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

[FlyingHacker]

There are different versions with different capabilities, for different prices. Sometimes you can hack the lower spec model into the higher spec, but this is not guaranteed. Unless you are really trying to save a few dollars just get the 24MHz model and be done with it.

[Aruck]

There are different versions with different capabilities, for different prices. Sometimes you can hack the lower spec model into the higher spec, but this is not guaranteed. Unless you are really trying to save a few dollars just get the 24MHz model and be done with it.

I agree with this, but I'd like to add that the 32xx series has arbitrary wave form generation while the 22xx series does not.

[nctnico]

I wonder if it would trip a GFCI protected circuit?

It has never tripped my GFCI, which it is plugged into.

I think for safety's sake I'm going to install a polarized plug to insure that cap is always on the neutral side.

First of all those capacitors are designed to be safe so no problems there. Secondly depending on a polarised plug to have the neutral where you expect it for safety is a bad idea. There are soooo many things which can go wrong with wiring! For example: what happens if the neutral gets interrupted somewhere? There is only one way out: ground the power supply!

[Aruck]

I wonder if it would trip a GFCI protected circuit?

It has never tripped my GFCI, which it is plugged into.

I think for safety's sake I'm going to install a polarized plug to insure that cap is always on the neutral side.

First of all those capacitors are designed to be safe so no problems there. Secondly depending on a polarised plug to have the neutral where you expect it for safety is a bad idea. There are soooo many things which can go wrong with wiring! For example: what happens if the neutral gets interrupted somewhere? There is only one way out: ground the power supply!

Yes the Y-Rated cap would be safe for connecting line to ground. The problem is that it isn't connected to ground, and therefor not currently safe (as seen by 115VRms on the floating ground).

The other issue I have is that if the non-polarized plug is flipped the wrong way the fuse is on neutral. This is not acceptable to me, if the fuse blows the power supply needs to be disconnected from line and not from neutral.

I agree with you regarding the grounding, I'd like to get this sucker on a proper 3 prong plug and grounded.

[nctnico]

There is not enough current flowing through the capacitors to cause harm to your body.

[med6753]

First of all those capacitors are designed to be safe so no problems there.

You really want to put your safety to a Chinese capacitor? Chances are nothing will go wrong but Mr Murphy is always lurking around the corner.

Secondly depending on a polarised plug to have the neutral where you expect it for safety is a bad idea. There are soooo many things which can go wrong with wiring! For example: what happens if the neutral gets interrupted somewhere? There is only one way out: ground the power supply!

If the neutral opens up and nothing is connected to the BNC's the unit won't power up. And I'm not even sure it will power up even if it's connected to a grounded device. Granted, a 3 wire grounded plug is the BEST solution. Next best is a polarized 2 prong plug as long as the fuse is on the hot side. That needs to be checked out.   

[nctnico]

The neutral doesn't open in the device ofcourse  Imagine this happening in your breaker box! In that case neutral will become live and this isn't a rare event!

[med6753]

The neutral doesn't open in the device ofcourse  Imagine this happening in your breaker box! In that case neutral will become live and this isn't a rare event!

I wasn't inferring that the neutral opens in the device. There are all kinds of scenarios that could take place with mains wiring and we could debate that ad nauseam. In 2 cases that I am personally familiar with a neutral open resulted in dead outlets. I had another case where I discovered by chance that the outlets were wired backwards and had been for years and nobody knew it. Ain't THAT a scary thought!     

[nctnico]

Dead sockets yes but harmless: definitely not!

[Aruck]

I pulled the 3 prong plug out of an old Dell ATX power supply, and it's perfect. It happens to have the same exact white connector on it, even in the right polarity, with a separate ground wire.

Time to run to harbor freight to pick up a dremel tool

[med6753]

Took some measurements today and got interesting results. See photos. Initially no matter which way I plugged the line cord I got 1.2 mv AC. Then I realized that the USB cable was plugged into my hub. Unplugged it and BINGO! 

Fluke 87 connected between the outside barrel of the BNC and electrical (earth) ground.

Power on, 56.9 VAC


Power off, 1.357 VAC


Then I reversed the line cord. Power off, 107.1 VAC


Power on, 59.7 VAC


I tried taking some current measurements and got nothing because the leakage voltage dropped to zero.

Next step is to open this bad boy up and see if there's a safe way to insure it's grounded. It does not appear that a 2 prong polarized plug is a complete solution.
 

[MikeLogix]

Greetings all, I decided to pull the heatsink from the output amps on the FY3200S front panel board. The parts appear to be TI parts, but I have not been able to find the datasheet anywhere on Google. I also tried the TI database as well. I can't figure this out, I was hoping someone might be able to come up with a part number/datasheet so a good schematic of the signal chain can be made. Of course I could guess and write down what I think is there, but I would rather be certain. See photo of the output stage amp below.

Almost forgot, here is a link for the TI part number lookup database that I tried.
http://www.ti.com/packaging/docs/partlookup.tsp#divline

[erpalma]

It's a THS3002.

http://pdf.datasheetcatalog.com/datasheets/700/322243_DS.pdf

[med6753]

I opened it up today.....

3 wire cord installed and grounding verified.....

[leonerd]

It's a THS3002.

http://pdf.datasheetcatalog.com/datasheets/700/322243_DS.pdf

Excellent work.

I see its headline stats are:
  420MHz GBW product
  6500V/µs slew rate

At a GBWP of 420MHz, I'd expect to see no more than 17.5 times gain around it if it's going to achieve 24MHz output.

A slew rate of 6500V/µs would allow a worst-case rise time of -10V to 10V in 3ns leading to a near-straightline triangle wave output at any period faster than 6ns, equating to 166MHz. We should therefore expect this amp to handle a nice-looking square wave at anywhere below about a tenth of that; 16MHz. Sounds to me, at least on these paper specs alone, that this chip is more than adequate to handle the quoted specs of the generator; at 24MHz sine and 6MHz square/triangle. Yet it's observed in practice that this machine just can't cope with this limit.

Do we have an idea of the surrounding circuit connectivity around it? How easily can it be traced without taking the heatsink off? I'm loathe to do that to mind just yet but if someone else has already done it could they trace out the periphery of this chip? I'd be interested to see what the gain is set to on this amp.

[leonerd]

I see its headline stats are:
  420MHz GBW product
  6500V/µs slew rate

A slew rate of 6500V/µs would allow a worst-case rise time of -10V to 10V in 3ns leading to a near-straightline triangle wave output at any period faster than 6ns, equating to 166MHz.

Oh except literally as a post that I see the next headline stat of
  40ns settle time to 0.1%

So even if it can slew -10 to 10V in a supposed 3ns, it'll be 40ns before it settles. A maximum square period of 80ns, coming out at 12.5MHz at absolute best. I think maybe it's that settle time that causes it not to cope so well at sharp edges.

[leonerd]

I notice that the manual is lacking pictures of any of the waveforms. So I captured them.

Here's some screenshots from my Rigol DS1054Z watching the output of the main channel (CH1) when set to 10kHz 1V pk.

SINE = Sine; SQUR = Square; TRNG = Triangle.
PRE1 = Lorentz; PRE2 = Multitone; PRE3 = Random noise
PRE4 = E.C.G.; PRE5 = Trapezoidal; PRE6 = Sinc
PRE7 = Narrow pulses; PRE8 = White Gaussian noise
PRE9 = "amplitude modulated"; PRE10 = "frequency modulated".

No point capturing the ARBs because those are user-uploaded anyway (and also they're all blanked to zeroes in my unit anyway )

[1N4148]

I doubt that it's a THS3002. Even if there is a datasheet, where the THS3002 is labeled "PRODUCT PREVIEW", I cannot find it at ti.com

[Texacate]

I just modded my FY3224S with a proper 3-wire grounded power cord... and managed to blow up the power supply board, (in spectacular fashion I might add), in the process of making this thing "safer".    I won't go too deep into my faux-pas, but suffice it to say that you are supposed to connect the SECONDARY-side-gound (low voltage DC ground) to mains-earth-ground.  Do NOT, under any circumstances, connect the PRIMARY-side "ground" (from the bridge rectifier) to mains-earth-ground.   That isn't a meant to be earth-ground.  It's the ground "reference" of the primary.

I've noticed that FeelTech is using whatever power supply they can get their hands on, as mine looked nothing like the ones posted by med6753 & tombddiver.  My unit came with a very nice SICO-650 power supply.  (Well, it WAS nice 'till I fried it).  I put in a "Universal DVD Replacement Power Supply" I managed to find on Amazon. It took some searching to find one with the required +5v & +/-12v  outputs.

My mod included a toggle switch in the grounding path, in case I ever have the need to "float" the BNC outputs, the way the unit is shipped from FeelTech.  Otherwise that switch stays in the safe/gounded position.

One more note;  The sticker on the bottom of the unit is printed on an ink-jet printer.  If it gets wet, it will bleed and smear.

[ledtester]

Speaking for alternate power supplies...

Does anyone have an idea of what the power requirements are for this unit?

I have this ready-made transformer board that produces (unregulated) +/15 and +5 - at least that's what the silk screening says.

The specs on the transformers read:

- 2 x 10.0 V, 2 x 5.0 VA
- 2 x 8.0 V, 2 x 1.25 VA

Do you think it will work if I just use some 78xx and 79xx regulators?

[med6753]

Speaking for alternate power supplies...

Does anyone have an idea of what the power requirements are for this unit?

I have this ready-made transformer board that produces (unregulated) +/15 and +5 - at least that's what the silk screening says.

The specs on the transformers read:

- 2 x 10.0 V, 2 x 5.0 VA
- 2 x 8.0 V, 2 x 1.25 VA

Do you think it will work if I just use some 78xx and 79xx regulators?

I doubt it's more than a few hundred ma and 78xx and 79xx regulators would work no sweat. There's plenty of room in the case to install a linear supply. I had initially considered doing the same thing if grounding the DC side didn't work out.

[pascal_sweden]

Has anybody in this thread contacted FeelTech about this voltage issue?

Maybe the OP that reported this voltage issue can take the initiative, as he has all the information and measurements at hand.

http://www.made-in-china.com/showroom/hnregong/product-detailFvAJyhUdhnYp/China-Fy2200s-Dual-Channel-Dds-Function-Signal-Generator.html

What about the software port of the MHS-5200 to the FeelTech FY3224S?
http://hackaday.com/2015/07/01/hacking-a-100-signal-generator/

It would be really nice to have this software available for the FY3224S as it is apparently compatible with the arbitrary waveform format of the easy to use Waveform Manager Plus software!
http://tti1.co.uk/downloads/waveman-plus.htm

Conclusion: The moment FeelTech has corrected this potential safety issue in the power supply,
and the moment the software port is completed, you can get a very cheap signal generator that can do an impressive job, including an easy way to generate your own waveform formats. This is rock&roll!

[Texacate]

In the process of replacing the power supply that I fried,  i connected the front panel electronics to a bench power power supply in order to verify I didn't ALSO fry the signal generator section.   I verified that it does not draw much from the +12v rails.  I didn't note the actual value but as med6753 has noted, it wasn't more than a few hundred milliamps.    This intuitively makes sense.  10v DC into 50 ohm load would be 200mA.    I might take mine apart again, and carefully measure the current draw on each rail.   But it's not much.  The replacement power supply was rated at 400ma on the +/- 12v rails, and 800mA on the 5v rail.  That's ~13W total output power.

http://www.amazon.com/Universal-Power-Supply-Circuit-Player/dp/B00NQ2OLB8

I did measure the AC power draw with a cheap AC power meter.  It draws ~2W with both channels on, but no load.  It went up to ~3.1W driving one channel into 50 ohm load with a 10v sine wave. Guessing it would draw ~4W if both channels were loaded at 50 ohms.  My meter isn't super accurate or anything, but it let me know that my 13W power supply isn't being brought to its knees by the function generator.

[Texacate]

BTW, studying the pictures of the function generator throughout this thread, Ive noticed two different function generator main-boards with respect to power (and thus two different types of power supplies). If you look at the photos posted by med6753, his power supply board appearently sends a single DC voltage to the main board via a single black & red wire pair.  But on my FG main-board (which looks like the one posted by tombdiver)  the power supply generates +5 and +/-12v, and sends it to the FG main board via a 5 wire pin header.  I'm assuming on med6753's style of FG, the analog and digital power is generated/separated on the main-board.  Mine clearly does that on the power supply board.

So beware.  They are not all the same.  There are at least two different variants.

[pascal_sweden]

Who is going to contact FeelTech about the power issue?
See in the thread above for contact details of the company FeelTech.

I don't have all the details of the power issue at hand, and believe it would be better if the OP of the power issue takes the initiative.

Who is going to port the software program of the MHS-5200 to the FeelTech FY3224S?
Maybe we can collect some money here all together, and buy a FeelTech unit for the software author of the software program. This way he can do it