Tektronix purchased Keithley, broke my instruments?

See "Cheap homemade 30 MHz - 6 GHz vector network analyzer"

formatting link

It would be easy to extend the range down to 30 Hz.

Reply to
Steve Wilson
Loading thread data ...

People laugh at me still using my mirror galvanometer, but at least I never get caught out by software issues. :-P Seriously, there's much to be said for classic test equipment.

--
This message may be freely reproduced without limit or charge only via  
the Usenet protocol. Reproduction in whole or part through other  
 Click to see the full signature
Reply to
Cursitor Doom

Why??

Reply to
Cursitor Doom

Of all the public designs on the internet, I think that is the best one.

To measure high isolation values (e.g. filters with high rejection in the stop band), some more serious shielding metalwork, and isolation of other coupling paths e.g. via supplies and via LO distribution, will be needed. It would be fun to work on that.

Reply to
Chris Jones

Yes, it looks pretty good. I'd go for a 120 dB log amp converter on the output. One tricky thing is the directional couplers. I'd look to Minicircuits. They have a promo going for components for a VNA. Not cheap.

Reply to
Steve Wilson

The couplers that follow Joel Dunsmore's invention are much more lossy, being resistive bridges, but if you can deal with that, it's easy to get good directionality from 300KHz up to 3GHz just by improving a $12 Chinese special. I know that, because I've done it - data on request I haven't published it yet. Versions of Dunsmore's coupler have been built that work from 100KHz to over 13GHz.

Clifford Heath.

Reply to
Clifford Heath

Thanks. That led to a wealth of downloads on Dunsmore's work. I have seen

6 GHz. So Dunsmore seems to ne the way to go.

BTW, the 120 dB log amp is the AD8307. See "Figure 42. 120 dB Measurement System:

formatting link
sheets/AD8307.pdf

Newer devices are available.

Reply to
Steve Wilson

FWIW I had a very similar problem with an Agilent DMM, the 34410A. This has an embedded web server, but it needs the Java runtime environment installed on the client machine. At some point the version of Java baked into the instrument got blocked by most of the browsers as it was full of security holes, initially you could add specific URLs to a white-list but eventually that was closed as well.

The DMM also talks SCPI over GPIB, so I bought a GPIB-USB adapter from Prologix, apart from the virtual-COM-port driver the only software needed is a serial terminal. Prologix also do a GPIB-Ethernet adapter

formatting link

It's supposed to be possible to use SCPI via the LAN port on the 34410A, but I've never figured out how. The manual is silent on this matter. The recommended way to use the LAN or USB ports (using Agilent's proprietary protocols, not SCPI) involves installing hundreds of megs of bloatware on your PC. No thanks.

If you are happy writing your own scripts to control the instruments this is the way I would go. If you want the 'virtual instrument display' offered by the Web interface this would be a lot more work to replicate.

Reply to
RBlack

Dunsmore's bridge has unequal legs to achieve low loss. One leg has non-ideal ferrite, and the matching leg cannot do anything to cancel that.

Paul McMahon VK3DIP figured out that you can cancel the non-ideal reactance of the ferrite, but it only works for an equal-sided bridge (6dB loss). This product I have is an incorrectly-constructed clone of a Ukrainian version of Paul's design. After fitting the missing link to correct the circuit, and changing the

100R resistor pairs to 50R, replacing the badly-fitted coax to semi-rigid line, etc, it had directivity better than 45dB from 300KHz to 320MHz, 40dB to 600MHz, 30dB to about 1GHz, and still almost 20dB at 2.4GHz. Gotta be pretty happy at that price! Smaller and better geometry (0402 resistors, etc) can push the top end way up - this was poorly constructed on FR4.

Clifford Heath.

Reply to
Clifford Heath

Yes, that's the way I see things too. If the instrument has a LAN port, use it. If the instrument has only GPIB, use a Prologix GPIB-Ethernet adapter and proceed as before. I'm very happy with that little gadget. I used to have a National Instruments GPIB ethernet bridge. That was expensive, required a proprietary driver and associated API library and stopped working after an OS update. Never again.

I write my own scripts. Some instruments have weirdnesses making life a little harder, but I always end up making things work.

Jeroen Belleman

Reply to
Jeroen Belleman

When did this happen ?

Now take that date and remember it.

Now go to the thingie in Windows and find out whenever whatever update happened that coincides with the date of your problem.

Then you can draw your own conclusions.

Want mine ?

Reply to
jurb6006

I'd be interested in some of your designs, then. Could you post some?

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

RBlack wrote in news: snipped-for-privacy@reader80.eternal-september.org:

I still say he is trying to connect thru IPV6, which at one point was set up as the java default, and getting rejected at his ISP.

IPV4 works.

Reply to
DecadentLinuxUserNumeroUno

Log amps are fine if you need to measure changes in power level with an accuracy of maybe 1dB or a bit better, but if you need to measure a change in power level with an accuracy of say 0.01dB then a mixer and ADC is a much better approach.

The reason is this: log amps like the AD8307 work by cascading a lot of amplifier stages and then putting a rectifier on each stage, and summing the outputs of the rectifiers. This works pretty well, but if you plot the output signal (voltage) vs. the input power in dB, then you see a periodic ripple, with a small bump every few dB, spaced according to how much gain per stage is used in the log amp. That is why the AD8307 has a "log conformance +/-0.3dB typical, +/-1.0dB max" specification.

In a manual scalar network analyser where you just plot the response on an oscilloscope, the log amp would be fine (at low frequencies anyway) as you probably don't care to measure the response of a filter better than a dB or so. Without calibration to correct for impedance mismatching on your VNA ports you are not going to be able to check the steps on an attenuator to 0.01dB or stuff like that anyway.

In a VNA where you want to use computer correction and calibration to take out the effects of the cables and imperfect matching, it is necessary to measure the relative amplitude of signals with much better accuracy. e.g. if your connectors produce reflections at -20dB, but you can measure this error and subtract it out digitally, then you can still measure the return loss of a DUT at say -40dB, perhaps with 0.1dB precision or better. This relies on being able to subtract two fairly large signals, and accurately determine the difference between them, which is much smaller. Having 0.3dB of ripple in the transfer function of the log amp really screws up this subtraction. There might be ways to characterise it, but it is temperature dependent too. You could put it in a thermstatic oven, but it is frequency dependent too. A mixer can be more linear with amplitude, provided it is not overdriven on the RF port, for which it helps to drive the LO port really hard with approximately a square wave.

Similarly, you can make a 16-bit audio DAC with 5% resistors, and if you toggle the LSB, the signal will be about 96dB smaller than if you toggle all of the bits together, so you could say that it has 96dB of dynamic range, but you will not like the distortion when you play music through it.

Log amps are also inherently wideband, so they measure the noise in a wide bandwidth, which limits dynamic range. Only by filtering part-way along the log-amp can this be fixed. Note the 10.7MHz bandpass filter in fig. 42, and so that circuit is only good for a 10.7MHz VNA! You will need a mixer for other frequencies, and then you might as well follow it with a nice, high resolution ADC which will give you very precise measurements across a very wide dynamic range.

For a hobbyist VNA, I see no need for a high IF frequency, and isolation on the PCB is easier to achieve at low frequencies. I was thinking of using an IF at a few kHz and an AD7608 ADC, because for VNAs you are interested in the ratio of the channel gains, and the relative phase shifts between receiver channels. Having all ADC channels on the one chip, with simultaneous sampling, and with the anti-aliasing filters on the same chip so that the cutoff frequencies are well matched, should be useful here. Also, there are enough channels to make a 4-port VNA with two mixers per port, using just one ADC chip which I think would help to provide good value for money.

I was intending to try using LT5560 for the mixers. I have a theory that using the two RF input pins separately rather than as a single differential port, would allow two RF signals to be directly subtracted, which is useful if you are building directional bridges with resistor networks rather than making stripline or Dunsmore style directional couplers. Resistive bridges are wideband and very small, which again helps with cost. I got this idea from:

formatting link
though I don't like his idea of driving mixers with simusoidal LO - this makes them more temperature sensitive - nor do I like the idea of mixing using DDS DAC images. Square wave LO and using harmonic mixing to extend the frequency range should give more stable performance vs. temperature. I suspect that with the LO port driven nice and hard with a differential square-ish wave, the LT5560 would mix to a usable extent at the 3rd harmonic of its LO frequency, perhaps extending coverage well above 4GHz.

I got someone to decap some LT5560 and image them for me, so I could check the internal connections - I wanted to drive the LO port differentially, and bias the RF port with DC coupling.

I think the LO distribution to the mixers could be done with something like a ADCLK948, though I need to check how much isolation is needed in the LO path between the ADCLK948 and each mixer. If one mixer is measuring a RF signal 100dB above the signal that another mixer is measuring, it is foreseeable that coupling from the RF port of a mixer to its LO port through the LO traces and between the bondwires of the ADCLK948 and then to the LO port of another mixer and then to its RF port, could limit accuracy. Adding an individual LO buffer for each mixer would be undesirable for cost/space reasons and also because any temperature difference between the separate LO buffers will cause phase skew between receiver channels.

Each mixer would probably want its own screening can. Maybe a Wuerth

36103205 frame and 36003200 lid.

I was intending to put an AD8253 between the output of each mixer and its corresponding ADC channel on the AD7608, because the channel to channel isolation of the AD7608 might otherwise limit dynamic range. The gain steps of the AD8253 are very accurate and stable. Each AD8253 could be inside the same can with its mixer, so that very tiny baseband signals don't have to traverse the whole PCB near bigger ones.

The LO could come from one of the ADF4351 type things, and be fed to one input of the ADCLK948. I haven't looked at the newer ones yet. To go to lower frequencies than the minimum that it can divide to internally, the extra division stages could be implemented in the system fpga, if necessary re-clocked to the ADF4351 output using a NB4L52 and then fed to the other input of the ADCLK948.

The RF stimulus is a more complicated problem, if you care about it being a sine wave. I'm not sure that it is important, but it would be nice. I was thinking of using an ADF4351 or its newer relatives, and a filter bank like the hforsten.com design, however below 50MHz I would instead feed the PLL chip output into the clock of a DDS like the AD9954, using the AD9954 only to generate integer fractions of the PLL frequency. Since the DDS is being used as a divide-by-N, there are no spurs other than harmonics, but at low frequencies (e.g. audio) the output is much closer to a sinewave than if the internal divide-by-2^N of the ADF4351 is used. This might mean that there is no need for filtering at low frequencies. (The AD9954 is nicer than some newer ones because it still has the DACBP pin - see:

formatting link
)

Since this whole PCB might end up running quite warm, I thought it would be worth putting some SMD thermistors on it and SMD heater resistors, to warm it quickly to its final temperature and to slightly ovenise it. The different mixers could have their own separate heaters and thermistors, to keep thermal gradients small or at least constant. A microcontroller could fairly easily run all of those PID loops.

Anyway I got very busy in the last few years so I had no time to make further progress on designing this VNA.

Reply to
Chris Jones

Some thought my browser couldn't reach tek.com. But Java, running the instrument's program, blocked.

OK, it's working now. Modern FireFox & Java no longer works, so I replicated the old environment: A laptop with 32-bit Windows 7, delete Firefox and Java. Then install old FireFox 7: Firefox Setup 7.0.1.exe and a middle-age version of Java 6: jre-6u45-windows-i586.exe Later, when Java asks to update itself, disable updates.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

As of last Sept 18th, Firefox no longer supports NPAPI, required for Java applets in the browser. And none of the 64-bit FireFox versions ever did.

formatting link

--
 Thanks, 
    - Win
Reply to
Winfield Hill

Chris Jones wrote: [...]

That's a very nice description. We'd love to see more like it on SED. This makes it worthwhile to stay.

Thanks, Jeroen Belleman

Reply to
Jeroen Belleman

Have you looked at something like 'LiveConnect'?

formatting link

I run the latest 64-bit version of Firefox on my Mac and have no problem viewing the tek.com web site. I do have java installed on my machine (java version "1.8.0_171") even though it appears (as you point out) that Firefox doesn't use it.

formatting link

Is Java actually needed to access tek.com? According to java.com it is just needed for some applets...

John :-#)#

Reply to
John Robertson

No, I'll check it out later, off to work now.

This is what was happening. 64-bit Firefox put up the instrument's web pages, but when I clicked on anything requiring a Java applet to run, it failed, and the web program then went to Keithley.com = tek.com for a copy of Java, but quickly put up a security error message.

Trying 32-bit FireFox and Java also didn't work, since the new picky FireFox no longer supported Java applets. It took some experimenting to find the right versions.

The Virtual front panel, and the TSP Express system in the machine, using the Java applets, is super useful. I don't believe they in fact create any security risk, but the world has moved on.

Loading this software combination into a laptop should work with any 2000 to 2015-era Java web applet machine.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

Thanks. Not yet. Later.

Reply to
Steve Wilson

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