Spectrum Analyzer - Rigol DSA815

[N8AUM]

Another use for "ramping up" the output this way is if lets say you know that you have loss in your connectors on the high end so now you can compensate for that loss using the power sweep function.
 

[PA0PBZ]

I wonder why it skips a one dB step every 5 steps.

[ted572]

I wonder why it skips a one dB step every 5 steps.

Its not skipping 1dB, its jumping 2 dB.
You should see the results of Power Sweep with its 1db amplitude steps over a 20dB range.  Although 1 or more of the steps may have 2 steps lumped together as a 2dB ramp up (probably a Bug). 

[N8AUM]

I wonder why it skips a one dB step every 5 steps.

Its not skipping 1dB, its jumping 2 dB.
You should see the results of Power Sweep with its 1db amplitude steps over a 20dB range.  Although 1 or more of the steps may have 2 steps lumped together as a 2dB ramp up (probably a Bug).

Its because of the resolution with the parts they are using and or the software. Normally when you sweep a wider range you will not notice these "steps". 

[Howardlong]

You can determine if Power Sweep works in your DSA815-TG by connecting a coax cable between the DSA815-TG's 'TG Output' and 'SA Input', and then selecting the following settings:

Preset (Green key), FREQ: 750MHz (the Test Frequency), SPAN: 100Hz, RBW: 100Hz (Required if 10Hz RBW Option is installed),  TG: On, TG Level: -20dB, Power Range: 20dB, Power Sweep: On, AMP: Auto Scale, Scale/Div: 2dB, Ref Level: -1dB (or A/R).

You should see the results of Power Sweep with its 1db amplitude steps over a 20dB range.  Although 1 or more of the steps may have 2 steps lumped together as a 2dB ramp up (probably a Bug).

Here's a video of mine not working: http://youtu.be/yRFq4LElbw0

It was shot in 1080p/50fps, it will take a few minutes to show at the full resolution.

[ted572]

You can determine if Power Sweep works in your DSA815-TG by connecting a coax cable between the DSA815-TG's 'TG Output' and 'SA Input', and then selecting the following settings:

Preset (Green key), FREQ: 750MHz (the Test Frequency), SPAN: 100Hz, RBW: 100Hz (Required if 10Hz RBW Option is installed),  TG: On, TG Level: -20dB, Power Range: 20dB, Power Sweep: On, AMP: Auto Scale, Scale/Div: 2dB, Ref Level: -1dB (or A/R).

You should see the results of Power Sweep with its 1db amplitude steps over a 20dB range.  Although 1 or more of the steps may have 2 steps lumped together as a 2dB ramp up (probably a Bug).

Here's a video of mine not working: http://youtu.be/yRFq4LElbw0

It was shot in 1080p/50fps, it will take a few minutes to show at the full resolution.

Have you contacted Rigol about Power Sweep not functioning on your unit?

It is interesting that the only other people here I know of that have reported Power Sweep not working is 'obiter' and 'N8AUM'.  There is one thing in common between the three of you, which is the U1105 modification.  Is there any chance that connecting pins 7 and 8 of U1105 could have effected this function?  Although it seems  unlikely to me.  Anyway at this time it would to nice to hear from others having this issue that have NOT touched U1105 in their DSA815-TG.  I'm very sorry for suggesting this could even be a possibility, and I truly hope that it is not.

[Teneyes]

I wonder why it skips a one dB step every 5 steps.

Its not skipping 1dB, its jumping 2 dB.
You should see the results of Power Sweep with its 1db amplitude steps over a 20dB range.  Although 1 or more of the steps may have 2 steps lumped together as a 2dB ramp up (probably a Bug).

Here is my b'11' .
I was going to suggest you get some cheese as I was going to do a lot of Whining

I played at 10 and 20 Seconds sweep time.
Playing with the 'Power Sweep' I see it as a function that combines at Power Stepping ramp with a frequency Sweep , from lower to higher. The Steps are normally in 1db steps and the Frequency is from a start to stop, (or center Frequency and span).
I do not know if this function is implemented on other  SA?
I suppose to test a Amplifier doing it by  changing 2 parameters at the same time does the test quickly. I would it one at a time. which can done with the DSA815, (set the span narrow)
As for the 2db jump and the no jump later, can be used to see if the response of the DUT is different between a 1 and 2 db step change, and also if the settling time after a power change can be observed for the long change spot.
(I once thought that an Engineer drew a step plot to explain to a programmer how the function was to be implemented, and then programmer copied the crude plot exactly with odd steps  )

One room for improvement is that the stepping ramp should hold the last step for the full period and NOT just a short time. That is for 20 db range change . divide the span into 21 parts.
Note there is an long hold at one power step, but a double 2db step to offset the long hold.

What is the cause of the sloping change of the power step (fast sweeps)?
Is it the:
   program? showing smooth ramping ( but why hold at slow time sweeps)
   response of the Func. Gen power change?
   response of the SA? averaging?
   response of the Display program?

In order to see in the time domain , I watched two 'Power Ramp Sweeps' with My DSO
Center= 100 Mhz , span 50MHz and 2 Mhz (Note displayed Freq.)

[Teneyes]

video of mine not working:

Is there any chance that connecting pins 7 and 8 of U1105 could have effected this function?  Although it seems  unlikely to me. 

Is there a preset condition for the tracking gen that is not set??
The Video shows the power drops? Does the output turn completely Off when the Ref Level was set?

[N8AUM]

You can determine if Power Sweep works in your DSA815-TG by connecting a coax cable between the DSA815-TG's 'TG Output' and 'SA Input', and then selecting the following settings:

Preset (Green key), FREQ: 750MHz (the Test Frequency), SPAN: 100Hz, RBW: 100Hz (Required if 10Hz RBW Option is installed),  TG: On, TG Level: -20dB, Power Range: 20dB, Power Sweep: On, AMP: Auto Scale, Scale/Div: 2dB, Ref Level: -1dB (or A/R).

You should see the results of Power Sweep with its 1db amplitude steps over a 20dB range.  Although 1 or more of the steps may have 2 steps lumped together as a 2dB ramp up (probably a Bug).

Here's a video of mine not working: http://youtu.be/yRFq4LElbw0

It was shot in 1080p/50fps, it will take a few minutes to show at the full resolution.

Have you contacted Rigol about Power Sweep not functioning on your unit?

It is interesting that the only other people here I know of that have reported Power Sweep not working is 'obiter' and 'N8AUM'.  There is one thing in common between the three of you, which is the U1105 modification.  Is there any chance that connecting pins 7 and 8 of U1105 could have effected this function?  Although it seems  unlikely to me.  Anyway at this time it would to nice to hear from others having this issue that have NOT touched U1105 in their DSA815-TG.  I'm very sorry for suggesting this could even be a possibility, and I truly hope that it is not.

My 815 is a virgin, no mods. It came with 1.09 firmware. I tried with 1.12 with same results. Unfortunately it also came with 1.04 bootloader so am not able to try it with older firmware. I spoke with Rigol and they will exchange unit for a new one once they come in. Because of the strike out west might be a while before new stock comes in.
 

[ted572]

video of mine not working:

Is there any chance that connecting pins 7 and 8 of U1105 could have effected this function?  Although it seems  unlikely to me. 

Is there a preset condition for the tracking gen that is not set??
The Video shows the power drops? Does the output turn completely Off when the Ref Level was set?

Teneyes:

Please understand that Power Sweep does do what it is expected to do for checking the linearity of an Amplifier, such as showing as the 1dB Compression Point, etc.  For myself and others it works.  And although sometimes it skips a 1dB increase step in amplitude and does 2dB instead, it still does the job.  But, yes the bug should be fixed to provide smoother display of an Amplifiers output.

It does not work at all for three people with version .01.12 firmware and BootLoader .04.  They had also modified  the U1105 DSA815-TG in their D815-TG.  In any case this is a separate issue.

See the PDf here -> https://www.eevblog.com/forum/testgear/spectrum-analyzer-rigol-dsa815/msg603946/#msg603946   and the way to test for Power Sweep here -> https://www.eevblog.com/forum/testgear/spectrum-analyzer-rigol-dsa815/msg604298/#msg604298   Other than that I feel you over analyzing this and need to pull back and understand what it is doing.  Remember it does work!  Keep in mind that it is testing at 750MHz (essentially as  CW signal), and that 100Hz span (sweep) is just a method to easily step the output amplitude up in 1 dB steps as much as 20dB maximum.  And the speed of this doesn't matter, because the SA stores each repetitive display.  The basic questions about this all came up because Rigol didn't fully explain its operation or what it is meant to do.

Edit:  I have got some cheese here, and I'll buy us the first two rounds of beer.  Do you have the time to join me?

[Howardlong]

It is interesting that the only other people here I know of that have reported Power Sweep not working is 'obiter' and 'N8AUM'.  There is one thing in common between the three of you, which is the U1105 modification.  Is there any chance that connecting pins 7 and 8 of U1105 could have effected this function?  Although it seems  unlikely to me.  Anyway at this time it would to nice to hear from others having this issue that have NOT touched U1105 in their DSA815-TG.  I'm very sorry for suggesting this could even be a possibility, and I truly hope that it is not.

FWIW, my unit doesn't have pins 7 & 8 connected, so the FRAM is writeable, but it does have the slightly more advanced fix that resets the eight FRAM time bytes on power up.

However, I agree that we need some further evidence from the non-tamperers!

It's not a big deal for me, I have a VNA for measuring this stuff anyway, but it would be good to get to the bottom of this.

[ted572]

It is interesting that the only other people here I know of that have reported Power Sweep not working is 'obiter' and 'N8AUM'.  There is one thing in common between the three of you, which is the U1105 modification.  Is there any chance that connecting pins 7 and 8 of U1105 could have effected this function?  Although it seems  unlikely to me.  Anyway at this time it would to nice to hear from others having this issue that have NOT touched U1105 in their DSA815-TG.  I'm very sorry for suggesting this could even be a possibility, and I truly hope that it is not.

FWIW, my unit doesn't have pins 7 & 8 connected, so the FRAM is writeable, but it does have the slightly more advanced fix that resets the eight FRAM time bytes on power up.

However, I agree that we need some further evidence from the non-tamperers!

It's not a big deal for me, I have a VNA for measuring this stuff anyway, but it would be good to get to the bottom of this.

I don't understand how a VNA can be used as a substitute for Power Sweep to measure an Amplifier's Linearity?

Yes, I hope we get a resolution to why a few DSA's don't work with Power Sweep.  Certainly your U1105 jumper 7/8 fix is very handy to keep the Trial Options active for new DSA815 owners, so I hope that there is no issue there with it.  And it seem unlikely.

Have you tried using UltraSigma to send SCPI Reset commands (such as Factory, etc) to the DSA to see if that may clear something up.  Its easy to do and the commands are listed in the DSA800 Programing Guide.  Or also using the new Sanitation mode in the unit.  I recall getting into Power Sweep initially seemed a little flaky to me(?), and now it has been working solid for me.  Did I clean something up playing with the Power On issue when applying AC power to the unit, or just zero in on the commands/sequence to use it reliably(?). 

Edit with a Question: What does FWIW stand for?  It must be something like FYI, IMHO(?).  I have been wondering about this for some time now.  When I first saw it I assumed you had a random finger/keyboard coordination glitch.  2nd edit one minute later:  Never mind, I just figured it out.  For Whatever It's Worth   

[Howardlong]

I don't understand how a VNA can be used as a substitute for Power Sweep to measure an Amplifier's Linearity?

Yes, I hope we get a resolution to why a few DSA's don't work with Power Sweep.  Certainly your U1105 jumper 7/8 fix is very handy to keep the Trial Options active for new DSA815 owners, so I hope that there is no issue there with it.  And it seem unlikely.

Have you tried using UltraSigma to send SCPI Reset commands (such as Factory, etc) to the DSA to see if that may clear something up.  Its easy to do and the commands are listed in the DSA800 Programing Guide.  Or also using the new Sanitation mode in the unit.  I recall getting into Power Sweep initially seemed a little flaky to me(?), and now it has been working solid for me.  Did I clean something up playing with the Power On issue when applying AC power to the unit, or just zero in on the commands/sequence to use it reliably(?). 

Edit with a Question: What does FWIW stand for?  It must be something like FYI, IMHO(?).  I have been wondering about this for some time now.  When I first saw it I assumed you had a random finger/keyboard coordination glitch.  2nd edit one minute later:  Never mind, I just figured it out.  For Whatever It's Worth   

My 30 year old 8753A does a power sweep - see http://course.ee.ust.hk/elec518/ampsmeasure_NA.pdf



I haven't done any SCPI with it yet I'm afraid, but when I get a moment I'll have a go.

[ted572]

I don't understand how a VNA can be used as a substitute for Power Sweep to measure an Amplifier's Linearity?

My 30 year old 8753A does a power sweep - see http://course.ee.ust.hk/elec518/ampsmeasure_NA.pdf

Oh Wow, when you said your VNA, I never imagined that you had one of these jewels.  At home I have just used a RF Signal Generator and a Tek 496 Spectrum Analyzer to test Amplifier Linearity.  At work I used Spectrum Analyzers with Power Sweep, actually they were Network Analyzers. 

[Teneyes]

My 30 year old 8753A does a power sweep - see http://course.ee.ust.hk/elec518/ampsmeasure_NA.pdf

Thanks Howard
 It was interesting to see that the HP Power ramp test was at a fixed (user) Frequency.
and the suggestion to use the frequency of a Frequency Sweep 1db drop point.

Where as the Rigol changes both frequency and power level at the same time.

The HP8753 is able to vary the power level continuously, But the Rigol changes the power level in 1db steps (mostly )

@Howard did you check when the output power drops off during your TG setup?
From your video it looks like the 'Ref level' change, or the video Editing?

[Howardlong]

I don't understand how a VNA can be used as a substitute for Power Sweep to measure an Amplifier's Linearity?

My 30 year old 8753A does a power sweep - see http://course.ee.ust.hk/elec518/ampsmeasure_NA.pdf

Oh Wow, when you said your VNA, I never imagined that you had one of these jewels.  At home I have just used a RF Signal Generator and a Tek 496 Spectrum Analyzer to test Amplifier Linearity.  At work I used Spectrum Analyzers with Power Sweep, actually they were Network Analyzers.

When  you replace the monochrome CRT with a NewScope-9 LCD, it suddenly seems to take 20 years off the unit. I wish I could take a bit of that medicine!

Just like with an SA with a TG, you still have to take care with your levels, having more than a few pads about is always wise.

[Howardlong]

My 30 year old 8753A does a power sweep - see http://course.ee.ust.hk/elec518/ampsmeasure_NA.pdf

Thanks Howard
 It was interesting to see that the HP Power ramp test was at a fixed (user) Frequency.
and the suggestion to use the frequency of a Frequency Sweep 1db drop point.

Where as the Rigol changes both frequency and power level at the same time.

The HP8753 is able to vary the power level continuously, But the Rigol changes the power level in 1db steps (mostly )

@Howard did you check when the output power drops off during your TG setup?
From your video it looks like the 'Ref level' change, or the video Editing?

I was deliberately following ted572's instructions to the letter in case of any misunderstanding ( https://www.eevblog.com/forum/testgear/spectrum-analyzer-rigol-dsa815/msg604298/#msg604298 )

When the Auto Scale is hit, as the TG has constant power the unit has min==max, so although the trace looks like it disappears, it's there, it's just at the bottom of the trace. Setting the scale/div to 2dB and ref to -1dBm doesn't help, as the TG is still fixed at -20dBm.

[N8AUM]

So I noticed if you set the scale to .1db/div and sweep is normalized it slowly starts dropping. The picture shows about  -.04db drop after 10 minutes. Its been about an hr now and it dropped to -.07db. Will let it run over night and see what happens. Also every couple of sweeps I get a blip than goes away, I think its something with internal clock and sweep timing ? Anybody working on a new keygen for the ones that have the 1.04 bootloader? I sure miss the 10Hz RBW !

[Howardlong]

So I noticed if you set the scale to .1db/div and sweep is normalized it slowly starts dropping. The picture shows about  -.04db drop after 10 minutes. Its been about an hr now and it dropped to -.07db. Will let it run over night and see what happens. Also every couple of sweeps I get a blip than goes away, I think its something with internal clock and sweep timing ? Anybody working on a new keygen for the ones that have the 1.04 bootloader? I sure miss the 10Hz RBW !

Over a span of 1.5GHz I'm not at all surprised.

Don't know about the blip, but yes a fiddle would be nice for the 10Hz, I am sure it's possible, but it's beyond what I'm realistically capable of that's for sure!

[atx]

It is interesting that the only other people here I know of that have reported Power Sweep not working is 'obiter' and 'N8AUM'.  There is one thing in common between the three of you, which is the U1105 modification.  Is there any chance that connecting pins 7 and 8 of U1105 could have effected this function?  Although it seems  unlikely to me.  Anyway at this time it would to nice to hear from others having this issue that have NOT touched U1105 in their DSA815-TG.  I'm very sorry for suggesting this could even be a possibility, and I truly hope that it is not.

I have an untouched 815TG with firmware 1.09 and boot 1.04 and I have the exact same problem.  However, I could not believe that this is a faulty unit and contacted Rigol.  I got a response today and confirmation from Rigol that they just found out about this DSA issue.  Rigol confirmed that this is a firmware bug with units that contain the newest version of hardware.  They further informed me that this would be high priority and is expected to be corrected soon, whatever that means as the issue I brought up months ago with the MSO2k Logic Analyzer section (see other thread) isn't fixed so far.

Thus, the hardware must have been improved to result in that lower noise level as my response curve with an unplugged input looks very similar to that of HowardLong's unit as shown earlier in this thread..

My 815 is a virgin, no mods. It came with 1.09 firmware. I tried with 1.12 with same results. Unfortunately it also came with 1.04 bootloader so am not able to try it with older firmware. I spoke with Rigol and they will exchange unit for a new one once they come in. Because of the strike out west might be a while before new stock comes in.
 

 

Was that a recent information by Rigol?  As stated above, they seem to have found out that it is a firmware bug.

My system info is:

Main Board: 00.08
RF Board FPGA: 00.05
Digital Board FPGA: 00.04
Firmware 00.01.09
Boot:00.01.04

atx

[Howardlong]

Thank you for the information atx, sounds more likely it is indeed a firmware/hardware compatibility feature.

Here are my hardware and firmware levels, the only difference appears to be the motherboard.

Mainboard 00.07
RF FPGA 00.05
Digital Board FPGA 00.04
Firmware 00.01.09
Boot 00.01.04

[ted572]

It is interesting that the only other people here I know of that have reported Power Sweep not working is . . . .

I have an untouched 815TG with firmware 1.09 and boot 1.04 and I have the exact same problem.  However, I could not believe that this is a faulty unit and contacted Rigol.  I got a response today and confirmation from Rigol that they just found out about this DSA issue.  Rigol confirmed that this is a firmware bug with units that contain the newest version of hardware.  They further informed me that this would be high priority and is expected to be corrected soon, whatever that means as the issue I brought up months ago with the MSO2k Logic Analyzer section (see other thread) isn't fixed so far.

Thus, the hardware must have been improved to result in that lower noise level as my response curve with an unplugged input looks very similar to that of HowardLong's unit as shown earlier in this thread..

Was that a recent information by Rigol?  As stated above, they seem to have found out that it is a firmware bug.

atx

This is great news!  It is nice that you looked into it with Rigol. . .     So hopefully soon everyone should be able to use Power Sweep.

Edit 2/18/2015:  Added Initial Settings for Configuring / Verifying Power Sweep Operation with the DSA815-TG.  See attached PDF.

[smgvbest]

I've looked but have not found a firm answer to several questions

Does the keygen work with the V1.09 or greater of the DSA815?
If it does work should I buy a DSA815 or DSA815-TG?   I look like the keys add the TG so I figure I so not need the -TG version

Thank you

[Howardlong]

I've looked but have not found a firm answer to several questions

Does the keygen work with the V1.09 or greater of the DSA815?
If it does work should I buy a DSA815 or DSA815-TG?   I look like the keys add the TG so I figure I so not need the -TG version

Thank you

The TG is hardware as well as firmware, and it's not available as an after market add on as far as I know.

The keygen only works on 1.08 of the firmware and below. However, this firmware only works on bootloaders below 1.04. 1.04 is the current shipping bootloader, and 1.09 or 1.12 is the current publicly released firmware. With the 1.04 bootloader, you cannot put on 1.08 firmware, so the keygen doesn't work with new units.

There are two workarounds which prolong the trials pf the options indefinitely.

The first is to short two adjacent pins on an FRAM chip on the board before the trials expire. This write protects the FRAM but has the side effect of disabling the ability to change LAN settings, and possibly other things like updating calibration settings, so it's wise to do a recal before doing the mod. Other than that, as yet there are no,other known side effects.

The second rather more complicated way is to add a small PIC with some firmware that sneaks in quickly at boot time to reset the time stored on the FRAM. This has the benefit that the LAN and any other settings stored in the FRAM can be modified as usual.

Note that in both cases, the time remaining on trials still appears to count down every 10s, but upon reboot it goes back to the time remaining when the mod was done. What happens if you happen to leave the SA on long enough to expire the trial times isn't known.

Both mods work with firmware 1.09 and 1.12.

I am sure someone smarter than me in the debugging department may well be able to come up with a rather more elegant way to achieve this.

[smgvbest]

Ok thank you now some of this makes sense.