Skip to main content
Topic: Test Release 2.03 and 2.04 - SPI mode. (Read 37790 times) previous topic - next topic

Re: Test Release 2.03 and 2.04 - SPI mode.

Reply #105
[quote author="bash"]
[quote author="alm"]
Yeah, the Intronix Logicport is by far the best
[/quote]

I have a Intronix LA1034 , but packed it and use my Saleae to work...  Because Intronix soft is so stupid (like 80´s one) and is very confused for simple things... LA1034 are a very good hardware, but have a very bad soft... [/quote]

sigrok says they will support the Intronix with their multi-platform software soon ... the bundle Intronix/sirok may give you the total killer LA package in the price range of under US$ 500 ;)

Quote
I checked only group 0..( use only 4 channels ( 0 to 3 )) ... checked Noise Filter...Internal 10Mhz.. and selected 2K size capture.. (configure my trigger too) ...
2K size...very good..
but above 2K capture are wrong... show me nothing ! .. What´s wrong? Some configuration that I don´t set?

FW v0.4 or v0.6/Bitstream rel. 2.04 currently  support only 3 main capture configurations:
   - up to 4k samples for 32 channels
   - up to 8k samples for 16 channels (if channel groups 3 and 4 are disabled in the capture menu)
   - test mode (needs to connect channels on the unbuffered header with channels on the buffered header with patch cables i.e.)

The RLE option does not work properly, captures will show wrong/invalid data.

But that doesn't explain why you will receive only 2K captures ... try to run following capture configuration:

Inside, Internal 10MHz, check only Channel Groups 1 & 2, Recording size 4k (and then 8k), Noise Filter Enable, RLE Disable (!Enable) 

[quote author="alm"]
Not sure if you really need a high voltage to damage a USB root hub. What if it gets +12V or -12V on its signal lines? What are the guaranteed max breakdown voltages? What if you accidentally touch the probe to some other voltage (eg. vacuum fluorescent display), or the power supply of your device shits itself (as in this case) and puts out a high voltage (high enough to kill a 7812)? The fact that you are debugging it means something is not working right, so you should expect more abuse than consumer devices like MP3 players.[/quote]
1. I expect the USB root hub to die when it gets +12V/-12V on the signal lines ... but then the OLS will die first ... the FPGA will for sure if you don't use the buffered header. But then the unbuffered header is actually intended for a wing like a DSO extension and any other protection of the lines but with serial resistors (which is no real protection against cases we talked about) would make it unsuitable for this.
2. Same when you touch VFD power with a probe - VFD voltag is typically over 50V and qualifies for "high-voltage" in the meaning of better using opto-isolated interfaces (not in the meaning of saftey regulations)
3. In the end even the best lab equipment is only protected to certain limits - it's up to the user not to abuse it (like I did) and it's not protected against (possibly) 60-110VAC (or more if the switcher in the wall mart went into resonance before it died) inducted into the circuitry (terrymeyer) because of a failing wall mart - like if you buy a B6 armoured Mercede-Benz GL but drive it over a cliff into an abyss or get run over by a 40 ton truck the armore won't keep it from getting damaged.

Quote
I think some of them have protection diode arrays and inline resistors (not sure about the rating), but no isolation, that's correct. Fast protection diodes and damping resistors are a probably a reasonable solution, the advantage of magnetic/optical isolation is that you get almost unlimited isolation (a few kV) for not much money (SI8431 is $1.65 for 100 at Mouser, although that doesn't take care of the power). Not sure if I'd consider any of them professional tools, although the LogicPort comes close. I'd expect the $$$ ones from Agilent and Tek to have better input protection, but in that case I probably would be more worried about the analyzer than my PC ;).
Yes, some have protection diodes and resistor arrays - it is one suggestion for the next redesign of the OLS to implement a protection diode at least in the USB power line. When saying they have no protection I was actually responding to your suggestion to use opto-couplers between PIC and FPGA (my bad, I didn't quote the exact reference). opto-isolation is more or less close to useless if you don't seperate your circuitry completely into two electrically independent sections (no electrical connections, two truely independent power sources, one for each section). For the OLS this would mean to add at least 2 high-speed 3 channel opto-couplers, a power plug and one more 3.3V VR plus a complete redesign of the PCB (separation), not to forget the extra wall mart/power supply the user will need regardless if he needs the opto-isolation or not. All this would raise the cost by at least US$ 10 .... the main problem will be PCB space ... Ian wants to stick with the free version of Eagle CAD for good reasons, I think.

For professional lab equipment there are usually two options (and in most cases they are not standard) to protect test equipment like LAs, scopes etc. OR the equipment they are connected to against over-currents and high-voltages:
a) magnetics or opto-isolated interface towards the PC/network - does not protect the test quipment itself
b) opto-isolated (active) probes - this is the recommended way to go if the opto-couplers can provide enough bandwith for the application because it provides the maximum protection for the test equipment as well as for the operator.
In case of the OLS this may lead to the option an opto-isolated probe project for I²C, SPI, RS-485/422, RS232 etc.

Quote
In industrial environments and with runs between multiple floors/buildings, ground potential differences (not sure if I'd call them loops) can be damaging to even not-so-sensitive equipment (I don't expect an RS-232 to 4-20 converter to be very delicate).
Today all RS232 <--> 20mA converters for industrial environments I know of are opto-isolated ... what I described was back in the 80's.

Quote
I was thinking more about ground loops on your workbench, the potential differences are usually below 100mV or so in that case.
I agree that "typical" ground loops encountered in home audio and TV equipment or on stanges in electronic audio equipment cause a potential difference of some 10mV to 1V, maybe 2V. The ground loops I described in industrial environments are extreme but not uncommon with line lengths of hundreds of meters or more ... Ian asked for examples of ground loops causing damage to electronics so I picked the example with the data lines in industrial environments ... I could have described ground loops I experienced with cable-based communication equipment running through a military communication headquarter with high-power HF transmitters not following basic grounding rules on some of the equipment ... the example from the industrial environment would pale in comparison. I have no idea what else to call an unwanted current in a conductor that's supposed to be on the same potential at both ends but a ground loop?

Re: Test Release 2.03 and 2.04 - SPI mode.

Reply #106
[quote author="IPenguin"]
sigrok says they will support the Intronix with their multi-platform software soon ... the bundle Intronix/sirok may give you the total killer LA package in the price range of under US$ 500 ;)
[/quote]
As soon as sigrok itself becomes usable ;). I tried it recently with a Saleae Logic, and it seemed really immature, but I would really like it to become better.

[quote author="IPenguin"]
1. I expect the USB root hub to die when it gets +12V/-12V on the signal lines ... but then the OLS will die first ... the FPGA will for sure if you don't use the buffered header. But then the unbuffered header is actually intended for a wing like a DSO extension and any other protection of the lines but with serial resistors (which is no real protection against cases we talked about) would make it unsuitable for this.
[/quote]
I was mainly referring to the buffered inputs. It's fine if the buffer or FPGA dies ($50 worst case to replace the OLS), but does that mean the root hub will survive (a much more expensive replacement, especially in a laptop)? Voltages like that are pretty common in analog circuits, and might be close to digital lines in mixed-signal designs.

[quote author="IPenguin"]
2. Same when you touch VFD power with a probe - VFD voltag is typically over 50V and qualifies for "high-voltage" in the meaning of better using opto-isolated interfaces (not in the meaning of saftey regulations)
[/quote]
It was just an example of not power-supply related work that might benefit from isolation, that you might encounter in regular hobbyist circuits.

[quote author="IPenguin"]
3. In the end even the best lab equipment is only protected to certain limits - it's up to the user not to abuse it (like I did) and it's not protected against (possibly) 60-110VAC (or more if the switcher in the wall mart went into resonance before it died) inducted into the circuitry (terrymeyer) because of a failing wall mart - like if you buy a B6 armoured Mercede-Benz GL but drive it over a cliff into an abyss or get run over by a 40 ton truck the armore won't keep it from getting damaged.
[/quote]
Yes, I don't expect a logic analyzer to survive hundreds of volts or even mains, although I would expect my computer to survive the latter (OLS connected to mains voltage). Lab equipment tends to be more robust than normal equipment, almost any signal source or power supply is short-circuit proof, unlike some wall warts ;).

[quote author="IPenguin"]
Yes, some have protection diodes and resistor arrays - it is one suggestion for the next redesign of the OLS to implement a protection diode at least in the USB power line. [...] For the OLS this would mean to add at least 2 high-speed 3 channel opto-couplers, a power plug and one more 3.3V VR plus a complete redesign of the PCB (separation), not to forget the extra wall mart/power supply the user will need regardless if he needs the opto-isolation or not. All this would raise the cost by at least US$ 10 .... the main problem will be PCB space ... Ian wants to stick with the free version of Eagle CAD for good reasons, I think.[/quote]
I was actually thinking about clamping diodes from any signal line to power/ground (I'd probably do that on the input side), I think at least the Saleae Logic has them. Diodes to prevent reverse current in the power line might cost too much voltage, and aren't that useful if you consider the OLS as an enclosed piece of equipment, since the power lines are not usually connected to the DUT. Some of the AD iCoupler devices (look at us, we're modern and using trendy names) have an integrated isolated power supply up to 100mA or so, but they are quite expensive, and any other form of isolated DC-to-DC converter could also be used. I think $10 is a bit pessimistic, the 3 channel (2 out, 1 in or visa versa) isolator that I mentioned is less than $2 (do you really need two of those?). A DC power jack is probably under $1. Add extra costs for assembly and PCB. The need for an extra power supply without requiring isolation could be fixed with some sort of jumper. But your point about extra board space is true, one of the reasons why I'd prefer that the design was done in something open source like kicad, but this isn't exactly the right time to change.

[quote author="IPenguin"]
For professional lab equipment there are usually two options (and in most cases they are not standard) to protect test equipment like LAs, scopes etc. OR the equipment they are connected to against over-currents and high-voltages:
a) magnetics or opto-isolated interface towards the PC/network - does not protect the test quipment itself
b) opto-isolated (active) probes - this is the recommended way to go if the opto-couplers can provide enough bandwith for the application because it provides the maximum protection for the test equipment as well as for the operator.
In case of the OLS this may lead to the option an opto-isolated probe project for I²C, SPI, RS-485/422, RS232 etc.
[/quote]
Yes, isolating at the probes is also a good solution. Not sure if I see the point of active probes for I[sup:]2[/sup:]C and SPI, since they don't have any particular electrical interface. And the GPIB/RS-232/LAN interfaces are often isolated (at least for function generators, oscilloscopes and multimeters) on professional equipment, they often have a separate isolated micro-controller for communication. With LAN, you even get some isolation for free since an isolation transformer is mandatory for 10baseT/100baseTX/1000baseT, although you probably shouldn't rely on this.

[quote author="IPenguin"]
Quote
In industrial environments and with runs between multiple floors/buildings, ground potential differences (not sure if I'd call them loops) can be damaging to even not-so-sensitive equipment (I don't expect an RS-232 to 4-20 converter to be very delicate).
Today all RS232 <--> 20mA converters for industrial environments I know of are opto-isolated ... what I described was back in the 80's.
[/quote]
But even without opto's, RS-232 transceivers can usually withstand at least +/- 25V or so, and I can't imagine a 4-to-20mA converter (which is mainly used in industrial environments) to be very delicate either. Not like CMOS inputs that blow up when you sneeze.

Re: Test Release 2.03 and 2.04 - SPI mode.

Reply #107
I can now report successful BP upgrade done, communication seem to be reliable with initial test.


[quote author="rhyde"]
...
No bootloader so I did not update the PIC, I am not sure if that was required.  I do have a bp so if Ian can point the script for that method I am willing to try that.


$ /cygdrive/c/Program Files/Java/jdk1.6.0_20/bin/java.exe -jar analyzer.jar
Stable Library
=========================================
Native lib Version = RXTX-2.1-7
Java lib Version   = RXTX-2.1-7
COM1
COM14
Device Controller found: org.sump.analyzer.devices.FpgaDeviceController
COM1
COM14
Device Controller found: org.sump.analyzer.devices.Hp16500DeviceController
Device Controller = FPGA Controller
Tool found: org.sump.analyzer.tools.I2CProtocolAnalysis
Tool found: org.sump.analyzer.tools.SPIProtocolAnalysis
Tool found: org.sump.analyzer.tools.StateAnalysis
Tool found: org.sump.analyzer.tools.UARTProtocolAnalysis
Attaching to: COM14 (115200bps)
Run started
Device ID: 0x0
Run aborted
java.io.IOException: Device not found.
        at org.sump.analyzer.devices.FpgaDevice.run(FpgaDevice.java:648)
        at org.sump.analyzer.devices.FpgaDeviceController.run(FpgaDeviceControll
er.java:546)
        at java.lang.Thread.run(Thread.java:619)
[/quote]

Test Release 2.04 - SPI mode - My Own Full test on my new received OLS

Reply #108
Hi,

I have just received my OLS v1.01 from Seed Studio and it seems all work fine (at least using Test Mode and with some previous test using BusPirate+PWM) !!!

I have logged everything in Test Mode (configuration HW/SW used ...) in the attached zip file (containing OpenOffice .odt full description).

Best Regards

TitanMKD
Best Regards
Benjamin

Re: Test Release 2.03 and 2.04 - SPI mode.

Reply #109
TitanMKD, thank you for the detailed report.

On a side note, you must have used the new SUMP Java client contained in the bitstream release 2.04 package (2.04TestRelease.zip).
Unfortunately the client (analyzer.jar) still shows the old version from 2008 (... analyzer.jar that comes in 2.04TestRelease.zip is the
correct one for bitstream release 2.04 from the same package).

Re: Test Release 2.03 and 2.04 - SPI mode.

Reply #110
Here is an experimental package that can program the bootloader into the Logic Sniffer using a Bus Pirate. This should help out a few people who got an OLS without a bootloader. Read more here.
Got a question? Please ask in the forum for the fastest answers.

Re: Test Release 2.03 and 2.04 - SPI mode.

Reply #111
A rescue package and instructions for reprogramming the Logic Sniffer bootloader with a Bus Pirate are now posted here.

If you don't have access to a Bus Pirate or PIC programmer to replace your bootloader please contact me. I'll put you in touch with someone nearby who can reprogram the bootloader, or send a Bus Pirate, depending on your location.

I'm sorry about this bug, it really hit us by surprise. We're still not sure how it happened, or the extent of the problem. Hopefully it was a minor mistake that doesn't effect too many people.
Got a question? Please ask in the forum for the fastest answers.