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

[classical]

Thanks. Posts Reply #296 and Reply #289 give some options.
At the end of the day I used


• 2 pieces  PRINTTRAFO (PCB Transformer) Gerth 2X30V 5,6VA #V050561 @ Voelkner.de (cheaper (5,47 EUR each) and I have shipping flat rate) or Conrad 1092928 - 62  ; EAN: 2050001920652
• 1 veroboard Epoxy PCB with dots (off the shelf)
• 1 fuse holder (off the shelf)
• some cables 0.5mm^2 230V approved
• 2 pieces cut of a threaded rod 3mm
• some plastic sleeves or hoses to isolate the rods and to give them a similar diameter as the posts
• some cable ties to fix the PCB to the rods
• some heat shrink
• some hot melt

Doing:

 
• solder the Transformers on the PCB
• remove copper dots around 230V pins to get a clearance of ca. 10mm
• connect primaries in parallel
• connect all 4 secondaries in series and connect to the input of the original swiching power supply
• remove the 2 cables of the on/off switch and short them together
• connect the primaries over the switch and the fuse to the mains inlet jack
• heatshrink the HV cables
• tap M3 threads into the 2 bottom plastic posts
• enlarge the holes of the 2 top plastic posts to 3mm or 3.5mm
• srew the threaded rods into the lower posts
• remove back plane fiddle all togehteher (the rods in the holes of the upper posts) and fix the PCBs to the rods using cable ties
• fix the rods to the PCB a little bit so the the direction of the rod is fixed and will meet the top stands during final assembly
• put backplane back and do final assembly
   

I am not sure if it is really worth while drawing a circuit diagram, up to now I did not.

I use no fan. The hottest temp I measured is between the two transformers and is 48°C.  You can increase the distance between the two transformers to 5 or 10mm to reduce this temp a few degrees.

Edit: Schematics added

[SteveyG]

Edit: Schematics added

That's actually a great solution - much simpler and just as effective! 

[classical]

much simpler

Thank you! Simpler and cheaper - I agree.

and just as effective!

Im am not sure because so I have to rely on the properties of the original switched mode power supply. And frankly spoken, I never measured this quality (e.g. ripple). Your solution delivers a very controlled high quality supply voltage.
Maybe I am spoiled by my daily job. So I could not overcome myself to invest that money and time in a 60EUR device even it is woth while doing. Sorry.
Currently I measure ca. 20V and some µA between GND and protection eart. So it is not as good as your solution but hopefully good enough for my purpose.

[Candid]

Thank you. Very good description and surely enough to "fix" the device with lowest costs and effort. A small inside isolation transformator. I am currently using a "little" bigger 1000VA isolation transformator for this on my bench but this is really not a very mobile solution ;-). So your modification would be my best guess for this device if I ever need to be mobile with the F"l"eeTech ;-). Thanks again!

[classical]

Thank you!
I also have an older chunky 630VA safety isolation transformer under my bench. It is a potted green monster, has an inrush current limiter and a grasp on top for better mobility 
But normally I do not use it for my test gear but more for the DUTs. 
So I was willing to spend 11EUR and some spare time for the FY3224S.

[masterx81]

I've received my fy3224s and after some some search seem that the output amplifier is this one:
http://pdf.datasheetcatalog.com/datasheets/700/322243_DS.pdf
Maybe can find a drop in alternative with more power...

I can confirm low +/-12v output on his psu. With 5.24v on the 5v line there are less than 11v on the +/-12v lines.

[masterx81]

Ordered 2 of this:
http://www.ti.com/lit/ds/symlink/ths3092.pdf

For the output buffer, must provide enough power. The other parameters are similar to the original THS3002 (slew rate only a bit slower).
First of all need to sort the poewr supply...

[electronauta]

Gby:

I have managed record two waveformes in the pos ARB1 and ARB2: 8 & 10 senoidal cicles. At high frecuencys, time and phase jitter appear. Amplitude fall -logical- but it's still useful. In RF RX adjusts, level required is low but precission and stability is critically.

Thanks

[Texacate]

Ordered 2 of this:
http://www.ti.com/lit/ds/symlink/ths3092.pdf

For the output buffer, must provide enough power. The other parameters are similar to the original THS3002 (slew rate only a bit slower).
First of all need to sort the poewr supply...

Interesting idea.  I'm trying to figure out which performance spec (or specs) this would improve on the function generator.  I looked at the data sheets you quoted.  Seems the biggest improvement of the THS3092 over the THS3002 it is ability to source 250ma vs 100ma.   Are you thinking with the THS3092 you could drive larger voltage swings past 6MHz where this generator starts attenuating the amplitude and starts rounding off every waveform shape into a sine wave? 

Mean-Well  PD-2512  or PD-2515 might be good power supplies.    Although they don't do +5v for the digital, they would supply sufficient "analog" drive current, and are reasonably priced.

Please let us know your thoughts.

[bitseeker]

Add a regulator or buck converter to get the +5.

[leonerd]

Add a regulator or buck converter to get the +5.

This is my plan. I've ordered an LM317/LM337 linear regulator board to generate the +12 and -12V lines, and then I was just going to tap from the pre-regulated 16-ishV output from the bridge rec. of that board into a 5V buck regulator of the LM2597 variety. See also

https://www.eevblog.com/forum/testgear/feeltech-fy3200-psu-upgrade/

[pascal_sweden]

Currently there is too much focus on the improvement of the power supply IMO.
When are we going to tackle the signal output instead?

What about the other improvements, like improving the quality of the signal output?
Power circuit as such, is a NICE to have improvement, but further improving the signal output is more MUST have

[pascal_sweden]

Anyone here who played around with the improved 14-bit D/A version?

The FeelTech FY3224S is a bench type signal generator.

FeelTech also produces a portable non-bench type signal generator with a colour screen, the FeelTech FY2300:

http://en.feeltech.net/html/Signal%20generator/show_17.html

I wanted to point out here that FeelTech has recently released an improved version of their portable non-bench type signal generator, the FeelTech FY2300H:

14-bit D/A converter, 250 MSa/s sample rate, 8 Mpoints sample memory, 0.1 deg phase resolution
Improved power supply design to reduce noise on the signal output

http://en.feeltech.net/html/dynamics/show_49.html

Personally I like the FeelTech FY3224S bench type signal generator better, as it better suited for a lab environment.

Maybe they will release a FeelTech FY3224H as well in the future
That's why I wanted to share more info about the non-bench type signal generator in this thread.

[pascal_sweden]

Who is going to add Android support to this signal generator, based on the FT311D or FT312D chip?

I think it should be fairly easy to write an Android application that uses Bluetooth to interface with the FeelTech signal generator.

Another alternative is to use the FT311D or FT312D chip from FTDI that provides a USB host to your Android phone or tablet device, based on the Android Open Accessory mode that is supported since Android version 3.1 and onwards.

As there is enough space available in the FeelTech signal generator, you could even build in the FT311D Evaluation Board, and make the USB port available on the outside of the device.

http://www.digikey.com/catalog/en/partgroup/ft311d-evaluation-board-umft311ev/33342

This would provide Android support to your FeelTech signal generator.
It would highly increase the value of your signal generator!

Here is a nice tutorial from DigiKey about Android Interface Solutions:
http://dkc1.digikey.com/us/en/TOD/FTDI/Android-Interface-Solutions/Android-Interface-Solutions.html

Enjoy!

[leonerd]

Currently there is too much focus on the improvement of the power supply IMO.
When are we going to tackle the signal output instead?

Once I have a power supply I actually feel safe around, that'll be next on my list.

(I'm planning to fully-encapsulate the mains side of the input/transformer circuitry, thus leaving no exposed mains inside the case while it's open)

[masterx81]

i'm waiting the THS3092D for try to improve the signal output....

[JoeO]

Who is going to add Android support to this signal generator, based on the FT311D or FT312D chip?

I think it should be fairly easy to write an Android application that uses Bluetooth to interface with the FeelTech signal generator.

Another alternative is to use the FT311D or FT312D chip from FTDI that provides a USB host to your Android phone or tablet device, based on the Android Open Accessory mode that is supported since Android version 3.1 and onwards.

As there is enough space available in the FeelTech signal generator, you could even build in the FT311D Evaluation Board, and make the USB port available on the outside of the device.

http://www.digikey.com/catalog/en/partgroup/ft311d-evaluation-board-umft311ev/33342

This would provide Android support to your FeelTech signal generator.
It would highly increase the value of your signal generator!

Here is a nice tutorial from DigiKey about Android Interface Solutions:
http://dkc1.digikey.com/us/en/TOD/FTDI/Android-Interface-Solutions/Android-Interface-Solutions.html

Enjoy!

You have a lot of requests of other people.

My question to you is what are YOU doing?

[pascal_sweden]

I am suggesting the new ideas here!

As the original poster of this thread, I do believe that I have the freedom of suggesting new ideas =)

And yes, I want to help out as well, but unfortunately not in a position right now, because of time constraints. Hopefully this will change in the future, so that I can be part of active contributions.

Note that I was in contact with FeelTech about the reported issues in the past, and I actually got hold of them once, where they even confirmed that they will put one of their engineers on it. But afterwards, there was no feedback anymore from them. Really hard to understand, why they don't see any need for further improving their product. If they only would strengthen themselves a bit more, the product could become so much better, without much additional cost. But it seems like Chinese companies simply don't think like this. Obviously every penny counts for them, but I think in this case, it is not related to pennies, but more related to an attitude problem. They just think that whatever they do, is good enough, and that there is no need to strengthen any further. This gives me the impression that they operate under a "we know, but don't care" policy! Unfortunately there are many Chinese companies out there that seem to operate under this policy. With that said, I think that we can forget about any actual support from FeelTech, at least in the near future. We will have to come up with our own solutions to further improve and perfectionize their products!

[Assafl]

Note that I was in contact with FeelTech about the reported issues in the past, and I actually got hold of them once, where they even confirmed that they will put one of their engineers on it. But afterwards, there was no feedback anymore from them. Really hard to understand, why they don't see any need for further improving their product. If they only would strengthen themselves a bit more, the product could become so much better, without much additional cost. But it seems like Chinese companies simply don't think like this. Obviously every penny counts for them, but I think in this case, it is not related to pennies, but more related to an attitude problem. They just think that whatever they do, is good enough, and that there is no need to strengthen any further. This gives me the impression that they operate under a "we know, but don't care" policy! Unfortunately there are many Chinese companies out there that seem to operate under this policy. With that said, I think that we can forget about any actual support from FeelTech, at least in the near future. We will have to come up with our own solutions to further improve and perfectionize their products!

That is a bit unfair - do you know their cost of production? Do you know if they get components at discount, or if they are still in the red to recoup design expenses?

Improving cheap stuff has been around the audio "phool" or "engineering" area for at least 30 years. Nowhere were the compromises made by engineering frowned upon. The senseless compromises (e.g. what is available behind the fish monger in Shenzen) enabled the cheap price and eventual progression towards acceptable products.

Eventually their resources invested in this line will clear the black (hopefully) and at that point resources may be freed up and the design may be improved. 

(on a side note: If you were the owner - wouldn't you wait for the hobbyists to come up with a stable solution - and then just copy it? - seems all the maker stuff like 3d printer controllers evolve like this...)

[GigaJoe]

Engl. soft V2.2 + manual 2016MAY11
http://en.feeltech.net/index.php?case=attachment&act=down&aid=47

[all_repair]

Note that I was in contact with FeelTech about the reported issues in the past, and I actually got hold of them once, where they even confirmed that they will put one of their engineers on it. But afterwards, there was no feedback anymore from them. Really hard to understand, why they don't see any need for further improving their product. If they only would strengthen themselves a bit more, the product could become so much better, without much additional cost. But it seems like Chinese companies simply don't think like this. Obviously every penny counts for them, but I think in this case, it is not related to pennies, but more related to an attitude problem. They just think that whatever they do, is good enough, and that there is no need to strengthen any further. This gives me the impression that they operate under a "we know, but don't care" policy! Unfortunately there are many Chinese companies out there that seem to operate under this policy. With that said, I think that we can forget about any actual support from FeelTech, at least in the near future. We will have to come up with our own solutions to further improve and perfectionize their products!

That is a bit unfair - do you know their cost of production? Do you know if they get components at discount, or if they are still in the red to recoup design expenses?

Improving cheap stuff has been around the audio "phool" or "engineering" area for at least 30 years. Nowhere were the compromises made by engineering frowned upon. The senseless compromises (e.g. what is available behind the fish monger in Shenzen) enabled the cheap price and eventual progression towards acceptable products.

Eventually their resources invested in this line will clear the black (hopefully) and at that point resources may be freed up and the design may be improved. 

(on a side note: If you were the owner - wouldn't you wait for the hobbyists to come up with a stable solution - and then just copy it? - seems all the maker stuff like 3d printer controllers evolve like this...)

The engineers that offered suggestions normally have good intentions and assume that it is very profitable business for the makers in China.  The seller knew about the good intentions but did not explain their difficulties due to many reasons: 1. Do not want to offend people with good intentions  2. Respect  3.  Language barrier  4.  How and what is the point of making outsiders understand their situations and problems.

The biggest problem the shop that came out with a product that start to sell is the invitations to many other copiers that jump on them almost immediately.  The competition is against the copiers that are using savaged parts and hardly making minimum wage while loaded with debts.  The eco system does not allow them to fulfill the request.  They know that.  But the requester probably did not know that most of the answers are just for courtesy and are for PR.  And the requsters definitely do not know that fulfilling the request shall mean a definite death to the seller operation and cash flow.  If they can come out that product, they have some minimum intelligent to know something.  And the ignorant party maybe someone else.

So it amaze to see people here trying to save on an already cheap universal programmer MiniPro . 

[orion411]

......serial port connector J4 lists the pin out on the silk screen as:

   5V
   GND
   TXD
   RXD

Depending on whether you name the signals for the device (DTE) or where they go (DCE) TXD might be the generator output or the input.  So, do you know if pin 3 labeled TXD is driven by the FY3200 generator or if it listens on that pin?

Second question, are the signals on these lines 3.3V level or 5V level?

I hooked up my scope to the USB to Serial card T pin on J2. That wire goes to the front panel J4 pin RXD. Then I change setting in the FY3200S v2.0 software. My scope measured 5volts on that pin. I got nothing on the TXD pin from J4. It looks like TXD driving and RXD is listening at 5 volts.

[JBourke]

I decided to go with the DIY on-board isolation transformer solution as well but in a primary - secondary - secondary - primary configuration. Hardly any voltage drop or heat generation so the surgery seems to be a success. Also added a IEC male socket, fuse holder and a mains earth switch. Switched on I get about ~17V between device ground and mains earth, turned off a couple. Its a basic 2Mhz model and came with firmware revision 3. Apart from SIN, SQR, and TRI it has RAMP, NRAMP, PULSE, 4-PULSE, 8-PULSE, STEP and DC selectable wave forms. There's a FOLLOWING option in MENU which let's you select AMPL, OFFS, DUTY and FREQ for CH1=CH2. Have to work out what this does exactly, the manual is for an older version without this option. A big thanks to Classical for the isolation suggestion.

Cheers,

John

[technogeeky]

I decided to go with the DIY on-board isolation transformer solution as well but in a primary - secondary - secondary - primary configuration. Hardly any voltage drop or heat generation so the surgery seems to be a success. Also added a IEC male socket, fuse holder and a mains earth switch. Switched on I get about ~17V between device ground and mains earth, turned off a couple. Its a basic 2Mhz model and came with firmware revision 3. Apart from SIN, SQR, and TRI it has RAMP, NRAMP, PULSE, 4-PULSE, 8-PULSE, STEP and DC selectable wave forms. There's a FOLLOWING option in MENU which let's you select AMPL, OFFS, DUTY and FREQ for CH1=CH2. Have to work out what this does exactly, the manual is for an older version without this option. A big thanks to Classical for the isolation suggestion.

Cheers,

John

Good looking work.

The FOLLOWING option sounds exactly like the MHS- series (I have the MHS-5200A) TRACE (or TRACK). It basically makes CH1 and CH2 have equal properties (except the waveform itself). In your case, you can make Frequency, Amplitude, Phase, and Offset be equal.

[pascal_sweden]

Has it been confirmed that the 2 MHz version is identical to the 24 MHz version in terms of hardware?

How to perform the software upgrade?

Is there a big difference in price between the entry model and the top of the line model?
Maybe it's not worth the trouble, if the prices are very similar anyhow.