Re: Probing deep and/or narrow into a given unit for troubleshooting
On 2024-09-19 23:34, Radu Bogdan Dicher wrote:
Hi all,
I was wondering of asking for thoughts and ideas on probing
deep into some units, or maybe just through a narrow clearance.
What I'm thinking of is primarily for use with a DMM (vanilla
troubleshooting), though maybe scope probes and others could be
considered. But the latter is not my priority.?
?
I have a variety of probes, from "witches hats," to "spring
hooks" to many others I don't really even know how to call. I
got a bunch and I'll never get tired to look for more and
diverse ones. I absolutely think there's never enough tools for
probing at a bench. My "silver bullet" is this: ,
but even this is far too large for some low clearance
situations. But its length allows me to reach depths nothing
else I have here comes close. Everything else is up to about 3"
reach.?
?
Sometimes it comes down to having extender cards, or made
ejecting a card, attaching (including soldering) a wire and then
probing it once the card is back in there. But that's not always
possible. Especially with microwave or mm modules, I am weary of
disturbing them too much, and there's always more mechanical
aspects to messing with them.?
?
What are some creative and crafty solutions others have found
useful? Maybe some miraculous tools you've found??
?
Thank you,
Radu.?
I had a NAND flash overwrite problem in one of my designs some 10
years ago. Mostly BGA chips, so lots of via's.
A connection was needed to the TEK logic analyzer cables.
See the solution below. Wire is transformer wire, isolation melts at
the soldering iron tip.
The bottom row (GND pins) was soldered to the GND plane.
After a few days of waiting for the bad write it was found to be a
pointer problem caused by a 3rd party library.
Arie
|
The Metcal units, at least the two Mx500 I have in my shack, have a conventional transformer. No toroid, no SMPSU. The extruded aluminium housing does not shield the stray magnetic field.
Another fun thing is that some 13.5 MHz RF gets leaked by their cables. Can be fun if you wonder were that QRM comes from..
Other than that: brilliant kit those Metcal.
Wilko
|
Re: Free to good home, HP 5423A system + manuals and spares
On 9/19/24 15:29, Rik Bos wrote: Some info:
I have working copies of the firmware tapes and can duplicate them to 3M QIC80 tapes.
With some minor modification you can let the 5423/5420 system work with QIC80 tapes.
The systems are build around a HP 1000 M-processor with 48/64/128kW memory installed.
Every time you cold start the system it will load it's OS / Firmware from tape, with the original HP tapes that won't work anymore.
That's why I modified my system and copied the firmware tapes to QIC80 tapes, in my case 3M DC2120.
If you need a working copy of this let me know. Hi Rik, yes, I would like to have copies, thank you. I was aware of the HP-1000 in the box; that's one of the reasons I'm interested in these. I would like to perform the QIC80 conversion on this unit. Someone did quite a lot of work on these several years ago; I can't find the web page offhand right now, was that you? -Dave -- Dave McGuire, AK4HZ New Kensington, PA
|
Re: Transferring waveforms from 54542a through GPIB
To solve the re=trigger problem, I put my scope in SINGLE sweep mode, that way, the scope does not lose the captured image that you want.
You don't say what operating system that you are running on your PC. I have a command line tool set that I wrote to control the scope and download all active channels to a .CSV file for postprocessing. The tool is written for Linux, but it can be compiled to run under Windows, with some effort.
The Linux version uses the open source linux-gpib package to deal with the interface. That package works with a number of GPIB interface boards.
Source code and build instructions can be found at:
The windows version uses NI libraries and assumes an NI GPIB interface. There is one Windows file that is needed, but I don't include it because it is copywrited by Microsoft.
I write this tool to use my 54542A as a data acquisition system for another project that I was working on.
|
On 9/19/24 21:59, Harvey White wrote: With the Linux heritage, IIRC. Oh good heavens no; emacs is way older than that, and it did not originate in the UNIX world. It dates back to the mid-1970s, on DEC PDP-10s. It has been ported to, or implemented on, every interactive OS on every platform that I can think of, including every functional implementation of UNIX. ?I did (and do) a lot of IDEs, unlike people who have a separate editor, separate compiler, and a makefile (I did something like that on a DG Nova computer, it's what the company had.) I don't use IDEs as most people think of them. But when developing software (mostly firmware), I never have to leave the emacs window(s). Editing, compiling, target programming. That by definition is an "IDE", but it's not what most people think of when they say "IDE". So I know of it, but never used it in its native form.? Perhaps some other editors were derived from it, but what I did?? Not that I'd know of. You'd know. ;) Dozens of editors were derived from it, some free, some commercial. But you'd definitely know. -Dave -- Dave McGuire, AK4HZ New Kensington, PA
|
With the Linux heritage, IIRC.
?I did (and do) a lot of IDEs, unlike people who have a separate editor, separate compiler, and a makefile (I did something like that on a DG Nova computer, it's what the company had.)
So I know of it, but never used it in its native form.? Perhaps some other editors were derived from it, but what I did?? Not that I'd know of.
Harvey
toggle quoted message
Show quoted text
On 9/19/2024 9:40 PM, Dave McGuire wrote: 'Twas a joke. But emacs has to be one of the most widely-ported pieces of software in history.
-Dave
On September 19, 2024 8:23:47 PM "Harvey White" <madyn@...> wrote:
Never used it, so... wasn't available for what I was doing.
Harvey
On 9/19/2024 7:57 PM, Dave McGuire wrote:
Emacs!
On 9/19/24 19:53, Harvey White wrote:
Unless you like VHDL...... <grin>
Harvey
On 9/19/2024 5:29 PM, Dave Daniel wrote:
Add in SystemVerilog if one is contemplating FPGA/ASIC work.
DaveD
KC0WJN
Thanks for all the fish.
==============================
All spelling mistakes are the responsibilty of the reader (Rick
Renz, STK, ca. 1994)
==============================
On Sep 19, 2024, at 16:03, Roy Thistle via groups.io
<roy.thistle@...> wrote:
?
On Thu, Sep 19, 2024 at 09:11 AM, Frank Mashockie wrote:
I've been considering going into an online master's program for
EE.
If your boss is not paying for it... It's expensive.
2 to 3 years to complete, on average.
Say 50,000USD total cost, on average... and up... for many good US
schools.
Georgia Tech is... about the lowest cost... 10,000USD total cost.
Everyone is different... but IMO... the average person... depending
on which specialty they want... needs to have a strong background in
calculus... as in vector, complex, and differential equations...
linear algebra... numerical methods... statistics.
Computer science skills are required too.
Programming in Python or C
Simulations in software packages like Pspice
HDL like Verilog, or VHDL
Not to mention a strong background in
circuit analysis
devices
probably other stuff I forgot too.
--
Dave McGuire, AK4HZ
New Kensington, PA
|
'Twas a joke. But emacs has to be one of the most widely-ported pieces of software in history. -Dave On September 19, 2024 8:23:47 PM "Harvey White" <madyn@...> wrote: Never used it, so... wasn't available for what I was doing.
Harvey
On 9/19/2024 7:57 PM, Dave McGuire wrote:
Emacs!
On 9/19/24 19:53, Harvey White wrote:
Unless you like VHDL...... <grin>
Harvey
On 9/19/2024 5:29 PM, Dave Daniel wrote:
Add in SystemVerilog if one is contemplating FPGA/ASIC work.
DaveD
KC0WJN
Thanks for all the fish.
==============================
All spelling mistakes are the responsibilty of the reader (Rick
Renz, STK, ca. 1994)
==============================
On Sep 19, 2024, at 16:03, Roy Thistle via groups.io
<roy.thistle@...> wrote:
?
On Thu, Sep 19, 2024 at 09:11 AM, Frank Mashockie wrote:
I've been considering going into an online master's program for
EE.
If your boss is not paying for it... It's expensive.
2 to 3 years to complete, on average.
Say 50,000USD total cost, on average... and up... for many good US
schools.
Georgia Tech is... about the lowest cost... 10,000USD total cost.
Everyone is different... but IMO... the average person... depending
on which specialty they want... needs to have a strong background in
calculus... as in vector, complex, and differential equations...
linear algebra... numerical methods... statistics.
Computer science skills are required too.
Programming in Python or C
Simulations in software packages like Pspice
HDL like Verilog, or VHDL
Not to mention a strong background in
circuit analysis
devices
probably other stuff I forgot too.
-- Dave McGuire, AK4HZ New Kensington, PA
|
Re: HP 141T Power Supply Help Needed
Yep, seems to be much happier after flipping a few knobs.? ?I'm "in the process" of unsticking the tuning knob, but once it started moving more then I was able to see more exciting things.? ? With a telescoping antenna attached, I let it sit for an hour centered on 100mhz for a while, and i'm pretty confident I was watching the broadcast stations there.? ? I wrapped a loop of? wire around that antenna and took it to an old sig gen and i made a line wiggle.? ??
Time to settle down with a cold one and the manual to celebrate!? ?
The knob is still pretty tight, if anyone has any low touch suggestions to get that going i'd be much obliged.? ?
Really appreciate all the?support from the group as well!? ?I could have binned it so many times in the past month, but here we are.? ?
That display is normal for an 8553B in FULL SCAN mode.? The dip is the marker that indicates the center frequency.
Dave Wise
Look at that. The only issue I see is that the display is upside down.
I guess you need to flip it over. Humor.
By the looks of it you have some signal. But your IF BW may be narrow.
Typically with a wider bandwidth you would see noise in the floor.
Or its telling you gain is set low something silly like that. Front panel stuff.
The 141T is a nice spectrum analyzer. It was the first one I used in the Navy circa 1975. Liked it a lot. But never owned one. Have newer units.
Paul
WB8TSL
|
Re: HP 141T Power Supply Help Needed
That display is normal for an 8553B in FULL SCAN mode.? The dip is the marker that indicates the center frequency.
Dave Wise
toggle quoted message
Show quoted text
Look at that. The only issue I see is that the display is upside down.
I guess you need to flip it over. Humor.
By the looks of it you have some signal. But your IF BW may be narrow.
Typically with a wider bandwidth you would see noise in the floor.
Or its telling you gain is set low something silly like that. Front panel stuff.
The 141T is a nice spectrum analyzer. It was the first one I used in the Navy circa 1975. Liked it a lot. But never owned one. Have newer units.
Paul
WB8TSL
|
Re: Probing deep and/or narrow into a given unit for troubleshooting
Depends on what the design is:
If I design it, I put in test points or small (0.05 inch) headers for logic analyzers (ditto for the cheapie clones).? Some boards have 0.1 inch headers that go to HP logic pods.
Sometimes a single pin or a hook terminal works.
I have breakout boards for projects that stack boards.? Those have logic analyzer breakouts.
For someone else's designs, you have dip clips of all varieties (go to hamfests, perhaps?), breakout boards as extenders that have lots of test points.? You have standard grabber hooks and sometimes the micrograbbers.? Grabbers and micrograbbers are rather fragile, though.
Self designed generic breakout boards can work with ribbon cables.
Harvey
toggle quoted message
Show quoted text
On 9/19/2024 5:34 PM, Radu Bogdan Dicher wrote: Hi all,
I was wondering of asking for thoughts and ideas on probing deep into some units, or maybe just through a narrow clearance. What I'm thinking of is primarily for use with a DMM (vanilla troubleshooting), though maybe scope probes and others could be considered. But the latter is not my priority.
I have a variety of probes, from "witches hats," to "spring hooks" to many others I don't really even know how to call. I got a bunch and I'll never get tired to look for more and diverse ones. I absolutely think there's never enough tools for probing at a bench. My "silver bullet" is this: , but even this is far too large for some low clearance situations. But its length allows me to reach depths nothing else I have here comes close. Everything else is up to about 3" reach.
Sometimes it comes down to having extender cards, or made ejecting a card, attaching (including soldering) a wire and then probing it once the card is back in there. But that's not always possible. Especially with microwave or mm modules, I am weary of disturbing them too much, and there's always more mechanical aspects to messing with them.
What are some creative and crafty solutions others have found useful? Maybe some miraculous tools you've found?
Thank you,
Radu.
|
On 9/19/24 15:53, Radu Bogdan Dicher wrote: when you align FM tuners (uV of signal) or do metrology applications, this noise can be a determining factor. Yeah, I've been thinking of a zero crossing and filtered and shielded and low frequency switcher supply to make and sell for LEDs in labs. The usual products are designed for "don't care" how much EMI, and low cost.
|
Never used it, so... wasn't available for what I was doing.
Harvey
toggle quoted message
Show quoted text
On 9/19/2024 7:57 PM, Dave McGuire wrote:
? Emacs!
On 9/19/24 19:53, Harvey White wrote:
Unless you like VHDL...... <grin>
Harvey
On 9/19/2024 5:29 PM, Dave Daniel wrote:
Add in SystemVerilog if one is contemplating FPGA/ASIC work.
DaveD
KC0WJN
Thanks for all the fish.
==============================
All spelling mistakes are the responsibilty of the reader (Rick Renz, STK, ca. 1994)
==============================
On Sep 19, 2024, at 16:03, Roy Thistle via groups.io <roy.thistle@...> wrote:
?
On Thu, Sep 19, 2024 at 09:11 AM, Frank Mashockie wrote:
??? I've been considering going into an online master's program for EE.
If your boss is not paying for it... It's expensive.
2 to 3 years to complete, on average.
Say 50,000USD total cost, on average... and up... for many good US schools.
Georgia Tech is... about the lowest cost... 10,000USD total cost.
Everyone is different... but IMO... the average person... depending on which? specialty they want... needs to have a strong background in
calculus... as in vector, complex, and differential equations... linear algebra... numerical methods... statistics.
Computer science skills are required too.
Programming in Python or C
Simulations in software packages like Pspice
HDL like Verilog, or VHDL
Not to mention a strong background in
circuit analysis
devices
probably other stuff I forgot too.
|
Use multi-color LEDS positioned so that you can get light on front panels but not glare on scope screens.
Harvey
toggle quoted message
Show quoted text
On 9/19/2024 6:14 PM, Jim via groups.io wrote: I agree with you completely. Over 50 awesome messages and a wealth of ideas. I also liked seeing the LED "mood" lighting and the flat screen monitors incorporated into some benches
Jim
|
Warm and friendly includes a well lit recliner and a table for a cup of coffee or tea.
Harvey
toggle quoted message
Show quoted text
On 9/19/2024 5:53 PM, Radu Bogdan Dicher wrote: As a general observation - how incredibly?useful this conversation has turned out. As with just about any conversations here. All this is making me think and rethink my bench and workspace configuration. Some great ideas, best practices, "do nots," etc.
I personally think I have a great "working model" - and some visiting friends were very positive of what I did with my limited space, so I may have hopefully?inspired?others (they said as much) - but there's always a lot of "room" for improvement (no pun intended).
One criteria that is very important to me is how warm and friendly this space is. I'd absolutely not trivialize that aspect. To most of us, I think it's just this "safe space" where we feel great having fun with our hobby (for all of us, very serious hobby!). A contorted, dysfunctional, hostile space does no one any favors.
Light is another aspect to consider. I am currently using very bright LEDs (there's no such?thing as too much light at the bench!), but EMI and other garbage and noise is a terrible byproduct of that. A whole lot of money and time can be spent on just mitigating that. And when you align FM tuners (uV of signal) or do metrology applications, this noise can be a determining factor.
Radu.
On Thu, Sep 19, 2024 at 2:41?PM Wilko Bulte via groups.io <> <wkb@...> wrote:
As for magnetic influence: keep your Rb reference oscillators away
from magnetic fields. It probably does not show in everyday use
but the Rb units are susceptible to magnetic fields.
Wilko
|
Emacs!
toggle quoted message
Show quoted text
On 9/19/24 19:53, Harvey White wrote: Unless you like VHDL...... <grin>
Harvey
On 9/19/2024 5:29 PM, Dave Daniel wrote:
Add in SystemVerilog if one is contemplating FPGA/ASIC work.
DaveD
KC0WJN
Thanks for all the fish.
==============================
All spelling mistakes are the responsibilty of the reader (Rick Renz, STK, ca. 1994)
==============================
On Sep 19, 2024, at 16:03, Roy Thistle via groups.io <roy.thistle@...> wrote:
?
On Thu, Sep 19, 2024 at 09:11 AM, Frank Mashockie wrote:
??? I've been considering going into an online master's program for EE.
If your boss is not paying for it... It's expensive.
2 to 3 years to complete, on average.
Say 50,000USD total cost, on average... and up... for many good US schools.
Georgia Tech is... about the lowest cost... 10,000USD total cost.
Everyone is different... but IMO... the average person... depending on which? specialty they want... needs to have a strong background in
calculus... as in vector, complex, and differential equations... linear algebra... numerical methods... statistics.
Computer science skills are required too.
Programming in Python or C
Simulations in software packages like Pspice
HDL like Verilog, or VHDL
Not to mention a strong background in
circuit analysis
devices
probably other stuff I forgot too.
--
Dave McGuire, AK4HZ
New Kensington, PA
|
Unless you like VHDL...... <grin>
Harvey
toggle quoted message
Show quoted text
On 9/19/2024 5:29 PM, Dave Daniel wrote: Add in SystemVerilog if one is contemplating FPGA/ASIC work.
DaveD
KC0WJN
Thanks for all the fish.
==============================
All spelling mistakes are the responsibilty of the reader (Rick Renz, STK, ca. 1994)
==============================
On Sep 19, 2024, at 16:03, Roy Thistle via groups.io <roy.thistle@...> wrote:
?
On Thu, Sep 19, 2024 at 09:11 AM, Frank Mashockie wrote:
I've been considering going into an online master's program for EE.
If your boss is not paying for it... It's expensive.
2 to 3 years to complete, on average.
Say 50,000USD total cost, on average... and up... for many good US schools.
Georgia Tech is... about the lowest cost... 10,000USD total cost.
Everyone is different... but IMO... the average person... depending on which? specialty they want... needs to have a strong background in
calculus... as in vector, complex, and differential equations...? linear algebra... numerical methods... statistics.
Computer science skills are required too.
Programming in Python or C
Simulations in software packages like Pspice
HDL like Verilog, or VHDL
Not to mention a strong background in
circuit analysis
devices
probably other stuff I forgot too.
|
Argue that a master's degree looks like more highly educated employees.
I have a Master's degree in EE, and do software as well as hardware. I'd suggest taking courses in both.? Depending on what you want to do, Microprocessors and C++ (and possibly assembly, it could be argued) are good.? Python I don't really use, since it's easy enough to code in C++ (for me).? CAD?? Definitely if you want to do something with it, I could suggest looking at OpenSCAD as an interesting system.
C++ allows both processor and windows (and Linux, I'd say) applications.
You might want to see if you are going to be interested in either the system level programming (applications support) or applications.? All of the above are interesting.
Harvey
toggle quoted message
Show quoted text
On 9/19/2024 5:03 PM, Frank Mashockie wrote: Thanks for the feedback Roy. I really appreciate it. So I have a solid foundation of math/physics through my chemistry degree. Calc III, dif equations, quantum chemistry. The math I'm not worried about. I also feel comfortable in using most of the hardware an EE student might come in contact with. Software is a different story. I have zero experience with python or CAD. But I know I can learn. My company does subsidize tuition. However, I might fall into a grey area. We are a biotech company so I'm not directly making the company money but I am keeping their costs down. I could argue a degree in EE could translate into better understanding of the equipment which could in turn mean more costs savings. But you bring up excellent points! -Frank
|
Re: HP 141T Power Supply Help Needed
Look at that. The only issue I see is that the display is upside down. I guess you need to flip it over. Humor. By the looks of it you have some signal. But your IF BW may be narrow. Typically with a wider bandwidth you would see noise in the floor. Or its telling you gain is set low something silly like that. Front panel stuff. The 141T is a nice spectrum analyzer. It was the first one I used in the Navy circa 1975. Liked it a lot. But never owned one. Have newer units. Paul WB8TSL
|
Re: Probing deep and/or narrow into a given unit for troubleshooting
I have done the tack-wires-on method many times too. A few nodes is easy to handle, with leads held safely with tape to keep from shorting. When there are lots, I usually use some kind of connector like a square pin header or plug at the ends to keep them all together and safer and organized. Then probe each with a test lead including a pin or socket that can stay connected if needed.
?
I happen to have in my pile of test leads the longest grabber hook I've ever seen, and I'd think even longer might be available. I got it decades ago along with other junk, and never knew who made it until just looking now. This one is around five inches from the finger hold to tip, labelled "maxi grabber." I looked up that name, and it seems to be a Pomona product - you'll see all sorts of hits with that. I think this is their model 4225-2, where the "2" is probably the code for red.
?
I've actually seldom used it except on bigger old-school circuitry. Even with longer reach, there are still the problems of manipulation to grab something, and visibility, so you'd still likely have to tack on some kind of node extensions to probe say in a card cage. It's also kind of fat, but good and strong - maybe there are skinnier, lower profile types too.
?
Ed
|
Hi Adri, ? Yes, I am curious, you can send me a video on the intermittent noise problem, by going through OneDrive for example, or Dropbox. It could be that one of the PLLs is the cause. ? For the 8595E on which I was trying to repair, the problem was not on the 130 option as the error message seemed to indicate, but rather an incorrect alignment on the basic filters. The AMP CAL stopped because it was unable to find the two -3dB points of a filter. ? Yes, I realigned all the filters following the procedure in the service manual. There are 4 crystal filter sections and 4 LC filters. Simple procedure but a bit long, and you must not make any mistakes, otherwise you have to start from the beginning. I did the job slowly; everything went well and the original problem was solved. I recalibrated the instrument and took the opportunity to improve my software for flatness alignment, now after the test, all data is entered automatically...which speeds up the work. ? In my opinion, someone opened this 8595E to try to make a repair, the second converter had been opened, 2 screws were missing and 2 others were reset to 50%. I had to realign the second converter after closing it correctly. ? For the filter section, it is clear that one touched the adjustments, because half were well aligned, and for the others, the adjustments were very far from the optimal value. ? In total, 8 screws were missing on this instrument. Probably an interrupted repair¡ ? Yves ? ? hello Yves, great repair that of the opt 130
In my case the opt 130 has always worked well and I have never needed to study it in detail
Did you do the realignment by adjusting the trimmers on the board itself?
The adventure with my 8595E continues
To sum up, after my last email I focused on a more accurate and uniform alignment of the SYTF. With the help of the microscope, a strong light source, a precision power supply and a fairly reliable frequency source I obtained an insertion loss as per the attached excel file
After reassembling the RF front-end everything worked fine again, including the MXR CAL routine that?used to?give the?error? "insufficient harmonics"
When I moved from the "messy" version (pic)? reinserting everything into its space (which is very crowded, also due to the existence, in my case, of the tracking gen)?and connected?the SA input to CAL OUT I observed a type of intermittent noise, in the form of spikes of intensity about -60dBm that appear and disappear in the area of ??the highest harmonics (I already checked the?CAL OUT signal; it?is clean and fine)
Now I think I will have to re-disassemble the front-end (ouch) and try to understand from where does this noise originate :-(? If you are interested I?can send you a short video of the problem (don't think I?can post videos here)
Thanks for asking, Yves
Adri
|
Arie de Muijnck
Dave McGuire