Author Topic: How to **Safely** and properly align a Transmitter with a spectrum analyzer  (Read 502 times)

0 Members and 1 Guest are viewing this topic.

Offline The_Spectrum.A_idiot

  • Contributor
  • Posts: 7
  • Country: cz
Dear All ,
I would like to thank everyone in advance that will answer my question/s and for their time ,

So I have a Spectrum analyzer a "Siglent SSA3021X" 

so I learned how to tune filters by doing and failing and doing some light reading , and finally succeeding
although I know that the front end of Spectrums are very sensitive into input power ,
so as per recommendation of a friend I bought this :"100W N Attenuator 50db male to female DC-3GHZ 50ohm RF"
https://www.ebay.com/itm/100W-N-Attenuator-50db-male-to-female-DC-3GHZ-50ohm-RF/272899586791

Although still I am very skeptical if I can plug a transmitter at its lowest power output at 2W with this attenuator in line and still be safe from the maximum power input the Spectrum can handle ,

Is there any other way ( mathematically to be able to calculate if I fall within the maximum acceptable range of my analyser and how to do this safely and properly )

I was thinking to get a signal sampler although they are too expensive and almost nowhere to be found in a price range around 100$ 200$ max 

So the question is how is the most proper and safe way to align and check a transmitter (health) and generally see the bandwidth etc 

If this question was answered elsewhere I am sorry in advance

thank you in advance

The_Spectrum.A_idiot

 

Offline radiolistener

  • Frequent Contributor
  • **
  • Posts: 542
  • Country: ua
yes, you can connect it.

2 W = 33 dBm, after 50 dB attenuator it will be 33 - 50 = -17 dBm, it should be safe for spectrum analyzer.

Before connect I suggest you to check if your attenuator works properly.


And note! These Chinese attenuators have just one side which is intended to apply high power. If you connect transmitter to the wrong side, your attenuator will burn out.

This is not trivial to understand which side is intended for high power. I found that Chinese manufacturers probably marking it with a small "SU" marking on connector, but this is not 100% information, just my finding. I burned out several attenuators in such way, because these attenuators don't have any instruction or manual or any kind of note about where is input and where is output and there is even no any mention that they are single way. And the worse thing is that sellers also don't know it  ;D So it's just your luck to find where is proper input  ;D

I think that your attenuator probably has high power input on the bottom connector (see first photo on the product description). Check if it has "SU" mark. If it is present, then this connector probably high power. But I cannot guarantee...  :D

May be someone can suggest more reliable way to check which connector is intended for high power input.
« Last Edit: August 14, 2019, 09:02:58 am by radiolistener »
 
The following users thanked this post: The_Spectrum.A_idiot

Offline hfleming

  • Contributor
  • Posts: 26
  • Country: no
Another trick that we used used years ago to protect spectrum analyzers, is to use a «protection box» at the input of the analyzer. All it was was a 375mA fast acting picofuse in line with the RF input. (In series with the fuse, you can also add a 100nF 1kV cap just to make sure no DC can get in, as well as a high value resistor to GND to drain static, if you forget to disconnect you analyzer  from your external antnnana). You can also use a PIN-diode limiter, that you can get on e-Bay. (Frequency respons might roll off, but you can easily compensate for it).

But in anycase, a 50dB pad will definitely be good enough, even if you forget to switch your radio to 2W. (I assume you don’t have a 1kW transmitter).

 
The following users thanked this post: The_Spectrum.A_idiot

Online fourfathom

  • Contributor
  • Posts: 45
  • Country: us
I bought a 50dB power attenuator (Chinese), and measured the actual attenuation at 39dB.  It was reasonably flat from DC to 1 GHz, so I just put a "39dB" label on it and still use it.  I sometimes use it for testing my ham transmitters.

Another test device I have is a -40dB tap.  This has two BNC connectors with a direct "through" connection: one goes to the device under test, and the other to a dummy load (a cheap 100W unit).  The third port is a SMA jack, with a 40dB "L-pad" attenuator, using a few low-power resistors.  My simple construction gave me a flat response out beyond 100MHz.
 
The following users thanked this post: The_Spectrum.A_idiot

Offline tautech

  • Super Contributor
  • ***
  • Posts: 15331
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Dear All ,
I would like to thank everyone in advance that will answer my question/s and for their time ,

So I have a Spectrum analyzer a "Siglent SSA3021X" 

so I learned how to tune filters by doing and failing and doing some light reading , and finally succeeding
although I know that the front end of Spectrums are very sensitive into input power ,
so as per recommendation of a friend I bought this :"100W N Attenuator 50db male to female DC-3GHZ 50ohm RF"
https://www.ebay.com/itm/100W-N-Attenuator-50db-male-to-female-DC-3GHZ-50ohm-RF/272899586791

Although still I am very skeptical if I can plug a transmitter at its lowest power output at 2W with this attenuator in line and still be safe from the maximum power input the Spectrum can handle ,

Is there any other way ( mathematically to be able to calculate if I fall within the maximum acceptable range of my analyser and how to do this safely and properly )

I was thinking to get a signal sampler although they are too expensive and almost nowhere to be found in a price range around 100$ 200$ max 

So the question is how is the most proper and safe way to align and check a transmitter (health) and generally see the bandwidth etc 
Welcome to the forum.

The datasheet on P5/9 sets out the maximum input parameters under Amplitude and Level:
https://www.siglenteu.com/wp-content/uploads/dlm_uploads/2017/10/SSA3000X_DataSheet_DS0703X-E04A.pdf
Of important note: Maximum average RF power 30 dBm, 3 minutes, fc≥10 MHz, attenuation >20 dBm, preamp off

Further in the Users manual in 2.1.3 Amplitude (P38-41/95) in 2.1.3.4 Units are the amplitude conversion relationship formulas.
https://www.siglenteu.com/download/7918/

There is 51dB of user input attenuation available but best practice with strong signals is to use external attenuation to get signals to low levels.

Avid Rabid Hobbyist
 
The following users thanked this post: The_Spectrum.A_idiot

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 4736
  • Country: au
yes, you can connect it.

2 W = 33 dBm, after 50 dB attenuator it will be 33 - 50 = -17 dBm, it should be safe for spectrum analyzer.

Before connect I suggest you to check if your attenuator works properly.


And note! These Chinese attenuators have just one side which is intended to apply high power. If you connect transmitter to the wrong side, your attenuator will burn out.

This is not trivial to understand which side is intended for high power. I found that Chinese manufacturers probably marking it with a small "SU" marking on connector, but this is not 100% information, just my finding. I burned out several attenuators in such way, because these attenuators don't have any instruction or manual or any kind of note about where is input and where is output and there is even no any mention that they are single way. And the worse thing is that sellers also don't know it  ;D So it's just your luck to find where is proper input  ;D


I think that your attenuator probably has high power input on the bottom connector (see first photo on the product description). Check if it has "SU" mark. If it is present, then this connector probably high power. But I cannot guarantee...  :D

May be someone can suggest more reliable way to check which connector is intended for high power input.

I dunno about Chinese stuff, but the normal convention with smaller attenuators is that the input to the attenuator is a female "N" connector, & the output a captive male "N", to make it easier to connect directly to a following test instrument.

Although the one linked to is getting a bit close to the size where it could damage the SA's input front panel connectors due to its mass & mechanical advantage, I'm pretty certain it would follow that convention.

Similar units from the USA, the UK, Japan, & Germany commonly have an arrow silkscreened onto the body, showing the signal path through the device, as do many of the large "thruline loads".

Higher power "thrulines" & the like, often have female "N" connectors on both ends--- maybe to stop morons hanging them off the front of SAs, but more probably, so they don't need to use an adaptor to connect to a standard cable.

At 2 watts  it is doubtful that it would make much difference if the OP got the attenuator input & output mixed up.
 
The following users thanked this post: The_Spectrum.A_idiot

Online 0culus

  • Super Contributor
  • ***
  • Posts: 1477
  • Country: us
  • Electronics, RF, and TEA Hobbyist
Also a quality directional coupler or RF sampling tee can be helpful. See:

 
The following users thanked this post: The_Spectrum.A_idiot

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 4736
  • Country: au
Also a quality directional coupler or RF sampling tee can be helpful. See:



Yes!

It is well worth keeping an eye out for stuff like this, even if the original intended frequency of use is not near the one you are using.

When my ex-employer scrapped their old Siemens TV Transmitters, we ferreted through them to find any useful RF  "bits".

There were several directional couplers, one of which was already "spoken for" for a project we had in mind.
The other one was "up for grabs", so I put my hand up.

Although the original application was around 200MHz, I have used it up to 470MHz, & sweeping it with a SA & tracking generator shows it is usable out to around 800MHz.

It came in handy in a later job, when we had to retune RF amplifier boards that the Chinese manufacturer conveniently "forgot" to tune to our operating frequency!


« Last Edit: August 14, 2019, 02:04:19 pm by vk6zgo »
 
The following users thanked this post: The_Spectrum.A_idiot

Offline borjam

  • Supporter
  • ****
  • Posts: 680
  • Country: es
  • EA2EKH
The best way is to use a proper dummy load which will dissipate the transmission power with a RF sampling port. That way you can obtain a low level sample of your signal which is safe to feed to a spectrum analyzer.
 
The following users thanked this post: The_Spectrum.A_idiot

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 9450
  • Country: gb
    • Having fun doing more, with less
Many spectrum analysers can't tolerate DC on their input. In this case a decent blocking capacitor is an extra safety measure.

A blocking capacitor is not a panacea. It will affect the lower frequency visible, if there are large transients when the transmitter is switched on/off they could reach the SA, and many ceramic caps are non-linear and so would create intermod product.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 
The following users thanked this post: The_Spectrum.A_idiot

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9742
  • Country: us
  • DavidH
A directional coupler between the amplifier and load is safer than a big attenuator and they are more common than samplers and often more rugged.
 
The following users thanked this post: The_Spectrum.A_idiot

Online fourfathom

  • Contributor
  • Posts: 45
  • Country: us
A directional coupler between the amplifier and load is safer than a big attenuator and they are more common than samplers and often more rugged.

Safer in what way?  A directional coupler still has to have an appropriate coupling ratio (or additional attenuators), and the frequency range is generally much narrower than a resistive attenuator.  I use directional couplers, but I do wonder about your "safer" comment.
 
The following users thanked this post: The_Spectrum.A_idiot

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9742
  • Country: us
  • DavidH
A directional coupler between the amplifier and load is safer than a big attenuator and they are more common than samplers and often more rugged.

Safer in what way?  A directional coupler still has to have an appropriate coupling ratio (or additional attenuators), and the frequency range is generally much narrower than a resistive attenuator.  I use directional couplers, but I do wonder about your "safer" comment.

A directional coupler does not have to dissipate the power so there are fewer issues with overheating.  This makes power attenuators and terminators less rugged or at least easier to misuse.  Samplers have the same advantage but are not as common as directional couplers.

Directional couplers and samplers can still of course fail catastrophically.
 
The following users thanked this post: The_Spectrum.A_idiot

Offline The_Spectrum.A_idiot

  • Contributor
  • Posts: 7
  • Country: cz
Dear radiolistener

thank you so very much for the help yes indeed after having the onboard attenuator at 50 DB and the external one and transmitting 2 W i get a very small signal well with in the range of the front end of the Spectrum ,

Thank you for your time

yes, you can connect it.

2 W = 33 dBm, after 50 dB attenuator it will be 33 - 50 = -17 dBm, it should be safe for spectrum analyzer.
This was very nice to know how to calculate everything

Before connect I suggest you to check if your attenuator works properly.


I did by using the onboard Tracking Gen ( i presumed is safe since it is on the instrument anyway )


This is not trivial to understand which side is intended for high power. I found that Chinese manufacturers probably marking it with a small "SU" marking on connector, but this is not 100% information, just my finding. I burned out several attenuators in such way, because these attenuators don't have any instruction or manual or any kind of note about where is input and where is output and there is even no any mention that they are single way. And the worse thing is that sellers also don't know it  ;D So it's just your luck to find where is proper input  ;D

I think that your attenuator probably has high power input on the bottom connector (see first photo on the product description). Check if it has "SU" mark. If it is present, then this connector probably high power. But I cannot guarantee...  :D

May be someone can suggest more reliable way to check which connector is intended for high power input.
No such marking was found
Although a friend with more experience told me that there are no such "sides" per say lastly no 'SU' mark exists I used it the way I found convenient the N type male on the instrument and N type female towards my transmitter (although data sheets of high end attenuators have sides per say in the datasheet )

anyway it was very helpful and enlightening ,

thank you for your time and effort and rely to my question

The_Spectrum.A_idiot
« Last Edit: August 15, 2019, 07:15:41 am by The_Spectrum.A_idiot »
 

Offline The_Spectrum.A_idiot

  • Contributor
  • Posts: 7
  • Country: cz
Dear hfleming
Thank you very much for responding to my post ,
Another trick that we used used years ago to protect spectrum analyzers, is to use a «protection box» at the input of the analyzer. All it was was a 375mA fast acting picofuse in line with the RF input. (In series with the fuse, you can also add a 100nF 1kV cap just to make sure no DC can get in, as well as a high value resistor to GND to drain static, if you forget to disconnect you analyzer  from your external antnnana). You can also use a PIN-diode limiter, that you can get on e-Bay. (Frequency respons might roll off, but you can easily compensate for it).

But in anycase, a 50dB pad will definitely be good enough, even if you forget to switch your radio to 2W. (I assume you don’t have a 1kW transmitter).
Fascinating box I may try to do that although for now that i am quite inexperienced with RF I would prefer a ready made commercial solution so that I won't kill anything by mistake ( as far as I know no warranty will cover idiot user mistake )

but still thank you so very much for your contribution

thank you for your time and effort and rely to my question

The_Spectrum.A_idiot

 

Offline The_Spectrum.A_idiot

  • Contributor
  • Posts: 7
  • Country: cz
Dear fourfathom

thank you very much for the contribution and the idea

I bought a 50dB power attenuator (Chinese), and measured the actual attenuation at 39dB.  It was reasonably flat from DC to 1 GHz, so I just put a "39dB" label on it and still use it.  I sometimes use it for testing my ham transmitters.

Another test device I have is a -40dB tap.  This has two BNC connectors with a direct "through" connection: one goes to the device under test, and the other to a dummy load (a cheap 100W unit).  The third port is a SMA jack, with a 40dB "L-pad" attenuator, using a few low-power resistors.  My simple construction gave me a flat response out beyond 100MHz.

hopefully soon I may try to make this so called (RF sampler lets call it ? ) although I would use N type all over for my convince although the response in my application is a bit more critical to go up to 500 mhz

anyhow when I will get more experience into making my own RF circuits I will try to do this ( seems very easy and a simple design with almost nothing could go wrong )

thank you for your time and effort and reply to my question

The_Spectrum.A_idiot

 

Offline The_Spectrum.A_idiot

  • Contributor
  • Posts: 7
  • Country: cz
Dear tautech

Thank you so very much for the links , I skimmed though most of them before buying the instrument although now it looks like I need to read it more and more thoroughly to get a better grasp of understanding my instrument

Dear All ,
I would like to thank everyone in advance that will answer my question/s and for their time ,

So I have a Spectrum analyzer a "Siglent SSA3021X" 

so I learned how to tune filters by doing and failing and doing some light reading , and finally succeeding
although I know that the front end of Spectrums are very sensitive into input power ,
so as per recommendation of a friend I bought this :"100W N Attenuator 50db male to female DC-3GHZ 50ohm RF"
https://www.ebay.com/itm/100W-N-Attenuator-50db-male-to-female-DC-3GHZ-50ohm-RF/272899586791

Although still I am very skeptical if I can plug a transmitter at its lowest power output at 2W with this attenuator in line and still be safe from the maximum power input the Spectrum can handle ,

Is there any other way ( mathematically to be able to calculate if I fall within the maximum acceptable range of my analyser and how to do this safely and properly )

I was thinking to get a signal sampler although they are too expensive and almost nowhere to be found in a price range around 100$ 200$ max 

So the question is how is the most proper and safe way to align and check a transmitter (health) and generally see the bandwidth etc 
Welcome to the forum.

The datasheet on P5/9 sets out the maximum input parameters under Amplitude and Level:
https://www.siglenteu.com/wp-content/uploads/dlm_uploads/2017/10/SSA3000X_DataSheet_DS0703X-E04A.pdf
Of important note: Maximum average RF power 30 dBm, 3 minutes, fc≥10 MHz, attenuation >20 dBm, preamp off

Further in the Users manual in 2.1.3 Amplitude (P38-41/95) in 2.1.3.4 Units are the amplitude conversion relationship formulas.
https://www.siglenteu.com/download/7918/

There is 51dB of user input attenuation available but best practice with strong signals is to use external attenuation to get signals to low levels.

now I will most probably buy a more reputable 50db attenuator and call it a day or maybe even an RF sampler ,

anyway it looks like I have my work cut out for me so I should get to reading :D Thank you again for pointing out the details i've missed and for the links too ,

Thank you in advance

The_Spectrum.A_idiot
 

Offline The_Spectrum.A_idiot

  • Contributor
  • Posts: 7
  • Country: cz
Dear vk6zgo

Thank you for your time and response to point out several things I didn't even thought about ,

yes, you can connect it.

2 W = 33 dBm, after 50 dB attenuator it will be 33 - 50 = -17 dBm, it should be safe for spectrum analyzer.

Before connect I suggest you to check if your attenuator works properly.


And note! These Chinese attenuators have just one side which is intended to apply high power. If you connect transmitter to the wrong side, your attenuator will burn out.

This is not trivial to understand which side is intended for high power. I found that Chinese manufacturers probably marking it with a small "SU" marking on connector, but this is not 100% information, just my finding. I burned out several attenuators in such way, because these attenuators don't have any instruction or manual or any kind of note about where is input and where is output and there is even no any mention that they are single way. And the worse thing is that sellers also don't know it  ;D So it's just your luck to find where is proper input  ;D


I think that your attenuator probably has high power input on the bottom connector (see first photo on the product description). Check if it has "SU" mark. If it is present, then this connector probably high power. But I cannot guarantee...  :D

May be someone can suggest more reliable way to check which connector is intended for high power input.

I dunno about Chinese stuff, but the normal convention with smaller attenuators is that the input to the attenuator is a female "N" connector, & the output a captive male "N", to make it easier to connect directly to a following test instrument.

Although the one linked to is getting a bit close to the size where it could damage the SA's input front panel connectors due to its mass & mechanical advantage, I'm pretty certain it would follow that convention.

Similar units from the USA, the UK, Japan, & Germany commonly have an arrow silkscreened onto the body, showing the signal path through the device, as do many of the large "thruline loads".

Higher power "thrulines" & the like, often have female "N" connectors on both ends--- maybe to stop morons hanging them off the front of SAs, but more probably, so they don't need to use an adaptor to connect to a standard cable.

At 2 watts  it is doubtful that it would make much difference if the OP got the attenuator input & output mixed up.

The direction of the attenuator ( yes indeed there are no markings , and indeed that is the most typical way i looked on  more expensive attenuators on the web that the male part is to the instrument and female part to the transmitter )

regarding the mechanical advantage once I plugged it on the SA I didn't think about it but now I am definitely buying a barrel connector and having a small patch cable for it

thank you in advance for your time and ideas :D
The_Spectrum.A_idiot
 

Offline The_Spectrum.A_idiot

  • Contributor
  • Posts: 7
  • Country: cz
Dear 0culus

Thank you very much for the youtube video amazing channel definitely going to watch all his videos :D
Also a quality directional coupler or RF sampling tee can be helpful. See:



Damn that capacitive sampler looks amazing ( definitely making one out of copper pipe as the person on the video suggested)

thank you so very much
The_Spectrum.A_idiot
 
The following users thanked this post: 0culus

Online 0culus

  • Super Contributor
  • ***
  • Posts: 1477
  • Country: us
  • Electronics, RF, and TEA Hobbyist
Alan is a forum member too; FWIW.
 
The following users thanked this post: The_Spectrum.A_idiot

Offline hfleming

  • Contributor
  • Posts: 26
  • Country: no
Siglent USA has an app note describing an input protection circuit.
https://siglentna.com/application-note/diy-spectrum-analyzer-input-protection/
 

Online fourfathom

  • Contributor
  • Posts: 45
  • Country: us
Dear fourfathom

thank you very much for the contribution and the idea

Another test device I have is a -40dB tap.  This has two BNC connectors with a direct "through" connection: one goes to the device under test, and the other to a dummy load (a cheap 100W unit).  The third port is a SMA jack, with a 40dB "L-pad" attenuator, using a few low-power resistors.  My simple construction gave me a flat response out beyond 100MHz.

hopefully soon I may try to make this so called (RF sampler lets call it ? ) although I would use N type all over for my convince although the response in my application is a bit more critical to go up to 500 mhz

The reason my tap was only flat to 100MHz was that I used 1/4/W resistors with long leads and paid no attention to high frequency layout.  If you do it right (say smt resistors on microstrip PCB) there's no reason your tap couldn't be usable into the GHz range.

And *thank you* for responding so nicely!  Many times someone gets an answer and then disappears.  We like to hear how it all comes out.
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 4736
  • Country: au
If your sampler varies in output overa wide frequency range, it isn't the end of the world, as sometimes you only need a small part of that range when testing.

If you havd a tracking generator with a Spec An, you can determine if the range you need is fairly flat, prior to using it to test a transmitter.
Even if there is a bit of a "slope" you can correct for that.
This is how we did things back in the old days, when the flatnesss of equipment was not always a given.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf