Author Topic: Rigol 815 TG Input attenuator inquiry, anyone ?  (Read 1788 times)

0 Members and 1 Guest are viewing this topic.

Offline ytterligare

  • Contributor
  • Posts: 12
  • Country: it
Rigol 815 TG Input attenuator inquiry, anyone ?
« on: April 01, 2016, 02:02:48 AM »
Hello, a quick request to 815's owners and not : I read from the Manual that there's an RF attenuator built-in, with a default 10 dB parameter, with 1 dB step to max 30 dB.

Also, the manual says that the Max Input Level CW RF power is +20 dBm ( 100 mW), but I don't understand if this figure is when the max attenuator is cranked up to max ( 30 dB ) or not.
In short : will I need an additional external 30 dB attenuator in order to test HAM RTX from 1 to 1300 Mhz with max 120 W output, or will the internal one be sufficient ?

Thank you !!

Andrea
 

Offline Performa01

  • Frequent Contributor
  • **
  • Posts: 344
  • Country: at
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #1 on: April 01, 2016, 02:26:49 AM »
No one needs to own an 815 to answer that question, since it is just basics. And looking at the datasheet would help indeed…

Even without any datasheet, I guess I will not be too far off in assuming the 1dB CP will be somewhere near 0dBm. From that it is clear that the level at the mixer input has to stay well below that and +20dB input power can only be applied with at least 20dB (more likely 30dB) input attenuation.

Furthermore, the IIP3 is most likely around +10dBm, so for any decent 3rd order dynamic range, the signal at the mixer input has to be -20dBm or lower. So if you have a max. 30dB input attenuator built in, the level at the input jack of the instrument cannot exceed +10dB for sensible measurements.

Finally, it should be obvious that the internal attenuator will not be able to handle 120W. In fact, it will be more like 1W (+30dBm), at which level hopefully some protection circuit will kick in and interrupt the input signal path. But even that is only guaranteed up to 50W at most and even higher levels will inevitably cause permanent damage.

So the final answer is yes, you need an external 30dB attenuator that is capable to handle at least 120W, which will be a rather big and expensive beast.
 

Offline K5HJ

  • Regular Contributor
  • *
  • Posts: 56
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #2 on: April 01, 2016, 06:07:03 AM »
I wouldn't get near the 815TG with a 120W transmitter without at least 40dB of attenuation.  And the attenuator must be capable of dissipating all of the 120W.
Otherwise, your poor 815TG will go up in smoke very quickly.

Running near the absolute maximum input level is never a good idea.

Randy
 

Offline wkb

  • Frequent Contributor
  • **
  • Posts: 519
  • Country: nl
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #3 on: April 01, 2016, 08:31:26 AM »
No one needs to own an 815 to answer that question, since it is just basics. And looking at the datasheet would help indeed…

Even without any datasheet, I guess I will not be too far off in assuming the 1dB CP will be somewhere near 0dBm. From that it is clear that the level at the mixer input has to stay well below that and +20dB input power can only be applied with at least 20dB (more likely 30dB) input attenuation.

Furthermore, the IIP3 is most likely around +10dBm, so for any decent 3rd order dynamic range, the signal at the mixer input has to be -20dBm or lower. So if you have a max. 30dB input attenuator built in, the level at the input jack of the instrument cannot exceed +10dB for sensible measurements.

Finally, it should be obvious that the internal attenuator will not be able to handle 120W. In fact, it will be more like 1W (+30dBm), at which level hopefully some protection circuit will kick in and interrupt the input signal path.

The 815 is max 20dBm input rated.  As for protection circuits.. I am not holding my breath.  And I am sure not gonna try in my 815  :-BROKE
 

Online TheSteve

  • Supporter
  • ****
  • Posts: 2150
  • Country: ca
  • GHz or bust
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #4 on: April 01, 2016, 08:41:22 AM »
If you need to see accurate signal levels then go with 40dB of attenuation as already mentioned. If you just need to see relative signal levels then you can buy a dummy load with a sample port. On the surplus market they are often much cheaper then a proper 120+ watt attenuator.
VE7FM
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 2231
  • Country: gb
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #5 on: April 01, 2016, 10:01:27 AM »
For that power level and frequency range I use one of these attenuators from Bird

http://www.birdrf.com/Products/Attenuators/1kW/1000-WA-Series_RF-Attenuators.aspx#.Vv2qMZXmqjE

Mine is a custom 40dB version and these things are not cheap! If you want a stable and accurate attenuator up to 1.5GHz at a continuous 120W then this is the kind of thing you would need to use. It's rated to 1kW but in my experience you need to use an attenuator rated at at least 5 times the power you want to put through it if you want the attenuator to be stable and reliable in the long term. You could try using a 300W attenuator to save cost but you would need to blow it with a fan. Never let the attenuator get 'very' hot. It shortens the useful life a lot.

If you explore other (cheaper, less accurate) methods, such as couplers or crude samplers then I recommend you fit a protection limiter ahead of the spectrum analyser input port as a means to protect against failures or mistakes wrt the sampler/coupler connections.

A typical limiter can protect up to an accidental hit of 5-10 Watts at the analyser input so it can be worth including it in your sampler/coupler setup.
« Last Edit: April 01, 2016, 10:11:59 AM by G0HZU »
 

Offline ywara

  • Contributor
  • Posts: 21
  • Country: us
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #6 on: April 01, 2016, 10:37:03 AM »
A coupler is the way to go. A decent coupler will be almost as flat as a bad attenuator. Any 120W attenuator you can afford to buy is going to be fairly bad. If you characterize the coupler across the range of interest, you can remove any flatness issues.
 

Offline ytterligare

  • Contributor
  • Posts: 12
  • Country: it
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #7 on: April 01, 2016, 09:03:44 PM »
Thank you gentlemen, plenty of info and ideas.
Being my first SA, and loving to tinker with such glowing & beeping devices as an amateur, I can only make good use of your experience.
First of all, yes : I cannot afford to get a 1KW Bird Attenuator, although it would be perfect, so I will have to lower my target.

The more powerful radio I own got this specifications :

Output power (cont. adj.)

HF/50 Mhz -> 2-100W (SSB/CW/RTTY/FM/DM) 2-25W (AM)
144 MHz -> 2-100W (SSB/CW/RTTY/FM/DV)
430/440 MHz -> 2-75W (SSB/CW/RTTY/FM/DV)
1200MHz Mhz - > 1-10W (SSB, CW, RTTY, FM, DV)

Most of the other radios will not go beyond 50 W @144 Mhz and/or 35 W @ 430 Mhz

As you can see, I will not ( for the moment ) need to go beyond 1200 Mhz/10W, no AMPS.

I was thinking about this attenuator : http://194.75.38.69/pdfs/BW-40N100W+.pdf

But from the data sheet I read : "Average power at 25°C ambient, derate linearly to 50W at 100°C, bi-directional"

Now, from your experience, to stay within amateur stability & reliability tolerances how much time it takes for a device like that to get "hot" as G0HZU suggest with the power and frequencies involved ?

Thank you

Andrea

Addendum :
As per G0HZU suggestion, I decided to add this Limiter  http://194.75.38.69/pdfs/VLM-33W-2W+.pdf to the setup....according to the specs, a MAX 14.49 dBm should be the output when +33 dBm are applied in the input.
Any idea how the limiter will behave for inputs below +12 dBm ?
« Last Edit: April 01, 2016, 11:54:50 PM by ytterligare »
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 2231
  • Country: gb
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #8 on: April 02, 2016, 12:06:36 AM »
A lot depends on how you plan to use the attenuator in terms of duty cycle. If you only want to do brief keyup tests then you could use the attenuator in your link.

Also, a lot of 100W ham transceivers are really only capable of moderate duty cycle operation anyway.

But if you wanted to transmit for 10 minutes at full 100W and then maybe transmit again for 10 minutes a few minutes later (etc, etc) then at the very least you would have to force air cool the attenuator.

At my place of work I've seen a lot of 25W, 100W and 300W attenuators (rated at several GHz from the top manuf) fail or go erratic after a certain time.

The cold>hot>cold>hot>cold thermal stress gets to them eventually and sometimes the attenuation become unpredictable over temperature/power and sometimes the internal connections become erratic. We must have thrown away dozens of them over the years and I don't think any of them were misused/abused other than the user (probably)expected the attenuator to be rated at the power level on the label.

If you do buy the 100W attenuator, then my advice would be to force air cool it even if you only want to use it for a few minutes at a time at anything over 50W.
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 2231
  • Country: gb
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #9 on: April 02, 2016, 12:24:09 AM »
Quote
Any idea how the limiter will behave for inputs below +12 dBm ?
I've not used that type of limiter but a typical limiter will be fine up to about +3dBm (PEP) and above this it will begin to show a lot of distortion on a two tone test (two tones at -3dBm each) on a spectrum analyser as it enters the limiting zone.

So I generally make sure I don't exceed 0dBm PEP at the analyser when using any of my limiters here at home.

I made my own limiters for LF though to 500MHz. Normally, you would want/expect the limiter to cope up to 5W before it becomes damaged and fails. Two of my limiters include an RF fuse that (hopefully) will blow before the limiter diodes do if I ever blast the limiter with power >>10W. 
« Last Edit: April 02, 2016, 12:36:41 AM by G0HZU »
 

Offline ywara

  • Contributor
  • Posts: 21
  • Country: us
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #10 on: April 02, 2016, 11:20:22 AM »
At my place of work I've seen a lot of 25W, 100W and 300W attenuators (rated at several GHz from the top manuf) fail or go erratic after a certain time.

The cold>hot>cold>hot>cold thermal stress gets to them eventually and sometimes the attenuation become unpredictable over temperature/power and sometimes the internal connections become erratic. We must have thrown away dozens of them over the years and I don't think any of them were misused/abused other than the user (probably)expected the attenuator to be rated at the power level on the label.

I can't say how your failed units were constructed, but traditionally these were manufactured with large pyrolytic carbon resistive elements on BeO, AlN, or occasionally alumina ceramic. This makes them fantastic at surviving high power, high temperature applications but incredibly unstable. Oil immersion helps somewhat with stability, but not enough. On top of that, getting the power from metal to carbon is always a challenge. These contact points are a frequent source of failure, and it is not always heat that is the culprit. These manufacturers are not generally selling overspecified parts, but the failure rates are unpredictable and much higher than would be acceptable for most normal consumers.
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 2231
  • Country: gb
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #11 on: April 04, 2016, 01:53:08 AM »
I did take a few of them apart and sometimes the failure occurs in the RF feed connection just behind the coaxial connector. I'm definitely no expert when it comes to the various design compromises that get made with these attenuators but it looks to me like they often use springy connections here and presumably this is to cope with thermal stress. Often, the attenuator will develop a poor connection here and this can give various symptoms. Other failure modes include a significant change in attenuation or return loss.

Some of the high power directional ones use a series of flanged attenuators that are bolted down in series along a rail. The idea is to use progressively larger attenuator values in each section along the rail until the final (40dB?) total is achieved. I've see these fail at the coax connector 'springy' connection and I've also seen the flanged resistor 'tab' connection suffer a fractured joint. It's as if the thermal stresses eventually break the soldered tab connection and it becomes erratic.

Some of our attenuators get used for several days continuous at high power when we soak test some of our amplifier products. Others are exposed to random duty cycle use for long periods and this can be for frequencies from VHF through several GHz for some of our amplifiers. So they probably see far more stresses than most. We also buy them from various manufacturers rather than just one source.

I wouldn't recommend anyone to buy 'used' attenuators like this because you won't know the history and they may prove to be unreliable straight away.
 

Offline Earendil

  • Regular Contributor
  • *
  • Posts: 89
  • Country: hu
Re: Rigol 815 TG Input attenuator inquiry, anyone ?
« Reply #12 on: April 04, 2016, 03:11:09 AM »
I have one of these puppies:
http://www.aliexpress.com/item/100W-N-type-RF-coaxial-fixed-attenuator-1-40dB-DC-3Ghz-free-shopping/32494488595.html

I know that people are cautious with cheap Chinese stuff and with a good reason but this seems to work reasonably fine.
I've tested it with a 3Ghz SA + RF signal source and also with a 75W E&I RF amplifier at 30Mhz.

It may not be completely flat but you can compensate it with the Rigol 815.
« Last Edit: April 04, 2016, 03:20:27 AM by Earendil »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf