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

[SteveyG]

How is the heat build up inside the case? I didn't see much in the way of ventilation. Are the slots on the side open? They looked decorative, but could be cut open.

I really need a FLIR One, but the calibrated finger says the heatsink was between 40 and 50°C at full load. The 15V transformer was probably the hottest part, after 30 mins at full output into 50R that transformer was a bit too hot to touch. I will make some measurements but I think a 40mm fan may be needed, however it definitely doesn't need to run very fast.

The vents are open, on the bottom half only, so kinda pointless on their own. You could probably drill a matrix of holes on the bottom and top which would be sufficient for convection to draw outside air in.

Discount code valid for May 2016 @banggood.com is here: https://sdgelectronics.co.uk/uncategorized/fy3224/

[FlyingHacker]

Back in the 90s I had a USR 33.6k modem (state of the art at that time) which would overheat and reduce its speed. I drilled a bunch of holes in the top and bottom and it worked a treat. It did look a little stupid 

Seriously, I was thinking the same thing, as I finalize my linear supply design. Would be nice to be able to get the regulators on the side of the case with a heat sink mounted on the back or side. I would really rather avoid another fan...

[nctnico]

Those small mains transformers are very inefficient to begin with so together with the linear regulators you'll need more airflow. Still you might get away with drilling some extra holes at the top in the rear cover plate so air can flow in from the bottom and flow out at the top of the instrument.

I'll just ground my FY2306 instead of replacing the PSU and be done with it. After all the USB interface isn't isolated either. While at it I'll also remove the tilt stand because it makes the box uneven.

[bitseeker]

The vents are open, on the bottom half only, so kinda pointless on their own. You could probably drill a matrix of holes on the bottom and top which would be sufficient for convection to draw outside air in.

At least it has them on the bottom. That's a good start.

Still you might get away with drilling some extra holes at the top in the rear cover plate so air can flow in from the bottom and flow out at the top of the instrument.

Venting the top of the rear plate was exactly what I was thinking. Functional without getting ugly. Not that it's a beautiful case, mind you, but drilling the top cover would be less aesthetically pleasing to my eye.

[nctnico]

Venting the top of the rear plate was exactly what I was thinking. Functional without getting ugly. Not that it's a beautiful case, mind you, but drilling the top cover would be less aesthetically pleasing to my eye.

Not just that but there is also the risk of stuff falling into the generator. I always try to prevent having holes in a top cover.

[SteveyG]

Those small mains transformers are very inefficient to begin with so together with the linear regulators you'll need more airflow. Still you might get away with drilling some extra holes at the top in the rear cover plate so air can flow in from the bottom and flow out at the top of the instrument.

I'll just ground my FY2306 instead of replacing the PSU and be done with it. After all the USB interface isn't isolated either. While at it I'll also remove the tilt stand because it makes the box uneven.

Earthing the PSU is definitely a good (cheap) option, and yes, the smaller transformers are not the best for efficiency.

My expensive TTi sig gen is obviously earthed as are all the BNCs. The USB interface will be isolated in the next video though - I think having this one able to float might be handy.

Venting the top of the rear plate was exactly what I was thinking. Functional without getting ugly. Not that it's a beautiful case, mind you, but drilling the top cover would be less aesthetically pleasing to my eye.

Not just that but there is also the risk of stuff falling into the generator. I always try to prevent having holes in a top cover.

I think I'll go for the noisy option of a fan on the rear panel. It doesn't seem like it'll need much airflow and half of the Agilent gear on the bench has whiny fans anyway 

Holes in the top cover could be an option if you put some speaker grille fabric on the inside for example, but that obviously restricts airflow.

[classical]

Reading about these heat issues I most probably will opt for an isolation transformer. I can start with my 60VA. But  a 26VA isolation transformer is <14 EUR new. And 26VA should be sufficient for a 5 to 6VA instrument. Insert it in an (old) housing, add a fuse holder and a power cord and that's it.

[nctnico]

How about putting a small isolation transformer inside the case? Or 2 low voltage transformers back-to-back.

[SteveyG]

Reading about these heat issues I most probably will opt for an isolation transformer. I can start with my 60VA. But  a 26VA isolation transformer is <14 EUR new. And 26VA should be sufficient for a 5 to 6VA instrument. Insert it in an (old) housing, add a fuse holder and a power cord and that's it.

To be honest, an isolation transformer is not required unless you specifically want the AWG floating. If you connect the 0V on the PCB to mains earth you're fine from a current leakage point of view.

2x13.5V transformers may be available too which would cut down some of the dissipation.


Lots of options :-)

[classical]

How about putting a small isolation transformer inside the case?

Yes, e.g. ebay 181990291398 or similar. But I do not know if  88 x 53 x 48mm will fit into the case.

[classical]

To be honest, an isolation transformer is not required unless you specifically want the AWG floating.

Yes, this can be helpful in avoiding ground loops, e.g. in audio applications.

[bitseeker]

I think I'll go for the noisy option of a fan on the rear panel. It doesn't seem like it'll need much airflow and half of the Agilent gear on the bench has whiny fans anyway 

Yeah, if you already have fans going in a bunch of instruments, adding a little one to the mix won't be noticeable.

Looking forward to the next vid, Steve.

[nctnico]

An alternative could be to use a transformer with an 85V (or more) secundary winding (or 2 lower voltages in series). My generator says it will run from 85VAC or more.

[bitseeker]

Ah, I see. Just put a 120->85+ or 240->85+ transformer before the existing supply. Yeah, it doesn't have to be 1:1.

[classical]

The closest (unexpensive but new and available in Germany) which I could find is
- 2 of Gerth  EAN 2050001920652;  38 mm x 32 mm x 45 mm; 5.6VA; 2x30V (=120V); @5,50 EUR each = 11 EUR  ; Safety-Transformer,  IEC 61558-1  and 2-6; protection index IP 00;  prepared for class I / II equipment; vacuum potted; max. admissible ambient temp. 70°C; 2 chamber safety-bobbin; test voltage 4.000 V ac; non short-circuit-proof; approval mark: VDE, ENEC 10
Looks good if it fits in the case.
otherwise the smaller "brother" (3 to 4 mm smaller in each dimension)
- 2 of Gerth EAN 2050001920355 ; 34 mm x 29 mm x    41 mm;  3.6VA (could be at the edge) @4,40 EUR each =8,80 EUR
Maybe you find something more appropriate or an option with a better availability in your country.

[calin]

Darn !!! ... I cold not stop ... just ordered one. Thanks for the discount code Steve !


It will probably make a nice addition to my old Tek FG504

[bitwelder]

If one wants the transformer isolation mod be as small/light as possible, how much can you optimize the transformers, so that it still provides the current required by the fun.gen. and its output voltages allow to scale down the heatsink as well?

[bitseeker]

@SteveyG

Regarding your request for questions in the latest vid:

Have you had a look at the generated waveform during transition to a different frequency? Does it still glitch with the latest firmware the way it did last year as reported in post #45? Hopefully, that was something that could be rectified via firmware update.

[SteveyG]

How about putting a small isolation transformer inside the case?

Yes, e.g. ebay 181990291398 or similar. But I do not know if  88 x 53 x 48mm will fit into the case.

Inside, you have about 63mm height. Based on that, with the power supply that was originally fitted inside the case to my signal generator, there is enough room. I see some people have slightly larger PSUs, but it should still fit.

[SteveyG]

@SteveyG

Regarding your request for questions in the latest vid:

Have you had a look at the generated waveform during transition to a different frequency? Does it still glitch with the latest firmware the way it did last year as reported in post #45? Hopefully, that was something that could be rectified via firmware update.

Unfortunately, yes it does. It is however fixable by FW - there is no hardware limitation that I can conceive of that would cause this.

[bitseeker]

Thanks for the confirmation, Steve. At least the transitions are smooth when it sweeps frequencies.

[classical]

How about putting a small isolation transformer inside the case?

Yes, e.g. ebay 181990291398 or similar. But I do not know if  88 x 53 x 48mm will fit into the case.

Inside, you have about 63mm height. Based on that, with the power supply that was originally fitted inside the case to my signal generator, there is enough room.

Thanks a lot for this info. So, the ebay 181990291398,  2 of Gerth  EAN 2050001920652 or an elma TT EAN 4016138043341; 30VA; 15 EUR http://www.produktinfo.conrad.com/datenblaetter/500000-524999/514810-da-01-en-Trenntrafo_30VA.pdf could  just fit. So, we have some feasable and reasonable options.
I will order a signal generator and we will see how it looks like and then I will decide for the transformer.

[classical]

Regarding the USB isolation: Did anybody try an USB-Isolator based on ADuM4160 like ebay #141998972441?

[classical]

My unit arrived today. The Y-capacitor is approximately 1 nF and the leakage current between GND and PE is about 0.2mA. Power consumption at full load is approximately 8VA. So the 2 x 5,6VA transformers should be sufficient.

[SteveyG]

Regarding the USB isolation: Did anybody try an USB-Isolator based on ADuM4160 like ebay #141998972441?

Probably no need to isolate the USB side of the UART-USB converter. I've been unwell last week so didn't get chance to do the next video on USB isolation - hopefully I'll get something out this week.

Steve