Author Topic: RIGOL DSA815-TG @ 10 MHz issue/bug  (Read 25695 times)

0 Members and 1 Guest are viewing this topic.

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
RIGOL DSA815-TG @ 10 MHz issue/bug
« on: August 04, 2015, 06:16:06 PM »
Hello everyone,

this is my first post in this forum even if I usually read it.

It's almost 1 year that I own the Rigol SA and since the first day I saw a strange behavior when I measure the response curve of a crystal filter with passband center @ 10 MHz. 

When the sweep pass exactly @ 10.000.000 Hz  the trace on the screen is distorted as there's a hop in the frequency sweep of the traking generator.

I attach a screenshot where the distortion is really bad.

Is there people with the same problem? If so, is it possible to know the hardware/firmware version of those with the issue and the hardware/firmware version of those without issue?

Thank you for any infos

« Last Edit: August 08, 2015, 10:55:57 PM by Giorgio »
 

Offline pauln

  • Contributor
  • Posts: 28
  • Country: au
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #1 on: August 04, 2015, 07:27:56 PM »
To start with with - are you sure the filter is OK to begin with? 
 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #2 on: August 04, 2015, 07:34:54 PM »
Yes, everything is set as Rigol asked when I opened a ticket with them.

 

Offline pauln

  • Contributor
  • Posts: 28
  • Country: au
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #3 on: August 04, 2015, 07:50:26 PM »
OK Giorgio.

I have made and measured many home-built Crystal filters for another project using the 815TG.  I have found that apart from the response of the filter itself, the impedance matching was critical and I have certainly seen responses such as yours when trying to match their in/out impedances.

I will rephrase my question:  are you certain that the filter under test does not have the blip built in to it?  and you are measuring it correctly?

Paul vk4apn 
 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #4 on: August 04, 2015, 08:23:54 PM »
Yes Pauln,

I tried all the possible settings, I already checked with onother brand new (1 year ago) SA by the distributor and it gave the same result. It's an issue presents in all units (at least some firmware/hardware combination)

Rigol even confirming that it's an issue was not able to find (at the time of writting) a solution.

Now I am looking for firmware/hardware combination to look for an used unit without issue.
 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #5 on: August 05, 2015, 05:13:44 AM »
Here a new screenshot with more evidence of the 10MHz distortion.

the setup:

First normalized with 0dBm TG -> SA Input

CF 10.000 MHz
BW 300 Hz
2 dB/division
Ref -40 dBm
TG level 0 dBm

then connected one 10 MHz crystal in series between TG output and SA input

the markers 2, 3 and 4 show the missing part of the trace shape.


 

Online Bud

  • Super Contributor
  • ***
  • Posts: 1771
  • Country: ca
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #6 on: August 05, 2015, 05:26:36 AM »
The change is rather abrupt and does look as a firmware issue
 

Offline K5HJ

  • Regular Contributor
  • *
  • Posts: 50
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #7 on: August 05, 2015, 05:27:52 AM »
Can you try a slower sweep time?

Randy
« Last Edit: August 05, 2015, 05:42:10 AM by K5HJ »
 

Online TheSteve

  • Supporter
  • ****
  • Posts: 2006
  • Country: ca
  • GHz or bust
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #8 on: August 05, 2015, 05:37:47 AM »
I don't have a DSA815 to experiment with any more but I do recall that the internal 10 MHz reference does leak into the SA input. I wonder if there is some interaction occurring.
VE7FM
 

Online Bud

  • Super Contributor
  • ***
  • Posts: 1771
  • Country: ca
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #9 on: August 05, 2015, 05:46:26 AM »
Seems ADC missing a sample or few of them. Also appears everything to the right of the blip is delayed and because of that readings are wrong.
 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #10 on: August 05, 2015, 05:48:09 AM »
Here is a 10 second sweep with the same setup as before.

Yes it's a firmware problem, Rigol said that this happen with the new hardware but they were not able to find a solution for now ( and for now I mean 1 year)

So I want to know when this flaw showed up for the first time, with your help. Maybe new hardware with the old firmware was not showing this bizare behavior that make my unit unuseful @ 10 MHz as tracking generator.

 

Online TheSteve

  • Supporter
  • ****
  • Posts: 2006
  • Country: ca
  • GHz or bust
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #11 on: August 05, 2015, 06:08:22 AM »
Perhaps the firmware omits some samples right around 10 MHz to hide the LO leakage spike.
VE7FM
 

Offline K5HJ

  • Regular Contributor
  • *
  • Posts: 50
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #12 on: August 05, 2015, 06:13:02 AM »
It might be that the generator or analyzer is going through a band switch point or a correction table lookup at that exact frequency.
I'll have to see if I can duplicate this myself.

Randy
 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #13 on: August 05, 2015, 06:28:30 AM »
Thank you Randy,

I don't know if you have a crystal or a filter close to 10 MHz. If not you can use a LC tuned circuit, even with very wide passband the trace should be distorted (I hope not for you).

By the way here is my unit information:

Version of Main Board: 00.07
Version of Radio Frequency Board FPGA: 00.05
Version of Digital Board FPGA: 00.05
Version of Firmware: 00.01.13
Version of Boot: 00.01.04


 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #14 on: August 05, 2015, 06:46:46 AM »
TheSteve,

well maybe there's something involved with the 10 MHz clock.Could be the reason why they can't find a solution?

What's the differences between the old hardware and the new one?

Ah and I found another way to check if the unit has the same problem as mine:

Instead to use a filter connected between the TG output and SA input, you can easly use an external generator @10 MHz and feed the SA input and in the mean time activating the tracking generator. without using the output.

This let me see that  the SA part is the one with the hop in frequency.

And if you switch on and off the TG when you feed the input with the external generator, the trace shift a few to the right and left of the center.

This is just to verify if the problem is present when you have no crystal or filter for 10 MHz.

Hope more people will check their units.

 

Offline cncjerry

  • Supporter
  • ****
  • Posts: 659
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #15 on: August 05, 2015, 11:57:30 AM »
If there was a problem with the SA or TG the glitch would be there for all DUTs.  Also, if the SA was normalized which would imply a straight line without a DUT it should normalize-out the glitch.  So it has to be amplitude related.  I wonder if your filter peak is exceeding the ADC at resonance.  Is the glitch there if you externally attenuate the input?  after you normalize the SA, try putting an attenuator after the filter and run the sweep.  On most SA you would see a rounded peak causing compression when you exceed the dynamic range of the SA.
 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #16 on: August 05, 2015, 04:29:27 PM »
Hi Cncjerry,

thank you for the hints, trust me, I already tried every setting possible to minimize it.

As I wrote before, Rigol support already confirmed that this problem is present in all their units and that they can't find a solution. They explained that at the frequency of 10.000 Mhz something happens because of the TG design and is hardware related and this is why they can't find a solution with the firmware. Maybe in the future they change again their hardware but for now they didn't gave me any hope.

This is why I ask if others can check the hardware/firmware and make the test to confirm if the problem is present or not.

I don't know but maybe the same "glitch" is on other frequencies and because it's visible only with narrow filter settings it's hard to find it in 1,5 GHz band sweep.

It's just because I had to build a 10 MHz filter that I found the anomaly in the response trace. You know.... when you are so lucky......you find a needle in a haystack  |O

Thank you
 

Offline Mark

  • Regular Contributor
  • *
  • Posts: 171
  • Country: gb
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #17 on: August 05, 2015, 09:29:17 PM »
I hooked up a simple parallel LC circuit and saw this at 10MHz:

 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #18 on: August 05, 2015, 10:14:31 PM »
Hi Mark,

I am sorry but you have a unit with the same problem as mine.

Can you post your system infos?

Thank you
 

Offline cncjerry

  • Supporter
  • ****
  • Posts: 659
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #19 on: August 06, 2015, 12:24:25 AM »
And you don't see it when you normalize with a straight cable?  That's weird.
 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue
« Reply #20 on: August 06, 2015, 12:46:18 AM »
Cncjerry,

when I normalize I see just the straight line, everything seems ok.

If you don't normalize and look the trace at 10.000.000 Hz you will see like a step of 2-3 pixels. That step is where the distortion will take place when connecting a DUT after nomalize.
 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue/bug
« Reply #21 on: August 08, 2015, 11:00:09 PM »
Is Mark the only one having the same problem. Anybody else who confirms the hardware/firmware pairs showing the bug?
 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 3694
  • Country: gb
Re: RIGOL DSA815-TG @ 10 MHz issue/bug
« Reply #22 on: August 08, 2015, 11:42:40 PM »
I built a BPF with a 33pF and 10uH in series, and show similar results.

(Gotta love HF, even BNC->croc clip leads work!)





 

Offline Howardlong

  • Super Contributor
  • ***
  • Posts: 3694
  • Country: gb
Re: RIGOL DSA815-TG @ 10 MHz issue/bug
« Reply #23 on: August 08, 2015, 11:51:23 PM »
Normalized:

 

Offline Giorgio

  • Contributor
  • Posts: 30
  • Country: 00
Re: RIGOL DSA815-TG @ 10 MHz issue/bug
« Reply #24 on: August 09, 2015, 01:25:33 AM »
Howardlong in your test the bug is more evident. Thank you for the test and infos.
Anybody else who wants to check with Howardlong like DUT?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf

 

http://opalkelly.com/