Author Topic: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator  (Read 315973 times)

0 Members and 5 Guests are viewing this topic.

Offline Ebel0410

  • Contributor
  • Posts: 40
  • Country: fr
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #250 on: November 26, 2017, 08:23:05 am »
So replace brand new components with old ones in unknown condition?
I never said that all my old stock is vintage, and I tested them before mounting on the PS pcb.
Reuse quality components for servicing is better than use brand new ones with low performances imho.

Quote
Why is that transformer at that ridiculous angle, surely you sorted that out?
Not at all, the FY6600 was delivered with the transformer in this position.
 

Offline bianchifan

  • Regular Contributor
  • *
  • Posts: 66
  • Country: de
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #251 on: November 26, 2017, 12:33:32 pm »
2. Then, I replaced the dodgy output amplifier op-amp. The original one is an unknown 30021. At first I thought this is a dual variant of the single THS3001 but some guy has inquired about it from Texas Instruments and they replied that they do not manufacture such.
Your dodgy crappy unknown 30021 are very well known 3002I -> THS3002 with extended operational temperature!
They are used frequently in FeelTech Signal Genarators (esp. 32xx) and often discussed in russian forums like mysku, radiokot, x-faq and so on...
For various reasons many people prefer them to THS3092, although 3092 can drive more current.
 
The following users thanked this post: sorin, xzswq21, jebem

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #252 on: November 26, 2017, 03:19:59 pm »
Mine came yesterday.  Being forewarned I plugged it in and took a meter to it.  50 V between BNC and earth ground.  FeelTech *really* should fix that.  It's a huge menace to anyone using it as a signal source for embedded work.  I've generally steered clear of such kit because of this.  The *only* reason I felt comfortable buying one was the information from this thread.  Hopefully FeelTech will take that to heart and modify the design appropriately.   I'm mostly going to be using it with MCUs, so the other mods are less important to me.
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 523
  • Country: nl
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #253 on: November 26, 2017, 04:29:31 pm »
I really still believe it is theoratically inherent to the 2-prong class II design. In other words, all non-ground connected SMPS have it.
How much they have it depends on the size of the Y-Cap, and that depends on how much they want to suppress EMI and how much power supplies
(Note to self, I need to check that 'real good' Samsung USB for EMI leakage (but I do not have the equipment) and floating while loaded)

So I believe a design change simply means, factory fitted with 3-prong cable.
Which is a good idea, for sure.

Please note: I am not saying its a good supply, but in my view, the leakage is not part of its badness.
What is part of its badness is being labeled at 12V but giving over 15V, unconnected 20V but having only 16V caps.
 

Offline don.r

  • Frequent Contributor
  • **
  • Posts: 730
  • Country: ca
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #254 on: November 26, 2017, 04:40:59 pm »
What is part of its badness is being labeled at 12V but giving over 15V, unconnected 20V but having only 16V caps.

Yes. This is not acceptable. The output caps should be 25V rated if the supply is giving 15V. The other problem is that the main board caps are all surface mount and 16V rated also. I'll have to probe around and see what the voltages are, especially around the output amplifiers.
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #255 on: November 26, 2017, 05:52:08 pm »
I just connected my scope to the ground of the unit.  Yikes!  Yeah, it's 50 Vrms, but it's 174 Vpk-pk!

The first scope image is with the probe connected to the BNC ground, no ground lead connected.  The second image is the BNC ground lead of the FeelTech connected to the scope ground via a 100K resistor and probe across the resistor.

It's so bad I'm reconsidering whether to keep it.  I notice that there are no approval marking of any kind.  It shipped from CA which raises the question of whether it's even legal to sell it in the US.

Edit:

Here's the drop across a 10K resistor.
« Last Edit: November 26, 2017, 06:39:44 pm by rhb »
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 523
  • Country: nl
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #256 on: November 26, 2017, 05:56:25 pm »
Now do the same for any other 2-prong device you have.
It will be better for smaller power supplies, but most of them look scary.

You are basically just measuring mains through a 1pF capacitor.

Whether you want to send back or not is up to you, but of you keep it, just do the 3-prong+y-cap mod.
I now have about 1Vac, and it is still floating. Or hard-link to the GND for 0V
 

Offline Specmaster

  • Super Contributor
  • ***
  • Posts: 10511
  • Country: gb
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #257 on: November 26, 2017, 06:01:22 pm »
Quote
Why is that transformer at that ridiculous angle, surely you sorted that out?
Not at all, the FY6600 was delivered with the transformer in this position.
WTF is that all about then, is there something under it like thermal cutout or what, if there's nothing there then I would have certainly have corrected that, its a reasonable sized lump flapping in the breeze there and producing untold stresses on the PCB due to inherent vibrations.  :palm:
Who let Murphy in?

Brymen-Fluke-Solartron-HP-Thurlby-Heathkit-Thander-Tek-Extech-Black Star-GW-Advance-Avo-Kyoritsu-Amprobe-ITT-Robin-TTi
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #258 on: November 26, 2017, 07:14:37 pm »
Now do the same for any other 2-prong device you have.
It will be better for smaller power supplies, but most of them look scary.

Interesting.  I just repeated the test with a 5V wall wart supply from a kit of Arduino stuff.  Very similar picture.  I'd never  considered the matter.  Good thing to be more aware of.  I'd never thought closely about ground loops in my bench setup, but clearly with a variety of power sources and grounds, it gets messy.

I'm looking it over closely.  I don't understand the main board SMD electrolytic markings.  Two are clearly marked 16 V, but most are marked "100 10S 3R7"  Has anyone got a decoder ring?

I shall certainly put a proper ground on it if I keep it.
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 523
  • Country: nl
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #259 on: November 26, 2017, 07:35:31 pm »
I'd never  considered the matter.

Same for me, till this thread. Learned a bit

I have a old ATX power supply that I use for easy and beefy +5V and +12V (as well as minor -12V)
I had isolated them from ground, yeaaaars ago, never had a problem.
Probably because I always hook up 0V first, and then my DUT is referenced to that 0V.

But just to feel happy, I just added a capacitor and 1 meg resistor to ground there too (I like to keep it floating)
Now also looks a lot nicer. from 60Vac floating to 1Vac floating


Basically, the build in Y-cap crossing the transformer has to be significant larger (10x) then the inherent capacitance of the transformer itself, to avoid EMI.
And to counter mid-point "leaky" ac voltage, the cap from secondary 0V to ground should be significant (10x) bigger then that Y-cap
I believe this 2nd cap does not need to be Y, I am using X. Because it is not connected to mains. Also because I have those.
 

Offline don.r

  • Frequent Contributor
  • **
  • Posts: 730
  • Country: ca
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #260 on: November 26, 2017, 07:52:21 pm »
Now do the same for any other 2-prong device you have.
It will be better for smaller power supplies, but most of them look scary.

Interesting.  I just repeated the test with a 5V wall wart supply from a kit of Arduino stuff.  Very similar picture.  I'd never  considered the matter.  Good thing to be more aware of.  I'd never thought closely about ground loops in my bench setup, but clearly with a variety of power sources and grounds, it gets messy.

I'm looking it over closely.  I don't understand the main board SMD electrolytic markings.  Two are clearly marked 16 V, but most are marked "100 10S 3R7"  Has anyone got a decoder ring?

I shall certainly put a proper ground on it if I keep it.
The caps marked 10S are 10V. I checked continuity and none are connected to the "12"V rails. They are connected to the 5V or come after some of the other linear regulators on the board. The 16V ones do have continuity to the 12V rails. Hopefully they hold.
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 523
  • Country: nl
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #261 on: November 26, 2017, 08:19:49 pm »
The 16V ones do have continuity to the 12V rails. Hopefully they hold.

When I worked on the supply, and had it disconnected, I measured 20V.
Before I connected to the AWG board again, I had the presence of mind to first discharge the power supply caps. Gave a bit of a zap. Imagine hooking up the AWG board with 20V precharged into the caps...
 

Offline don.r

  • Frequent Contributor
  • **
  • Posts: 730
  • Country: ca
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #262 on: November 26, 2017, 08:26:31 pm »
My supply measured 15.8V when unloaded. Appears to be a lot of variability, at least when measured unloaded.
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #263 on: November 27, 2017, 02:21:41 am »
I worked up the nerve to connect it to my new scope.  The Instek spectrum analyzer demo allows the user to control the SA settings by pressing the lower right corner button  by the screen.  I've set it up for 0-100 MHz span.  As far as I can tell, this is the MDO line SA function.  It's a bit flaky, but seems pretty usable if you  can accept an RBW of 10KHz.  The white trace is a 16 point average of the spectrum.

The plots are 0.1 V, 1.0 V and 10 V output settings.  At low output voltages it looks quite good, but the harmonics get really bad above 1 V.  The obvious question is, is it the PS or the PA?

It seems to me the appropriate next step is to hook it up to my bench supply.

The SG is set to 5 MHz.  I'm using an MSO-2204EA.
« Last Edit: November 27, 2017, 02:29:26 am by rhb »
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #264 on: November 27, 2017, 02:43:49 am »
For the sake of comparison here is the internal AWG on the Instek MSO-2204EA at 1 Vpk-pk and 5 V pk-pk.

Edit:

For completeness I've added the 0.1 V case.  The Instek built in AWG pretty much goes to hell around 1.5 V.  5 V is the max.

All the plots had a 6 dB attenuator inline to provide a 50 ohm match for the source.
« Last Edit: November 27, 2017, 03:04:20 am by rhb »
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 523
  • Country: nl
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #265 on: November 27, 2017, 06:18:05 am »
That is undesirable. I do not have a SA myself (and my  rigol FFT is non-usable) but looking at this link it should be much better.
Vpp is not mentioned, so I am assuming 5Vpp default.

Maybe your log steps of 0.1 - 1 - 10 are a bit too much, and the supply only falls short over 9Vpp
So mid-voltages are still good.

But this guy does indeed claims both a better Op-Amp and a better supply are needed at 20Vpp.

Interested to see what it does when hooked to your bench supply instead of internal, but smaller steps please.
« Last Edit: November 27, 2017, 06:48:33 am by cybermaus »
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #266 on: November 27, 2017, 01:32:13 pm »
The harmonics grow rapidly above about 1.5 V.  5 V was the limit on the Instek AWG.  I am unable to control the input attenuator in the SA demo.  So some of this may be sampling artifacts.  I'll repeat later today with the regular FFT function.  Unfortunately, that does not have the ability to super impose a running average of the spectrum.
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #267 on: November 27, 2017, 10:41:06 pm »
Here are screen shots with the generator running from an Instek linear bench supply feeding +- 12 & +5 V.  The harmonics start growing around 6-7 V.

For these I used the regular FFT function rather than the SA demo.  I think a lot of the harmonic distortion in the previous screen shots was the scope attenuator being set too low and clipping the ADC input.  The SA demo does not allow any adjustment of the input channel gain that I can tell.

There's a 6 dB attenuator inline to provide a 50 ohm termination to the FeelTech.

I'll repeat the same tests using the internal PS later.
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #268 on: November 27, 2017, 11:19:32 pm »
Here they are using the internal PS.  I can't see a significant difference toggling between the images.
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 523
  • Country: nl
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #269 on: November 28, 2017, 05:57:26 am »
Here they are using the internal PS.  I can't see a significant difference toggling between the images.

So if I read that correct:

+01Vpp -55dB
+05Vpp -55dB
+10Vpp -44dB
+15Vpp -31dB
+20Vpp -25dB

I actually think that is pretty good. That is still 25dB down at full 20Vpp swing....

I'd argue the old argument that it is only a $100 device, comparable features to a to a $900 Rigol DG1062Z
But to be quite honest, I think that Rigol may not fare much, if any, better at +20Vpp

For comparison, look at the remainder of this video on the previous generation of cheap awg.
If I can eyeball the distortion, imagine the spectrum on that?
 
The following users thanked this post: tjxkeith

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #270 on: November 28, 2017, 01:47:14 pm »
I think it is quite good, especially for the price. And with my embedded & SDR  focus I can't imagine a need to go above 5 V.  I've noticed that there is a relay click when going from 5 to 6 V.

For my purposes I think all that's needed is to suppress the AC leakage.  Which raises a question.  One Y cap or two?  I'd never even heard of them, but the last time I was heavily active in electronics (i.e. had a bench set up) most supplies were linear.

While fooling around last night I noticed some curious behavior for ARB28 when I varied the frequency setting.  So an investigation of the frequency axis seems in order for this evening.  Unfortunately, it appears that either the dual display FFT or the V 1.32 FW update removed the ability to display an FFT all the way to Nyquist which is rather annoying.  So I got a better UI at the price of reduced functionality.
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #271 on: November 29, 2017, 01:19:25 am »
Here is a plot of 1 MHz at 1 V.  Cursor measurements give the following:

2nd  -60
3rd   -64
4th  -60
5th  -60
6th  -52
7th  -62
8th  -60
9th  -60

Noise floor -72

The Instek SA demo is set up with the internal function generator a 1 V.  The input attenuator setting is unknown, however, I should expect that Instek made there demo look as good as possible.  So I think the results are reasonably accurate within the limitations of 8 bit sampling.  Looks quite good to me. In fact, amazing.  I don't want to think about what it would take to build an analog 1 MHz oscillator with equal performance.
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #272 on: November 29, 2017, 02:17:53 am »
Here are spectra at 1 V for 10-60 MHz in 10 MHz steps.
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #273 on: November 29, 2017, 02:39:52 am »
Same setup, but 1 MHz span
 

Offline cybermaus

  • Frequent Contributor
  • **
  • Posts: 523
  • Country: nl
Re: FeelTech FY6600 60MHz 2-Ch VCO Function Arbitrary Waveform Signal Generator
« Reply #274 on: November 29, 2017, 07:11:49 am »

I wonder if those sidebands (at roughly 30KHz / 300pSec offset) are jitter from the oscillator.

Would be interesting to see if they still exists after replacing that oscillator with that NB3N502 PLL that supposedly only has 15pSec jitter
Either ArthurDent needs to check that, or I need to buy a decent Spectrum Analyser.


 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf