On Thu, Jul 25, 2019 at 12:12 PM, Dave McGuire wrote:

It sounds like Keysight's upper management has forgotten the nature of
the business that they're in, and the nature of their own company.
Suits often forget (or want to erase) where things come from.

More likely they never knew.

There's also the fact that they live under the tyranny of the financial quarter and get any long term thinking beat mercilessly out of them.

I've looked through the thread and not found a like to a bill of materials anywhere.? Is there one?? I'm particulalry interested in where you are getting the main connector.

A year or two ago, I bought an ICS Electronics 8065 Ethernet-to-GPIB bridge on the usual auction site.? It has an interesting approach.? On the Ethernet side, it implements and publishes a VXI-11 interface (which is based on a documented RPC mechanism).? You can open a TCP/IP connection to this interface, tell it to address a specific GPIB device address, and then use the RPCs to communicate.? It handles multiple GPIB targets at once, via individual VXI-11 connection.? As far as I can tell it has full-strength bus drivers.

It can be used with any code library that implements the necessary RPCs... either commercial products, or open-source implementations such as python-vxi11.? I threw together a little Python script which lets me control an HP signal generator via a Web CGI interface... a bit of the code follows.

instr = vxi11.Instrument("ics8065", "gpib0,7")
...
??? if deviation != "":
??????? instr.write("fm:stat off")
??????? instr.write("fm:dev " + deviation)
??????? instr.write("fm:stat on")
??? if action == "Modulate":
??????? instr.write("am:stat off")
??????? instr.write("pm:stat off")
??????? instr.write("fm:sour int")
??????? instr.write("fm:int:freq 400 hz")

It seems to me that a network-to-GPIB bridge of this source wouldn't be too terribly difficult to open-source.? A Raspberry Pi or similar Linux-based SBC, and a "hat" or similar add-on card with bus drivers wired to the GPIOs, and the proper connector would be about all of the hardware one needed.? A modest wall-wart for power would be plenty.

It's not as compact as a USB-based interface, and quite not as portable, but it would "un-tie" the developer from needing to have a laptop right on the workbench near the test instruments.? You can access the network connection remotely from anywhere on your LAN (wireless if you wish), or from anywhere in the world if you have properly-secured remote access of some sort.? (I used the Web CGI I'd thrown together to turn the signal generator into a remotely-controlled fleapower transmitter... with a single-button "Send my amateur-radio callsign via CW" to keep me legal :-) )

The VXI-11 "server" software would need to be written, of course... something to implement the VXI-11 RPCs and multiplex access to the bus via the GPIOs.? As I understand it, it's not terribly difficult to access the Pi GPIOs at fairly high speeds via memory-mapped access, so one should be able to get pretty good throughput without needing an FPGA or dedicated protocol chip.

I have two HP 410C meters. The probe (HP 11036A) in my first one uses the EA53 which is designed for a 6.3VAC filament. The probe for my newest one has the 2-01C which is designed for a 5.0VAC filament.

I know that the 410B (I have two of those as well) has an adjustment for the filament voltage so it can use either tube; however, the 410C doesn't have that adjustment.

I always thought that all the 11036A probes were compatible with all 410Cs but perhaps not. Were some 410Cs made for a 2-01C and others for the EA53? If so, is it possible this is tied to a serial number prefix and the EA53 probes are only to be used with certain prefixes and the 2-01C used for the other prefixes? All the schematics I've found show 6.0VAC as the heater voltage supply so it seems odd to me that the 2-01C was used at that voltage but maybe so?

Thanks,
Barry - N4BUQ

I believe it is called "progress" , what used to be designed by thinking engineers are mostly done by the kids and their dumb phones IMHO.
this drives me nuts too!
as far as linux may have been easily solved by booting from a disk or usb.
no more win stuff for me 2 yrs ago went to linux! there is a learning curve. I have a 98,xp and 7 machines that never go on line..for things that must be win
搁别苍é别

The state of software is a royal mess and is getting worse.

I had an old Thinkpad T60 with Win7 which I was using as a lab bench machine, mostly downloading pdf files of parts for troubleshooting. The disk started to have problems and I was unable to reinstall the OS, despite the worthless product key. I installed Linux on a new drive and all was well, until Linux one day booted to some command line. Nobody could help me recover the OS and guess what? To get the files I downloaded required buying some $65 utility. Gee, I was able to very easily do that from the Windows disk. Start from scratch using another laptop, scrap that one.

Last night my wife asked me to make a copy of a CD with a bunch of files. Easy, right? All built into Win10. I first copy the files to a temp directory, then put a blank CD in. Then I get the option to make a data CD, name it, then move the files to that window. So far so good, and it tells me the files are to be copied to the CD. But there is no button or menu item to start the process! I Google it and it shows the option, but it’s not there on my system. So I remove the CD and put it back in as a last resort and sure enough the option is now shown. WTF? How much do they check this stuff? Making a copy in XP or 7 was foolproof, why did they break it?


Peter

On 26/07/19 14:50, Peter Gottlieb wrote:

Support for an OS has several angles.

I would hope the negotiated deal for a piece of equipment with an embedded OS would have a lifetime, fully transferrable license for that piece of equipment. So your Win2k scope should have the rights to use that OS until the instrument is scrapped.

I don’t think Microsoft should be forced into infinite support in the form of security updates, although they sort of secretly do for Win2k and XP, if you set the registry flag indicating they are in a bank ATM. But even that can’t last forever. However, if the scope is operated standalone or on a very well protected network as a practical matter a lack of updates doesn’t matter.

In reality it is worse than that; here's three Microsoft horror stories...

Once, while WinXP was still fully supported, I had an "interesting" experience with a Samsung netbook after its hard disk failed. Printed on the underside of the netbook were the WinXP licence and the xxxxx-xxxxx-xxxxx-xxxx product key.

I bought a new hard disk and installed it, then used an external CDROM drive to re-install WinXP from my own WinXP installation CDROM, using the product key on the bottom of the netbook.

WinXP installed in the usual way, quite successfully, as expected.

But when I first rebooted I was confronted by an unfamiliar DOS-box window, which said "Microsoft" at the top and "shan't boot into WinXP because the product key was wrong". What the...?! WinXP had just installed correctly.

I contacted Microsoft, and their service person said it was a Samsung problem, even though this was clearly a Microsoft display.

I contacted Samsung, and their service person said it was a Microsoft problem, quite reasonably IMHO. (But they did try to weasel out by asking the disk drive's manufacturer; the original was Western Digital (!) and by chance the replacement was Samsung)

In the end the only solution would have been to buy another a new hard disk from Samsung, with WinXP pre-installed.

Bugger that; I installed Linux and the 10 year old netbook continues to work well with an excellent battery life since the I set the BIOS to only charge it to 80% of capacity.



The second horror story is Microsoft's PlaysForSure (TM) music. Since Microsoft have turned their servers, you can't move the music you purchased to another hard disk. Makes a mockery of "PlaysForSure", despite Microsoft claiming it was merely "pining for the fjords"


The third horror story happened this month: Microsoft closed down its ebooks platform. It has offered a refund, plus a bit if your carefully curated bookmarks have gone up in aetherial smoke.


See a pattern?

Support for an OS has several angles.

I would hope the negotiated deal for a piece of equipment with an embedded OS would have a lifetime, fully transferrable license for that piece of equipment. So your Win2k scope should have the rights to use that OS until the instrument is scrapped.

I don’t think Microsoft should be forced into infinite support in the form of security updates, although they sort of secretly do for Win2k and XP, if you set the registry flag indicating they are in a bank ATM. But even that can’t last forever. However, if the scope is operated standalone or on a very well protected network as a practical matter a lack of updates doesn’t matter.

As for Tek supporting older instruments, there should be reasonable limits. It will be increasingly expensive to support older gear as fewer and fewer personnel are familiar with the older technology. Should Tek still support 545 vacuum tube scopes? What about parts which are no longer available? But, from the customer perspective, what is reasonable? Should 10 years of replacement ASICs be stocked? What about increasingly complex and expensive test setups, which themselves may have unobtanium repair parts? 10 years would be satisfactory to me for standard instruments, maybe 5 for highly specialized ones, but I would definitely balk if less than that, and try to avoid a company that hinders third party repairs. I’ve worked at many places that have older gear which is maintained and calibrated by these third party firms and if gear must be sent back to inly the manufacturer then there is little difference to Chinese or Korean test gear.


Peter

<bump>

And: A reminder that I will be more than happy to help scan the SM's - heck, I'll pay (reasonable) postage to send back and forth.

My personal preference for the associated SN's (but not limited to) are the HP 3663A, 5335A, 5345A, 5245L, 3325A, 5105A, 5110B, 8640B, 8903A, 3456A, 3455A.? Again, while those are on the top of my personal list, I will of course help with any others that anyone wants.

Thank you all again,

Hal

Problem is, and not wishing to sound arrogant in any way...., In their eyes you have had this scope too long and
Tek is expecting you (like Keysight/HP and others) to roll over and buy a new one.

I wonder how they would feel if their shiny new BMW, Mercedes was deemed unrepairable/unserviceable after only 2 years..

Most companies think a lifetime of between 5-10 years is more than enough to keep top end/high tech equipment.

The use of an embedded PC style OS is a real help to them as the moment M/soft or whoever declares end of support then
the test equipment supplier rubs his hands with glee and can happily refuse to support it.

So many companies now are calling EOS for their equipment as soon as they can.
Not nice when it is hard to justify replacements especially in a teaching or research environment.

What is needed is for the salesmen to be nailed to the floor with written statements that they will support and honour
any stated equipment lifetime such that it makes it really hard for them to sell kit.

That will hurt them (the manufacturers the most) - only then will they consider change or lose a sale to up coming
Instrument makers (eg. Korean or Chinese) gradually their performance is improving and hopefully the more advantageous
pricing will help all of us.

Unfortunately gone are the days of beautifully prepared manuals and information about everything to do with the item from
well known TE companies such are HP, Marconi, Tek etc.

Regards

Nigel

Dear All,

As Alwyn says: "What really annoys me is the lack of support of equipment which uses older versions of Windows- knowing the policies of that company, I now insist on hearing how long term support will be assured before I buy."

Indeed!!!

I have a Tek TDS7000B series in our group that I am trying to "repair" because its Windows 2000 OS has gone AWOL from a hard drive failure. Tek refuse to supply the Microsoft Windows restore discs "because of licensing issues" which to me seems entity bogus for a piece of equipment that cost well over GBP20,000 in 2003 (i.e. more than a Hi-End VW Golf Turbo).

We have happenchance found an original Restore Manual and set of discs on e$ay which fingers crossed will resolve the issue, but the lack of support from Tek is astounding for such expensive equipment (TDS7154B).

I am now recommanding Pico Technology scopes wherever possible, even though I personally prefer a 'scope with knobs on it the price differentials are to great to ignore.

Best,
Susan.

Susan Parker, Laser Consortium, Department of Physics, Imperial College London, UK.

开云体育

On 7/25/2019 3:50 PM, saipan59 wrote:

I want to suggest that the discussion should include the trade-offs of an FPGA implementation vs an MCU [I have very limited experience with the former, but a lot with the latter], as it relates to the applications we're talking about.

I'll weigh in here.

FPGA:? lots faster, more expensive, extensive I/O.? True hardware implementation, simultaneous operation of functions.? Often 3.3 volts for I/O.? Cost (including CPLDS, which are more limited in scope, cheaper, and generally nonvolatile), will run from about 2 dollars to 16 dollars for Xilinx parts (what I'm familiar with).? VHDL is essentially specifying hardware.? If you want it to respond to different conditions (as in pretending to have a program in it) then? you need either a state machine or a processor IP (Picoblaze is possible).? RAM resources are somewhat limited depending on chip (CPLDs have none).? Best use: smart hardware and hardware subsystems.? SPI interface may be programmed in for processor interface.? Programming is in VHDL or Verilog.?

Microprocessor:? No hardware systems need to be designed.? Programming is sequential, slower, although DMA and some subsystems can operate independently.? 3.3 and 5.0 volts depending on processor.? Cost may run from less than one dollar to 20 dollars depending on family and manufacturer.? Microprocessors need external memory, external I/O subsystems.? Microcontrollers take program and data memory inside, and provide subsystems (I2C, SPI, Serial, etc.).? Programming is in any high level language available or assembly.? Operating system can give illusion of simultaneous operation, but unless more than one processor core is available, operation is time shared.

I'll add some points about FPGAs:

• short latency and guaranteed latency between input and/or signals changing, i.e. hard realtime operation. The operations are performed directly by hardware, so there are no interrupt service routines, multitasking etc to "get in the way"
• Xylinx Zync devices have lots of FPGA fabric plus two onboard very capable ARM A9 processors. The processors can run an RTOS or Linux, or RTOS+Linux. This allows tight integration between the processor and custom hardware

There is one processor that is half-way between a standard MCU and FPGA in terms of performance: the XMOS xCORE devices, available from Digikey. They have a long and solid pedigree, being directly descended from Transputers and Occam and Hoare's Communicating Sequential Processes (CSP) formalism.

A single chip can have up to 32 cores (expandable with chips), run at 4000MIPS, have onboard RAM and flash memory, "FPGA like" IO (e.g. SERDES etc)? - but that is useless unless you can program them quickly and effectively. The xC language enables that, being designed for multiprocessing from the ground up - unlike C. This makes it surprisingly easy and fun to actually use the devices.

As for hard realtime, the IDE will guarantee the execution times and latency. That avoids the need to run the code, make measurements, and piously hope you have stumbled across worst-case execution times. An example is that you can use some of the cores to directly grab and process 100Mb/s serial data from an ethernet or USB interface, while other cores simultaneously get on with the real work.

Yes, I'm a fan of them, for the right application.

Altera was purchased by Intel, so Oracle might have more luck if they wanted to try again.

Xilinx requires a license for their free edition, 1 year long as I recall, which always seems to expire in the middle of a hot project and have to pause to wait for renewal.

Intel offers the Quartus software in 3 versions, Intel Quartus Prime Lite Edition as a free download with no license file required, now up to version 19.1, and Prime Standard and Prime Pro as paid versions.

Xilinx has several different programming modes, so the learning curve to get from compiled VHDL or Verilog code to a programmed part seems so much simpler with Altera.

I also like that the Altera Max 2 parts I use EPM240-EMP1270 don't require a separate memory chip to store the bit map, although that's not a problem if using one of the Chinese FPGA complete modules.

What sort of processing needs to occur in the FPGA? With each different hardware adapter requiring a different driver, and only talking to different software, It's not as simple as simply converting from the serial USB data stream to parallel HPIB.

John

开云体育

Well, I don't have a problem with that, but I'm just a contributor to the thread.

I have a lot (~43 years) experience with both embedded hardware (including large FPGA designs) and embedded software, as well as embedded system architecture, so I'll add my $.02.

Partitioning an architecture into the hardware and software components is not always easy (in fact, it almost always is not easy).

Remember that, at 70,000 ft., and ignoring all of the development methodology issues, an FPGA or other fixed-hardware portion of a design is different topologically than the part that the processor handles. For example, the FPGA design is fixed, topologically and (ignoring re-programming with different images for different applications) may be designed to be deterministic in it's behavior. Aside from design faults that cause timing and metastability issues. A schematic or other fixed-topological view of the hardware is the only model needed to depict the hardware (deciphering it is another issue).

The processor-controlled portion of the design necessarily involves the execution of code, which has no fixed topology at run-time, and is not necessarily deterministic. A software based design requires, at a minimum, both a structural model of the code and a behavioral model (think UML or SysML) to understand.

It is not necessary to have both elements in an embedded system, but many embedded systems do have both.

My point is that the determination of whether to use an FPGA or other fixed-hardware design, or a processor-based design, or a combination of both, is entirely dependent on the requirements that dictate how the system behaves and what it is supposed to do. In the commercial world, that determination also must take into account fixed production costs and maintenance costs of the system.

Most (?) very small embedded systems designed by "hobbyists" today use a single SBC such as an Arduino, Raspberry Pi, BeagleBone, etc.and there is no need for the amount of fixed hardware that would make a CPLD or FPGA a reasonable design option, But, as the GPIB-USB discussion shows, sometimes maybe using both is a good idea.

Hmmm, well, I may have transgressed "simple".? LED controller (pulse, pwm brightness), timer and trigger for ultrasonic distance measurement, Neopixel encoder, SPI interface, programmable pins, FIFO, usart, I2C to parallel decoder (in the works).

also GPIB encoder, replacement for an 8259 keyboard/led driver, and support circuitry for a DM5010.?

Harvey

DaveD

On 7/25/2019 3:50 PM, saipan59 wrote:

I want to suggest that the discussion should include the trade-offs of an FPGA implementation vs an MCU [I have very limited experience with the former, but a lot with the latter], as it relates to the applications we're talking about.

Anecdote: When designing the 'Power-On Reset' features for a product a few years ago, I planned to use a small MCU. When the Architect found out, he vetoed it, saying "We can't rely on running code for a critical start-up function". He was fine with an FPGA or CPLD (at higher cost, more board real estate, less flexible behaviors, and HUGE complexity in the 'code' that runs under the covers). My belief is that he was equating an FPGA to the old bipolar PAL's used in the 70's/80's, where logic functions were literally burned in, so it was considered super-reliable.
In the end, I was able to meet the requirements with just a few little timer (1-shot) chips and a bit of glue logic.

Pete

Virus-free.

开云体育

I want to suggest that the discussion should include the trade-offs of an FPGA implementation vs an MCU [I have very limited experience with the former, but a lot with the latter], as it relates to the applications we're talking about.

I'll weigh in here.

FPGA:? lots faster, more expensive, extensive I/O.? True hardware implementation, simultaneous operation of functions.? Often 3.3 volts for I/O.? Cost (including CPLDS, which are more limited in scope, cheaper, and generally nonvolatile), will run from about 2 dollars to 16 dollars for Xilinx parts (what I'm familiar with).? VHDL is essentially specifying hardware.? If you want it to respond to different conditions (as in pretending to have a program in it) then? you need either a state machine or a processor IP (Picoblaze is possible).? RAM resources are somewhat limited depending on chip (CPLDs have none).? Best use: smart hardware and hardware subsystems.? SPI interface may be programmed in for processor interface.? Programming is in VHDL or Verilog.?

Microprocessor:? No hardware systems need to be designed.? Programming is sequential, slower, although DMA and some subsystems can operate independently.? 3.3 and 5.0 volts depending on processor.? Cost may run from less than one dollar to 20 dollars depending on family and manufacturer.? Microprocessors need external memory, external I/O subsystems.? Microcontrollers take program and data memory inside, and provide subsystems (I2C, SPI, Serial, etc.).? Programming is in any high level language available or assembly.? Operating system can give illusion of simultaneous operation, but unless more than one processor core is available, operation is time shared.

.

Anecdote: When designing the 'Power-On Reset' features for a product a few years ago, I planned to use a small MCU. When the Architect found out, he vetoed it, saying "We can't rely on running code for a critical start-up function".

Note to architect, it's been done in military qualified systems for quite a few years.? Tektronix (and others) had extensive processor diagnostics before the horse was allowed outside the barn.

He was fine with an FPGA or CPLD (at higher cost, more board real estate, less flexible behaviors, and HUGE complexity in the 'code' that runs under the covers). My belief is that he was equating an FPGA to the old bipolar PAL's used in the 70's/80's, where logic functions were literally burned in, so it was considered super-reliable.

And he may have been thinking of early microprocessors as well.

Harvey

In the end, I was able to meet the requirements with just a few little timer (1-shot) chips and a bit of glue logic.

Pete

On 7/25/2019 10:58 AM, Dave Daniel wrote:

Also, don’t foget issues of cost of buying and/or finding “free” Verilog/VHDL copiler/synthesis and simulation tools.

ISE webpack is the free edition.? All it wants to do is phone home to mommy to tell them what kind of device you're using.? No worse than Microsoft or symantec.

simulator, editor, fitter, programming software, pin planner, etc.

Good for the larger projects.

Harvey

DaveD

Sent from a small flat thingy

On Jul 25, 2019, at 10:08, Harvey White <madyn@...> wrote:

On 7/25/2019 12:58 AM, John kolb wrote:

CPLD's are less complex than FPGA's but not necessarily cheaper as the newer design parts get a lot more logic in a lot less silicon, at the cost of using lower voltages.

Xilinx. 32 cell = 1.50, 64 cell = 3.25, 128 cell = 7.00, 256 cell = 15 dollars

Spartan 3AN (built in eeprom) 50K gates, about 14 dollars.

Spartan 6, (no eeprom, use winbond = 0.75) about 12 dollars, well over 50 K gates, has a 16 dollar pin compatible version with twice capacity.

Have standardized on the Spartan 6.

Pay careful attention to the bypass capacitors and clock run out to the configuration eeprom. Programmer available through amazon, web edition (local compile, report back to daddy) is free.

3.3 volts but then again was using them with Xmega and now with ARM.

All chips in non-BGA versions, TQFP-144, so therefore solderable.

SPI interface available through open cores, suggest register structure with read/write registers and go from there.

(and yes, done that....)

Harvey

Lattice 7000S series parts were good for a 5V part, but no longer available new, although still plenty on ebay.

An Altera Max 2 series part (3.3V), mounted on a PC board along with a clock oscillator and 5V to 3.3V voltage regulator would be a good start.
EPM240 series along with a USB Blaster programmer for about $12


Similar modules are available for other CPLDs and FPGAs if the logic doesn't fit the EPM240.

Using a premade module gets one away from having to deal with surface mount parts, at least as long as the interface IC's are still available as thru-hole parts.

I'm still hoping I can figure out how to use the UGPlus to talk to my HP 3456.

John

On 7/24/2019 1:12 PM, Harvey White wrote:

On 7/24/2019 2:01 PM, saipan59 wrote:

I once did a design that had the upD7210 in it (includes controller), driven by an XMEGA. I've got a similar design (unproven yet) that is driven by an ARM processor. Since the 75160/75162 are readily available (or at least, were....) I'd stick with them. I also have a possibility of using an FPGA to do the same. I'd stick with the FPGA for a new design since it can always be tweaked. I suspect that a CPLD may be too expensive for the complexity needed at that level.

FPGA goes through level shifters to the 488 bus drivers, so it's all well protected.

After I do a bunch of software (OS/graphics redesign) that's on the list (it says here......)

Harvey

I've been watching this thread with interest, and the group as a whole has identified most of the issues I dealt with in developing KISS-488. I did in fact use a PIC microcontroller with the high-current drivers on the pins going out to the bus (with series protection resistors). I also realized that the standard GPIB terminators are the Thevinin equivalent of 2K to 3.3V. Downright convenient using a modern 3.3V power supply for all the logic, and 3.3V is well above the standard TTL threshold of 2.4V for a logic high.

That said, I'm sure many on this list are dyed-in-the-wool hackers (in the older and more respectful sense of the word), and would enjoy the challenge of building your own. But if you just want to be able to get a screen dump, or log data via Telnet, or just run your instrument from a PC, please give KISS-488 a look.

Steve Hendrix

I've been meaning to read into the difference between the protocols of the SCSI PATA adapters for CD-ROM drives and hard drives since I see there are on the market adapters that are more cost effective for those devices.?

Wondering and thinking that there is some sort of firmware modification to enable... otherwise I guess some other circuit with chip is required.?

Guessing is just a firmware mod however there might be something more complex with the linked list data structure of read/write protocol and needing a SOIC.

- James Finch?

开云体育

Well, I don't have a problem with that, but I'm just a contributor to the thread.

I have a lot (~43 years) experience with both embedded hardware (including large FPGA designs) and embedded software, as well as embedded system architecture, so I'll add my $.02.

Partitioning an architecture into the hardware and software components is not always easy (in fact, it almost always is not easy).

Remember that, at 70,000 ft., and ignoring all of the development methodology issues, an FPGA or other fixed-hardware portion of a design is different topologically than the part that the processor handles. For example, the FPGA design is fixed, topologically and (ignoring re-programming with different images for different applications) may be designed to be deterministic in it's behavior. Aside from design faults that cause timing and metastability issues. A schematic or other fixed-topological view of the hardware is the only model needed to depict the hardware (deciphering it is another issue).

The processor-controlled portion of the design necessarily involves the execution of code, which has no fixed topology at run-time, and is not necessarily deterministic. A software based design requires, at a minimum, both a structural model of the code and a behavioral model (think UML or SysML) to understand.

It is not necessary to have both elements in an embedded system, but many embedded systems do have both.

My point is that the determination of whether to use an FPGA or other fixed-hardware design, or a processor-based design, or a combination of both, is entirely dependent on the requirements that dictate how the system behaves and what it is supposed to do. In the commercial world, that determination also must take into account fixed production costs and maintenance costs of the system.

Most (?) very small embedded systems designed by "hobbyists" today use a single SBC such as an Arduino, Raspberry Pi, BeagleBone, etc.and there is no need for the amount of fixed hardware that would make a CPLD or FPGA a reasonable design option, But, as the GPIB-USB discussion shows, sometimes maybe using both is a good idea.

DaveD

I want to suggest that the discussion should include the trade-offs of an FPGA implementation vs an MCU [I have very limited experience with the former, but a lot with the latter], as it relates to the applications we're talking about.

Anecdote: When designing the 'Power-On Reset' features for a product a few years ago, I planned to use a small MCU. When the Architect found out, he vetoed it, saying "We can't rely on running code for a critical start-up function". He was fine with an FPGA or CPLD (at higher cost, more board real estate, less flexible behaviors, and HUGE complexity in the 'code' that runs under the covers). My belief is that he was equating an FPGA to the old bipolar PAL's used in the 70's/80's, where logic functions were literally burned in, so it was considered super-reliable.
In the end, I was able to meet the requirements with just a few little timer (1-shot) chips and a bit of glue logic.

Pete