¿ªÔÆÌåÓý

Date

Re: Looking for HP rotary encoder datasheet/info

 

¿ªÔÆÌåÓý

Try a cheap standard one.?

Peter

On Jul 4, 2024, at 5:07?PM, Andy Warner via groups.io <andyw@...> wrote:

?
I have an HP 1672E logic analyser that works great apart from the shaft of the front panel rotary encoder is broken (and not really repairable.)

The device in question is a QRPG-A482, and is listed as a rotary pulse encoder.
I have verified that it does indeed produce pulses, and is not a straightforward encoder.

I have spectacularly failed to find either a replacement part, or a datasheet for the device itself.

Given that these are unobtainium, I am currently planning to build a little circuit that takes a standard rotary encoder and synthesizes appropriate?pulses. Looking at the QRPG-A482 outputs, it looks like there are simply two pulses whose relative phase determines?the direction of rotation. The QRPG-A482 appears to have variable timing. I doubt very much that the timing matters and so I can likely just create pulses that look similar to the fastest ones I captured (approx 1mS wide.)

Having a datasheet for the original part might help me avoid any subtleties - even advice such as whether the outputs are likely to be open drain/collector would be appreciated.

I wonder if anyone here has travelled this path before me, and has any advice to offer.
--
Andy


Looking for HP rotary encoder datasheet/info

 

I have an HP 1672E logic analyser that works great apart from the shaft of the front panel rotary encoder is broken (and not really repairable.)

The device in question is a QRPG-A482, and is listed as a rotary pulse encoder.
I have verified that it does indeed produce pulses, and is not a straightforward encoder.

I have spectacularly failed to find either a replacement part, or a datasheet for the device itself.

Given that these are unobtainium, I am currently planning to build a little circuit that takes a standard rotary encoder and synthesizes appropriate?pulses. Looking at the QRPG-A482 outputs, it looks like there are simply two pulses whose relative phase determines?the direction of rotation. The QRPG-A482 appears to have variable timing. I doubt very much that the timing matters and so I can likely just create pulses that look similar to the fastest ones I captured (approx 1mS wide.)

Having a datasheet for the original part might help me avoid any subtleties - even advice such as whether the outputs are likely to be open drain/collector would be appreciated.

I wonder if anyone here has travelled this path before me, and has any advice to offer.
--
Andy


Fault-tracing with KSR

 

My HP spectum analyzer (8566B) has the KSR feature built in which is intended to make troubleshooting easier.? Essentially, KSR will display a load of frequency values on screen. Now this could be enormously helpful, but in the service manual, it is not precisely explained how much reliance one can place on the displayed values. AISI, there are 2 main possibilities here:

1) KSR is displaying a list of what would be the correct frequencies to expect if the analyzer is functioning properly and it is for the serviceperson to check that the displayed on-screen values coincide with the values actually measured.

2) KSR has its own built-in frequency counter(s) and is displaying the actual output of the various modules in real-time.

Now my question is: is it 1) or 2) above, or something else?

Many thanks,

J.


Re: Spline or Bristol Set Screws

 



Although the 10 pcs Spline key set on eBay seems the least expensive;

Direct from Bristol Wrench Co.
-check the sizes that you might need:

?9-Key Set: $25


10-Key Set: $35
https://bristolwrench.com/product/spline-short-arm-10-key-kit-ss-408/

13-Key Set: S150? ( yikes !! ), etc.


Re: A stupid problem is jamming me (8443A, but is rather general)

 

It's fairly common for pot wipers to get intermittent after long times. You can replace the pot or try to fix it in place. You can usually clean the wiper by running the pot end to end lots of times. It helps to also have a load to provide some current (like a few mA) through the contact - just the Q's base won't take very much when the circuit is active. If lots of mechanical working doesn't clean it enough, it may be easier to take out the whole board and rig it so you get some cleaning current through the pot, from an external low voltage source. You can tack up a temporary circuit for this. Just avoid reverse-biasing the Q's B-E junction beyond a few volts.

Ed


Re: Spline or Bristol Set Screws

 


For??? Collins??? &??? Hallicrafters??? Spline key knobs:

"? GENIUNE USA MADE 10 PIECE BRISTOL WRENCH SET FOR COLLINS & HALLICRAFTERS KNOBS "

10 pcs Bristol spline set on eBay which might (?)? be less expensive than the individual McMaster Spline keys:



$32.99 with shipping @$5.
( no affiliation)



Re: Help needed programming 34972A Data Acquisition/Switch Unit

 

Been playing around a bit tonight, and found something...

When I apply my config file from the USB stick and press scan, it all works as expected, but I never get the 'Scanning...' display as is apparently expected, only the 'mux VDC' as shown before the scan started.

So I made a simple 3 channel program from the front panel set to automatically loop a few times and found it doesn't say 'Scanning' on the first run through, but does so on subsequent scans. It also does the same no matter how I load the program into the instrument, via PC connection, USB memory stick or front panel.
As my program runs only once through at a time, it never shows 'Scanning' as it never gets to a second scan. I wonder why it never says 'Scanning' on the first run-through? A bug or my misconfiguration somewhere? (I'm running v1.17 firmware, the latest I think?)


Re: 70420A based phase noise test system

 

The highlights of my journey were as follows:

If you¡¯re considering a 70000 MMS SA for wideband phase noise measurement, a 70900A LO won¡¯t work, or at least not with whatever software version mine had. Mine with a 70900B worked fine.

the E1430A VXI ADC is on the supported hardware list for E550xA systems, but does not work. It doesn¡¯t matter if you¡¯re using MXI-2 or FireWire VXI bridges to a host PC. The hardware is seen, the soft front panel is grayed out.

The E1437A I haven¡¯t physically had one of to test, but a guy in France did have one he never got working either.

The E5504B systems use an NI PCI6111 digitizer card, with a specific module on it that I don¡¯t recall the part number of at the moment. The adapter on the back just provides shielded cable connections with presumably low pass filters.

I haven¡¯t tried the NI PCI digitizer card, because at the time I could buy an 89410A or even a whole 89441A for less than the $800+ the PCI card showed in sold items history.

The original install CDs for the E5500 are something I¡¯ve never laid my hands on. Within them, there may be special drivers or support libraries necessary to make the NI PCI card actually work. The NI directories on the CD likely contain a necessary dependency for the E1430/E1437 digitizer to work as well.

I¡¯d be very interested if anyone has an original install CD or a CD image for E5500 systems.

I¡¯d recommend an 89410A as the analog front end unless you already own one of the other supported FFT analyzers(3561/3582).

Even if you have a 3561 or 3582 you might want to consider the impact of 100KHz or 25KHz vs 25.6MHz of an 89410A sample rate and the impact on measurement time for close in phase noise. Since the DUT and references all have their own phase noise, beyond a certain measurement time, the time alone may limit noise floor.

Moving on to software¡­

The ¡°old¡± windows 7 and lower software is 32 bit only. If you install in a 64 bit system you¡¯re limited to viewer functionality only. The licensing is based on your test set serial number. Keysight has a help article out there with a generic key for test set serial number 0000A00000 and resetting the test set serial number is east. To install the software, you¡¯ll need the pre-requisites bundle and the latest software bundle.

The new software is W10 officially(probably will install fine in W11) and 64 bit only. It uses host based licensing.


1st generation ISA system with probably windows NT 4.0 and a no longer supported ISA digitizer card
2nd generation windows 7 32 bit, 2U rack mount, GPIB card and digitizer card lay parallel to motherboard
3rd generation windows 10 64 bit, Kontron 4U rack mount PC with cards standing up.

For E550xB systems, there were 3 generations of PCs they supplied:
1st generation ISA system with probably windows NT 4.0 and a no longer supported ISA digitizer card
2nd generation windows 7 32 bit, 2U rack mount, GPIB card and digitizer card lay parallel to motherboard
3rd generation windows 10 64 bit, Kontron 4U rack mount PC with cards standing up.

4th generation PXIe used a 4 slot wide embedded windows controller in a mainframe holding a digitizer with FPGA Hardware and phase detector(s) cards.

Based on several inquiries over more than a year with Keysight, I¡¯ve run into a hard wall with respect to licensing for the W10+ software.

As near as I can tell, support can probably retrieve previously issued licenses, but generating a new license has to go through manufacturing.

I have no idea if there¡¯s software automation that prevents issuance of a new key, or corporate policy that will NOT be violated¡­but a system serial number or sales order number to look up the system serial number is what they require.

Anyway, that means piecing together a PCIe frame even with only the widows embedded controller is a dead end unless you¡¯re sure your frame has a system serial number on it.

I did get the impression since this is keyed off the system serial number that you could supply a different host ID for your board if you replaced a failing one. Being they may be expecting a Kontron board with a specific number of digits not just any board might do trying to build up a system with hopes of getting a license.

A best case scenario would be figuring out how to get Keysight to generate a $0 sales order (or some nominal fee affordable for an individual not using a company checkbook) and maybe getting the system serial number for the test set assigned as an E550xA upgrade. Being they can no longer sell a windows 10 3rd generation system, as components are no longer available, they¡¯re not as risk of losing revenue. Maybe the right person can find a way, but I haven¡¯t found them yet. If you do get this far, you¡¯d need to make sure your license includes LTUs for ever piece in your system, and might as well make sure it supports upgrades.

And finally the last topic:
What tunable reference (signal generator) you use matters because from bad to good to excellent, at say 1/1k/10k/100k there¡¯s probably at least a 30-40db improvement even ignoring what your intentions are for wideband measurement.

In my system, based on cost, I used a non-HP generator that cost me at least $1000 less than a comparable HP.

a few final thoughts:

When ¡°systems¡± are being advertised, make sure you understand exactly what a system includes. A test set + computer certainly isn¡¯t all you need, nor is it necessarily false advertising. You just need lots of extra parts to make a useful instrument.

If a ¡°system¡± is supplied with a computer, consider utility of what said computer will be whether that computer is serviceable or not.

To me this is a perfect case for right to repair and highlighting all of the problems with licensing¡­.being unable to upgrade hardware to something on my own with no support from the company to be able to run an OS for which security patches are available is a problem. I¡¯d even like to say the overall pattern is a problem for national security. Maybe someday I¡¯ll find some people in congress that will be in a mood to put some exceptions in place to carve outs industries lobbied for.


The latest 64 bit software if you extract pieces from the installer actually still supports or appears to support using an external FFT analyzer and the older hardware even though it was designed for the newer AXIe chassis.

In principle, the 32 bit software should be able to run on 64 bit system, but the installers are broken in ways that prevent doing that directly. I haven¡¯t had time lately, due to work, to revisit that and seeing if I can build a new installer with 64 bit support or a mix of 32 and 64 bit dependencies to install on 64 bit only.


Re: Spline or Bristol Set Screws

 

A word of warning regarding Spline keys, Bristol ones are not the only types of Spline keys used.

I worked in the US in the early 1970¡¯s and know that Honeywell used a type all of their own on chart recorders as did the manufacturers of multi turn potentiometer dials.

G Edmonds

On Thursday 4 July 2024 at 16:38:08 BST, gren <hrgerson@...> wrote:



There is a 10 pcs Bristol spline set on eBay which might (?)? be less expensive than the individual McMaster Spline keys:



$32.99 with shipping @$5.
( no affiliation)

Uncertain if this set is cheaper than McMaster-Carr individual Spline key prices,
or if McMaster offers a full Spline Set ( > 10 keys ) at a lower cost ?

Would be interested in knowing..
--? eventually I will need a set to repair a Collins ART-13

rick


Re: Spline or Bristol Set Screws

 


There is a 10 pcs Bristol spline set on eBay which might (?)? be less expensive than the individual McMaster Spline keys:



$32.99 with shipping @$5.
( no affiliation)

Uncertain if this set is cheaper than McMaster-Carr individual Spline key prices,
or if McMaster offers a full Spline Set ( > 10 keys ) at a lower cost ?

Would be interested in knowing..
--? eventually I will need a set to repair a Collins ART-13

rick


Re: KE5FX GPIB Toolkit

 

I'm running VirtualBox on Linux Mint 21.? It didn't really require any special messing around to get the NI USB adapter recognized using the usual VBox hardware capture procedure.

John
On Jul 4, 2024, at 12:00 AM, "Tim Tuck via " <skybase.net@groups.io target=_blank>[email protected]> wrote:

Hi John,

I'm about to embark down this path and I was wondering if you are using
KVM or VirtualBox or something else.

thanks

Tim

On 4/7/24 08:45, John Ackermann N8UR via wrote:
I'm running KE5FX tools using an NI USB adapter on a Windows 10 VM
under Linux with no problems at all.

John







Re: Multiple VISA packages on the same computer?....

 

Dave,
That's a great idea (though my IT/IS days are long in the past... I'd need to figure out doing this in my environment). In fact, the multiple partitions were such that I'd be able to run both Windows 8 and XP at that time, as I could only use the AP PCI interface on XP. I've upgraded since to an APIB-to-USB adapter and so I can do all of it under WIndows 10 at this point. Anyway, I digress.?
I'd prefer, if possible, though, to run all of this from the same system (VM or not). Switching between VMs to access different instruments would be pretty inconvenient. But what is an interesting scenario is creating a VM and rebuilding everything on it, this time with the benefit of hindsight.?
Rolling back the project here though, it sounds like running multiple VISA environments on the same machine is not an option. Also, that non-destructively replacing the R&S VISA with the NI is not a simple operation either.?
Thank you,
Radu.?

On Thu, Jul 4, 2024 at 7:12?AM Dave_G0WBX via <g8kbvdave=[email protected]> wrote:
At least with a VM, subject to enough local storage, you can clone the entire VM (as a backup) after each new addition. Giving you a "get out of jail free card" if the next addition goes bad.

Regards.

Dave B





Re: 70420A based phase noise test system

 

Not at all. But if they do, we can move it to the HP 70000 list. ;)

Bob

On Jul 4, 2024, at 10:09, Dave McGuire via groups.io <mcguire@...> wrote:

?
I think it's safe to say that no one will mind such discussion here on the list. I myself would certainly like to follow along.

-Dave

On 7/3/24 22:50, Ed Marciniak wrote:
I put together a working E5504 system(26.5GHz variant) and I¡¯d be happy to take you though all of the hardware and software choices. There's "supported" hardware on their list that doesn't actually work.
Send me a PM with your phone number and I'd be happy to call on my dime.
Should anyone else be interested, I'd be happy to do the same or we could reply all in a small group format.
73,
Ed
NB0M
------------------------------------------------------------------------
*From:* [email protected] <[email protected]> on behalf of Jason <jason0m@...>
*Sent:* Wednesday, July 3, 2024 7:54:24 PM
*To:* [email protected] <[email protected]>
*Subject:* [HP-Agilent-Keysight-equipment] 70420A based phase noise test system
Hello,
I have started on a project to put together a microwave phase noise test system based on a 70420a, which Hewlett-Agi-sight called the E5500 series. I am curious if any other hobbyists have assembled one of these systems before and have any advice. My interest is as much in teaching myself how phase noise measurement works as it is to yield a working system, and I expect this to be a long term project. My previous experience with phase noise measurement is limited to software running on a spectrum/signal analyzer.
So far I have collected a 70420A opt 001 test set and a 70422a downconverter, plus most of the basics for a 70000 system (70004a and 70001a; parts for diy hp-msib cables are in the mail). The last major part that I have to decide which way to go, then find, is the DSA or Digitizer. I plan to use my e4435b as the ref source, at least until i get through the learning curve. The somewhat questionable N9020a discussed some weeks ago on a different thread will be the spectrum analyzer.
The biggest question I have concerns the software. My initial impression is that it will not be possible to run the original application, and that i will be signing up to reimplement a limited version of it, for just the tests that i want to perform. Has someone already done this?
Otherwise, i'm just curious to hear from anyone else who has experience with these systems and may have any advice as i figure out how to put this system together.
Thanks in advance,
-Jason
--
Dave McGuire, AK4HZ
New Kensington, PA






Re: Multiple VISA packages on the same computer?....

 

Sounds like you need a good whole drive rolling backup facility with that lot.

Dave B


Re: Multiple VISA packages on the same computer?....

 

At least with a VM, subject to enough local storage, you can clone the entire VM (as a backup) after each new addition. Giving you a "get out of jail free card" if the next addition goes bad.

Regards.

Dave B





Re: 70420A based phase noise test system

 

I think it's safe to say that no one will mind such discussion here on the list. I myself would certainly like to follow along.

-Dave

On 7/3/24 22:50, Ed Marciniak wrote:
I put together a working E5504 system(26.5GHz variant) and I¡¯d be happy to take you though all of the hardware and software choices. There's "supported" hardware on their list that doesn't actually work.
Send me a PM with your phone number and I'd be happy to call on my dime.
Should anyone else be interested, I'd be happy to? do the same or we could reply all in a small group format.
73,
Ed
NB0M
------------------------------------------------------------------------
*From:* [email protected] <[email protected]> on behalf of Jason <jason0m@...>
*Sent:* Wednesday, July 3, 2024 7:54:24 PM
*To:* [email protected] <[email protected]>
*Subject:* [HP-Agilent-Keysight-equipment] 70420A based phase noise test system
Hello,
I have started on a project to put together a microwave phase noise test system based on a 70420a, which Hewlett-Agi-sight called the E5500 series.? I am curious if any other hobbyists have assembled one of these systems before and have any advice.? My interest is as much in teaching myself how phase noise measurement works as it is to yield a working system, and I expect this to be a long term project.? My previous experience with phase noise measurement is limited to software running on a spectrum/signal analyzer.
So far I have collected a 70420A opt 001 test set and a 70422a downconverter, plus most of the basics for a 70000 system (70004a and 70001a; parts for diy hp-msib cables are in the mail).? The last major part that I have to decide which way to go, then find, is the DSA or Digitizer.? I plan to use my e4435b as the ref source, at least until i get through the learning curve.? The somewhat questionable N9020a discussed some weeks ago on a different thread will be the spectrum analyzer.
The biggest question I have concerns the software.? My initial impression is that it will not be possible to run the original application, and that i will be signing up to reimplement a limited version of it, for just the tests that i want to perform. Has someone already done this?
Otherwise, i'm just curious to hear from anyone else who has experience with these systems and may have any advice as i figure out how to put this system together.
Thanks in advance,
-Jason
--
Dave McGuire, AK4HZ
New Kensington, PA


Re: 8510C operation using NI usb GPIB control device and Pyvisa

 

Well done Jason.

Dave B


Re: KE5FX GPIB Toolkit

 

From memory, I think NI Max will find non SCPI & non IEEE488.2 instruments, but just shows that "there is something" at that address, but cant identify it.

Not used WINE for an age now, but yes, serial ports can be mapped to COM ports, but I think that's about all.

No reason (I think) why a Prologix (or clone) adapter should not work with Wine.? But some software is fussy about its ID string, and/or, some other functionality of the virtual COM port.

73.

Dave B.


Re: A stupid problem is jamming me (8443A, but is rather general)

 

STOP, I GOT IT!

A difficult test for you: I checked every capacitor, resistor, diode, transistor, SCR¡­ but not trimmer R50.

Reckon: where was the problem?


Re: A stupid problem is jamming me (8443A, but is rather general)