FeelTech FY3224S 24MHz 2-Channel DDS AW Function Signal Generator

[gwu]

My version of using transformer isolation to reduce residual voltage and leakage current.
initialy i had 46vac to ground and 30 ua leakage current. now 0

The isolation transformer i used is a 50va (yes it's overkill) triad N-68X
you can source it from most of the usual places, mouser, newark, digikey and so on for about 17 usd.

note: the center 'black' wire you see on the picture on the secondary is NOT a centertap, that is the electrostatic shield connector, it is usual it is white but i had to bodge on a wire as it was chopped short when i received it.

Hi innkeeper ,

You have board version 1.5.  What is your voltage rail outputs? +5,+5,+18 or +12?,-18 or -12?

Mine are +18,-18 but I am wondering if they sig gen will run of +12 -12 anyway. 

How successful is your mod?  I am considering replacing the whole supply board but curious if your style is still as good at reducing the output AC.

Cheers

[innkeeper]

Since i used a real isolation transformer, it was 100% successful.
0 leakage current
I'm still very satisfied on how it turned out.

I don't know what my rail voltages are, id have to measure.

[gwu]

Would you be able to check the labels on the rails?

No leakage current, and not leakage AC voltage? I measured no leakage DC on mine but there was AC. To be honest I wasn't sure where the current was being measured. Between the centre of BNC to outside of BNC?

I am trying to understand if the issue is the power board not isolating or if it's the board outputting when it shouldn't be.

Last, your EIC inlet filter, do you know if it improves anything? My question is if some of the noise on AC power is coming through the whole system and if in your situation the filter provided cleaner power and then isolation provided the safety. As that really would be a simpler approach then building a whole new supply.

Thanks.

[Kleinstein]

The rather high AC voltage / leakage current is not specific to the Feeltech unit. It is a general properties of switched mode supplies with only a 2 pin plug. To meet EMI regulation limits this essentially requires a small capacitor from the output to the input side. Without a dedicated ground or neutral   (so UK and Ausstralian plug might have an advantage here - but not the more common US or EU one, so units are often not takin advantage of knowing which pin is neutral), there is no good point on the input side. The best they can do an usually do is having 2 cap to both pins. So normally one will see half the mains voltage and a small but measurable leakage current to ground.

The easy fix would be to have a 3 Pin plug and cable and use a suitable filter toward PE. Connecting the GND to PE would be an other option - though it could cause trouble with ground loops, as most scopes are grounded too.

The leakage current to ground is usually measured form the outside of the BNC to GND (PE). However it does not make a significant difference measuring from the BNC center pin. Usually the current through the class Y caps should be relatively small and well build other instrument should be tolerate the possible discharge current on connection. So the voltage measured is not such aa big deal.

[gwu]

The rather high AC voltage / leakage current is not specific to the Feeltech unit. It is a general properties of switched mode supplies with only a 2 pin plug. To meet EMI regulation limits this essentially requires a small capacitor from the output to the input side. Without a dedicated ground or neutral   (so UK and Ausstralian plug might have an advantage here - but not the more common US or EU one, so units are often not takin advantage of knowing which pin is neutral), there is no good point on the input side. The best they can do an usually do is having 2 cap to both pins. So normally one will see half the mains voltage and a small but measurable leakage current to ground.

The easy fix would be to have a 3 Pin plug and cable and use a suitable filter toward PE. Connecting the GND to PE would be an other option - though it could cause trouble with ground loops, as most scopes are grounded too.

The leakage current to ground is usually measured form the outside of the BNC to GND (PE). However it does not make a significant difference measuring from the BNC center pin. Usually the current through the class Y caps should be relatively small and well build other instrument should be tolerate the possible discharge current on connection. So the voltage measured is not such aa big deal.

Kleinstein, I'd like to check that I understand all the parts. I know a little about mains power but not a lot.

"To meet EMI regulation limits this essentially requires a small capacitor from the output to the input side".  Output and input of the psu?

"The best they can do an usually do is having 2 cap to both pins".  Does this mean literally wire from the mains plug -> cap -> psu? 1 cap on neutral and 1 on live?

"So normally one will see half the mains voltage and a small but measurable leakage current to ground." Could you explain why half?

"The easy fix would be to have a 3 Pin plug and cable and use a suitable filter toward PE." What is PE?

So is the linear power supply different because the transformer crates the decoupling and thus no capacitor is needed?

[Kleinstein]

There are litteraly 2 small caps: one from neutral to the output and one from the life to the the output. These 2 caps form a capacitive divider and thus one see's half the voltage.  For safety reasons these caps are Y-rates types that a guarantied not to fail short. Usual capacitance is about 4.7 nF.

In principle a transformer power supply will show a similar voltage. Here it is from capacitive coupling in the transformer. However the capacitance is usually much smaller (e.g. maybe 1%) than the Y-caps. A linear supply does not need the extra caps because there are essential not RF emissions. Even a SMPS behind the conventional transformer can usually get away without.

PE is the short for protective earth, so the safty ground from the 3 rd. pin. If available one could have the capacitors from L to PE and N to PE and than either a direct connection from PE to the BNC or just a capacitor.

[Unordung]

Hey guys,
Does anyone have a schematic for this device? I recently damaged what I suspect will be the output amplifier for Chanel one. I accidentally connected the output to 30vdc forgetting the de-coupling capacitor..... 
Everything still works ie it produces the signals they just have very little amplitude ie 200mV rather than 20v etc.

[Daruosha]

I received mine today. The output signal looks pretty clean and reasonable, however the unit itself crashed too often (especially when I turn on both channels and try to witch between channels and change the settings frequently). Is this normal (if yes, WTF!!) or is it possible to update the firmware and make it more stable?

[Dutch RC]

Question before I push the buy button, I was about to buy a XR2206 7 bucks signal generator and then it was out of stock.
Now I don't want to do anything high tech fancy stuff right now but if I spend 10* the money and get this one https://www.banggood.com/DANIU-FY3224S-24MHz-Dual-channel-Arbitrary-Waveform-DDS-Function-Signal-Generator-Sine-Square-Wave-Sweep-Counter-p-1157268.html?p=ZW051057416242016088 will it be worth the money?

For now what I want to do is making a led blink and if possible trigger the scope through the back output of this.

I need a blinking led to test camera latency.

[bitseeker]

For blinking an LED, you could just make an oscillator circuit very inexpensively to drive it (LC, crystal, 555 timer, etc.). So, spending $70 for that purpose is overkill. However, if you believe that you'll have use for a variety of waveforms at a selectable frequency, and sometimes even two at the same time with different phase or frequency, then the FY3224S is an inexpensive option that works fine in general.

Keep in mind that this is not a high-end instrument and has its downsides, though some can be worked around: no user-accessible firmware updating, poor low-amplitude output (use an external attenuator), distortion near zero-crossing, high jitter between channels depending on frequencies chosen, high-voltage AC leakage (as covered extensively in this thread), etc.

For general hobby electronics use, these downsides aren't showstoppers, but as always, caveat emptor with any tool you buy. Know what it does and doesn't do well, use it accordingly, and you'll be just fine.

As for the price, unless you're in a hurry, you may want to shop around. I've seen the 24 MHz version occasionally for around $50.

[don.r]

Question before I push the buy button, I was about to buy a XR2206 7 bucks signal generator and then it was out of stock.
Now I don't want to do anything high tech fancy stuff right now but if I spend 10* the money and get this one https://www.banggood.com/DANIU-FY3224S-24MHz-Dual-channel-Arbitrary-Waveform-DDS-Function-Signal-Generator-Sine-Square-Wave-Sweep-Counter-p-1157268.html?p=ZW051057416242016088 will it be worth the money?

For now what I want to do is making a led blink and if possible trigger the scope through the back output of this.

I need a blinking led to test camera latency.

Depending on the speed you wanted you could do this by programming a small Arduino. For less than $5 you have extreme versatility and a built in LED along with USB programmibility.

[g2]

Hello
I bought two weeks ago FeelTech FY3220S. Firmware v.4.
It has another FPGA of 144 pins.
I have drawn a partial schematic of the Feeltech FY3224S,
using drawings from eevblog (masterx81) and from these links:
http://kiedontaa.blogspot.dk/2017/09/feeltech-fy3224s-reverse-engineering.html
It is a very good article.
See next post for Schematic.

[g2]

Hello
I have drawn a new partial schematic of the Feeltech FY3224S,
because in the first schematic I had made mistakes in the attenuator (HC4053).
So I hope you see this here.
New schematic is attached.

[bitseeker]

Thanks for documenting this, g2, and following up with corrections.

[g2]

Hello, thank you to bitseeker for nice words.
New schematic is attached.

[jcrubin]

Can anyone supply from their own test equipment if capable, the harmonic distortion from their device at 1khz 1V unladen so that I can get a consensus as to the values people are getting.
I know what the documentation states.  Im interested in real world values.

If your test gear input has a specific impedance, please specify

thanx


j

[g0mgx]

He is a measurement using a Rigol DSA815

I can't do 1KHz so here is sinewave 1MHz -7.8dBm into 50 ohms.

The unit is the FY3224S 24MHz version firmware is 3.5.

Let me know if you want any other measurements.

Mark
G0MGX

[g0mgx]

and here is a 1MHz signal at as close to 0dBm as I can get.

Mark

[JohnPen]

Here are some traces using my Scotties Spectrum Analyser.  The first trace shows 'No signal' to the MSA which illustrates the various 'spurs' that exist in the design albeit  <-75dBm.
The second is from a home built Wien AF generator, uses an RA53 for old timers, set for 1K at ~0.2v o/p.  The lower level is to avoid driving the MSA  input mixer too much and creating IM distortion.  The third trace is the Feeltech set at 1K and 0.2v o/p.  Comparing the traces should help you see any differences and get a visual picture of the quality of the Feeltech's output.  The 'Hump' at the beginning of the traces is the Zero frequency hump typical of Spectrum analysers and is the breakthrough of the first oscillator in the front end mixer.

Hope this helps.

John

Edit:  NB. The latter traces show the rise in the Noise floor due mixing first oscillator phase noise.

[jcrubin]

Could you do 10k or possibly 20K, I need a value I can match and 1mhz is too high.  Sorry

[cdev]

You should be able to find it for much less than $70 now..

Question before I push the buy button, I was about to buy a XR2206 7 bucks signal generator and then it was out of stock.
Now I don't want to do anything high tech fancy stuff right now but if I spend 10* the money and get this one https://www.banggood.com/DANIU-FY3224S-24MHz-Dual-channel-Arbitrary-Waveform-DDS-Function-Signal-Generator-Sine-Square-Wave-Sweep-Counter-p-1157268.html?p=ZW051057416242016088 will it be worth the money?

For now what I want to do is making a led blink and if possible trigger the scope through the back output of this.

I need a blinking led to test camera latency.

[cowasaki]

I bought one of these and have since built an isolating power supply and USB board.  The isolator ICs arrived yesterday so I'm hoping to get the first board made later today but the power supply is right and the USB isolator is a standard circuit so should work.

NOTE:  The 3D rendering has the positive electrolytic drawn reversed AND the 3 smaller electrolytic should be 100nF ceramics  

So once I've got it built and working I should have some spare boards as it was only a few £ more for 10 than 2 and they sent me 11 anyway.

[bitseeker]

Thanks for sharing, cowasaki. Is J6 for a switch to enable/disable earth ground?

[cowasaki]

Thanks for sharing, cowasaki. Is J6 for a switch to enable/disable earth ground?

Yes it is. I thought there might be a use for it so added it just in case

[bitseeker]

Yep, it's good to have so that you can float the generator.