Help me recommend a VNA.

[dindondee]

Greetings wizards,

I need your help. The time has come for me to invest in a dedicated VNA and I need some of your recommendations/advice. I'm digging deeper into RF and it is now time to invest in equipment that will work well for me. It's been two years since I started and I now find myself needing to experiment with power amplifiers, match antenna's and measure impedance from PCB's amongst other things. And while I consider myself an absolute noob, I can say I know enough to require the following

- Max Frequency range: 3Ghz, for now. I anticipate working on projects with 5.8Ghz before the end of next year so maybe I will need to just get something with max 6Ghz(or 12GHz?).
Correct me if I am wrong, I have read somewhere to get equipment that is double the frequency you will work with and I have forgotten the reason why... something to do with harmonics or the noise floor.

- Dynamic range : Anything reasonable, 90-120dB

- Trace/Floor noise : enough to get accurate signals through filters.

My research has lead me to the following models- HP 8563E, HP 8566A/B, HP 8753C/D/E and the SVA1075X(I hear I can get educational discounts on siglent products). If you have better recommendations or perhaps need to school me on whatever, please do.  Cal kits, pros/cons, etc...

Or do you think I should consider a spectrum analyzer with a tracking generator?(please say no)
I have a limited budget (around $2,000 not including calibration kit) being that I am still an undergraduate and I consider this purchase an investment into my future career. I will save up for the rest of the year if I have to raise my budget. And I would also love to know your favorite dealers to purchase from. I just need enough recommendations/guidance to steer me into what will be my choice to purchase and where to find this.

PS- I do not mind taking a trip to the Bay area where I hear people get great deals, please show me the way.

Thank you for your time and help.   

[Mechatrommer]

the only usable VNA 3-6GHz so far the cheapest is Deepace VNA http://www.deepace.net/ usable means, we dont need PC as a monitor. added bonus is this thing portable can bring to broadcast tower to do testing. yes when we talk about harmonics, we'll need to see at least one or 2 higher harmonics, so if working on 3GHz fundamental we need to see 6-12GHz harmonics to be comfortable, but its a tough call to the pocket. your researched model HP is guaranteed used/old unit if $2000 budget and you'll need luck to wait in ebay. the good thing about Deepace, they added fringing inductance parameter for CAL kit as we've requested, older branded VNA only provided fringing capacitance. the bad thing is they shut down their bbs forum site.

[hendorog]

Welcome to the rabbit hole

You may be aware, but HP 8563E, HP 8566A/B are not VNA's. They are spectrum analysers only.

I'd suggest you probably want both a spectrum analyser _and_ a VNA.
I don't think you want to go down the SA+TG path at those frequencies. It will perform worse than a VNA, and will probably not be much cheaper.

As such the SVA range is pretty hard to beat as it has both a spec analyser, and a VNA in one unit. (The SSA is an SA + TG for what its worth)

There is also the GenComm GC700A and GC780A which were mentioned here by Wuerstchenhund: https://www.eevblog.com/forum/testgear/advice-needed-spectrum-analyzer-purchase/msg2934404/#msg2934404

Another option, which is very affordable, is a NanoVNA. You definately should just go and buy one of these anyway. They are so cheap and a great way to get your hands dirty and learn on a real VNA without risking much cash. Also it is good to have a cheap instrument to experiment on where the cost of a mistake is almost nothing. Comes with everything you need to get started. Frequency range is too limited for your intended use, but it's still a no brainer for learning.

Cal kits are expensive, but you can at least get started with the one included with the NanoVNA.
A cal kit is only as good as the characterisation data that you have for it. In the case of the Nano kit, there is no characterisation data.
A better kit, which is still cheap, is available from SDR kits - these are partially characterised and have good quality parts, but the characterisation data is only intended for use up to 1.3 GHz.
Siglent supply cal kits for the SVA range. The characterisation data is in the SVA itself, and might not be supplied seperately for use on other VNA's(?)
Next step up is probably Kirkby Microwave's kits, which are getting expensive, but are a 'proper' cal kit and come with a complete set of data and a verification standard.
Then there are second hand ones from eBay, which likely vary wildly in quality and price, and then new from the like of Keysight etc which will cost much money.

You could DIY of course, but you will need to have access to a real kit to characterise your DIY kit at those frequencies. It is likely a better option to just buy a cheap one and characterise that instead.

[dindondee]

the only usable VNA 3-6GHz so far the cheapest is Deepace VNA http://www.deepace.net/ usable means, we dont need PC as a monitor. added bonus is this thing portable can bring to broadcast tower to do testing. yes when we talk about harmonics, we'll need to see at least one or 2 higher harmonics, so if working on 3GHz fundamental we need to see 6-12GHz harmonics to be comfortable, but its a tough call to the pocket. your researched model HP is guaranteed used/old unit if $2000 budget and you'll need luck to wait in ebay. the good thing about Deepace, they added fringing inductance parameter for CAL kit as we've required, older branded VNA only provided fringing capacitance. the bad thing is they shut down their bbs forum site.

the KC901Q 20GHz looks like something I will add to the list of my considerations, thanks. I also understand the frequency range will increase the price, but this is okay.

Welcome to the rabbit hole

You may be aware, but HP 8563E, HP 8566A/B are not VNA's. They are spectrum analysers only.

I'd suggest you probably want both a spectrum analyser _and_ a VNA.
I don't think you want to go down the SA+TG path at those frequencies. It will perform worse than a VNA, and will probably not be much cheaper.

As such the SVA range is pretty hard to beat as it has both a spec analyser, and a VNA in one unit. (The SSA is an SA + TG for what its worth)

There is also the GenComm GC700A and GC780A which were mentioned here by Wuerstchenhund: https://www.eevblog.com/forum/testgear/advice-needed-spectrum-analyzer-purchase/msg2934404/#msg2934404

Another option, which is very affordable, is a NanoVNA. You definately should just go and buy one of these anyway. They are so cheap and a great way to get your hands dirty and learn on a real VNA without risking much cash. Also it is good to have a cheap instrument to experiment on where the cost of a mistake is almost nothing. Comes with everything you need to get started. Frequency range is too limited for your intended use, but it's still a no brainer for learning.

Cal kits are expensive, but you can at least get started with the one included with the NanoVNA.
A cal kit is only as good as the characterisation data that you have for it. In the case of the Nano kit, there is no characterisation data.
A better kit, which is still cheap, is available from SDR kits - these are partially characterised and have good quality parts, but the characterisation data is only intended for use up to 1.3 GHz.
Siglent supply cal kits for the SVA range. The characterisation data is in the SVA itself, and might not be supplied seperately for use on other VNA's(?)
Next step up is probably Kirkby Microwave's kits, which are getting expensive, but are a 'proper' cal kit and come with a complete set of data and a verification standard.
Then there are second hand ones from eBay, which likely vary wildly in quality and price, and then new from the like of Keysight etc which will cost much money.

You could DIY of course, but you will need to have access to a real kit to characterise your DIY kit at those frequencies. It is likely a better option to just buy a cheap one and characterise that instead.

hey hendorog, thanks. this rabbit hole is scary. i have actually been playing with the NanoVNA and it is what has given me the courage to step up to something serious. I had no prior knowledge about using a vna- taking measurements and the need for calibration before taking measurements until I started with the nanovna. yes, it is limited and for my applications, the accuracy i get vary wildly from repeated measurements on the same circuit  to different measurements on other circuits with the same elements.

I am also aware of the Kirby cal kits, they are what i intend to purchase along with the corresponding HP equipment i decide to purchase.
https://www.kirkbymicrowave.co.uk/Everything-you-wanted-to-know-about-the-HP-8753-VNA/

So,it seems my budget is the limiting factor to getting recommendations. could you please ignore that budget and tell me what would work with my needs/specs that has a reasonable price? And where I can find such, navigating ebay seems terrifying. Like I said I am making an investment in this for my career and willing to save up for it.

[Mechatrommer]

the KC901Q 20GHz looks like something I will add to the list of my considerations, thanks. I also understand the frequency range will increase the price, but this is okay.

be aware that the 20GHz model wont work on its own, it needs external bridge/coupler or what iirc the last time i read it. and double digit GHz is really another rabbit hole in the rabbit hole. designing and pcb cost manufacturing head ache alone is awaiting ahead. not worth for me at least.

here again from the site about KC901Q..

The KC901Q is significantly different in function from other KC901 series instruments. KC901Q has NO Built-in coupler, can NOT measurement S11 parameter directly, The built-in output attenuator is ONLY work below 7GHz, More than 7GHz it will bypass the attenuator. Special attention should be paid when measuring amplifiers.

[tautech]

@dindondee
FYI, a couple of threads with some SVA1000X examples of use that might give you a bit more info on these models. Initially I only had the 1.5 GHz model but now have upgraded to the SVA1032X.

https://www.eevblog.com/forum/rf-microwave/antenna-project-log/
https://www.eevblog.com/forum/rf-microwave/what-really-is-this-antenna/

FWIW these are one of my fav bits of kit such is the feature set they offer.

[TheSteve]

Check those spec sheets carefully, generally getting 90-120 dB of dynamic range excludes lower end equipment.

Between 10 and 1500 MHz the SVA1032X is only 60 dB and that drops to 55 dB by 3.2 GHz. This isn't even close to the same league as an older HP machine. It also makes it useless for tuning duplexers and such.

[tautech]

Check those spec sheets carefully, generally getting 90-120 dB of dynamic range excludes lower end equipment.

Between 10 and 1500 MHz the SVA1032X is only 60 dB and that drops to 55 dB by 3.2 GHz. This isn't even close to the same league as an older HP machine. It also makes it useless for tuning duplexers and such.

Err well, yes. But better still read the OP's requirements more carefully as there is already a stated interest in SVA1075X that has entirely different and much better dynamic range specs.
Check for yourself on P15.
https://siglentna.com/wp-content/uploads/dlm_uploads/2018/06/SVA1000X_DataSheet_DS0701X_E03A.pdf

[joeqsmith]

Beyond what's been posted,  a few things you may want to consider when looking for a VNA:
Do you want all 4 S-parameters.   
What sort of software features do you want.   
Do you want to interface it with a PC, possibly write your own software to automate tests.  Could it run headless (requires a PC for the display). 
If you want an interface, is GPIB going to be good enough.  Is 10,100,1000Mb good enough?   
Do you want to save data to your network, USB thumb drive.   
Is SOLT good enough or do you want TRL (other).  Do you want 2 receivers or 4. 
Do you want the system to include bias Ts, step attenuators. 
At 6GHz, consider the cables and fixtures. 
Is size/weight a problem.   
If you buy used, can you actually repair it.  Consider the cost and your existing lab and your skill level.   

6GHz at $2000, considering cables, cal standards, fixtures.... I'm not so sure.   Personally, I suggest you wait for some reviews on the Nano V2 and see if it looks like it would carry you over.   <$100 if you order it with cables and the standards.   If you really want to work with amplifiers, you may find it would be a better fit as I understand they do not use harmonics.  With the Nano, you would need to consider the fundamental. 

[joeqsmith]

https://coppermountaintech.com/vna/r60-1-port/

https://www.picotech.com/vector-network-analyzer/picovna-106/picovna-series

I like Copper Mountain's software but wasn't impressed with PICO's.  If you are interested in looking at the Pico, maybe start here:
https://www.eevblog.com/forum/testgear/picovna/

[Mechatrommer]

having complete 4 parameters S11,S21,S22,S12 is good if not so much addition of money. but if double the price and dependent on PC/laptop i'm not so sure for beginner. capability to bring to open field to test probably like a large tower antenna installation setup will be a bonus usually in HAM-like hobby work imho. but if only testing large antenna S11, there are few cheap small devices out there to buy, not worth risking $2000 device to deal with high voltage statics of large antenna.

otoh as a beginner, i still struggling to think or find a way how a VNA can measure S22,S12 for things like active amplifier or an opamp, its good to measure output impedance of an opamp (diy active amplifier) but how its possible to inject signal to the output of an amplifier/opamp and measure back its S22? let alone S12, signal injected to its output and get back a signal leaked at its input? istr putting an output of an opamp/active device to S11 port of my VNA as a hack to measure S22, i'm not sure if this is a good idea. although i was being carefull not to output signal from the opamp so not to damage the S11 port. but eventually later when i got a (used) SA, my VNA S11 port signal generator inside got damaged when i measure its signal output at the SA (very low signal output level). i did send my VNA back to china for repair, luckily they didnt charge for the repair, the only hiccup is i have to pay the trackable (safe) shipping cost myself for sending it back. i'm not sure its the hack i did that got it damaged or some other reason. i did being carefull as i mentioned.

[joeqsmith]

I would imagine the amateur radio group knows what steps to take to avoid damaging their sensitive equipment.  Guessing they understand they should discharge everything before connecting to sensitive equipment.  When experimenting with high voltages, I will use a GDT, attenuator and diode clamp combo.  For the GDT, I use the Citel parts. 
 
When I demo'ed a small amplifier on the Nano, I avoided the issue with harmonics.  If the fundamental makes it to the amplifier you could imagine it over driving the VNAs input, possibly causing damage.  Not to mention the obvious problems with the measurement.  In the case of the $50 Nano, not so much a big deal.  As I understand it, the V2 avoids this. 

First hit on Google for "Hot S22":
http://na.support.keysight.com/pna/help/latest/Tutorials/Hot_S22_Measurement.htm

[dindondee]

Beyond what's been posted,  a few things you may want to consider when looking for a VNA:
Do you want all 4 S-parameters.   
What sort of software features do you want.   
Do you want to interface it with a PC, possibly write your own software to automate tests.  Could it run headless (requires a PC for the display). 
If you want an interface, is GPIB going to be good enough.  Is 10,100,1000Mb good enough?   
Do you want to save data to your network, USB thumb drive.   
Is SOLT good enough or do you want TRL (other).  Do you want 2 receivers or 4. 
Do you want the system to include bias Ts, step attenuators. 
At 6GHz, consider the cables and fixtures. 
Is size/weight a problem.   
If you buy used, can you actually repair it.  Consider the cost and your existing lab and your skill level.   

6GHz at $2000, considering cables, cal standards, fixtures.... I'm not so sure.   Personally, I suggest you wait for some reviews on the Nano V2 and see if it looks like it would carry you over.   <$100 if you order it with cables and the standards.   If you really want to work with amplifiers, you may find it would be a better fit as I understand they do not use harmonics.  With the Nano, you would need to consider the fundamental. 

Yes, all 4 s-parameters, definitely would need to save data , I prefer solt with 2 receivers, size and weight is not a problem, i prefer to purchase used and calibrated from a reputable user/dealer especially since i have no experience repairing test equipment and for now- ram of the gbib port does not really matter, I dont care for pc interface , I'm not really big on features yet because like I said I'm just getting into the field and have just reached the point where I require accurate measurements and perform projects/experiments to go along with simulations, smith charts- barest minimum. I am well informed of cable and connector compatibility, sensitivity and requirements  with ranging frequency and the importance of cal kits. Eventually some of these features being talked about may come in handy...

I am sensing that being a beginner, most of the responses would suggest that i am probably inexperienced and will damage the equipment. Also, the cost has been a major limiting factor in getting recommendations. I agree to an extent that i am a noob however i have a mentor in faculty who has shown me the basics using an agilent  e5080b and would not mind putting me through with whatever i end up purchasing. problem is that we are both limited on how much access/time we have on the e5080 and since i see myself with a major interest in rf and may most likely end up working on rf portions, i want to get one.

as for the cost, just like i said it is an investment and what i have is not enough but i am willing to save up for. i am just looking for recommendations and the experience of you seasoned wizards to point me in the right direction.

my opinion of the nano vna, it is quite good however it is grossly inaccurate when it comes to my current needs and even with the latest firmware updates, my experience with it would not use it past 1.5ghz. i have high hopes of the nano v2, i plan to stilll get that to learn but sooner then later it will be inadequate.

having complete 4 parameters S11,S21,S22,S12 is good if not so much addition of money. but if double the price and dependent on PC/laptop i'm not so sure for beginner. capability to bring to open field to test probably like a large tower antenna installation setup will be a bonus usually in HAM-like hobby work imho. but if only testing large antenna S11, there are few cheap small devices out there to buy, not worth risking $2000 device to deal with high voltage statics of large antenna.

otoh as a beginner, i still struggling to think or find a way how a VNA can measure S22,S12 for things like active amplifier or an opamp, its good to measure output impedance of an opamp (diy active amplifier) but how its possible to inject signal to the output of an amplifier/opamp and measure back its S22? let alone S12, signal injected to its output and get back a signal leaked at its input? istr putting an output of an opamp/active device to S11 port of my VNA as a hack to measure S22, i'm not sure if this is a good idea. although i was being carefull not to output signal from the opamp so not to damage the S11 port. but eventually later when i got a (used) SA, my VNA S11 port signal generator inside got damaged when i measure its signal output at the SA (very low signal output level). i did send my VNA back to china for repair, luckily they didnt charge for the repair, the only hiccup is i have to pay the trackable (safe) shipping cost myself for sending it back. i'm not sure its the hack i did that got it damaged or some other reason. i did being carefull as i mentioned.

i know how these things work the more they can do the higher the cost. my work is with opamps is mostly following well documented examples and some ham articles with guidance from my mentor and some folks at my local ham club. what vna was that you had to send back if i may ask?

-----------------------------------------------------------

the 8563E is what i am considering since it will be useful for the upcoming 5.8ghz projects , anything thats 6ghz max freq range also i just have not narrowed it down yet and i'm here to get recommendations for others and lessons/tips for where to purchase and how much i should expect to shell.

ps - a graduate student i work with told me he got the hp 8753D for around $3k with the help of a friend who sadly passed on most recently.. and he was going to help me with this so I appreciate any help i get.

[Gandalf_Sr]

I have an HP8753 VNA with options 006 and 010 (6GHz and Time Domain).  It's in great shape and has a new LCD that's bright and clear.  PM me if you're interested.  I'm in Michigan.

[Mechatrommer]

my work is with opamps is mostly following well documented examples and some ham articles with guidance from my mentor and some folks at my local ham club.

i will anticipate making my own rf amplifier, so characterizing its output may be helpfull. but i'm not reaching that stage yet due to many other tasks. and some opamp in the market do not publish their output impedance vs freq plot, so sometime its difficult to choose or design matching output/load. in the end i have to test it myself.

what vna was that you had to send back if i may ask?

Deepace KC901V. i bought it to deal with aftermarket 5.8GHz video transmitter/receiver that i bought and burnt many years ago. still have no time to repair but the VNA has been helpful to characterize some of my antenna and built devices' S11, S21.

I have an HP8753 VNA with options 006 and 010 (6GHz and Time Domain).  It's in great shape and has a new LCD that's bright and clear.  PM me if you're interested.  I'm in Michigan.

i waited for brand name used VNA that can go to 6GHz within my budget spec (~$2000) for a long time, unfortunately i was not patient enough and give the Deepace unit a go after thesignalpath review. those Agilent brand are highly recommended if it works in complete set.

[joeqsmith]

For a 6GHz system, I am not sure I would trust anyone outside the manufacture to test and repair one.  If you could work a deal with the seller to have it sent to the manufacture for checkout you may get a better idea of how much your going to be into it.  Maybe make it part of the deal that if the system requires repairs, you are off the hook.    Even if you covered the costs for shipping to the Mfg and having them test it, you may be further ahead.   Then again, its really your comfort level.   

Based on your last post, the 8753 may be a decent fit.  I looked at getting an 6GHz 8753E with the built-in test set but really wanted some of the features of the newer systems.   

If you're are looking for something newer there is a member Pete Manfre who sells a lot of stuff in the following group.  He wrote me about having an old 9GHz PNA with TDR option, 500MHz CPU, cables and cal set.  You could see if he still has it. 
https://groups.io/g/test-equip-trader/topics?p=recentpostdate%2Fsticky,,,20,1,40,0&jump=1

[Howardlong]

Does the OP know why they need 90-120dB of dynamic range in a VNA?

[Gandalf_Sr]

Mine has the built in test set.

[joeqsmith]

Mine has the built in test set.

Nice!  Where were you last year when I was looking to upgrade?      I've used the 8753 and like how they drive.  Are you letting the APC7 adapters and any other bits go with it?   

On the subject of dynamic range,  a few more things OP may want to consider are IF bandwidth, support for averaging, how fast they want it to sweep and the number of samples. 

Another article on the Pico:
https://www.electronicspecifier.com/products/test-and-measurement/low-cost-vna-boasts-dynamic-range-of-118db-at-10hz

[Gandalf_Sr]

Mine has the built in test set.

Nice!  Where were you last year when I was looking to upgrade?      I've used the 8753 and like how they drive.  Are you letting the APC7 adapters and any other bits go with it?   

On the subject of dynamic range,  a few more things OP may want to consider are IF bandwidth, support for averaging, how fast they want it to sweep and the number of samples. 

Another article on the Pico:
https://www.electronicspecifier.com/products/test-and-measurement/low-cost-vna-boasts-dynamic-range-of-118db-at-10hz

I could let the APC7 adapters go, I also have a Kirby Cal kit with FDD for it.

I've been thinking of selling it for a while as I don't use it enough to justify keeping it.

[joeqsmith]

I could let the APC7 adapters go, I also have a Kirby Cal kit with FDD for it.

I've been thinking of selling it for a while as I don't use it enough to justify keeping it.

This is the reason I struggled with buying another.   I use my VNA a hand full of times a year.   

That would make a really nice setup for the OP by the sound of it. 

Here is an article that talks about some of the key advantages to the system Pete Manfre was selling compared with the HP8753. 
https://www.microwavejournal.com/articles/print/3031-a-novel-design-approach-for-an-rf-vna

[Gandalf_Sr]

One exactly like mine that I was watching on eBay just sold for $3,500 + $50 shipping, I could let mine go for that kind of money.  If the OP isn't interested, I think I'll list on eBay tomorrow.

[dindondee]

Mine has the built in test set.

Interested. Sent you a PM.

Does the OP know why they need 90-120dB of dynamic range in a VNA?

While I currently do not need it and it may not be necessary for the time being, it would be nice to find something that has that range and I have some projects that do. Experiments on rf signal using an FPGA+ADC, with a very tight constraint on the LNA. Currently use the one at school for that. If you have any recommendations and dynamic range is less than 90db please still give me your recommendation, thanks.

[OwO]

If you can wait for nano V3 (ETA 1 year), it is most likely going to reach at least 90dB dynamic range and will be full two port (all 4 S parameters at once).

[joeqsmith]

One exactly like mine that I was watching on eBay just sold for $3,500 + $50 shipping, I could let mine go for that kind of money.  If the OP isn't interested, I think I'll list on eBay tomorrow.

I am curious about your picture.  The unit is sitting on a shelf with no depth.   This is similar to my home office where my equipment sits on wire shelves.  There's lots of air flow and it takes up little space but there's no place to work.   What are you using for a workspace when you use equipment like this?   

At home, I have one of my mom's old portable wooden tables from her garden but in many cases, I slide a sheet of FR4 under the equipment.  A while back I was running some tests and had three DMMs on a sheet and it was really sagging.   I've since soldered some wedges to add some strength.  Then parts roll off onto the floor so my latest one has a lip soldered around the outside edge.  I normally will run a ground strap to it but it's now evolved so the wrist strap will plug into the table.   A real hack job but strong enough to support meters and various small test jigs.