Anritsu MS8609A 9kHz - 13.2GHz Spectrum/Signal Analyzer - Short Review

[eevcandies]

Of course the years of working on TV's and touching 20KV probably kept you very alert!!

[eevcandies]

OUCH!
Anritsu factory Repair and Calibration Pricing
Model Number MS8609A-13GHz TRANSMIT TESTER
Standard Calibration Z540                $1524.00
Premium Cal (Pre and post cal data) $1948.00

Standard Repair                                                $3690.00
Full Service (repair/std calibration)                   $5277.00
Premium Full Service (Repair/pre/post cal data) $5723.00

What was the list price on the MS8609A unit & options?

[Mechatrommer]

right.. i paid for the unit is $1000, calibration due 2016, maybe if i need recalibration, i'll just simply browse for another used unit in ebay, hopefully still within calibration spec, or... i will improvise the Z540 rig, gain more knowledge and all.

[eevcandies]

right.. i paid for the unit is $1000, calibration due 2016,
That seems pretty good for a unit with options 1/4/8/31...the essentials.  I have seen almost none with option 2 ever odffered--must not be popular.

[Mechatrommer]

but it has 11 years on the counter, and doesnt have rubidium clock. finger crossed. i was really hoping the seller didnt accept that fast, now i'm locked. i checked back the manufacturing date of Advantest R3465 is 1994, 25 years ago, now it start to show deterioration at input atten, YIG or i'm not sure what (failed built-in calibration but still shows reading). i hope things like this are built for 20+ years of life, so i can still enjoy 9+ years clock of the yet to come Anritsu MS8609A. i cant find though when MS8609A first manufactured. 2 years counter and rubidium @ $1450 is much more bargain, even like $2K imho, but unfortunately that kind of figure will really pressing my budget (i'm currently eyeing for the used 1D X Mark II as well, i only can drink my tears).

[eevcandies]

I think these things will last a long long time...11 years is a spit in the bucket.  Actually not having rubiduim is better for you, since that is a "lamp" that will wear out line an old neon bulb. 
Of course you might keep it for a while & maybe even sell it for more than you paid for it..see, get paid to have fun. 

Looks like it came out around 2002, here is an introduction from then...note the fewer options

http://www.test-italy.com/Specifiche%20strumenti/ANRITSU/ANRITSU%20MS8608A-MS8609A/MS8608A%20-%20MS8609A.pdf

Look at all the "old" equipment from the 80's that is still in use!

[YU7C]

Yes, I found it as new product in Anritsu catalog from 2002.

[eevcandies]

Yes, I found it as new product in Anritsu catalog from 2002.

 Did they give a price---I wonder how much they were & for the options.

[YU7C]

Yes, I found it as new product in Anritsu catalog from 2002.

 Did they give a price---I wonder how much they were & for the options.

Sorry, no price in my copy.

[Mechatrommer]

we dont ask such prices in new, it can easily reach 5-6 figures, i guess they dont usually show the price because no mere mortal will be able to afford it, if you can it means you'll have the courage to contact them by phone or email. just to get the idea if you google players in the league...

an active product.. only 2GHz BW... $13K (still used)
https://www.anritsu.com/en-us/test-measurement/products/mt8820c
https://www.ebay.com/p/1818844324

https://www.tequipment.net/Rohde-&-Schwarz/FPC1000/Spectrum-Analyzers/
(This item is not available for export to your country: MALAYSIA, bugger. but maybe good for me)
Rohde SA will be just like toy,, 1GHz BW $2K how much do you think 13GHz will cost? simple math will tell 13 x 2K? not in reality, its usually exponential...

https://www.testunlimited.com/productDetail.aspx?product_id=931
Rohde and Schwarz FSW26 2 Hz to 26.5 GHz Spectrum Analyzer (Signal Analyzer) $86K
13GHz? half of it? maybe, but i'm guessing more. what is it for something specialized in radio mobile tester (constellation, phase or whatever capable tester?)

so we dont usually talk about the price new, so people can have less appreciation of what are available in ebay, and sellers too will have less temptation to increase the price. if they have to base on currency bubbling or shrinking or whatever the financial term is, no ebay inhabitants will be able to hit their listing anyway they only can dust the items in their shelf for so very long time.

[Mechatrommer]

my 8609 arrived (1 week Fedex is quite fast according to my scale), quickly turned it on, so far so good 10 minutes playing with it, except in SA, its showing -60dB noise floor (3rd pic) but in ebay listing, the seller can test it down to -110db floor (last pic) is this something wrong? or that low noise feature should be activated manually? i know very little on how to use this thing yet.

physical wise, it feels much lighter than Advantest R3465 albeit about the same size, and LeCroy WaveMaster is much heavier than both, having exercised on LeCroy unit, not much problem lifting this 8609. and if anyone want to mention the cooling fan sounds like jet engine, they should listen to my Epson D700 printer (barely visible at the back in 1st picture), it has double jet engine on both sides and noise is also double or triple, its just a printer.

and the boot screen showing copyright 2000, so the worst case, this is a 19 years old unit. does "Anritsu" means "Big Tree"?

[Mechatrommer]

managed to get the low noise floor, by turning on pre-amp under "amplitude" button, but... something looks ill (or normal?) on 3.148 GHz at larger span (last 3 pics)... and it has UNCAL tag on it next time, find how to CAL..

[eevcandies]

I was able to clearly see a 500MHz -140dBm signal pretty cleanly (limit of my R&S  SMHU gen (very nice gen, low cost), using -100dBm ref level and narrow span & BW.
Spectrum also measured within 1 dBm of the SMHU generator set to -140 dBm. 

for option 04:
Outline：Real RMS detection with digital calculation
RBW ：10Hz ~ 1MHz
Average noise level： RBW 10Hz, ATT 0dB
<-136.5dBm+1.5f(GHｚ) ；1MHz ~ 2.5GHz
< -132.5dBm+1.5f(GHｚ) ； 2.5GHz ~ 3.2GHz
< -128.5dBm ； 3.15GHz ~ 7.8GHz
< -119.5dBm ； 7.7GHz ~ 13.2GHz

for option 02:
Outline : Narrow resolution bandwidth with FFT
RBW : 1Hz ~ 1MHz
Average Noise Level： RBW=1Hz, ATT=0dB
< -146.5dBm+1.5f(GHｚ) ;1MHz ~ 2.5GHz
< -142.5dBm+1.5f(GHｚ); 2.5GHz ~ 3.2GHz
< -137.5dBm ; 3.15GHz ~ 7.8G
< -129.5dBm ; 7.7GHz ~ 13.2GHz

TO CALIBRATE JUST PRESS SHIFT CAL (LOOK AT THE NUMERIC KEYS)...it will take a few minutes...apply NO signal
THE UNCAL message is something else due to your settings...I had that too & at first thought it was a "CAL" issue, found it in the manual but can't remember what the correction was

Mine is a newer unit & also shows the 2000 year copyright   but ver is 4.12  ...Anritsu website has download for core software (doesn't mention which version)

found it for you (in operator manual #2, page 137)
The spectrum traces on the screen are displayed as shown at the left according to the sweep time. The optimum sweep time gives a waveform like [1]. However,
a sweep time that is too fast decreases the waveform amplitude on the display as shown in [2] and [3]. Therefore, the apparent bandwidth gets wider, and the
frequency also shifts. When waveform [1] cannot be maintained, “UNCAL” is displayed.

[YU7C]

... except in SA, its showing -60dB noise floor (3rd pic) but in ebay listing, the seller can test it down to -110db floor (last pic) is this something wrong? or that low noise feature should be activated manually? i

I believe that -60dBm is quite normal with 12GHz SPAN!

[Mechatrommer]

i've done shift cal, CAL ALL, and the weird shape still there, o well what should i expect from a 19 yrs old unit? i spent sometime just now just to get the darn dBc/Hz thing (instead of dBc/ch ) even managed to hang the OS few times (need soft button reboot) so i learnt a few things its intoxicating, maybe keep the pre-amp OFF is better to keep its longer life? maybe, only when i need low noise design then i will turn it ON, i dont even know how to design -60dB clean signal. i think this unit is better than my R3465, maybe its first job will be to repair the R3465 next time, now get back to work thanks guys for giving moral support.

[YU7C]

Is there any level correction turned on?

[eevcandies]

and the weird shape still there
 WHAT weird shape?

[Mechatrommer]

and the weird shape still there
 WHAT weird shape?

there.. (attached)

Is there any level correction turned on?

i've put the 8609 to the side, actually i'm in the middle of something else, i was just sneaking the time just to check the unit is ok before giving +ve feedback to the seller and close the tracking/accounting record (yes i do have them). i will check this again later, thanks for the idea.

[tmbinc]

This is how it looks on mine.  (Note: Mine doesn't have option 31, probably explaining the high noise floor in the Band 2+)

[Mechatrommer]

that looks about exactly the same, so i guess its normal, some kind of shifting in frequency sweep, different signal path or who knows what? good to know and have friend sharing similar behaviour, so i know its not so weird/broken. further its still -80dB down, still good enough compared to what i'm capable of.

This is how it looks on mine.  (Note: Mine doesn't have option 31, probably explaining the high noise floor in the Band 2+)

if i disable the pre amp, the weird "cliff" shape is reduced due to higher noise floor like attached.. but when pre amp ON, only after 3.148GHz go to lower noise, less than that still the same, i guess (low noise) pre amp only work upper than 3.148GHz..

btw, how do you people get nice screen captures from it? i dont have PCMCIA card attached. can i buy the usual PCMCIA/ATA to CF Card adapter?

[tmbinc]

(note 5dB vs. 10dB scale)

The pre-amp afaik only covers a low-frequency range up to 3GHz (can't find the exact number). What happens if you enable the pre-amp is that the UI shifts down the readings (i.e. scale) by the amplification of the pre-amp (to compensate/make readings correct), so you see the noise floor being reduced. However only in the actual area where the pre-amp is actually amplifying (<3GHz or whatever) the readings will be correct. That likely explains the case where the higher range noise drops visibly - but as the signal is not amplified, the SNR in that range will not be actually better. The device should show UNCAL in higher regions but I'm not in front of my device right now so I can't check.

The screenshots I did via PCMCIA -> CompactFlash adapter -> SD Card adapter -> MicroSD adapter -> microSD card. Feels like a Matryoshka doll adapter but actually works.

Unfortunately my instrument doesn't have ethernet enabled. It also doesn't have the power sensor, which is marked as "standard/always installed" in all Anritsu documents, but it's not active (or not present).

I've started to disassemble the firmware, found the menu description structures, and there's a special maintenance menu that allows to enable/disable options, but I failed to actually find the path that gets me into this module. (I could patch the software but the flash is not easily accessible so I didn't want to go that route) I _think_ there's a special key combination you have to press probably on the system screen.

If anyone has a service manual or any additional information, that'd be awesome.

[eevcandies]

You have the power meter...activate the W-CDMA module , then press TX TESTER--big button, then at bottom of menu goto page 2, at bottom of page 2 , select "POWER METER" .
This is also the same for EV-DO module (perhaps same for all modules).

MS8608A-08/MS8609A-08 Pre-Amplifier
Outline ：Built-in pre-amplifier with 20dB gain
Gain ：Gain 20dB(Typ.)
Noise exponent： 6.5dB(Typ.)；Input frequency <= 2GHz
Noise exponent:  12dB(Typ.)；Input frequency > 2GHz
Frequency range：100kHz ~ 3GHz
Average noise level： RBW 300Hz,VBW 1Hz,ATT 0dB
<= -137dBm+2.0xf(GHｚ) ;1MHz ~ 3GHz

[eevcandies]

Some extra tidbits of information:

https://manualzz.com/doc/21046659/mx268x30a-mx860x30a-product-introduction-wireless-lan-mea...

this might be useful for remote programming:
https://manualzz.com/doc/13932384/using-matlab-with-anritsu-ms2690a--ms2691a--and-ms2692a-s...

[eloyex]

hello friend, do you still have the ms8609 ?     is still working fine ?
no regrets so far ?

I am about to buy one for WIFI testing and WIFI related job , plus some analog traditional RF measures and would like to hear/comfirm your ideas about that unit.
I was looking for a low cost spectrum up to 6.2ghz, but this is way better range …  what do you think ?   any new feedback about this unit ?

Eloy / Miami

[eloyex]

hi   may I know how much you paid for the unit ?    it was LOADED with features !!!     I am about to buy one and need to know how is the price point on this product ..
thks !!