

Mark Zenier snipped-for-privacy@eskimo.com Googleproofaddress(account:mzenier provider:eskimo domain:com)

Reply to
Mark Zenier

I've never used a logic analyzer. It seems to me that there's a lot of setup, and what it usually does is just force you to review a hardware or software design that should have been checked more carefully at the design stage, before it was fired up.

Just yesterday I did check the data stream from a 16M serial flash chip that's configuring a Xilinx Spartan6 FPGA, to see if I'd got the bit order right and such. It was easy to read on a digital scope.

A LA could be useful there, reverse engineering a complex and undocumented serial protocol.

My guys, who like this sort of thing, tend towards small USB pod things.

John

Reply to
John Larkin

There universe of errors is far larger than design.

For simple problems like that I use a scope too. Many problems are a lot deeper and some only happen once a blue moon. Logic analyzers are invaluable for tracking these down. As you point out, they're a lot of work to get set up so I rarely use them but when they're called for they can be life savers.

Or violations of protocol (parallel or serial).

Never used one of those, but perhaps it's time. ;-)

Reply to
krw

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.