RIGOL DSA815-TG @ 10 MHz issue/bug

RIGOL DSA815-TG @ 10 MHz issue/bug
Posted by Giorgio on 04 Aug, 2015 08:16
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

#1 Reply
Posted by pauln on 04 Aug, 2015 09:27
To start with with - are you sure the filter is OK to begin with?

#2 Reply
Posted by Giorgio on 04 Aug, 2015 09:34
Yes, everything is set as Rigol asked when I opened a ticket with them.

#3 Reply
Posted by pauln on 04 Aug, 2015 09:50
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

#4 Reply
Posted by Giorgio on 04 Aug, 2015 10:23
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.

#5 Reply
Posted by Giorgio on 04 Aug, 2015 19:13
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.

#6 Reply
Posted by Bud on 04 Aug, 2015 19:26
The change is rather abrupt and does look as a firmware issue

#7 Reply
Posted by K5HJ on 04 Aug, 2015 19:27
Can you try a slower sweep time?

Randy

#8 Reply
Posted by TheSteve on 04 Aug, 2015 19:37
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.

#9 Reply
Posted by Bud on 04 Aug, 2015 19:46
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.

#10 Reply
Posted by Giorgio on 04 Aug, 2015 19:48
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.

#11 Reply
Posted by TheSteve on 04 Aug, 2015 20:08
Perhaps the firmware omits some samples right around 10 MHz to hide the LO leakage spike.

#12 Reply
Posted by K5HJ on 04 Aug, 2015 20:13
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

#13 Reply
Posted by Giorgio on 04 Aug, 2015 20:28
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

#14 Reply
Posted by Giorgio on 04 Aug, 2015 20:46
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.

#15 Reply
Posted by cncjerry on 05 Aug, 2015 01:57
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.

#16 Reply
Posted by Giorgio on 05 Aug, 2015 06:29
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

Thank you

#17 Reply
Posted by Mark on 05 Aug, 2015 11:29
I hooked up a simple parallel LC circuit and saw this at 10MHz:

#18 Reply
Posted by Giorgio on 05 Aug, 2015 12:14
Hi Mark,

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

Can you post your system infos?

Thank you

#19 Reply
Posted by cncjerry on 05 Aug, 2015 14:24
And you don't see it when you normalize with a straight cable? That's weird.

#20 Reply
Posted by Giorgio on 05 Aug, 2015 14:46
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.

#21 Reply
Posted by Giorgio on 08 Aug, 2015 13:00
Is Mark the only one having the same problem. Anybody else who confirms the hardware/firmware pairs showing the bug?

#22 Reply
Posted by Howardlong on 08 Aug, 2015 13:42
I built a BPF with a 33pF and 10uH in series, and show similar results.

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

#23 Reply
Posted by Howardlong on 08 Aug, 2015 13:51
Normalized:

#24 Reply
Posted by Giorgio on 08 Aug, 2015 15:25
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?

#25 Reply
Posted by Howardlong on 08 Aug, 2015 18:18
If you reduce the TG drive level to, say -10dBm or -20dBm, then normalise, the results are much better. -5dBm isn't enough but it's better than 0dBm.

-5dBm:

-10dBm:

-20dBm:

#26 Reply
Posted by mcinque on 09 Aug, 2015 12:14
Same behavior here.

main board 00.08
radio fpga 00.05
digital fpga 00.05
firmware 01.12
boot 01.04

#27 Reply
Posted by cncjerry on 15 Aug, 2015 05:19
Is the bug in the TG or SA? if you just run a 10Mhz signal in does the glitch show? I've been thinking about buying this setup with the TG but that bug would drive me crazy.

#28 Reply
Posted by Giorgio on 17 Aug, 2015 07:22
Quote from: cncjerry on 15 Aug, 2015 05:19
Is the bug in the TG or SA? if you just run a 10Mhz signal in does the glitch show? I've been thinking about buying this setup with the TG but that bug would drive me crazy.

The bug is in the SA WHEN TG is activated. When the TG is OFF the SA works flawless.

I did this check to confirm this (I explained in a replay above, to be able to check the bug without having a passband filter on 10 MHz):

- Connected an external signal generator @ 10.000.000 Hz to the SA input.
- Center frequency at 10.000 kHz to have the trace of the external generator .

When I switch the TG ON (without connecting it to the SA input) I get the trace peak position shifting to the right (or left) a few pixels.
If you use a narrow bandwidth it's more evident.

This small shift is the part missing of the trace when measuring a filter response @ 10 MHz

As I said, maybe this bug is repeated at other frequencies I didn't check, maybe someone will find it.

If you use the instrument on the HF portion (as I usually do) you will hit this bug for sure. I often need to build and trim 10 MHz filters and this bug make it impossible to do and for me this is not an instrument and Rigol is not giving any solution. When I met the german Rigol support at Friedrichshafen they gave me no hopes.

Cncjerry, better you buy a used HP/Agilent unit, you will be happier.

I renew here the question:

Is there anyone else who can confirm the Hardware/Firmware pair showing the bug?

#29 Reply
Posted by Howardlong on 17 Aug, 2015 07:49
Do you get any improvement by reducing the TG power to -10dBm or -20dBm and re-normalising? That seemed to largely resolve it in my limited testing.

It's difficult to say what the problem is. At first I thought it was something to do with the frequency scheme either for the SA or TG, in conjunction with mixer non-linearity, but now I don't know: I took a look at the TG output on a sweep triggered to a scope, but although there is a definite smallish blip around 10MHz, I didn't see any difference in behaviour between 0dBm, -10dBm or -20dBm, other than a smaller amplitude across the span. That would point to mixer non-linearity in the SA rather than the TG at higher power levels, but why it's at this spot frequency I can't explain.

#30 Reply
Posted by Giorgio on 17 Aug, 2015 09:38
Quote from: Howardlong on 17 Aug, 2015 07:49
Do you get any improvement by reducing the TG power to -10dBm or -20dBm and re-normalising? That seemed to largely resolve it in my limited testing.

From the tests I made when I reduce the TG level the distortion get less evident but it's always there and impossible to make measure if the glitch fall exactly where the -3 dB is.
Anyway some filters with a high loss need a higher TG level to get off enough from the noise floor, I find the 0 dBm level already low for a TG for some applications.

I checked the TG output too with another SA to see if there's a frequency hop passing on 10 MHz but to me it seems ok. Of course the level go up and down but there's Normalize for this.
It's the SA that make a frequency hop in the sweep when the TG is activated. It's clearly visible if you make the test I was explaining in the previous replay.

#31 Reply
Posted by Giorgio on 31 Aug, 2015 19:21
Just upgraded to the new FW 1.14 and the problem is still there.

They are really unable to solve this problem

#32 Reply
Posted by mcinque on 31 Aug, 2015 20:01
I asked Rigol EU about this and they replied telling that there is a beta firmware that moves the issue from 10 to 8MHz as a temporary fix, do you know anything about it?
In 1.14 firmware release notes there is nothing about that issue fix, so why you were expecting to have improvements by installing it?

#33 Reply
Posted by Giorgio on 01 Sep, 2015 10:13
Yes I already got the beta at the beginning of this year but was of no use, I use the TG mostly on HF region and to have at 8 MHz or at 10 MHz makes no differences and don't want to switch from a FW to another just to make a new measure.

The fact is that they are not able to find a solution and this is really silly said by a company producing instruments to measure.

Don't you agree?

#34 Reply
Posted by mcinque on 01 Sep, 2015 11:10
I agree. If the unit can't be fixed by the firmware, it MUST be replaced with a different unit with different hardware, in the same way they replaced the DP832. There are no alternatives.

#35 Reply
Posted by ted572 on 01 Sep, 2015 17:01
Hello Giorgio:
I think that it is very possibly an internal problem with the TG On while using a TG setting of 10MHz at a level greater than -20dBm, and not related to the actual SA Input level.
I suggest trying the following:
1. Set the TG level at -20dBm (default) and add an external 20 to 30dB gain amplifier on the TG Output. Connect the External amplifier's Output to the SA Input. You should now have 0dBm to +10dBm for testing a DUT (your 10 MHz Filter, etc.) due to the gain of the external amplifier.
2. Normalize the system.
3. Connect your DUT/10MHz Filter between the External TG Amplifier's Output and the SA Input.
4. Does this provide the similar results as 'Howardlong' got with his -20dBm TG level? -> https://www.eevblog.com/forum/testgear/rigol-dsa815-tg-10-mhz-issue/msg726899/#msg726899 But of course you now have a higher TG Level for testing a DUC that has too much loss for testing at -20dBm.
If this helps, the credit should go to 'Howardlong' from the work he did in the referenced link above. And if it does help, them this may be a useful alternative until this issue is properly addressed by Rigol.
Edit: If your External Amplifier doesn't have a solid 50 ohm output impedance use a 6dB (Min.) to 10dB (preferred) 50 Ohm Attenuator between the Output of your Amplifier and the Input to your DUT. Of course you will need to Normalized the Spectrum Analyzer (Step 2.) with the 50 Ohm Attenuator on External Amplifier's output.

#36 Reply
Posted by TheSteve on 01 Sep, 2015 17:17
For building an amplifier this part may be handy - ERA-1sm+ datasheet here - http://www.minicircuits.com/pdfs/ERA-1SM+.pdf
I have considered buying a few myself, it would be as simple as it gets for a nice homebrew amp.

#37 Reply
Posted by Giorgio on 02 Sep, 2015 16:53
Thank you Ted572 and TheSteve.

I will give a try as soon as I get a nice buffer amp for the TG. I think I have some widebad amps chips in my junk box.

If not I will follow the TheSteve hint.

In the mean time I hope that Rigol comes with a solution to this.

#38 Reply
Posted by ted572 on 02 Sep, 2015 18:03
Quote from: Giorgio on 02 Sep, 2015 16:53
Thank you Ted572 and TheSteve.
I will give a try as soon as I get a nice buffer amp for the TG. I think I have some widebad amps chips in my junk box.
If not I will follow the TheSteve hint.
In the mean time I hope that Rigol comes with a solution to this.
The Mini-Circuits ERA-1SM+ recommended by 'The Steve' isn't suitable because it doesn't have anywhere near enough gain (also has a minimum order of 20 units).
If I had a DSA815-TG with the 'newer hardware' as yours I would test it myself as I have a RF Amplifier with 20/40 dB of Gain with 50 Ohm I/O impedance. Hopefully you can come up with approximately 20dB (min.) of gain at 10MHz with a output level of 0dBm (min.). I guess that this may be a tall order. So perhaps someone with the newer DSA815-TG that also currently has a 20 - 30 dB gain amplifier will be interested in trying this test for you and everyone else with a newer DSA815-TG.

#39 Reply
Posted by N8AUM on 02 Sep, 2015 19:06
I was fairly happy till I found this thread
As far as I can tell the SA shifts its frequency ever so slightly and its only noticeable when the RBW is fairly narrow. Using my ol Boonton the SA would show a shift in freq of approximately 156Hz. I tried from 100Khz to 1Ghz. Only at 10Mhz do I see the "missing" portion. I suspect this bug might be another reason why they stopped offering the 10Hz RBW feature
In my case it made NO difference of the TG output level that was terminated. For shits n giggles I verified the 10Mhz ref clock out and it was AOK.
Made No difference if I used firmware ver 12 or 14
My sa has board ver .08

I think if A LOT of people complain to Rigol than they might come out with a fix within a couple years

Another thing I noticed a while back was the 10Mhz bleed thru if your looking at signals below -125dB but hay, what do you want for $1500 ?

#40 Reply
Posted by mcinque on 02 Sep, 2015 19:46
Rigol knows about the problem, probably they are not receiving enough complaints/pressure from clients.

I think that only a video (maybe from Dave, which has an outstanding visibility) exposing publicly this issue can speed up Rigol's solution, as what happened with the DP832's heatsink.
If I would have known about this issue, I wouldn't have purchased this SA, and so many other clients: a popular channel's video about the bug would expose the issue so that many clients would think twice about the purchase of this model.

This IMO would create a quick response from the manufacturer, since many unit would remain on the retailer's shelfs.

#41 Reply
Posted by ted572 on 02 Sep, 2015 20:47
Quote from: N8AUM on 02 Sep, 2015 19:06
I was fairly happy till I found this thread
As far as I can tell the SA shifts its frequency ever so slightly and its only noticeable when the RBW is fairly narrow. Using my ol Boonton the SA would show a shift in freq of approximately 156Hz. I tried from 100Khz to 1Ghz. Only at 10Mhz do I see the "missing" portion. I suspect this bug might be another reason why they stopped offering the 10Hz RBW feature
In my case it made NO difference of the TG output level that was terminated. For shits n giggles I verified the 10Mhz ref clock out and it was AOK.
Made No difference if I used firmware ver 12 or 14
My sa has board ver .08

I think if A LOT of people complain to Rigol than they might come out with a fix within a couple years

Another thing I noticed a while back was the 10Mhz bleed thru if your looking at signals below -125dB but hay, what do you want for $1500 ?

They never ever offered 10Hz RBW. It was in their FW, and for some strange reason they provided it as a Trial on few units in 2012 or 21013 (?), but never planed for it to be released for the DSA815. Yes, WTF were they thinking! Give us a teaser and then pull it back. Sorry, its NOT for you guys (suckers) to buy! They were at the time going to use common code along with the next versions of the DSA800 Series. Someone here found it lurking in the firmware while looking for a method to activate the offered options, and fortunately we were able to get it on our original DSA815s. But now its been pulled out (I guess) and we may never be able to get on the newer DSA815 hardware versions (?). Yet 'To Be Determined' (TBD). I'm lucky that I have the original version 5 main board in my older unit.

Congratulations: You made a nice find with the ~156Hz Freq. shift at 10MHz. The more that is discovered and reported will certainly help in getting Rigol to take action on this. And a video Rant (as suggested by 'mcinque') would be a real boost to push this along.

#42 Reply
Posted by TheSteve on 02 Sep, 2015 20:51
I was sure that I read here that Rigol offered the 10 Hz RBW but only for the chinese market.

#43 Reply
Posted by Howardlong on 02 Sep, 2015 21:32
The weirdness is that the glitch is significantly less apparent at lower TG power levels, assuming re-normalisation after power adjustment. If it were just a synthesiser/frequency chain discontinuity, Imwould (a) expect the discontinuity to be the same in relative terms and (b) fixable in firmware - assuming they use the same reference for generating below 10MHz as they do above. Or maybe that's the problem.

#44 Reply
Posted by N8AUM on 02 Sep, 2015 22:11
I noticed absolutely NO change regardless of the TG's level with my toy.
It would be nice if Dave or Shahriar could find some time and investigate this "bug".
Since I dont plan on launching any deep space probes any time soon I can live with all the little Bugs. It's good to know about all of them so you can hopefully find a work around? Thanks to Giorgio for bringing this Bug to everyone's attention.
Anyways, I am not going to hold my breath for Rigol to come up with a fix just to satisfy 10 or 20 customers.
Here's another thing that I noticed with my toy, when I do a screen dump using the LAN connection the speed is super slow, about 30KB/s.

#45 Reply
Posted by ted572 on 02 Sep, 2015 22:33
Quote from: Howardlong on 02 Sep, 2015 21:32
The weirdness is that the glitch is significantly less apparent at lower TG power levels, assuming re-normalisation after power adjustment. If it were just a synthesiser/frequency chain discontinuity, Imwould (a) expect the discontinuity to be the same in relative terms and (b) fixable in firmware - assuming they use the same reference for generating below 10MHz as they do above. Or maybe that's the problem.
I think it is due to some internal coupling at the higher TG levels, and unrelated to the SA Input Level. That is why I suggest running the TG at the lowest level and externally amplifying it back up to see if this could be a possible workaround. Or at least help us understand it more than we do now.

#46 Reply
Posted by N8AUM on 03 Sep, 2015 04:26
I hooked up the 300Mhz cal source from my HP SA and did some more poking around. Other than the slight shift in frequency when the tracking gen is on the wave looked normal REGARDLESS of the SA power output. I decided to try it with the power sweep on and got some strange results.
I just picked 3 power ranges 3, 9 and 20dB

#47 Reply
Posted by Giorgio on 03 Sep, 2015 16:01
Hi Gents,

I am sorry if I made unhappy lot of you opening this thread. You were thinking the 815-TG was nice and precise (for it's price) but by my side I am happy that now you know that it has a problem and maybe, as said here, with a lot of requests to Rigol we could have a solution.

As far as I know I was the only one who opened a ticket about this problem 1 year ago with Rigol and now maybe there's more unhappy customer opening their ticket to force Rigol to do something about it.

I didn't wrote here or anywhere else just because I was trusty with Rigol support finding a solution but after many months without solution I decided that this had to be public.

In April I wrote to Dave asking if he could analize the problem but he replied it was too busy for that and said to post here in the forum.

Now I hope Dave will make a video of this so more people will know what they will get if they buy the 815-TG at this time.

About the shift in the frequency sweep at 10 MHz when the TG is activated: that hop is the missing part of the trace when you analize the response of a filter.

#48 Reply
Posted by mcinque on 03 Sep, 2015 21:20
Totally agree with you Giorgio.

#49 Reply
Posted by Howardlong on 03 Sep, 2015 21:59
I did a couple more tests, but the results are not what I expected.

First, I set the TG to sweep 9.999MHz to 10.001MHz with a sweep time of about 4s, 100Hz RBW & VBW, TG at 0dBm, and listened to the result on a communications receiver on upper sideband (have to be careful about using "USB" on here, could easily mean different things to differnent people!) There was no discontinuity that I'd expect if there was a jump in the TG. I also tried it at different TG power levels, down to -20dBm, and again no audible frequency jump, it was a clean sweep. Not a very scientific test, but I can be very sure I'd be able to hear a 150Hz frequency jump if there was one coming from the TG.

Without any normalisation, there is a definite bump down of 0.73dB at 10MHz with 0dBm TG. An external 20dB attenuator brings the bump to 0.78dB.

Back to direct TG to SA connection with no external attenuator, dropping the TG power to -5dBm gives a bump down of 0.50dB at 10MHz. -10dBm TG gives a 0.53dB bump. -20dBm TG gives a 0.32dB bump. -30dBm TG gives a 0.32dB bump also. Again, this is without normalisation.

Now, a new test.

Setup an external RF sig gen to 9.9999MHz unmodulated carrier, -30dBm and plug it into the SA. Nothing's connected to the TG output.

Preset the SA to default settings. Set the centre freq to 10MHz and the span to 1kHz. You should see a nice bell curve. The carrier is deliberately offset from the SA centre freq to get the curve slope rather than the peak on 10MHz.

Now switch on the TG (still without it connected to anything). You'll see (a) there's a big discontinuity appears and (b) the curve moves.

Weird!

Here's a quick video of it. Sorry it's a bit shaky.

#50 Reply
Posted by Giorgio on 04 Sep, 2015 06:48
Howardlong,

Finally a video on youtube, I didn't because my english is not so fluent.

That's the test I suggested to do too early in this thread if someone has no narrow filter on 10 MHz as DUT.

It's the test I sent 1 year ago to Rigol to have evidence there's something strange happening at 10 MHZ, their reply was they were not interested in such test because bla bla bla... you need to read the how to use TG on the Rigol site etc etc..... and this test proves nothing because it's not the way the instrument has to be used and so on.

For me it was the simplest way to check if the problem is present in a unit. Period.

It's clear that the problem is within the SA when the TG part is activated, beside the different output level of the TG that one can normalize before to measure, the SA has a hop exactly at 10.000.000 and the hop value varies with the sweep and bandwidth values settings. This suggest me there's a bunch of missing bits lost during the pass through one stage to another along the digital circuits. Maybe cut on purpose to mask some other anomalies at 10 MHz.

The german support guy I was in touch via email and I met at Friedrichshafen hamfest said that they can't solve the issue because exactly at that frequency in this hardware version of the instrument there's a switch between circuits (if I understood well, sorry if I am wrong) and that they just can try to shift the problem on a different frequency (this was made with the beta FW they sent me shifting the problem at 8 MHz and something) and that the old hardware version has not this problem.

Ciao

#51 Reply
Posted by Gertjan on 04 Sep, 2015 07:19
I found it very easy to reproduce Howardlong's video.
The phenomenon is there, at all signal levels and frequency / span / RBW / variations.
I used a Rigol DG1032z as a source. Gen and DSA815 were both locked to the same 10MHz reference.

I must say, in practice I don't have much problems with this phenomenon.
Recently I did a lot of measuring frequency curves of 10MHz transformers.
I noticed the amplitude discontinuity at 10 MHz, but with a new normalise it would disappear.
The problem was worst at a TG level of 0dBm. Taking the TG level a few dBm down made the problem much less bothersome.

Here is an example, comparing frequency curves of three transformers:

Regards, Gertjan.

#52 Reply
Posted by pa3bca on 04 Sep, 2015 07:31
I tried to replicate Howardlong's findings, exactly following his test setup.
On my DSA815-TG I cannot reproduce the amplitude discontinuity. The 156 Hz frequency upshift of the displayed spectrum however does appear as soon as I activate the TG.
Yellow: 9.9999 MHz (from a DG1032Z) without TG active,
Purple: with TG activated (level of TG does not seem to have any influence), note 156 Hz shift.
Also notice the 1.5 dBm lower peak level... what gives?

#53 Reply
Posted by Gertjan on 04 Sep, 2015 08:27
I reproduced exactly the same screen shots as pa3bca:

It appears the difference is in the version of the Main Board.....
pa3bca has an older Main Board, version 00.04 (with boot 01.03)
I have a newer Main Board, version 00.08 (with boot 01.04)

So it seems that the problem was not in the original design of the DSA815, but introduced in a later revision......

regards, Gertjan.

#54 Reply
Posted by mcinque on 04 Sep, 2015 11:26
Thanks for the video!

Interesting test, so it's not an issue in the signal that the TG produces! It's something that happens inside (mixer?) just when it's been activated! Really weird.

Thanks again for the video.

#55 Reply
Posted by N8AUM on 04 Sep, 2015 22:27
Well this is really starting to suck now! I applied a 10.8Mhz sig as if I was messing around with a typical IF frequency and found that the SA shifts ONLY 6Khz when I have the RBW set at 3K and the span at 100Khz !!!!! I used an external 10Mhz GPSDO to eliminate if the internal clock farts or not when the TG is on.
Being off 156Hz is not a big deal but 6K ??
I also noticed the screen grab program ONLY works when I have the SA setup in a certain way. If I try screen grab with the RBW set at 1K the LAN is AFU but it works fine if I change the RBW to 300 or 100Hz
Can someone with the hardware mod try and duplicate this ?
Also, when the screen grab is working how long does it take using LAN ?

#56 Reply
Posted by johnny_canuck on 05 Sep, 2015 02:48
Quote from: pa3bca on 04 Sep, 2015 07:31
I tried to replicate Howardlong's findings, exactly following his test setup.
On my DSA815-TG I cannot reproduce the amplitude discontinuity. The 156 Hz frequency upshift of the displayed spectrum however does appear as soon as I activate the TG.
Yellow: 9.9999 MHz (from a DG1032Z) without TG active,
Purple: with TG activated (level of TG does not seem to have any influence), note 156 Hz shift.
Also notice the 1.5 dBm lower peak level... what gives?

I have a 00.03 main board. Otherwise, the same configuration as PA3BCA, with the same results.
Ken, VE3FIT

#57 Reply
Posted by pa3bca on 05 Sep, 2015 09:57
Quote from: N8AUM on 04 Sep, 2015 22:27
Well this is really starting to suck now! I applied a 10.8Mhz sig as if I was messing around with a typical IF frequency and found that the SA shifts ONLY 6Khz when I have the RBW set at 3K and the span at 100Khz !!!!! I used an external 10Mhz GPSDO to eliminate if the internal clock farts or not when the TG is on.
Being off 156Hz is not a big deal but 6K ??
I also noticed the screen grab program ONLY works when I have the SA setup in a certain way. If I try screen grab with the RBW set at 1K the LAN is AFU but it works fine if I change the RBW to 300 or 100Hz
Can someone with the hardware mod try and duplicate this ?
Also, when the screen grab is working how long does it take using LAN ?
I can replicate these (disturbing) findings.
There always seems to be a frequency shift when the TG is activated. The shift depends on the RBW and the SPAN. for instance if I decrease the span in abovementioned test from 100 KHz to 50 KHz, the shift halves to 3 KHz. The shift also decreases if the RBW is decreased. But there is always a shift...
Looks like a simple to fix SW bug. Let's see how long it takes Rigol to notice and (after realizing what a serious issue this is) to fix it.

#58 Reply
Posted by H.O on 05 Sep, 2015 10:47
Quote
Let's see how long it takes Rigol to notice and (after realizing what a serious issue this is) to fix it.
They are not monitoring or searching forums for potential issues to be reported so you need to actually tell them directly.
I don't have a SA and I don't know much about them but this, unlike some other obscure "bugs" that gets totaly blown out of proportion, looks like a pretty serious issue which I'd be surprised (and seriously pissed of if it was my SA) if they didn't look into properly but again, you need to tell them directly.

And even though it has been reported it's probably better that everyone who does have an issue with this does so that they're not thinking it's just "one guy".... The more people complainging about it the larger the chanse of it getting fixed, I would think.

#59 Reply
Posted by KD0RC on 05 Sep, 2015 15:07
It seems like the severity of the issue needs to be determined. The normal use for the TG is to sweep filters and that kind of thing. It would be interesting to compare with a high end SA to see if the shift is intentional. In other words, is the SA display intentionally shifted to properly display the TG freq as it sweeps? If someone could verify that one way or the other, it would help the discussion.

Len

#60 Reply
Posted by Howardlong on 05 Sep, 2015 15:29
Quote from: KD0RC on 05 Sep, 2015 15:07
It seems like the severity of the issue needs to be determined. The normal use for the TG is to sweep filters and that kind of thing. It would be interesting to compare with a high end SA to see if the shift is intentional. In other words, is the SA display intentionally shifted to properly display the TG freq as it sweeps? If someone could verify that one way or the other, it would help the discussion.

Len

I have started doing that. I have an HP 8753A here and some narrow 10.7MHz crystal filters that I'm matching up to 50 ohms for some tests. It will be interesting to see what happens.

#61 Reply
Posted by Howardlong on 05 Sep, 2015 21:52
...and here is the result:

#62 Reply
Posted by georges80 on 05 Sep, 2015 22:04
Thanks Howard for the video. I presume the frequency 'offset' changes depending on the RBW of the DSA ??

I notice the frequency shift changes if you feed say 10MHz into the DSA from an external signal generator and you vary the RBW of the DSA with the TG turned on. With the TG off, the peak is spot on at 10MHz.

cheers,
george.

edit: dumb typo.

#63 Reply
Posted by N8AUM on 06 Sep, 2015 02:33
Nice job Howard

I found a few crystals in my scrap box and trying to find one that works !
Do you have couple of 3 or 6dB pads that you can stick on the the input and outputs so the crystal sees the same regardless of which instrument your hooked up to ?
I see the 1.6Khz shift with my Rigol and will try to see what my old HP 8591E comes up with.

Thanks for posting video, reading results are ok but seeing is much better LOL !

#64 Reply
Posted by Howardlong on 06 Sep, 2015 13:10
Three more vidja's for you.

The first does the same as before but with 6dB pads. The result is the same: I spent some time matching the filter for the previous video as it was, if it had been significantly different I would have had some explaining to do!

The second video follows on showing the impact of RBW. Eeeewww, not nice.

The final video is more educational, and shows how I came up with the matching network. This kind of thing is all too often shrouded in mystery and the black magic voodoo of Smith Charts, but it's also the bread and butter of RF engineers. In this video I show how to calculate your matching network without any of that scary nonsense. While I am not dissing traditional methods, you'd be nuts to want to do this kind of thing on a Smith Chart these days.

#65 Reply
Posted by N8AUM on 06 Sep, 2015 13:45
EXCELLENT videos Howard, many tnx !!!

#66 Reply
Posted by georges80 on 06 Sep, 2015 19:22
More fun.... Feed in your signal (10MHz etc), turn on the TG, set the RBW and note the frequency offset (from 10MHz).

Now, push the Sweep/Trig button and adjust the "Time" parameter. Note how the center frequency moves closer to the 10MHz (or whatever you set) as you increase the Sweep Time.

So, there is definitely something weird going on across the board when the TG is turned on.

The real question becomes "are these effects due to some purposeful software that compensates/aligns for the TG as a source"? i.e. if the TG is actually feeding a circuit that the DSA then gets input from is the frequency aligned correctly? And we are only seeing frequency shift anomalies due to feeding externally a frequency source that is not related to the internal TG timing.

Of course the 'discontinuity' in the newer mother boards adds in yet another quirk (a more irritating one).

cheers,
george.

#67 Reply
Posted by Howardlong on 06 Sep, 2015 19:46
The more I look at it the more I believe this is fixable in firmware, particularly the frequency offset thing.

#68 Reply
Posted by KD0RC on 06 Sep, 2015 20:01
Nicely done Howard! Thanks for the effort on this. I concur - it looks like a bug, not an intentional shift. For my hobbyist purposes, this is not terrible, but good to know so that I don't try to adjust a filter that is actually already in spec... I hope Rigol addresses this, but I am not holding my breath waiting for them...

Len

#69 Reply
Posted by AF6LJ on 06 Sep, 2015 20:03
Quote from: Howardlong on 06 Sep, 2015 19:46
The more I look at it the more I believe this is fixable in firmware, particularly the frequency offset thing.

I am sure you are right.
My question is why hasn't Rigol addressed this?

#70 Reply
Posted by AF6LJ on 06 Sep, 2015 20:05
Quote from: KD0RC on 06 Sep, 2015 20:01
Nicely done Howard! Thanks for the effort on this. I concur - it looks like a bug, not an intentional shift. For my hobbyist purposes, this is not terrible, but good to know so that I don't try to adjust a filter that is actually already in spec... I hope Rigol addresses this, but I am not holding my breath waiting for them...

Len

You do have to be careful; your test equipment will lie to you from time to time and often at the most critical of times.

I still think for the money I am better off with a well cared for HP-141T or an HP-8569.
Back in the old days we used seperate sweepers and diode detectors, they seemed to work just fine.

#71 Reply
Posted by KD0RC on 06 Sep, 2015 22:21
Quote from: AF6LJ on 06 Sep, 2015 20:03
Quote from: Howardlong on 06 Sep, 2015 19:46
The more I look at it the more I believe this is fixable in firmware, particularly the frequency offset thing.

I am sure you are right.
My question is why hasn't Rigol addressed this?

My guess is that they either don't know about it, or do know about it but have not heard enough complaints to do anything about it yet.

#72 Reply
Posted by AF6LJ on 06 Sep, 2015 22:43
Quote from: KD0RC on 06 Sep, 2015 22:21
Quote from: AF6LJ on 06 Sep, 2015 20:03
Quote from: Howardlong on 06 Sep, 2015 19:46
The more I look at it the more I believe this is fixable in firmware, particularly the frequency offset thing.

I am sure you are right.
My question is why hasn't Rigol addressed this?

My guess is that they either don't know about it, or do know about it but have not heard enough complaints to do anything about it yet.
The third and fourth possibilities;

They don't feel it is worth spending ham-hours to resolve a firmware issue that is in their mind a minor issue.

They have in some way painted themselves into some kind of engineering corner and cannot band-aid or patch their way out.

#73 Reply
Posted by Bud on 06 Sep, 2015 22:57
And the most obvious possibilty that everyone have missed is they are simply incompetent .

#74 Reply
Posted by AF6LJ on 06 Sep, 2015 23:11
Quote from: Bud on 06 Sep, 2015 22:57
And the most obvious possibilty that everyone have missed is they are simply incompetent .
I wasn't going to say that considering they seem to make things that work more often than not.

#75 Reply
Posted by Giorgio on 07 Sep, 2015 07:36
Quote from: H.O on 05 Sep, 2015 10:47
Quote
Let's see how long it takes Rigol to notice and (after realizing what a serious issue this is) to fix it.
They are not monitoring or searching forums for potential issues to be reported so you need to actually tell them directly.

They do monitor this forum for sure, there's evidences everywhere in other Rigol relate threads, at least the german support.

Happy to see see my problem with Rigol is getting everybody problem..... nice job guys.

We hope now they will take seriously my initial complain, the "just one guy rant.."

Please open tickets to Rigol support or we can stay here talking for the rest of our lives without a Rigol response.

Dave answered again he has no time for this so if there's more people spreading videos about it would be great.

Use the filter DUT test not the external generator because Rigol already answered me it's not proving anything because the TG and SA are directly connected and there's always some kind of normalization on the signal to compensate known levels and frequencies anomalies.

So stick a crystal filter or a simple 10,7 MHz IF transformer and make tests and videos with it.

Ciao.

#76 Reply
Posted by H.O on 07 Sep, 2015 12:33
Quote
They do monitor this forum for sure, there's evidences everywhere in other Rigol relate threads, at least the german support.
Last time I was in contact with them (and it was the German office) they said they were not. But even if they drop by now and then it can not be expected from them to pick up anything and everything that gets written here - or on any forum.

They may or may not fix this issue and it and other issues should definitely be discussed here but the correct way to let the manufacturer know about it is thru their official channels. If they act on it based on a forum post that's great but they really can't be expected to. They CAN be expected to act on it, or at least tell you then won't act on it (which would suck), if it gets reported to them directly.

#77 Reply
Posted by N8AUM on 07 Sep, 2015 15:07
Jason from Rigol response:

"The tracking generator does effect the measurements when it is active. This is noted in the specifications:
From: User’s Guide for DSA800 Series section 5-2

But, I want to check and see if your reported shift is out of the ordinary."

I sent him a pic showing a 12.5Khz shift at 10.7Mhz
we will see what happens, probably nothing

#78 Reply
Posted by wkb on 07 Sep, 2015 15:23
Quote from: N8AUM on 07 Sep, 2015 15:07
Jason from Rigol response:

"The tracking generator does effect the measurements when it is active. This is noted in the specifications:
From: User’s Guide for DSA800 Series section 5-2

But, I want to check and see if your reported shift is out of the ordinary."

I sent him a pic showing a 12.5Khz shift at 10.7Mhz
we will see what happens, probably nothing

That remains to be seen of course. Especially if 'all' DSA815-TG owners file a problem report. Rigol have invested considerable
time and money into establishing themselves in the T&M marketplace, this is the sort of anti-marketing most companies are seriously unhappy about.

(/me wonders where to find the time to test my own DSA815-TG for this issue)

#79 Reply
Posted by Howardlong on 07 Sep, 2015 15:33
The reason I chose a crystal filter is that they tend to be narrow band and have very steep skirts, so showing frequency shifts like this is demonstrated much more clearly.

For a relatively gentle LC filter, even of several orders, you'd be hard pushed to show a 1.5kHz deviation at 10MHz. Indeed, for fiddling with LC filters at HF or above, I'm happy to be educated otherwise, but I wouldn't think the frequency shift thing is much of a problem practically speaking, a gentle breeze on the coil probably might have more effect.

For tuning a narrow band IF stage though, this is definitely going to be an issue, and I'd hazard a guess that at its price point, many of the purchasers of the DSA815-TG are ham radio operators who'd use it for just this kind of thing.

#80 Reply
Posted by ted572 on 07 Sep, 2015 18:43
In case there are any doubts: Every Single DSA815-TG, NOT Just Those With the Newer Hardware (i.e. Main Board .07/8) have a problem with the SA Measured frequency using a Resolution Bandwidth (RBW) greater than 100 Hz or so, when the Tracking Generator is turned On.

This is NOT just a problem at 10 MHz, but over the whole frequency range of the DSA815-TG. I typically see a +10 kHz offset using a 30kHz RBW. Of course this frequency offset can be different depending on the Frequency Span/Scan Time used.

To see this, simply insert a CW Signal between 1 MHz and 1.5 GHz from a Signal Generator and use the SA Marker's Peak Search function to readout the frequency with various RBWs. When you turn the TG from Off to On you will see the SA's peak maker frequency increase with the wider RBWs. Note: TG Normalization does NOT help correct the SA frequency readout error at all!

On the other hand the Glitch reported by 'Giorgio' while measuring a 10 MHz Filter seems to be only on DSA815-TGs with the Newer Hardware. And as far as I know this is only at 10 MHz. I don't have a unit with the Newer Hardware, so I can not verify this myself. Although all the reports from other users so far would indicate that this is probably the case.

So I believe that we ALL have a problem with our DSA815-TG's (even with Main Board .04/5),. Was it also there with older firmware? I don't know, or think that it matters at this time.

Please re-frame from making negative comments about Regal. As this won't get us anywhere, and I'm sure that they will correct this issue as they have with others in the past. And we can thank ' Giorgio' for discovering it. Hopefully it will be corrected soon, and how urgent can it be with it taking this long for us to uncover. As long as it is corrected I will be happy.

Edit: Underlined - when the Tracking Generator is turned On

#81 Reply
Posted by Giorgio on 07 Sep, 2015 20:33
Ted572,
thank you for the credits about discovering the bug. I really hope they will fix it. Again Gents, please fill your requests to your Rigol support, it's the only way to get it fixed.

And yes the shift in frequency readout is present from 10MHz to 1,5 GHz when TG is active.

As wrote before they were working on a new firmware they sent me in beta version that moved the problem from 10 to 8 and something MHz. When I spoke to the german support guy he said that the only way to solve it is to try to move the anomaly outside the display view because the problem is related to the hardware and at 10 MHz or 8 or outside the screen has to be there.

After that beta FW they stopped to give me any news. I think because I was the only one complaining.

So please send support requests and let's wait.

I am curious to know if maybe this problem is present also in the 3 GHz version, the DSA832-TG, anybody want to test it? I don't own one.

Edit: sorry I meant the 832, corrected.

#82 Reply
Posted by N8AUM on 07 Sep, 2015 21:15
The frequency shifting with TG on is present on ALL frequencies and the shift in frequency is dependent on SPAN and RBW combination.
As far as I can tell the worse case is a frequency shift of 5.833Khz with the 3Khz RBW
This amount of frequency shift might not even be noticed at high frequencies but it sure is noticeable on the low end, like at 455Khz !

I hope everyone ( all 10 of us LOL ) contacts Rigol about this bug.
Too bad all the schools that have thousands of these dont do the same, than for sure they will come up with a fix!

Its still a nice little toy for the money and I'v been playing with it almost every day for the past 8 months.

#83 Reply
Posted by pa3bca on 08 Sep, 2015 14:08
Well this is getting more and more disturbing. I can't understand why I (or anybody else) has not noticed this before.
I tried to measure a 475 KHz bandpass filter I built some time ago, with the TG.
Span 200 KHz (400-600 KHz), RBW 3 KHz (Yellow), 1 KHz (Purple) and 100 Hz (Magenta).
WTF?
1. The shift is there. The higher the RBW, the more the graph is shifted to the right. The filter is not steep enough to accurately measure the shift, but no doubt it is the same as measured by others and myself with other tests.
2. BUT with a RBW set to 3 KHz there is a 10 dB amplitude discontinuity at exactly 500 KHz!! RBW at 1 KHz also shows a small discontinuity at 500 KHz (hard to see in this screenshot).
Yes, I will definitely complain to Rigol. I urge others to do this too
(oh and by the way, My DS815 does NOT show the 10 MHz discontinuity....)

#84 Reply
Posted by Giorgio on 08 Sep, 2015 14:58
Quote from: pa3bca on 08 Sep, 2015 14:08
(oh and by the way, My DS815 does NOT show the 10 MHz discontinuity....)

Please post your System Information, maybe you are the "lucky" one with old hardware not showing the 10 MHz glitch.

Maybe we discovered that the old hardware has the glitch at 500 KHz........ well, if this is the case ALL (old and new) 815-TG in the world have some kind of problem with the discontinuity using the TG.

Cool

#85 Reply
Posted by pa3bca on 08 Sep, 2015 15:06
Quote from: Giorgio on 08 Sep, 2015 14:58

Please post your System Information, maybe you are the "lucky" one with old hardware not showing the 10 MHz glitch.

Maybe we discovered that the old hardware has the glitch at 500 KHz........ well, if this is the case ALL (old and new) 815-TG in the world have some kind of problem with the discontinuity using the TG.

Cool
I have the 'older' hardware, already posted my system information, but for the sake of completeness here it is again:

#86 Reply
Posted by Orange on 08 Sep, 2015 15:37
If you increase the sweep time to 300 ms the freq shift is gone and also distorted filter curve

The Auto SWT setting can be set to normal and accy, whatever that means, probably accurate

I have feeling that the sweep times are a little to high for the given RBW when the TG is on. The processor probably does not have enough time to deal with the SA and TG at the same time.

My analyser is an 'old' model revision 0.4, firmware 1.09

#87 Reply
Posted by pa3bca on 08 Sep, 2015 15:51
Quote from: Orange on 08 Sep, 2015 15:37
If you increase the sweep time to 300 ms the freq shift is gone and also distorted filter curve

The Auto SWT setting can be set to normal and accy, whatever that means, probably accurate

I have feeling that the sweep times are a little to high for the given RBW when the TG is on. The processor probably does not have enough time to deal with the SA and TG at the same time.

My analyser is an 'old' model revision 0.4, firmware 1.09
YES!
I think you are on to something.
The default sweeptime (for my 400-600 KHz span and 3 KHz RBW) is 22.22 ms. Setting accy to on increases the sweeptime to 66.66 ms and the amp dip decreases.. around 200 ms sweeptime the dip is into the noise.
Yellow: sweeptime default 22 ms, purple sweeptime manually increased to 200 ms. Both frequency shift and amp dip are gone.
So, for now we have a workaround.
Thanks!
now: can anybody with a new board & the 10 MHz discontinuity try the same?

#88 Reply
Posted by orbiter on 08 Sep, 2015 16:13
I have just had a response from a contact at Rigol after pointing them in the direction of this thread and the SA issue. They do indeed know about the Issues we're having and R&D are now investigating.

#89 Reply
Posted by Orange on 08 Sep, 2015 17:11
Quote from: PeDre on 08 Sep, 2015 15:53
Hello,

these steps are the tracking generator calibration data.
If you turn off the tracking generator calibration in service menu, these steps disappear.
I have documented the format of the calibration data.
There are more frequencies with these errors.
e.g. 500 kHz, 600 kHz, 700 kHz, 1.1 MHz, 1.3 MHz, 2 MHz, ...
There are 21 calibration groups with my DSA.

Peter

There are definitely anomalies around the calibration points. Some of them cause really bad curves the 500kHz one is a good example. I'm not going to mess with the service menu, as some of us have had bad experience with this.

This issue should be fixable for RIGOL. It is not a discontinuity in the HW I think

#90 Reply
Posted by cncjerry on 09 Sep, 2015 01:59
I'm trying to figure out what is worse, the RBW shift or that TG glitch. I think the RBW is a problem in their FFT algorithm dropping the results in the wrong bins or something. That would drive me crazy tuning a filter and I recently did a lot of people or around 10Mhz. I was planning to purchase this setup as a backup to my Advantest with TG but now will keep looking for an older HP.

#91 Reply
Posted by pa3bca on 09 Sep, 2015 19:15
Quote from: cncjerry on 09 Sep, 2015 01:59
I'm trying to figure out what is worse, the RBW shift or that TG glitch. I think the RBW is a problem in their FFT algorithm dropping the results in the wrong bins or something. That would drive me crazy tuning a filter and I recently did a lot of people or around 10Mhz. I was planning to purchase this setup as a backup to my Advantest with TG but now will keep looking for an older HP.
The freq shift and the amplitude glitch have the same root cause. I think Orange https://www.eevblog.com/forum/testgear/rigol-dsa815-tg-10-mhz-issue/msg749828/#msg749828 nailed it. It should be fairly easy for Rigol to fix.
Until then the work-around would be to check if the sweep time is 200-300 ms or more. Adjust manually if required.
The default 22.22 ms sweep time resulting from the settings I used (200 KHz span, default 3 KHz RBW) does not result in a significant faster update rate than a 200 ms sweeptime anyway so not much is lost. No Idea why they would default to this 22.22 ms sweeptime as the screen update rate does not seem to be more than 5 times/second anyway

Sanity check:
1: Yellow trace: Sweeping my 475 KHz filter+amp with the TG and sweeptime set manually to 200 ms
2: Purple trace: TG off, and using my DG1032Z on the input, generating a 1 sec sweep from 400 to 600 KHz @ -20dBm and the DSA815 on max hold. wait a few seconds.
Perfect match. So reliable enough.

#92 Reply
Posted by Giorgio on 10 Sep, 2015 06:23
Quote from: pa3bca on 09 Sep, 2015 19:15

Until then the work-around would be to check if the sweep time is 200-300 ms or more. Adjust manually if required.

I think this is valid for the old hardware with the 500 KHz glitch not for the new one. I already tried to slow down the sweep time but the 10 MHz glitch is always there and if it would be so easy for Rigol to Fix it should be done 1 year ago when I let them know about the problem.

73

#93 Reply
Posted by Giorgio on 10 Oct, 2015 07:32
Hi,

Any news from people who opened a ticket with the Rigol support, if any did?

#94 Reply
Posted by N8AUM on 10 Oct, 2015 16:40
I submitted a "ticket" one over a month ago, received a reply several weeks ago to supply them with my hardware and firmware revs WHICH I DID THE 1st time !
Anybody having better luck ?

#95 Reply
Posted by orbiter on 10 Oct, 2015 17:43
No reply here from Rigol either, other than the confirmation and agreement of the issues, although I'm still hopeful they'll sort this out.

Sent from my HTC One using Tapatalk

#96 Reply
Posted by mcinque on 10 Oct, 2015 20:43
Quote from: N8AUM on 10 Oct, 2015 16:40
received a reply several weeks ago to supply them with my hardware and firmware revs WHICH I DID THE 1st time !
They don't know how to reply and they are taking time...

#97 Reply
Posted by Pinkus on 08 Nov, 2015 13:56
the latest firmware of Rigol at http://int.rigol.com/File/ModelSoftWare/20150929/DSA815%20Software%20solidification.rar
is showing as the latest version: 1.14. Unfortunately the change log is not included to the download too, so I am attaching it here.
However, the 10Mhz issue still exists.

DSA815 Firmware 1.14 Release Notes.pdf

#98 Reply
Posted by Gertjan on 08 Nov, 2015 15:55
firmware version 1.14 was released on the 19th of June, 2015.........
I installed it months ago...

Regards, Gertjan.

#99 Reply
Posted by cncjerry on 13 Nov, 2015 04:55
wrt the glitch: Do you have a switchable frequency counter on the SA? The reason I ask is that I've been playing around with my HP 8568b and I noticed a similar glitch when the frequency counter is turned on and it makes sense on this older unit. For instance, as the trace reaches the peak (or wherever you have the marker) the counter is triggered and counts the frequency. During that period of time (at least on the 8568b), the counter is running and the trace is still being drawn. So the trace is normal, then the processor gets busy at the peak and when the trace starts writing again the time base is further along the X axis so you get an 'L' right at the top. The size of the 'L' is related to RBW, VBW, etc. Since the processor on the 8568 is relatively slow, I would expect this from a box that old. This all makes sense to me and I started wondering if it also had something to do with the Rigol's glitch.

So other than turning off the TG, can you turn off other functions? Is the glitch the same?

Also, on the frequency shift problem, both my analyzers experience this as you change the RBW unless I have signal tracking turned on. It's like the analyzer doesn't center the signal in the RBW bandwidth perfectly. It is a simple task to hit marker->peak and then marker->cf.

Jerry

#100 Reply
Posted by DD4DA on 03 Dec, 2015 20:38
Rigol had released the V1.15 at 12/02/2015 but removed them reasoned by some new problems - error message: "431 - intermediate frequency out of range" at the finish of the init of the dsa and just before the device is ready.
The noise-level seems to get more bad - about 30dB. I opend a ticket and they are quick in germany. They allways answer mostly the same day.
Some other issues are in the remote operation. My device (HW 0.4) got a currious behavior around the hardcopy and SCPI command response time. As longer i work with the dsa, as less i able to reproduce them again - that thrills me.
Just keep in touch with it.

#101 Reply
Posted by pascal_sweden on 03 Dec, 2015 21:17
There is a "new kid on the block" contender coming out in "spectrum analyzer land".
This time from Siglent! Will be interesting to compare it against Rigol DSA815-TG

https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/msg777458/#msg777458

#102 Reply
Posted by N8AUM on 04 Dec, 2015 06:58
Jerry,
I just picked one of those BEAST (8568B) couple weeks ago and it also acts the same when the RBW is 10Hz and Freq counter on. With the RBW at 30Hz its not really noticeable unless I look for it. Guess I will see what my 8591E does when I turn on the counter.
On a cheery note, my brother who is very active with meteor scatter and EME (moonbounce) came down for a visit and we were messing around with my SA`s. He looked at the specs of the 815 and said we might be able to see some signals being reflected off the moon if theirs any "big guns" transmitting. To our surprise we were able to see couple signals on 2 meters !!!

Has ANYBODY heard back from Rigol yet ?

#103 Reply
Posted by cncjerry on 05 Dec, 2015 20:40
N8,funny you should mention EME. I have a 22 element beam cut for 432Mhz SSB. I belonged to a net years ago. I live on a sloping hill and noticed that as the moon rises this time of year the beam is pointing right at it and now that you mention it, I am going to hook it to my analyzer and see if there is anything there. I recently added the 85685 RF preselector which has tracking filters and a 20dB preamp so this will be an interesting test.

#104 Reply
Posted by Psi on 05 Dec, 2015 21:14
Quote from: pascal_sweden on 03 Dec, 2015 21:17
There is a "new kid on the block" contender coming out in "spectrum analyzer land".
This time from Siglent! Will be interesting to compare it against Rigol DSA815-TG

https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/msg777458/#msg777458

aww, it doesnt do 5.8ghz. This is not relivent to my interests

#105 Reply
Posted by ted572 on 27 Jan, 2016 00:24
DSA815 Update 00.01.15.01.00 dated 2016-01-22 is available at http://int.rigol.com/Support/SoftDownload/3
Under 'Product Series' select 'Spectrum Analyzer', then 'DSA815 Firmware_00.01.15.01.00'.
It is not yet available on the Rigol North American web site.

I don't know yet what it is proposed to fix, but we have been waiting a long time now for Rigol to fix the issues with the Tracking Generator. Hopefully this release will help. . .

#106 Reply
Posted by xrunner on 27 Jan, 2016 00:41
Quote from: ted572 on 27 Jan, 2016 00:24
I don't know yet what it is proposed to fix, but we have been waiting a long time now for Rigol to fix the issues with the Tracking Generator. Hopefully this release will help. . .

Well one issue with this release is that it saves .jpg screen captures "upside-down".

#107 Reply
Posted by ted572 on 27 Jan, 2016 11:40
Quote from: xrunner on 27 Jan, 2016 00:41
Quote from: ted572 on 27 Jan, 2016 00:24
I don't know yet what it is proposed to fix, but we have been waiting a long time now for Rigol to fix the issues with the Tracking Generator. Hopefully this release will help. . .

Well one issue with this release is that it saves .jpg screen captures "upside-down".

Hi xrunner:
When you said: saves .jpg screen captures "upside -down", did you mean -> Prints them upside-down to a PictBridge printer? I understand that the DSA815 can only save its screen to a USB thumb drive as a BMP file. Have you discovered a way to save (not print) a .jpg file directly from the DSA815?

#108 Reply
Posted by xrunner on 27 Jan, 2016 13:20
Quote from: ted572 on 27 Jan, 2016 11:40
Hi xrunner:
When you said: saves .jpg screen captures "upside -down", did you mean -> Prints them upside-down to a PictBridge printer?

No, I mean it saves a .jpg upside-down to a USB thumb drive. That doesn't happen when you use .png or .bmp, just .jpg. I have confirmed this bug with another person I know who had a DSA815.

Quote
I understand that the DSA815 can only save its screen to a USB thumb drive as a BMP file. Have you discovered a way to save (not print) a .jpg file directly from the DSA815?

No, it can save it in three different type - .jpg, .bmp, and .png. Go to System, 3rd page, screenshots.

#109 Reply
Posted by ted572 on 27 Jan, 2016 15:05
Quote
I understand that the DSA815 can only save its screen to a USB thumb drive as a BMP file. Have you discovered a way to save (not print) a .jpg file directly from the DSA815?

No, it can save it in three different type - .jpg, .bmp, and .png. Go to System, 3rd page, screenshots.
[/quote]

Thanks so much. I don't know how I missed this capability. I learned something from you and I'm grateful. And it turns out that this capability was added in this new firmware 00.01.15.01.00. Ted

Edit: Added ref. to this being new in 00.01.15.01.00 Firmware

#110 Reply
Posted by pa3bca on 27 Jan, 2016 16:49
Just curious. Why would you want to save the image as .jpg?
.png filesize is just as small (even smaller probably) and .png is lossless, at least I think the Rigol .png is.
.png Looks much better than the somewhat fuzzy compressed .jpg's

#111 Reply
Posted by ted572 on 27 Jan, 2016 17:56
Quote from: pa3bca on 27 Jan, 2016 16:49
Just curious. Why would you want to save the image as .jpg?
.png filesize is just as small (even smaller probably) and .png is lossless, at least I think the Rigol .png is.
.png Looks much better than the somewhat fuzzy compressed .jpg's

Yes, I also prefer the PNG file format and it is around half the size of the JPG. The BMP and PNG files are not flipped over like the JPG is. BUT, it does take the DSA815 a lot longer to save a .png file to USB.
Comparison of DSA815 File Size and Time to Save (Size/Time): BMP ~1,125KB/~10sec. JPG ~110KB/~5 sec. PNG ~55KB/~23sec.

#112 Reply
Posted by pa3bca on 27 Jan, 2016 18:02
Quote from: ted572 on 27 Jan, 2016 17:56
Yes, I also prefer the PNG file format and it is around half the size of the JPG. The BMP and PNG files are not flipped over like the JPG is. BUT, it does take the DSA815 a lot longer to save a .png file to USB.
Comparison of DSA815 File Size and Time to Save (Size/Time): BMP ~1,125KB/~10sec. JPG ~110KB/~5 sec. PNG ~55KB/~23sec.
That long huh? 23 sec is crazy, and I can see no good reason for it to take that long. Well I know saving to USB takes a long time, so I always use a PC for screenshots. Using Ultra Sigma it only takes a few seconds for a screenshot and that ends up directly on the target PC anyway. (Ultra Sigma is next to useless for anything else though, the only other time I used it was for entering the magic codes )

#113 Reply
Posted by xrunner on 27 Jan, 2016 23:08
Regarding the upside-down .jpg issue, I submitted a report to Rigol and got this back -

Quote

My name is Jason Chonko and I am an Applications Engineer at Rigol Technologies USA. Thank you for writing in.

Yes. You are correct.

I will submit a bug to our R&D department.

When they have fixed the issue, we will send you an email with the new firmware revision.

Thank you for choosing Rigol.

Sincerely,
Jason

#114 Reply
Posted by ted572 on 28 Jan, 2016 00:00
Good news for those of us with the original DSA815 hardware versions: DSA815 Firmware Update 00.01.15.01.00 dated 2016-01-22 (available at -> http://int.rigol.com/Support/SoftDownload/3w (Under 'Product Series' select 'Spectrum Analyzer', then 'DSA815 Firmware_00.01.15.01.00') has corrected the issue I reported here -> https://www.eevblog.com/forum/testgear/rigol-dsa815-tg-10-mhz-issue/msg749136/#msg749136 And yes all of the options are still in place, and so far I haven't found any other issues.

Note: Edited above (in bold type) to add the full path to the new Firmware. From https://www.eevblog.com/forum/testgear/rigol-dsa815-tg-10-mhz-issue/msg851904/#msg851904

I'm happy, but how about those of you with the newer DSA815 hardware that 'Giorgio' (https://www.eevblog.com/forum/testgear/rigol-dsa815-tg-10-mhz-issue/msg723692/#msg723692) originally discovered and started this thread for. Has this been fixed?

#115 Reply
Posted by georges80 on 28 Jan, 2016 01:32
Thanks. Your link doesn't work for me, but it gets you into the system and then a few clicks will find the dsa815 firmware.

Agree, on the older hardware (which is what I have), the frequency shift is now gone when turning on the TG. So, that's good news for some of us. Just need to hear from folk with the newer hardware.

cheers,
george.

#116 Reply
Posted by Melt-O-Tronic on 28 Jan, 2016 01:46
Quote from: xrunner on 27 Jan, 2016 13:20
Quote from: ted572 on 27 Jan, 2016 11:40
Hi xrunner:
When you said: saves .jpg screen captures "upside -down", did you mean -> Prints them upside-down to a PictBridge printer?

No, I mean it saves a .jpg upside-down to a USB thumb drive. That doesn't happen when you use .png or .bmp, just .jpg. I have confirmed this bug with another person I know who had a DSA815.

Quote
I understand that the DSA815 can only save its screen to a USB thumb drive as a BMP file. Have you discovered a way to save (not print) a .jpg file directly from the DSA815?

No, it can save it in three different type - .jpg, .bmp, and .png. Go to System, 3rd page, screenshots.

What??? My System menu only has 2 pages and I'm on FW 1.14. What am I missing here? I spent an embarrassing amount of time over the weekend trying to figure out how to save screen shots to JPG.

#117 Reply
Posted by ted572 on 28 Jan, 2016 01:59
Quote from: Melt-O-Tronic on 28 Jan, 2016 01:46
What??? My System menu only has 2 pages and I'm on FW 1.14. What am I missing here? I spent an embarrassing amount of time over the weekend trying to figure out how to save screen shots to JPG.

Then please install the new version firmware to see if it shows up. I also didn't know where to find it until I was pointed to it, so maybe it was just added in this new Firmware 1.15 (?). Please let me know if you now find it. Thank you.

#118 Reply
Posted by Melt-O-Tronic on 28 Jan, 2016 03:36
Thanks! You're right -- it is v1.15 only. This is terrific!

I get the upside-down JPG with a white border too, but at least there is JPG finally! PNG is very slow as stated above, but works perfectly and has better compression than JPG.

#119 Reply
Posted by vk6zgo on 28 Jan, 2016 06:18
.
Weird results aren't just a Rigol problem:

We encountered a strange one with an IFR unit a couple of years ago.

The DUT had a PLL which would "let go" & revert to a default frequency about 30MHz higher than required.
This happened quite fast,& the IFR displayed it as a bunch of carriers,leading us to chase non-existent spurious oscillations.

We bought an "El Cheapo" SA,(which was an analog type) to constantly monitor 4 Transmitters (in an ISM application),& free up the IFR for other work.

The analog SA showed the PLL frequency change clearly.

Once the IFR's "bug" was known,we could work around it.

#120 Reply
Posted by ted572 on 28 Jan, 2016 17:56
Features of DSA815 Firmware Version 00.01.15.01.00, 12/25/2015:
1. Match performance/operation for newer hardware revision 05 / 07.
2. Add JPEG and PNG formats to picture save options.
3. Add CSV Correction file loading function.
4. Solved the problem of invalid data loading command.
5. Solved the problem of power average math result error.
6. Solved the problem of LAN communication connection errors.
7. Solved the problem of the amplitude of the output from the TG jumping at 10MHz.
8. Solved the problem with the peak frequency position changing after activating the TG.
9. Solved the problem with the EMI filter changing after calibration.
10. LAN library update, supports multi interface communication.
11. Solve the issue with garbled file names after opening 4 traces.

Edit: Revised item 1. from '1.05' to '07 and 05'.

#121 Reply
Posted by Howardlong on 28 Jan, 2016 20:31
Quote from: georges80 on 28 Jan, 2016 01:32
Thanks. Your link doesn't work for me, but it gets you into the system and then a few clicks will find the dsa815 firmware.

Agree, on the older hardware (which is what I have), the frequency shift is now gone when turning on the TG. So, that's good news for some of us. Just need to hear from folk with the newer hardware.

cheers,
george.

Very quick check on a "newer" unit suggests the 10MHz bug is gone. The trials appear to remain in place, but this is with the smarter version that resets the counter on power up via I2C, not the short. I'll leave that to some other brave soul to test.

I haven't tried anything else.

Main board 00.07
FPGA 00.05
Digital board FPGA 00.05
Firmware 00.01.15
Boot 00.01.04

#122 Reply
Posted by Wall-E on 29 Jan, 2016 16:56
New hardware unit with short on I1105 looks good with all fixes and still have all options. Could not be happier.

#123 Reply
Posted by N8AUM on 29 Jan, 2016 17:15
Quote from: Wall-E on 29 Jan, 2016 16:56
New hardware unit with short on I1105 looks good with all fixes and still have all options. Could not be happier.

Way to be the 1st !

Just upgraded my "moded" 815 (latest board rev) and all the extras remained in tacked !

#124 Reply
Posted by ted572 on 29 Jan, 2016 20:30
Quote from: N8AUM on 29 Jan, 2016 17:15
Quote from: Wall-E on 29 Jan, 2016 16:56
New hardware unit with short on I1105 looks good with all fixes and still have all options. Could not be happier.
Just upgraded my "moded" 815 (latest board rev) and all the extras remained in tacked !

Great, this is good news! Did you try the Power Sweep mode to see if they fixed it in your DSA815 with the new hardware?

#125 Reply
Posted by pa3bca on 29 Jan, 2016 22:13
Well the frequency shift with the TG (as a function of the RBW/sweeptime) seems to have been fixed with the .15 firmware.
However I still have the amplitude jump just below 500 KHz (also see my previous postings)
With all settings default and the RBW set at 10 KHz (span from 300-700 KHz) the jump is very present:

RBW reduced to 1 KHz.

No frequency shift (fixed, good!) but the amp jump at 10 KHz RBWE.. not so good.
Did they only look at & fix the 10 MHz issue? Funny thing is my 815 never had the 10 MHz amp jump issue.

#126 Reply
Posted by N8AUM on 30 Jan, 2016 01:51
Power sweep seems to work ok with mine. Havnt had time to play with it. Just got more boat anchors in that am currently playing with, 8566B, 8350B with 83595 plugin, 54502A and a 5342A

#127 Reply
Posted by Howardlong on 30 Jan, 2016 06:58
Quote from: N8AUM on 30 Jan, 2016 01:51

Power sweep seems to work ok with mine. Havnt had time to play with it. Just got more boat anchors in that am currently playing with, 8566B, 8350B with 83595 plugin, 54502A and a 5342A

I did a quick test, but it looked like the sweep didn't cover the range and/or base supplied in the UI: the maximum I could get was from about -16 to -2dBm, and only over about 1/2 of the span with it flatflining at the start and end, but that is compared to no change at all as it once was.

#128 Reply
Posted by N8AUM on 30 Jan, 2016 13:51
Here's a screen shot of power sweep enabled with the latest firmware.

#129 Reply
Posted by N8AUM on 30 Jan, 2016 19:01
Thanks for the PM Ted.

Quick observation with my 815. With the RBW set at 100 Hz I can go as low as 524 Hz span and still have the full 20db swing from low end to high end. That shows up as 9 steps on the sweep.

#130 Reply
Posted by Howardlong on 30 Jan, 2016 19:04
Quote from: ted572 on 30 Jan, 2016 13:51
Quote from: Howardlong on 30 Jan, 2016 06:58
I did a quick test, but it looked like the sweep didn't cover the range and/or base supplied in the UI: the maximum I could get was from about -16 to -2dBm, and only over about 1/2 of the span with it flatflining at the start and end, but that is compared to no change at all as it once was.
Howard: The original "DSA815-TG 'Power Sweep' Mode" instructions (.pdf file) that I put together and posted worked fine for me and others with the original DSA815 hardware, but doesn't work the same now with the new 1.15 FW. The big thing is that Power Sweep doesn't seem to work properly with 100 Hz SPAN, so I changed it to 300 Hz. It will also work with larger Spans but results are shifted horizontally. For now I recommend using 300 Hz. Please play with it and let me know what you think. Anyway hopefully it will now work for you, or at least get you better in the ball park.

Please use the revised procedure attacked below. Thank you, Ted

I used your originally ;-)

On the SA I have here under 1.15, I need a span of at least 1.315kHz to be able to see all 20 x 1dB steps of resolution in use. Beyond that, at low spans, the steps become non-equidistant, but above a few kHz it's probably not a problem... unless you're trying to measure an amp with a narrow band filter!

#131 Reply
Posted by Howardlong on 30 Jan, 2016 19:06
Quote from: N8AUM on 30 Jan, 2016 19:01
Thanks for the PM Ted.

Quick observation with my 815. With the RBW set at 100 Hz I can go as low as 524 Hz span and still have the full 20db swing from low end to high end. That shows up as 9 steps on the sweep.

Same here, but there are only 10 steps of 1 or 2 dB rather than 20 steps of 1dB. 1.315kHz span gets you the full 20 x 1 dB steps.

#132 Reply
Posted by N8AUM on 31 Jan, 2016 07:35
My toy had 34hrs left on its trails when I did the hardware mod. I left it on to run out the clock which of course disables the trails, and it did. Power cycled it and trails came back.

#133 Reply
Posted by ted572 on 31 Jan, 2016 12:57
Configuring the DSA815-TG for the Power Sweep mode

The changes Rigol made in DSA815 Firmware 00.01.15.01.00 required some changes to the Power Sweep mode information that I previously posted. Please feel free to let me know if you have any questions or issues with this. Thank you

DSA815-TG Power Sweep Mode .pdf