Author Topic: Re: FeelElec New Arrival FY-6900 Signal Generator  (Read 9205 times)

0 Members and 1 Guest are viewing this topic.

Offline evava

  • Regular Contributor
  • *
  • Posts: 146
  • Country: cz
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #75 on: August 03, 2019, 09:31:35 am »
pictures:

Go figure!
« Last Edit: August 03, 2019, 09:35:25 am by evava »
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2240
  • Country: se
  • Country: Broken Britain
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #76 on: August 03, 2019, 10:14:36 am »
"If I set both channels to square wave this triple curling appears in both channels!"

Are the curly edges still there with both channels set to frequencies that align exactly with the 4ns sampling times, 1Mhz and 2MHz would do.

15Mhz, yellow trigger, both square, both trippled.jpg
If the scope's 1GSa/s there's only 2.5 samples per division, most of those curls will be the scope's Sinx doing the best it can with so few samples on the edge.
« Last Edit: August 03, 2019, 10:23:58 am by StillTrying »
CML+  That took much longer than I thought it would.
 

Offline evava

  • Regular Contributor
  • *
  • Posts: 146
  • Country: cz
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #77 on: August 03, 2019, 10:51:05 am »
"If I set both channels to square wave this triple curling appears in both channels!"

Are the curly edges still there with both channels set to frequencies that align exactly with the 4ns sampling times, 1Mhz and 2MHz would do.

Blue: 2Mhz, not curling
Yellow: trigger, 30mHz next to, curling.
No jitter.
 

Offline evava

  • Regular Contributor
  • *
  • Posts: 146
  • Country: cz
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #78 on: August 03, 2019, 10:56:49 am »
And trigger vice versa:

Blue: trigger, 2Mhz, not curling
Yellow: 30mHz next to, curling.
Again, no jitter.

Instead of jitter there seems to be curling?  :-//
 

Offline evava

  • Regular Contributor
  • *
  • Posts: 146
  • Country: cz
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #79 on: August 03, 2019, 12:40:42 pm »
And 100kHz:

- time base 5ns per division, you see leading edges
- yellow trace 100kHz,
- blue trace trigger, 300uHz next to, curling
yellow slips slowly, then skip (jitter?), and again slips slowly, while blue trace is constantly curling!

They are trying so desperately to suppress the jitter..or?

I wonder, how Siglent or Rigol cope with that?
« Last Edit: August 03, 2019, 12:44:38 pm by evava »
 

Offline Johnny B Good

  • Frequent Contributor
  • **
  • Posts: 254
  • Country: gb
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #80 on: August 03, 2019, 03:45:55 pm »
Thank you JBG for your answer.
Indeed it must have been some crosstalk or crossmodulation between channels or external modulation  of a special kind.
If I set both channels to square wawe this triple curling appears in both channels!

(you can set your player to continuously replaying this short clip, so you can better see what is going on)

If I push Single on scope, there are always just thin lines (yellow and blue ones), but always they have another shape.
Photos in next post.

 That was a more informative video clip. I tried setting VLC to play those 1 second clips on a continuous loop but the best I could do was to get it to play two passes at a time. I had more luck with this 3 second video and the trick appears to be to set the B point a few frames short of the bitter end to achieve continuous loop play. I guess the brevity of those 1 second clips was the reason for my limited success with loop play.

 It's surprising what a difference reducing the resolution from full HD down to a 1280x720 clip does for the playing time you can get out of a 4.5MB clip. I'd expect just mounting the camera onto a tripod, locked off to eliminate any camera wobbles would let you get more time for a given file size.

 All the common video formats use lossy compression to minimise storage requirements (and reduce writing speed to the media), so the less changes, frame to frame, that have to be recorded, the less the amount of picture data that has to be sent in each group of pictures (GoP), typically half to one second's worth of frames between a total refresh of each key frame during which only frame to frame differences are recorded.

 "Wobbly-cam" recordings aren't much of an issue when you're hosting your home movies on Youtube but when it comes to trying to cram a few second's worth of video into a 4 or 5 MB attachment limit, this sort of care in the camera work becomes rather more critical. If you don't have a tripod, you can arrange some sort of stable support to rest the camera upon (a stack of hardback books for height adjustment on a chair or small table for example).

 Anyway, having noted those points regarding video technique, I can comment on the effect of this "Magic Pulse" you'd managed to capture. The "Magic" has obviously eliminated the gross 4ns jumps you'd otherwise be seeing with a 6600 or 6800 function generator, lending the movement a more liquid, if ripply, appearance.

 It looks like the 'jumps' in phase are now around the 1.5ns mark. The tripling up looks like it could be the effect of the 'scope using some sort of persistence (a one second persistence setting perhaps?), otherwise I'd expect this tripling up effect to produce a blurred image rather than the crystal sharp images of seemingly three horizontal scans worth overlaid on top of each other.

 It's quite obvious that you're terminating the 'scope end of those BNC cables with 50 ohms (5Vpp settings with the scope showing circa 2.5Vpp values) so we can't blame badly matched cables for any of these effects. It would be interesting to see whether others in your priviledged position of FY6900 ownership would see the same effect using a different make and/or model of DSO like a Siglent SDS1202X-E for example  ;)

 Anyway, that's something to ponder on and I thank you for your contribution, both for this longer clip :) and for reminding me about the forgotten loop playback option in VLC. :palm:

JBG
 

Offline evava

  • Regular Contributor
  • *
  • Posts: 146
  • Country: cz
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #81 on: August 04, 2019, 09:36:11 am »
Hi JBG,
thank you for your kind words and new information in every your post.
And thank you for pointing me to Persistence setting on the scope, I have never thought it has much use. (I was wrong, obviously).
And support for camera is now ready  :)

Quote
It's quite obvious that you're terminating the 'scope end of those BNC cables with 50 ohms (5Vpp settings with the scope showing circa 2.5Vpp values) so we can't blame badly matched cables for any of these effects
Yes, I use "Banggood Special" ones, 50(?!) Ohm feed-through terminators. They work just ok.

Quote
The tripling up looks like it could be the effect of the 'scope using some sort of persistence (a one second persistence setting perhaps?), otherwise I'd expect this tripling up effect to produce a blurred image rather than the crystal sharp images of seemingly three horizontal scans worth overlaid on top of each other.

In all previous pictures and videos I had scope persistence set to "Min".

For better clarification I am now attaching pictures a videos with persistence set to "Min" and then to "Infinite", to maybe allow you to discern what is the feature of my Rigol and what not.

P.S. I like your polite English very much :)
 

Offline evava

  • Regular Contributor
  • *
  • Posts: 146
  • Country: cz
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #82 on: August 04, 2019, 09:38:20 am »
15MHz, persistence to "Min" - like in all previous videos.
 

Offline evava

  • Regular Contributor
  • *
  • Posts: 146
  • Country: cz
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #83 on: August 04, 2019, 09:39:31 am »
15MHz, persistence set to "Infinite".

It seems to me that "Infinite" setting behaves rather like "Min" setting should behave...  :-//

Other settings (except Min and Infinite) behave like expected.
« Last Edit: August 04, 2019, 09:53:52 am by evava »
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2240
  • Country: se
  • Country: Broken Britain
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #84 on: August 04, 2019, 11:43:26 am »
The "Magic" has obviously eliminated the gross 4ns jumps you'd otherwise be seeing with a 6600 or 6800 function generator, lending the movement a more liquid, if ripply, appearance.

Don't forget that the sine waves will still have the up to +/- 2ns sample timing jitter, it's just that the sine's amplitude can be corrected to fit the +/- 2ns off position. So there'll still be some timing wobble depending on the analogue BW between the DAC and squarer.

"It looks like the 'jumps' in phase are now around the 1.5ns mark."

I think at the mid level of the square waves it's better than that. I still don't know what their video was supposed to be showing. :)
Going by the threads on here it seem that many don't trust the Rigol even in dot mode when it gets down to only 2 samples on a fast edge.
« Last Edit: August 04, 2019, 11:49:18 am by StillTrying »
CML+  That took much longer than I thought it would.
 

Offline evava

  • Regular Contributor
  • *
  • Posts: 146
  • Country: cz
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #85 on: August 04, 2019, 12:20:10 pm »
Both traces square, both 150Hz , persistence Infinite:

And - jitter 4ns, finally!

Is this FY6900 better than FY6600/6800, can someone compare?
Is it upgrade or downgrade?
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2240
  • Country: se
  • Country: Broken Britain
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #86 on: August 05, 2019, 03:24:20 pm »
"Both traces square, both 150Hz , persistence Infinite:
And - jitter 4ns, finally!"


If they're squaring up a sine wave I'd expect some jitter to still be there at 150 kHz if it's a none 4ns frequency, when you get up to MHz the analogue BW will smooth the gaps in the sine's samples, the same as sinx draws between samples.
CML+  That took much longer than I thought it would.
 

Offline Johnny B Good

  • Frequent Contributor
  • **
  • Posts: 254
  • Country: gb
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #87 on: August 06, 2019, 03:59:24 am »
Hi JBG,
thank you for your kind words and new information in every your post.
And thank you for pointing me to Persistence setting on the scope, I have never thought it has much use. (I was wrong, obviously).
And support for camera is now ready  :)

Quote
It's quite obvious that you're terminating the 'scope end of those BNC cables with 50 ohms (5Vpp settings with the scope showing circa 2.5Vpp values) so we can't blame badly matched cables for any of these effects
Yes, I use "Banggood Special" ones, 50(?!) Ohm feed-through terminators. They work just ok.

Quote
The tripling up looks like it could be the effect of the 'scope using some sort of persistence (a one second persistence setting perhaps?), otherwise I'd expect this tripling up effect to produce a blurred image rather than the crystal sharp images of seemingly three horizontal scans worth overlaid on top of each other.

In all previous pictures and videos I had scope persistence set to "Min".

For better clarification I am now attaching pictures a videos with persistence set to "Min" and then to "Infinite", to maybe allow you to discern what is the feature of my Rigol and what not.

P.S. I like your polite English very much :)

 Regarding the double and triple vision effect, I've just had a thought that it might simply be an artefact of the camera due to the shutter speed allowing two or three screen refreshes to be captured. On second thought, after having another look at the latest 30MHz clip, perhaps not.

 Persistence is usually used to emulate a CRO's averaging effect through the long persistence phosphors used in the CRT which provided an indication of where the trace would be painted most often and where it would paint infrequently when using a high horizontal scan frequency, giving a better visual impression of the jitter noise and just plain noise on the waveforms being displayed.

 Early DSOs didn't include such CRT persistence effects which were missed by users who had upgraded from a CRO to a DSO and had been relying on this extra 'information' they'd been able to glean from this 'defect of persistence' that had been eliminated in the earlier DSO models. Suffice to say, FeelElec's 'scope settings must be making use of persistence to demonstrate the reduction of the 4ns jitter on square waves in exactly the same way that Siglent have seemingly done in their own 'scope trace captures shown in the data sheet for their own generators blessed with their own "Easy Pulse"(tm) feature.

 Multiple traces like the ones in your own screenshots and movie clips, ime, tend to be down to the DSO's trigger processing becoming confused by a level of randomness in the triggering signal, typically seen when trying to examine the HF ripple on the output rails of an SMPSU.

 I suspect the technique being used is a close relative to the concept of adding 'dither noise' to a music CD recording to convert the more objectional quantisisation noise to a less objection 'tape hiss' like background noise. In this case, it would appear that the 4ns jitter has been traded in for a 1ns randomised jitter in this case, judging as best as I can from those traces. I can't blame Siglent for using persistance in their own trace captures. Without such persistance, the crystal clarity is actually masking the true random nature of the phase dither noise that's now replacing the gross 4 and 6.7ns clock jitter that would otherwise show up without the Magic Pulse and Easy Pulse processing in each respective case.

 Of course, I could be completely off the mark but that's my best guess as to what's happening here. You probably only need a 1 second's worth of persistence to give a more accurate impression of the effectiveness of FeelElec's "Magic Pulse" processing. The only troubling thing about FeelElec's version of Siglent's Easy Pulse technique being the evidence of the occasional rogue 4ns jitter transient.

 I suspect that a frame by frame search of a minute or more's worth of movie footage without the use of persistence, using a Siglent generator in this test at its "Golden Frequency" of 3MHz, would reveal a similar result but without any rogue transient 6.6667ns ghost jitter events (or not - who's to say Siglent have done any better a job of this? - as far as anyone knows, such a test has never been done before).

JBG
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2240
  • Country: se
  • Country: Broken Britain
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #88 on: August 06, 2019, 06:54:20 am »
JBG you're describing 'waveform averaging' rather than persistence. I still think they're squaring up a DDS sine wave, but I won't be buying one just to do some tests. :)
CML+  That took much longer than I thought it would.
 

Offline Johnny B Good

  • Frequent Contributor
  • **
  • Posts: 254
  • Country: gb
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #89 on: August 06, 2019, 12:43:19 pm »
JBG you're describing 'waveform averaging' rather than persistence. I still think they're squaring up a DDS sine wave, but I won't be buying one just to do some tests. :)

 You and me, both! :)

 TBH, I can't imagine that many owners of the FY6600 and 6800 models would be tempted to go for such a modest "upgrade" as represented by the improvements offered in the FY6900. That's not to say that when it comes to a hobbyist's choice of first time AWG, the FY6900 would be the one to go for despite the modest price premium over the now cheaper offerings of the older models.

 The PSU must surely be an "improved" +/-15 volt design simply to allow it to meet its DC to 5MHz 24Vpp specification alone. A 20% boost over the "Industry Standard" P2P limit is more marketing gimmick than any genuine asked for improvement. The traditional solution to obtaining higher voltage test signals has always been provided by the simple expedient of inserting a linear amplifier into the signal chain to serve this less common requirement. What self respecting EE doesn't keep a suitable linear amp to hand in their test lab for just such unusual testing scenarios? ;)

 I'm looking forward to seeing the first comprehensive review and tear down youtube video, even if it's simply to satisfy prurient curiosity.  :popcorn:

JBG
 

Offline Adrian_Arg.

  • Regular Contributor
  • *
  • Posts: 209
  • Country: ar
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #90 on: August 06, 2019, 02:01:14 pm »
I am waiting for my KKMOOM 6900 60 MHz ;D
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2240
  • Country: se
  • Country: Broken Britain
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #91 on: August 06, 2019, 02:50:17 pm »
Why do they market these things with names like KABOOM and JUNKTEK.

It doesn't matter how magical and mysterious the pulses are, I wouldn't buy one while they show 8kHz as 00'008,000.000'000'000'000  >:D
CML+  That took much longer than I thought it would.
 

Offline excitedbox

  • Regular Contributor
  • *
  • Posts: 50
  • Country: de
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #92 on: August 11, 2019, 05:05:16 pm »
My biggest problem is the website. As a web designer I don´t understand how any business can neglect their store and image. Your website is essentially a store location like a brick and mortar store at 1% of the cost but 1000x the value. If you walked into a pizza place with outdated or missing menus, broken tables and chairs that didn´t give you what you ordered (broken links), you would never go back.

A website is also a representation of your company and it´s values. A neglected website whether broken, or outdated, reflects the service you can expect from that company. Yet maintaining a clean and professional website is a one time investment of a couple hundred for most businesses compared to the monthly rent and huge construction costs of a brick and mortar business.

It is crazy to me that many companies that spend 10s of thousands on store fixtures and thousands a month on rent are not willing to spend $500-2000 once to build an entire second store.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2240
  • Country: se
  • Country: Broken Britain
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #93 on: August 12, 2019, 11:17:26 am »
My biggest problem is the website. As a web designer I don´t understand how any business can neglect their store and image.

I keep thinking that FT's website, marketing and technical support are just a student project for stuff they buy in.
http://feelelec.com
http://en.feeltech.net

CML+  That took much longer than I thought it would.
 

Offline excitedbox

  • Regular Contributor
  • *
  • Posts: 50
  • Country: de
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #94 on: August 12, 2019, 05:20:32 pm »
They are not the only ones though. So many businesses just don´t get it. They literally pay as much a month for 1 employee as an entire website costs and is as if you opened up a store in every town around the world.

I offered someone a website out of pity last week. Website with SEO for 250 EUROS literally 10% of my normal rate since I just wanted some extra cash and haven´t heard from them since.
« Last Edit: August 12, 2019, 05:29:54 pm by excitedbox »
 

Offline dunkemhigh

  • Super Contributor
  • ***
  • Posts: 1471
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #95 on: August 12, 2019, 07:59:32 pm »
What's the website for? No-one is going to buy from it. Also, it may be a (relatively) trivial cost to set up, but it has to be maintained by someone, so that's an ongoing cost.

Quote
As a web designer I don´t understand how any business can neglect their store and image

MRDA :)
 

Offline excitedbox

  • Regular Contributor
  • *
  • Posts: 50
  • Country: de
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #96 on: August 13, 2019, 09:03:43 am »
Most websites don´t need to be maintained besides answering emails or maybe updating a few words of text every few months which takes literally minutes. Ongoing costs for a basic website are under 50 a year in hosting fees and domain name registration fees.
 

Offline Heszu

  • Contributor
  • Posts: 20
  • Country: vu
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #97 on: August 14, 2019, 02:02:42 pm »
Hi, there..
Is FY6900 equipped with Fan in the back panel ??
 

Offline Adrian_Arg.

  • Regular Contributor
  • *
  • Posts: 209
  • Country: ar
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #98 on: August 14, 2019, 03:43:00 pm »
Hi, there..
Is FY6900 equipped with Fan in the back panel ??

No, it is only the hole so that if you want you put one. |O |O
 
The following users thanked this post: Heszu

Offline Jon.C

  • Supporter
  • ****
  • Posts: 118
  • Country: ad
Re: FeelElec New Arrival FY-6900 Signal Generator
« Reply #99 on: August 15, 2019, 06:39:06 pm »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf