TinySA Ultra launched

[switchabl]

In part this is because if you're using a particularly expensive piece of equipment, it's generally going to be presumed that you know what you're doing (if only because as a general rule, such equipment is owned by companies and no sane company is going to let anyone near such equipment unless they've demonstrated that they know what they're doing).

Yes, and legend has it that these special people never make mistakes either (or so they say). That is why you never see expensive RF test equipment with fried inputs or butchered connectors. 

More seriously, robustness is definitely still a desirable quality in high-end equipment. If nothing else, working with a special snowflake instrument consumes extra time and mental bandwidth just to make sure you don't break it.

It's true that performance requirements may limit your options when it comes to input protection. But then you don't have other constraints on big, expensive lab equipment. For example, the maximum input power on the TinySA is pretty much limited by its low supply voltage. And if a SA has a dedicated baseband path (direct to the ADC) for low frequencies anyway, like most high-end instruments do, then the cut-off frequency in the RF path can be a lot higher and protecting it from DC bias actually becomes easier.

[pdenisowski]

In part this is because if you're using a particularly expensive piece of equipment, it's generally going to be presumed that you know what you're doing.

Yes, and legend has it that these special people never make mistakes either (or so they say). That is why you never see expensive RF test equipment with fried inputs or butchered connectors. 

No, "special people" also make mistakes - trust me

The two main ways that spec ans get damaged are either (a) too much power applied to the input connector or (b) physical damage from shipping, dropping, liquid spills, etc.  In a professional lab environment, users typically are either using power levels that won't damage a modern spec an (+30 dBm or less) and/or are well aware of the need for an attenuator, etc. between the spec an and their particular device under test. 

[kcbrown]

Yes, and legend has it that these special people never make mistakes either (or so they say). That is why you never see expensive RF test equipment with fried inputs or butchered connectors. 

LOL.  Well, yeah, but the flip side is that they can at least presume that the probability of mistakes is a lot smaller for the buyers of such equipment than it would be for people who actually don't know what they're doing.

Nobody's perfect, everyone makes mistakes, so it's really just a question of whether the effort to put the protection in is worth the probability reduction.

Bear in mind that input protection is necessarily limited in its capability, so it just represents a reduction in the probability and not an elimination of it.

More seriously, robustness is definitely still a desirable quality in high-end equipment. If nothing else, working with a special snowflake instrument consumes extra time and mental bandwidth just to make sure you don't break it.

That's true.  An interesting question is whether or not you get an average productivity drop as a result.  I've found for myself that if I'm more careful, I tend to reduce other mistakes I might otherwise make as well.  The fewer the mistakes, the lower the chance that you'll have to go back and figure out what you did wrong, or at least correct what you did wrong.  There may be something to be said for forcing yourself to apply extra care.

It's true that performance requirements may limit your options when it comes to input protection. But then you don't have other constraints on big, expensive lab equipment. For example, the maximum input power on the TinySA is pretty much limited by its low supply voltage. And if a SA has a dedicated baseband path (direct to the ADC) for low frequencies anyway, like most high-end instruments do, then the cut-off frequency in the RF path can be a lot higher and protecting it from DC bias actually becomes easier.

If it's relatively easy and inexpensive to design in some input protection without compromising the performance, then it certainly makes all kind of sense to do it.  So I suppose that, like for so many other things, the answer here is "it just depends".   

[Bicurico]

Input protection lowers the performance of the device and is not part of the deal.

Everybody should know the input rating of the SA and protect it accordingly.

If you want to measure a subtle RF signal, you don't want unnecessary input protection raise the noise floor or attenuate the signal.

On the other hand, if you want to connect your ham 100W output stage directly to the RF input of your SA, then no input protection will prevent you from destroying the input stage.

If you now say, it should be protected up to 100W, what about a user connecting a 1000W signal? What about attaching a 100V DC? The more protection you have on the device, the less attention users will have on what they are doing.

A SA is a precision instrument and has to be dealt with as such. The input protection is the user's responsability and that is why they should use DC blockers and attenuators.

The lack of further input protection does not make sense in this context.

[MathWizard]

Ok thanks guys, I'll have a look for the extra accessories before I commit to buying anything. Just watching a video on the do's and dont's for SA's. So they really can't have over a few uW's of power inputted ? The SA in the video says -30 dBmW max, so isn't that 1uW ?

So is that just to minimize the loading on the DUT, and be able to work into the GHz ? Or what's the real reason the front-ends of SA and VNA are so easy to blow up?

I would be using these for pretty low frequencies, like FM radio circuit experiments. Or much lower, like probing control circuits on a SMPS. Maybe they won't be the right tools for that.

[erikka]

There are different reasons for different maximum input levels:

• Avoiding Damage
• Avoiding internally generated harmonics

When talking about a -30dBm level it is probably about avoiding internally generated harmonics when doing harmonics measurements.
The damage level to avoid is probably much higher, staying below 0dBm is safe for all SA's I know.
And yes, an SA is not a scope, don't probe an SMPS without a lot of protection

[gf]

So they really can't have over a few uW's of power inputted ?

TinySA spec sais "Absolute maximum input level of +6dBm with 0dB internal attenuation" - thats more than just µW.
https://tinysa.org/wiki/pmwiki.php?n=TinySA4.Specification

Anyway, think of it as a radio receiver. You also wouldn't feed 100W into the antenna input of a radio and expect no damage, would you?

[Bicurico]

Imagine this analogy:

A sea has rising or lowering water level. You use a multimeter to determine the water level. It can measure a water level of several meters.

An oscilloscope will measure the height of the waves - the big waves like the ones on the beach.

A spectrum analyser is used to measure the many tiny waves that move on top of the water surface. These waves are just a few mm in height - like when you throw a rock into the pond.

To measure these tiny waves on top of the water, you use a very tiny and sensible floater. It is of course very fragile.

If a huge wave comes or the sea level rises too much, the fragile floater gets broken.

It is difficult to make it less fragile as it becomes less precise the more robust it is build - remember you want it to measure the tiniest water movements on the surface.

I hope this analogy makes it clearer.

[pdenisowski]

There are different reasons for different maximum input levels:

• Avoiding Damage
• Avoiding internally generated harmonics

When talking about a -30dBm level it is probably about avoiding internally generated harmonics when doing harmonics measurements.
The damage level to avoid is probably much higher, staying below 0dBm is safe for all SA's I know.

Yes, there is a (often considerable) difference between maximum safe input power level (to avoid damage to the spec an) and the levels at which compression and intermodulation / harmonic distortion occur.  I talk about this in my video on dynamic range:  (from about 3:50 to about 5:45)

[pdenisowski]

Just watching a video on the do's and dont's for SA's. So they really can't have over a few uW's of power inputted ? The SA in the video says -30 dBmW max, so isn't that 1uW ?

Most modern "professional" spec ans have a +30 dBm (1 watt) max input power, although I've also seen some (especially handhelds) that only go up to +20 dBm. 

When we released a "high power" option (up to +38 dBm) for our SMA100B analog signal generator, we actually included a  configurable "output power limit" parameter that defaulted to +30 dBm because this is the normal "maximum" power for most modern spec ans.

[Messtechniker]

Can I use the tiny SA Ultra to adjust coax cable tv frequency tilt?
One problem is: coax cable tv is a 75 Ohms system whereas
the tiny Ultra SA input impedance is 50 Ohms. So i will need a
suitable balun. Any recommendations?

[horo]

Can I use the tiny SA Ultra to adjust coax cable tv frequency tilt?
One problem is: coax cable tv is a 75 Ohms system whereas
the tiny Ultra SA input impedance is 50 Ohms. So i will need a
suitable balun. Any recommendations?

Simplest solution is a minimum loss pad (MLP):

[baldurn]

Can I use the tiny SA Ultra to adjust coax cable tv frequency tilt?
One problem is: coax cable tv is a 75 Ohms system whereas
the tiny Ultra SA input impedance is 50 Ohms. So i will need a
suitable balun. Any recommendations?

Use a 75 Ohm 10 dB or more attenuator. The reflected signal will die in the attenuator. The Ultra SA will measure a little less. I am unsure exactly how much, maybe 3 dB? So you could enter the attenuator as 13 dB.

[ltwin8]

Is there any comparison between LiteVNA64/62 and the new tinySA ultra?

Seems the tinySA has the bigger battery and the pen hold for it, but aren’t there firmware updates for the liteVNA enabling peakhold?

I would be interested in peak hold for measuring Hobby Videolink and controllink

Best regards

[Bicurico]

There is no comparison, because these are two different devices!

The TinySA is a spectrum analyser, that shows you signals in the frequency domain (an oscillocope will show you the signal in the time domain).

The LiteVNA64 (or any other like NanoVNA, etc.) are Vector Network Analysers, used to show you the impedance of a DUT over a frequency range. Most common form of visualization is the Smith chart. This is used for instance for antenna matching, helping you to get the correct impedance for the desired frequency.

Think of it like this: the Spectrum Analyzer shows you what signals are present over a given frequency range, while the VNA measures the behaviour of a component or antenna over a given frequency range.

You cannot use a VNA as a SA (unless it features both functionality, which the LiteVNA or NanoVNA, doesn't) nor can you use a SA as a VNA (which is not 100% true as you can do some measurements when a Tracking Generator is present or when the SA features VNA functionality - which is not the case with the TinySA).

[rhb]

Ok thanks guys, I'll have a look for the extra accessories before I commit to buying anything. Just watching a video on the do's and dont's for SA's. So they really can't have over a few uW's of power inputted ? The SA in the video says -30 dBmW max, so isn't that 1uW ?

So is that just to minimize the loading on the DUT, and be able to work into the GHz ? Or what's the real reason the front-ends of SA and VNA are so easy to blow up?

I would be using these for pretty low frequencies, like FM radio circuit experiments. Or much lower, like probing control circuits on a SMPS. Maybe they won't be the right tools for that.

The way to use an SA with an external signal source is to put an adjustable attenuator set at maximum attenuation and then watch the SA as you reduce the attenuation and set it give the best performance which will be well below damage level.  I have a 60 dB x 1 dB 2 GHz rotary  attenuator I use with some 20-50 W fixed attenuators and a DC block to protect my 8560A.  I've ordered an ebay Chinese step attenuator and will be investigating its performance very closely.

BTW You probe something like an SMPS with an H field probe, not an electrical connection.

Reg

[DaneLaw]

Received my unit today from "shenzhen colourful Store" an Ali express store.
Stores like the official "Zeenko" is out of stock on the Ultra (atm)  // 13-7 They have stocked up again.!

What arrived.

Has there been any indication that clones of the Ultra with inferior components' had hit the market yet.?

[RAPo]

Kerry Wong just published a very on making a tracking generator for the tinySA Ultra Spectrum Analyzer.

[EC8010]

I see there's been a lot of talk about robustness. I bought a TinySA because I don't need an RF spectrum analyser very often so I can't justify something far more expensive. I accept that to achieve the performance it does, some corners have had to be cut, and that's fine by me. In general, you get what you pay for, and in the TinySA, rather more. Bravo!

[DaneLaw]

The Ultra is quite a step up vs the older TinySA version, but one thing that is a shame is the otherwise great waterfall and it disappears /cycles the moment you do anything..
fx you take a running measurement of the RF span, and you pause the sweep to analyze and wanna move the cursor on top, to align with visual indications on the waterfall, I cant.. for some reason it seems to remove the waterfall when you do anything, but maybe its me that does something wrong?

- that would also make something a simple as taking a basic screenshot to the incl. MicroSD problematic, as the moment you hit "save-capture" the waterfall cycles to nothing & the screenshot, will be all black, where the waterfall should have been.
I am on the fw it arrived with, so maybe is been fixed on later fw updates.?

Some wide sweeps. (avg)
0 to 12.072G
0 to 9.999G
0 to 6.2G

The fact that I just purchased mine TinySA Ultra & got it delivered a few days ago, you can bet there will likely come a new better version very soon..As that cycle is as sure as they say "amen" in the church, & me in-a-nutshell to display how close you can cut it - and of course that is also what is happening.
A new TinySA Ultra revision is about to reach the market.

As far as I can tell, no significant RF enhancements, it's mainly some hardware tweaking to the chip, with a new IPS LCD display, while the older one had TFT LCD.
Hardware revisions on the new revision of TinySA Ultra' will go under the label V0.4.5.1.1 from serial number SU-23070001 while the older was V0.4.5.1
My unit SN SU-23060705 & HW V0.4.5.1
The Hugen production of IPS variant will ramp up next week, though when it will be in stock.. anybody's guess.

Btw these cases works okay to hold the TinySA and TinySA Ultra in lack of better, - atm cost is 1.9 US delivered (incl. 25% VAT & shipping) though only a few left, just remember to get that -1US /+1.01 discount, which will choke the price down to 1.9 US from 2.9 US.) https://aliexpress.com/item/32835458985.html?
// the 1.9 US delivered price, gone

[shapirus]

Moving the discussion concerning the output frequency glitches in the signal generator mode from another topic to here.

Someone please check if it reproduces with your TinySA Ultra to see if it's not only my unit that has these glitches.

Latest firmware update was 5 days ago on the TinySA Ultra. I wonder if they fixed that bug. There's been significant updates apparently since the version I had. I was using 1.4-120, now they're at 1.4-159.

Nope, same exact thing with 1.4-159.

Here's a demonstration with increased persistence (to see the glitches) and the lowest memory length for the max waveform update rate.

First, there's a 20 sec-long frequency sweep from 20 MHz to 120 MHz, then (00:30) frequency is changed manually in +100 kHz steps and finally (01:25) amplitude is changed manually.

As we can see, on any change in frequency or amplitude it has a brief period during which it outputs all sorts of crap. Frequency sweep output is the easiest way of catching it.

It's not the scope: it displays frequency sweeps from a different generator just fine, nice and clean.

[pdenisowski]

As we can see, on any change in frequency or amplitude it has a brief period during which it outputs all sorts of crap. Frequency sweep output is the easiest way of catching it.

I was going to suggest making an Allan variance measurement, and then I looked at the screenshot more closely ....

[KungFuJosh]

Moving the discussion concerning the output frequency glitches in the signal generator mode from another topic to here.

Someone please check if it reproduces with your TinySA Ultra to see if it's not only my unit that has these glitches.

Glitches galore:




I wonder if this is something they care about or not since it's not the primary function of the device. Cleaner would be nicer though.

[DaneLaw]

I wonder if this is something they care about or not since it's not the primary function of the device. Cleaner would be nicer though.

Sure, the TinySA & SA4 creator Erik Kaashoeck are very willing to fix bugs.
He often push out new firmware with new features & abilities, and it will from time to time, create issues in the back-catalog that then need to be addressed.
Been very impressed of the support these TinySA models are getting.

Bug fixes etc. forum TinySA & SA4 forum https://groups.io/g/tinysa/messages

The only thing that won't get support for obvious reasons, are issues from clones, and in units that ain't produced by Hugen.
The original TinySA-market was flooded with clones with inferior components, the TinySA Ultra has been in the clear for those issues over the last few years, but it has changed here in 2024, where it seems that numerous TinySA Ultra clones are getting to market, so it goes without saying, use official sale-channels if you in the market for one.

[KungFuJosh]

I vote @shapirus reports the bug, he found it, I was just confirming it.

Here's a clean sweep for comparison with the same setup, but using the SDG2122X:



I wish Siglent would have a full frame rate option for the web console.