Recomendation wanted for Function generator

[Electronic]

Bought the well documented RIGOL DS1052E a while ago. Planning to do the hacks after warrenty expires.

The thing is that I need a function generator to go with it.
Budget is max $200. Including shipping to sweden.
I also need about 20Vpp to simulate rpm-signals in 12V-systems (13,6-14.4V)

Any ideas are welcome!

Edit:
37 views and no one has any suggestions?

[Richard W.]

Hello,

it's not possible to recommend a model when we don't know what do you want to do with the FG.

Do you want to do audio stuff, digital stuff or even both?
Which frequency range?
Do you want to remote control the FG?
Which stability of frequency and amplitude do you need?
Do you want a new one or a used unit?

Please consider: FGs are expensive. Even a used one in good condition costs more than the Rigol scope.
I'm very contented with my HP 8111A Pulse/Function Generator.
But second hand prices for these >20 year old units are usually between 400 and 1000€.


Richard

[saturation]

For general purpose versatility check the popular thread on the Hantek 3x25 USB based FB.

[alm]

Agreed with Richard that more information would be helpful. Don't agree that used FGs tend co cost more than a Rigol scope, this depends on your spec. Without even trying I see a bunch of Wavetek function generators on ebay for around $100, you can function gens for less if you put some effort in it, depending on your location. GW-Instek SFG-1003 is an example of a 3MHz DDS generator for about $150 new, which will do 20Vp-p in open circuit. High bandwidth, arbitrary waveforms and other features can drive up the price.

Saturation: How close does the Hantek 3x25 get to the 20V p-p requirement?

[Electronic]

Here is some more input variables:
At the moment I don't see any use above 2-3MHz. My development is at the moment based on Arduino Uno platform. And it will do until I can afford a total upgrade.
Foremost it will be functions related to enginecontrol. Max 8000rpm

3x25 does look intresting until you see the output. 3,5 Vpp is way to low.

The ones I found and wondering about is:

• Two types of barebone on eBay, really cheap ones
• VC2002
• SFG-1003
• UNI-T 9000 series (UFG9005x)

Uni-T seams like a good buy but I can't find any reseller other than one person on EBay who only got one C-model.

[Mechatrommer]

3x25 does look intresting until you see the output. 3,5 Vpp is way to low.

no worries. you can buy my 3x25 "add-on" for say 20 bucks? excluding shipping. is that ok with you? to get the required 20Vpp. 3x25 = 3Vmax, here 10x25 (3x25 add-on) is 10Vmax, ie 20Vpp but if you only need lower freq, i think 3x25 is too much for you, might not be suitable... too much freq (and too much noise and jitter ... by audio and T&M standard).
ps: i can give for 10 bucks if i do mass production. but since i only do small/personal batch, 20 is minimum i believe but if you're really up to the EE business, you might not need to buy from me after all

[Electronic]

3x25 does look intresting until you see the output. 3,5 Vpp is way to low.

no worries. you can buy my 3x25 "add-on" for say 20 bucks? excluding shipping. is that ok with you? to get the required 20Vpp. 3x25 = 3Vmax, here 10x25 (3x25 add-on) is 10Vmax, ie 20Vpp but if you only need lower freq, i think 3x25 is too much for you, might not be suitable... too much freq (and too much noise and jitter ... by audio and T&M standard).
ps: i can give for 10 bucks if i do mass production. but since i only do small/personal batch, 20 is minimum i believe but if you're really up to the EE business, you might not need to buy from me after all

Dude, you really need some soldering lessons. First tip, less tin and more heat. Yes, i'm up to the ee business. Just a thesis from a Bachelor of Science degree containing about 260 credits. Just that I don't have the time or energy at the moment to build a FG or add-ons. My main concern is to get at job in my field or find the perfect thesis.

About the soldering. The reason that I pointed it out is so you can learn and get better at it. When dealing with high freq. the soldering is main source of distrostion and faults. Have you opened a electronics device produced at a factory? Cellphones are good exampels of how it should look like. With at steady hand, thin tin and the right heat and tip its not that hard to do it by hand.
Think nice and clean. And good luck!

Looking forward to see pictures of version 2. 

[Electronic]

this one looks cute. but you have to pay double for that. i never have experience with it, just having fun ebaying for you

http://www.ebay.com.my/itm/ATTEN-ATF20B-FUNCTION-GENERATOR-20MHZ-/300563644155?pt=BI_Signal_Sources&hash=item45fafd3efb

Looks nice but a little bit too much in price. Bedtime here in Sweden now.
I will think about it. Maybe I could cut some expensives and wait a while until my budget is better.

[alm]

Do you need the extra bandwidth and features? There's always something better, but you could also spend the extra money on something else. Waiting also means you won't be able to use it for some time, which is what it's all about.

Try to find a review or user experiences if you're considering the Atten, Atten is not a company who's products I'd buy sight unseen. I believe there are some negative experiences on this forum about Atten products with usability issues.

[olsenn]

I've had numerous problems with my Rigol DG2041A arb. generator, but none the less I would have to say the new Rigol DG4062 looks to be a fantastic device for the money.

Check it out: http://www.rigolna.com/products/waveform-generators/dg4000/

[kaz911]

I've had numerous problems with my Rigol DG2041A arb. generator, but none the less I would have to say the new Rigol DG4062 looks to be a fantastic device for the money.

Check it out: http://www.rigolna.com/products/waveform-generators/dg4000/

I do not think that is available for sale yet :-) Maybe scareware to keep competition on their toes.

[alm]

the SFG-1003 got 16Vpp only

Do you have a reference for the 16Vpp figure? The datasheet claims 10Vpp in 50 ohm, and 50 ohm output impedance. I'm wondering which of the two specs is wrong.

[Mechatrommer]

duh must be late night mixed up. post corrected. but i keep wondering, iirc saturation own the SFG-1003 FG but keep recommending the USB based 3x25 AWG in many threads where FG question ever popped up. so maybe can ask him for hands on experience with it.

[saturation]

The Instek SFG1003 meets most of those requirements.  At ~$150 and the company has a better reputation.  I own both it and the Hantek 3x25.

To get any voltage higher that the base output of the Hantek, see mechatrommer's mod for a front end op amp; it also adds a layer of protection against blowing the Hantek's onboard op amp.  He has done great work in improving the device's software but its USB based but you can pull 100 MHz out of it for a $130-170 price.

Here is some more input variables:
At the moment I don't see any use above 2-3MHz. My development is at the moment based on Arduino Uno platform. And it will do until I can afford a total upgrade.
Foremost it will be functions related to enginecontrol. Max 8000rpm

3x25 does look intresting until you see the output. 3,5 Vpp is way to low.

The ones I found and wondering about is:

• Two types of barebone on eBay, really cheap ones
• VC2002
• SFG-1003
• UNI-T 9000 series (UFG9005x)

Uni-T seams like a good buy but I can't find any reseller other than one person on EBay who only got one C-model.

[saturation]

A unique question; the Hantek seems to be dwindling in number and its thread discusses its pluses and minuses well.  If a reader doesn't get one before its gone it could be very hard to get a hold off.  I've don't recall seeing one sold on eBay as second hand.  So either everyone finds it useful who owns one or its being trashed  

From my limited tests with the Rigols s/n limits using FFT, the sine wave harmonics are on par with the 1003 to 3 MHz, at least.  There are other reviews on the Hantek thread of its output characteristics.

The 1003 is better in that it has all hardware controls  and native higher output voltage + DC offset [alm's Vout estimate is right but not sure of the exact value] and published specs are from a reliable maker whereas the Hantek improved specs are user based.

duh must be late night mixed up. post corrected. but i keep wondering, iirc saturation own the SFG-1003 FG but keep recommending the USB based 3x25 AWG in many threads where FG question ever popped up. so maybe can ask him for hands on experience with it.

I think soldering lessons are a lot easier to learn than hacking the software; if the prototype didn't reveal issues as an working amp suboptimal solder joints most often just reduce long term reliability and shock/vibration resilience.  At the very least, you've pointed the critical fixes for it, and let the reader add their skills to improve it further, just as in the Hantek thread there are folks who have done further performance tests on the software mods to confirm its improved capabilities.

The 3x25 is a popular thread based on page views:

https://www.eevblog.com/forum/index.php?action=stats

...see mechatrommer's mod for a front end op amp

before anything... you humiliate me even further cheers mate.

3x25 does look intresting until you see the output. 3,5 Vpp is way to low.

no worries.

Dude, you really need some soldering lessons...

[Mechatrommer]

wow very detailed stat there, i wish i could see Top 10 Lurkers around here. btw, i sense some resistance that go against USB based "anything", so i will hold my breath on this, but no worries mate, i will be on your back

[Electronic]

wow very detailed stat there, i wish i could see Top 10 Lurkers around here. btw, i sense some resistance that go against USB based "anything", so i will hold my breath on this, but no worries mate, i will be on your back

I like the idea of USB-connectivity. But I also like to read the frequency directly on the generator. Thats why I having a hard time choosing.
And it looks nice when you stack the instruments on the shelf in the lab.

A lot of credits for the developement of the 3x25 to you! 

[Mechatrommer]

I like the idea of USB-connectivity. But I also like to read the frequency directly on the generator.

i was about to give Hantek name a big thumbs up for somehow "Open Source"ing their 3x25 HW through SDK at least so we can develop our own program, for me the hardware is good, no other HW that can do the same, price the same, as you already know/searched, the same priced "dedicated" bench type only can go up to 2-3MHz, 3x25 is hackable to 100MHz, and low Vpp is not a problem even for people like me.

and i repeat... i was about to give Hantek name a big thumbs up, until i'm searching for LA, i expected their LA5034 (same case as 3x25) can do the same (superb HW imo, open SDK), ekk, wrong! so the morale? dont trust the Brand name esp from China. we rely on reviews that people have made in the net, if you cannot find the review, then you are the one that "responsible" for it, esp if the device is broken, so other people wont do the same mistake and make the crook rich.

[saturation]

Its been in the top 10 threads for ~  1 year.  Good work mecha! While I prefer non-USB instruments for ease of use, if a USB instrument is better spec wise its a worthy trade off.  I think using virtual instruments will be less of a bother once we get them working on platforms like Android or iPhone/iPad because those platforms boot faster and touch screens are faster than working with only a mouse.

wow very detailed stat there, i wish i could see Top 10 Lurkers around here. btw, i sense some resistance that go against USB based "anything", so i will hold my breath on this, but no worries mate, i will be on your back

Its a good idea to test every instrument individually for merits and flaws, even the better names have bugs and annoyances, but often better names [ Agilent, Fluke etc., ] will respond to fix these issues if possible, OTAH the Chinese brand names often make errors in their designs that make it easier to hack [ Rigol, Hantek,  for example.]   

and i repeat... i was about to give Hantek name a big thumbs up, until i'm searching for LA, i expected their LA5034 (same case as 3x25) can do the same (superb HW imo, open SDK), ekk, wrong! so the morale? dont trust the Brand name esp from China.

[Mechatrommer]

Its a good idea to test every instrument individually for merits and flaws, even the better names have bugs and annoyances, but often better names [ Agilent, Fluke etc., ] will respond to fix these issues if possible, OTAH the Chinese brand names often make errors in their designs that make it easier to hack [ Rigol, Hantek,  for example.]   

yes true. in case of 3x25, the problem is in built in software, workable, but not so friendly (for me). but in case of LA5034, i think there is major problem i suspect down to FPGA firmware level, that they refused to respond customers requesting SDK. this link is from their web http://www.hantek.org/asken/iaskdetail.aspx?id=2010030500440655, just a simple "sorry this is no SDK", i bet they just left the project unfinished, move on and never look back, they left the sellers who still got the stock to sell that broken item waiting for innocent victim to buy, i'm really pissed off.

[alm]

i was about to give Hantek name a big thumbs up for somehow "Open Source"ing their 3x25 HW through SDK at least so we can develop our own program, for me the hardware is good, no other HW that can do the same, price the same, as you already know/searched, the same priced "dedicated" bench type only can go up to 2-3MHz, 3x25 is hackable to 100MHz

I think open sourcing the 3x25 is a big word for providing an SDK. All reputable vendors have been shipping VISA drivers for years, and usually also document the SCPI protocol. Just because some of the low-end vendors suck at documentation and software doesn't make Hantek special in this case.

and low Vpp is not a problem even for people like me.

It is a problem if the application requires 20Vp-p.

Its been in the top 10 threads for ~  1 year.  Good work mecha! While I prefer non-USB instruments for ease of use, if a USB instrument is better spec wise its a worthy trade off.  I think using virtual instruments will be less of a bother once we get them working on platforms like Android or iPhone/iPad because those platforms boot faster and touch screens are faster than working with only a mouse.

Given Apple's history the latter will probably only happen if Hantek ponies up a lot of cash for Apple's "designed for iPod' or whatever that program is called. Regarding Android drivers, is the protocol and hardware interface documented? Do they follow some (awful) industry standard like USBTMC? Or should the Hantek Windows SDK somehow be able to run on Android? This is exactly the reason why I prefer a completely documented protocol, not just a Windows SDK that will probably stop working once the major version of Windows comes along. Did I miss something and did Hantek provide protocol documentation?

Obsolescence is an issue with virtual instruments. A twenty years old piece of stand-alone equipment is still perfectly usable. A twenty years old virtual instrument, like a ISA DAQ interface, is useless. Some NI ISA cards only have drivers up to Windows '98. Their GPIB-ENET adapter only up to Windows XP. A life span of a few years may be acceptable for a $100 function generator, but kind of sucks if it's an oscilloscope costing a few thousands of dollars. Stand alone is a much safer bet, LCDs, knobs and buttons are more likely to keep working. NI have been predicting the death of the bench instruments for years. There still are bench scopes, multimeters power supplies and function generators on most benches, even those at NI. I consider virtual instruments more specialized instruments when automation is the prime objective and when space is at a premium. Virtual instruments tend to take up less rack space.

[Mechatrommer]

and low Vpp is not a problem even for people like me.

It is a problem if the application requires 20Vp-p.

i mean i have fixed the problem with my "10x25" albeit i'm not the intelligent one around here. (but dont talk about signal noise please, i know i will lose on this)

Or should the Hantek Windows SDK somehow be able to run on Android? This is exactly the reason why I prefer a completely documented protocol, not just a Windows SDK

ok maybe "open source" is not the right word by simply providing SDK. what i mean is exposing their HW for future (or users) development so the users will not stuck with the limitation of the built in software interface. i'm not sure if thats the right word either.

Stand alone is a much safer bet

2bfrank, i wish i could get all standalone on the bench devices, but considering my "environment" and my "status" and the return i can get from it, i dont think i can afford it, the same case with the OP, maybe he will get the benchtop 2MHz (3 or 5 whatever) VC2002 for his current project for this reason, but soon when he finished his "thesis" come out for the job, he will find out 5MHz is simply not enough for him. then it will goes back to square1. it will be alot of different story if i'm working in a big company or a fully professional EE designer and the core business is this stuff. and also partly, i believe the OP asked in the wrong place, where "engineers" are floating around. and partly, by right he should not ask for $200 device recommendation. or at least he will get cold replies. YMMV.

[alm]

I don't think any of the replies in this thread were looking down on a $200 function generator. In my opinion some of the <$200 DDS function gens (like the GW-Instek, not familiar with the competitors) are decent equipment. Main limit is the bandwidth. No arb either, but this is not a critical feature in my opinion. If he needs more bandwidth in the future, he'll have to reevaluate his requirements. Maybe he needs fast pulses, than a pulse gen may be a better option. Maybe he needs it for RF, than a stable signal generator may be better. No point in spending extra $$ if you don't have any idea if you need the extra features. Models with more bandwidth are often significantly more expensive. A (used) analog function gen may be a cheaper way to get more bandwidth, some go to 10MHz or beyond. Of course stability and spectral purity will be worse.

[Mechatrommer]

I don't think any of the replies in this thread were looking down on a $200 function generator.

only few (3 person if i'm correct including you, out of thousands EE members here) giving suggestion. the 2nd one suggesting USB, and the other one talking his broken AWG and suggesting 4x priced unit which eventually made it off spec'ed, like the 2nd person's low Vpp). so i'm "assuming" the low BW (within the $ limit) is hardly can be recommended, and also again "assuming" that everybody are working on higher BW. but i dont say $200 low BW FG is a bad thing, its actually better if its within your working spec. we was talking about "usability life span" of it.

[saturation]

Very good points alm, particularly the one on obsolescence.  I was writing more as a future of VI over it current existence on PCs versus traditional screen and knobs devices.

Buses and serial ports have advanced more rapidly and good instruments have very long shelf lives, devices that are bus based are often stuck with old hardware.  However, serial devices have more longevity, ye old RS232 can still be emulated well in a USB environment, and one can find GPIB to USB converters about.

Development using iPhones and Android as controller platforms for any instrument is in the exploratory phase and few production items exists, YTD.

http://www.redfishinstruments.com/

Agilent has this project:

http://www.home.agilent.com/agilent/editorial.jspx?cc=US&lc=eng&ckey=2075645&nid=-35714.384321.02&id=2075645&cmpid=zzfindsmartdevice

Its been in the top 10 threads for ~  1 year.  Good work mecha! While I prefer non-USB instruments for ease of use, if a USB instrument is better spec wise its a worthy trade off.  I think using virtual instruments will be less of a bother once we get them working on platforms like Android or iPhone/iPad because those platforms boot faster and touch screens are faster than working with only a mouse.

Given Apple's history the latter will probably only happen if Hantek ponies up a lot of cash for Apple's "designed for iPod' or whatever that program is called. Regarding Android drivers, is the protocol and hardware interface documented? Do they follow some (awful) industry standard like USBTMC? Or should the Hantek Windows SDK somehow be able to run on Android? This is exactly the reason why I prefer a completely documented protocol, not just a Windows SDK that will probably stop working once the major version of Windows comes along. Did I miss something and did Hantek provide protocol documentation?

Obsolescence is an issue with virtual instruments. A twenty years old piece of stand-alone equipment is still perfectly usable. A twenty years old virtual instrument, like a ISA DAQ interface, is useless. Some NI ISA cards only have drivers up to Windows '98. Their GPIB-ENET adapter only up to Windows XP. A life span of a few years may be acceptable for a $100 function generator, but kind of sucks if it's an oscilloscope costing a few thousands of dollars. Stand alone is a much safer bet, LCDs, knobs and buttons are more likely to keep working. NI have been predicting the death of the bench instruments for years. There still are bench scopes, multimeters power supplies and function generators on most benches, even those at NI. I consider virtual instruments more specialized instruments when automation is the prime objective and when space is at a premium. Virtual instruments tend to take up less rack space.