SG1005 0.01- 5MHz DDS Function Signal Generator who got one ? :-)

[voidptr]

I just received my new  SG1005   0.01- 5MHz DDS Function Signal Generator, I bought it on ebay.

http://www.ebay.ca/itm/0-01-5MHz-DDS-Function-Signal-Generator-Module-Wave-New-/270864768941?pt=LH_DefaultDomain_0&hash=item3f10cc5fad

Some people on an old threat seams to be satified with thier so ...

I tried it right away but darn is this just crap or I got a busted one ?

Here a vid I will send to the seller ...
http://s1114.photobucket.com/albums/k539/n45w73/dev%20stuff/?action=view&current=fctgen.mp4

showing first my SG1005 distorded 2MHz square wave, and comparing it with my simple square wave board i had...
I then reset it and show distortion on the sinewave ...

So here my question, is those SG1005 just junk like that ?
 what kind of wave form distortion do you have on yours ?

 
 

[typeglob]

Could you enable infinite persistence (page 45-46 of the Owon manual) and post an imagine of that after running it for at least a few seconds? Please zoom out a bit so we can at least see 2-3 whole cycles of that square wave.

I'm curious how much jitter that thing has, in addition to the terrible waveform it generates. Btw. it might have a TTL output, which should be a bit better.

I did some quick googling, but wasn't able to find if this thing uses a micro controller to generate the signal (which it probably does). Not too long ago I did some testing with a ATmega2560 generating a sine wave via DDS. While the shape of the sine looked ok, the jitter was pretty horrible (probably due to interrupt latency).

Anyway, the boards using the AD9850/AD9851 chips (which are meant for this purpose) seem pretty okay.

[voidptr]

I took time to take a deeper look at the square waveform ...

top part of the pic, SW at 4MHz infinite exposition ....   it seems really like a design flaw  some kind of constant jitter appear time to time...

bottom part, SW at 2MHz I took some measurement, 1 cycle of wave take 490 nS (2.041MHz) or 521nS (1.919MHz)  so jitter of 30nS
so an error of 30ns/500ns -    6% this kind of bad ! 

jitter is present on the ramp wave form also ( made from the triangle wave form menu)

the manual say i suppose to have a frequency sweep, but i do not have that ...

it really seem to be a bad product at many point of view !!!!
don't buy that !

[voidptr]

so anyone has good suggestions for a cheap, but working function generator ?   

[T4P]

so anyone has good suggestions for a cheap, but working function generator ?   

Hantek DDS3x25 ? 

[voidptr]

Hantek DDS3x25 ? 

i gonna read about that guy,  more expensive,  but i like thing that work, and the io pattern generator can be pretty handy :-)


my bad gen looks really like this one but without the plastic case ...
it also got the same kind of jitter !

https://www.eevblog.com/forum/product-reviews-photos-and-discussion/review-of-cheap-dds-function-generator-udb-series-%28udb1102-udb1103-udb1105%29/

[voidptr]

back on my silly gen for few minutes...

the cpu  is a STM8S103K3  (16 MHz)
and it used a CPLD  Altera Max II for driving 8 bits waveform and i guest all other stuff it can do...
there is a clock, seems an 34.546  MHz,  might by the clock for the CPLD generator... maybe...

so if  i suppose it is true, 34.546MHz is near 29 nSec, 
it could explain the jitter in a way to resync the waveform by adding or removing one sample
in a way to have a near correct global frequency waveform.

so it might not be that bad, but i will really prefer a more greatter output frequency,  so the
adjusting jitter is less present ...

i might have been interested to keep it has an hackable toy but i don't know if STM cpu  and CPLD Altera can
be locked.   if it is the case there no fun there ...

anyone had hack their SG1005 ? 

[ltaber]

I have a USB version, the USB transmit and receive are connected to the STM8S103K3.  The linux lsusb dos not show it as a USB device.  The parts in it are quite cheap in quantity.  Digi-key lists the processor at about 80 cents.  The Altera Max II epm240 is about $6.  It could be an interesting device to hack if it is not locked down.   The D/A converter is a discreet  R-2R ladder driven directly from the CPLD. 

The jitter is most obvious in the square wave.  There are only finite transition times from the main clock, apparently no PLL to up the frequency.         

[nukie]

I have one and it's crap. Unfortunately it's more like a toy... waste of $$$ 

http://iteadstudio.com/evaluation/hack-the-sg100x-dds-signal-generator/

[Dread]

Scrap all of what I said below, I did not watch the video, I assumed you meant a Siglent SDG1005.
What you have is a really low end ebay generator and they are crap.  I am not even sure what they are intended for as you can't trust anything that they output.

Try a Siglent Generator, the are fairly low in price and work very well.

Sorry about the Below post.
__________________________________________

I have been really happy with my 1010.
I have not encountered the problem described by the OP but I will do some further testing tommorow.  So far it has been very solid and the built in freq counter is useful.

At first I had some issues like a missing dsb Output option and a unusual high drop in output as I ramped up the freq but a firmware update from the company fixed the issues.

I will give an update as I do more tests

[eevfun]

Hi all,

I have one of those SG1005/UBD1005 (not the S) and noticed that one needs to be careful how you connect the scope.

Ground on the LEFT and Signal on the RIGHT and the Traces look pretty good for both,
the TTL and Analogue Outputs.

(otherwise my Traces do look just like itead's and yours ... )

There is some Jitter at higher Frequencies, but heck for the roughly 30 bucks it sure is OK for many uses.