Review: Hantek DDS 3X25. Anyone own one?

[saturation]

Except for wirewound and carbon composites, where only certain special types or conditions can be used, most film types are easily used to the GHz, its more a question of better and best, and then what is good enough to reduce parts cost.

How about these MELF resistors?  Vishay claims they're good to 10 GHz+
http://www.vishay.com/docs/28718/melfhf.pdf

They say low inductance, but no mention of capacitance.

[saturation]

Yes, just slightly amplified, it just shows that 75 vs 50 ohm for this application, the differences are not so striking.  But note the frequency count of your image capture, its 10.0210 MHz versus 10.0000 MHz on mine, for the same waveform.  The Rigol might be reading glitches as the slight difference, as frequency rises those glitches could cause more effects that can be eliminated with proper connections. 

 

Most striking is a negative glitch is half the magnitude.

its there, i just amplified it using improper connection (ringing?)

[saturation]

As always, alm, you pose the most interesting questions.  Thanks for those spec sheets.  Yes, saw the mention of 'high frequency' again.  The other part of resistor articles from those references say that the frequency response deteriorates at increasingly higher resistance ratings but moreso for carbon composition resistors, CCR [excluding standard wire wound for obvious reasons.]   Probably the best thing we can do is get a spec sheet from a CCR manufacturer, I'm sure they won't deceive us!

http://www.koaspeer.com/pdfs/RC_SS-263_R4.pdf

See the frequency spec with its X axis being MHz*MOhm.

I think that if a device worked with variable frequencies, one could compensate for the changes in impedance by choosing resistors of lowest values, whenever possible.  This causes higher power consumption but more consistent response.

Likewise, if the device worked at a narrower range of frequencies and required high value carbon composition resistors, then even if actual resistances are reduced, a designer could work with that predictable characteristic and simply choose a higher ohm resistor to compensate for the reduction.

Here's a US Navy document regarding application of CCR in USN electronics:

http://www.google.com/url?sa=t&source=web&cd=17&ved=0CDsQFjAGOAo&url=http%3A%2F%2Fwww.everyspec.com%2FUSN%2FNAVSEA%2Fdownload.php%3Fspec%3DTE000-AB-GTP-010_R1-CHG-A.030031.pdf&rct=j&q=variations%20of%20carbon%20composition%20resistor%20high%20frequency%20use&ei=lnuCTZXOA474gAeBy7jbCA&usg=AFQjCNGTf2y90OvaLGY6o3qA0mLGOYEyPg&cad=rja


In this rather large image, a curious enthusiast note that the only way to confirm a CCR is really to break in open, stating that some CCR sold since the 1970s actually contained film technology:

http://www.w8ji.com/beverages.htm

Thanks for the info, I wasn't aware of the parasitic capacitance. Most sources I've seen (eg. this and this, just random datasheets) seem to consider them superior to metal film for high frequency (thin film is obviously superior). The first datasheet recommends carbon composition for a part that has transition times of 20-30ns, not exactly GHz speed, but no 10MHz either.

I find it suspicious that the graph has only one curve for carbon composition, I can't imagine parasitic capacitance having the same influence at say 1ohm and 1Mohm. I also doubt that all carbon composition resistors are useless beyond 10MHz, lots of equipment from the seventies and earlier was full of carbon composition (wirewound was inductive, and metal film expensive), many of it can go beyond 10MHz. I'd have to check if for example the Tek TM500 series (which uses lots of carbon composition and multiple plugins go up to 250MHz) use metal film in all places that need to pass high frequency. I believe carbon comp was frequently used as termination resistor, but if they'd only work up to <10MHz, they'd be almost useless, since you need fairly long cables to even notice the transmission line effect at that frequency.

I just checked, the Tek PG-502 (250MHz pulse generator with a rise time of <1ns) uses two ~50ohm carbon composition resistors (as indicated in the electrical parts list) as switchable rear termination, with no reactive components nearby to compensate for parasitics. It would need about 350MHz bandwidth to reproduce the 1ns edge, can't imagine that the termination has a zero impedance (extrapolating from the graph you posted) at those frequencies. These were just the first resistors I checked, there are many more carbon composition resistors in there (and in similar equipment of that vintage). There are some 1% metal film resistors, but they're there because of accuracy/stability. This doesn't mean that carbon comp is superior to metal film for HF, but it does suggest that carbon comp works well beyond 10MHz.

Maybe there are different ways of producing a carbon composition resistor? That would explain the conflicting data, although I've never seen something like a 'low-capacitance carbon comp resistor'.

[ElektroQuark]

Please, reduce image. Thank you.

[saturation]

Its linked, not my own.  I've removed it, you can see it at the website.

Please, reduce image. Thank you.

[alm]

The other part of resistor articles from those references say that the frequency response deteriorates at increasingly higher resistance ratings but moreso for carbon composition resistors, CCR [excluding standard wire wound for obvious reasons.]

That sounds more believable.

Probably the best thing we can do is get a spec sheet from a CCR manufacturer, I'm sure they won't deceive us!

http://www.koaspeer.com/pdfs/RC_SS-263_R4.pdf

Thanks for the real data, most manufacturers don't give any data beyond DC.

See the frequency spec with its X axis being MHz*MOhm.

That would explain the conflicting information. The 50 ohm resistors I quoted would be good up to 10GHz or so. Unlikely, since lead inductance will be an issue, but never mind that, 350MHz should be fine. A 1Mohm resistor would only be good to 0.5MHz or so.

I think that if a device worked with variable frequencies, one could compensate for the changes in impedance by choosing resistors of lowest values, whenever possible.  This causes higher power consumption but more consistent response.

You usually want lower impedance circuits at higher frequencies anyway, because otherwise parasitic capacitance (eg. between traces) will swap your resistors, even with ideal resistors. Plus 1Mohm transmission lines are kinda impractical .

Likewise, if the device worked at a narrower range of frequencies and required high value carbon composition resistors, then even if actual resistances are reduced, a designer could work with that predictable characteristic and simply choose a higher ohm resistor to compensate for the reduction.

I'm not sure how stable the parasitic capacitance is, both over time and between samples. The figures give just typical data. But there's really no reason not to use something superior like thin film these days, you don't want those inductive leads at high frequencies anyway.

Here's a US Navy document regarding application of CCR in USN electronics:

http://www.google.com/url?sa=t&source=web&cd=17&ved=0CDsQFjAGOAo&url=http%3A%2F%2Fwww.everyspec.com%2FUSN%2FNAVSEA%2Fdownload.php%3Fspec%3DTE000-AB-GTP-010_R1-CHG-A.030031.pdf&rct=j&q=variations%20of%20carbon%20composition%20resistor%20high%20frequency%20use&ei=lnuCTZXOA474gAeBy7jbCA&usg=AFQjCNGTf2y90OvaLGY6o3qA0mLGOYEyPg&cad=rja

They suggest up to 1MHz, which agrees with Koa's data with resistances up to 1Mohm or so.

In this rather large image, a curious enthusiast note that the only way to confirm a CCR is really to break in open, stating that some CCR sold since the 1970s actually contained film technology:

So that's who the Chinese distributors selling carbon film as metal film learned from .

[saturation]

Here's an interesting article about the history of film vs carbon composition, in the context of Japan's emergence from WWII.  Not many historical articles exist about component development at an industrial level.

http://ieeeghn.com/wiki/images/c/c2/Takahashi,_Japanese_Electronics_after_WWII.pdf

[Mechatrommer]

this is an update. and partly a correction to my previous post/report on "leap year", "dajones glitch" and "non-synched non-multiple" effect. now i'm not sure whether the internal firmware/software in hantek 3x25 is problematic, but i think there is workaround. my last report is worsen by the fact that the pc software dds-3x25 usb (3.2.1.7) is not done properly to handle this "buggy" effect. hence uploading a flawed data to the 3x25 and revealing the "buggy" effect especially the worst out-of-synch "signal" vs "synch out" output.

with proper calculation on the data length to be sent to the 3x25, from my incomplete observation, all the three unwelcomed effect seem to be resolved (dissapear), as demonstrated by this demo tool. there is no more "not-multiple and glitch/out of synch", except the 3x25 "synch out" signal seems to have lower resolution at high speed signal ie you cannot type in freq 35MHz and expect the synch signal to be exactly that, there is gap that it will only generate signal either 33.3MHz or 39.999MHz etc. the same to the reported capability of 75MHz, its not, even if the signal can generate 75MHz sine, but its synch output will only be 100MHz at some signal freq range (70-75MHz), this explained why "out-of-synch" occured. and lastly the sine attenuation at that (fast) frequency is also a pc-software artifact, not entirely hardware.

bottom line. all the reported bugs so far, are easily avoided by pc software means. so up to now i will give hantek 3x25 AWG a full "four" thumbs up!
ps: i forgot to mention that i found out something like "golden frequency" for this device is 50KHz or 200MHz, but i cannot explain what or how, its just there.

edit: demo.zip is just a simple tool with nothing much to do, browse further this thread for more complete software tool. cheers

[Mechatrommer]

i've been lack in sleep for the past several hours. tried to work hard to get my hantek synch stable and in phase. i ask help if any hantek owner can make a test on this little demo/tester program if it can generate stable and in-phase signal vs synch output on their 3x25 unit. any reply and pictures will be appreciated. it use a re-calibration table (calib.csv) so you can adjust the shift angle value in there. just dont abuse the little tool, such as entering alphabet in frequency value etc. i havent made any proper checking for invalid input. hope will hear my program works.

[saturation]

AWESOME WORK Mecha, will give it a try and report back.

[Bored@Work]

Has someone seen a Linux driver for that thing, or at least some project where the USB interface is reverse-engineered?

[saturation]

It has to be seen in video.  10/10 terrific job, mecha!  You're done to Hantek what eevblog has done to Rigol 1052e.

[Mechatrommer]

man! i didnt expect it will be on video. and pls dont exagerate pal, its embarassing, i kind of over flattered, but thanx for the video effort. but i still dissapointed you didnt mention that now 3x25 support sawtooth wave nevermind its 100% ok. there is no magic in it, just a simple "remainder theorem" i think. and all rectification are done in software. several thing i want to point out...

1) the glitch and irregularity in "square synch out" is the property of the hardware, i've done nothing to repair it. its just automatic, and from my observation, "square synch" irregularity is affected by signal shape and phase just as shown in pictures below (note the notches location on square)

2) i made different 50ohm BNC cable and now i can read both signal at terminated setting (my previous pictures showed non terminated signal, but terminated synch output). with different cable and termination setup, my phase calibration no longer works. as also shown in the same pictures below (synch_vs_signal.jpg), by right they should be 0 degree phase shift, bt not anymore. so i think its the instability in the hardware, not much i can do. worst in highest freq range 100MHz, at 2 points per period, i will no longer able to shift between 0 and 180 degree. so its a bad news.

i would love to expand the software to include arbitrary point editing, to make it more flexible compared to hantek bundled software. but i think i need more vote on it since it will take much more time, and its still workable in the bundled software. i believe people like saturation will have no problem solving problem with the bundled hantek software. as for now, i still concentrate on how to (1) make the signal in-phase and (2) minimize signal attenuation by using some algorithm. about the former (1) since fom my test and observation... different cable setup and highly probably different circuit in real application will need different calibration value, so it will be difficult and i think we still need to look closely the "signal vs synch phase difference" during our real work.

and lastly, its not a genius work, its just a devotion to make a better thing and an appreciation on information sharing, so our one sleepless night will not just be benefited by us alone, but by many others and propagate throughout... Cheers.

[Mechatrommer]

and lastly, but not least. this hantek can be made to work down to 0.5Hz (extended lower limit. capable 0.75, 0.63 and any decimal between sub 1Hz and 0.5Hz) the spec said only down to 1Hz, so another exploit. but the synch output only can go down to 1Hz, so less than 1Hz signal is not synched. in the little tool you can see less than 1Hz signal by entering the freq and disable the "Stable Synch / Jitter" option.
ps: if you found out something you stumbled (cannot do) with existing (bundled and the little tester) software. be glad to ring me a bell.

[Mechatrommer]

fooling around with sub hz, i found interesting thing. the fpga tries its best to synch at any frequency by means of previously mentioned "dajones glitch" and "leap year effect". now we can study those in great stability at sub hz frequency as shown in the picture.

[saturation]

You still need the applause mecha, as you've substantially improved the functionality of this device!  Yes, I did notice the other improvements, but leave those comments for later.  The core problem of dajones, you've addressed very well.  Even if the phase starts shifting, its consistent, so you can still use the sync signal output for triggering.



More later but what I would love to see fixed are:

1. Sweep function: the factory software only allows sweeping to 100kHz.  Its a shame since it can go to 100 MHz.

I'd like to see it capable of sweeping from 0 Hz to maximum Hz supported, as a single sweep.

You can even run this through the client software so it goes from X-Hz to Y-Hz in Z-Hz increment with a delay from milliseconds to seconds.

That would be fantastic!

2.  amplitude and frequency modulation: similarly limited in maximum frequency.  If you can allow it to use any modulation frequency and a carrier up to the maximum supported frequency, that would be great.  As far as any artifacts this creates, I think the user should manage any DDS generated artifacts from modulation.

[Mechatrommer]

1) 0Hz is impossible in hardware. a pc software dirty trick (processing power) is needed for this.
2) what do you mean "client software"?
3) modulation: there is limit to carrier frequency depending on signal frequency. will study more on this.
4) will keep the suggestion for the next upgrade

[saturation]

0 Hz  ! Of course, what I meant was support as low as supported by the device.

The 'client' is the software on the PC use to control the Hantek

No rush, do what you feel in good time.  

Sweeping function, if that is easier to do, would be a more important concern as AM or FM modulation is less used today except for radio.

1) 0Hz is impossible in hardware. a pc software dirty trick (processing power) is needed for this.
2) what do you mean "client software"?
3) modulation: there is limit to carrier frequency depending on signal frequency. will study more on this.
4) will keep the suggestion for the next upgrade

[Mechatrommer]

0 Hz  ! Of course, what I meant was support as low as supported by the device.

1) 0Hz is impossible in hardware. a pc software dirty trick (processing power) is needed for this.

sorry i mean 0 < Hz < 0.5, 0Hz is actually super easy.
for the sweeping, i have a dilemma since i got this thing. signal transition, seems not so smooth, there is signal overlap (eg. transiting from sine to triangle, or sine changing frequency etc) but it was rough observation, and its hard to get the capture on screen. i will look more closely next time.

[Mechatrommer]

ok, i've done with sweep function (f,v,offset capable) for primitive signals (hantek.zip). i believe its far more superior than the bundled software
you are free to download and use it, except you cannot make a bomb with it!

[ElektroQuark]

Awesome work, Shafry.
Thanks.

You can try sellint it to Hantek

[saturation]

Thanks mecha!  Will report back.  Hantek should give you a commission or buy your software 

[Mechatrommer]

thanx guys. the bad thing is, i come late in this party, they have discontinued this product, otherwise they have to do something on the software side. i dont understand how they didnt manage to harness the hardware to its full potential (0.5Hz-100MHz, fast high freq sweep etc). on the hardware side, imo is done excellently, its just the most important part of interfacing with it on the software side is done the other way around (not very excellent). maybe their software guy is too busy. but there is a plus to this con, ie they cannot boost the product price due to their crappy software. all its needed is several night of "thoughtfull pressure'less seclusion" and many glasses of nescafe and cigarettes.

ps: there is one kind of signal i want to put in the little tool, but i still cannot figure out how. a single pulse, if anyone can give advice on how to do single pulse, i will appreciate it.
there is DDSSetSingleWave Lib "DDS3X25Dll.dll" in the library, but dont know what it does, it will not generate any wave if i activate it.

[saturation]

It is still in the catalog on their website, and a lot for sale on google.com

IIRC Saelig.com was the store specifically who said they were discontinuing the product [from their catalog?] and had it one sale until inventory was gone.

https://www.mortoncontrols.com/index.php?main_page=product_info&cPath=2_9&products_id=232

If you search google, most everything about Hantek's capabilities [ other than vendors selling it and Hantek's own website] is through the eevblog forum, and our threads.  So, it was a good find on our part and we made a good team.

I guess one real problem is they didn't have you on their design team!

thanx guys. the bad thing is, i come late in this party, they have discontinued this product, otherwise they have to do something on the software side. i dont understand how they didnt manage to harness the hardware to its full potential (0.5Hz-100MHz, fast high freq sweep etc). on the hardware side, imo is done excellently, its just the most important part of interfacing with it on the software side is done the other way around (not very excellent). maybe their software guy is too busy. but there is a plus to this con, ie they cannot boost the product price due to their crappy software. all its needed is several night of "thoughtfull pressure'less seclusion" and many glasses of nescafe and cigarettes.

ps: there is one kind of signal i want to put in the little tool, but i still cannot figure out how. a single pulse, if anyone can give advice on how to do single pulse, i will appreciate it.
there is DDSSetSingleWave Lib "DDS3X25Dll.dll" in the library, but dont know what it does, it will not generate any wave if i activate it.

[Mechatrommer]

I guess one real problem is they didn't have you on their design team!

i can work part time online, but i will not go to china to work with them . well its not needed, all they need is to listen and make effort if they want to make a good business. if not then i will still be happy to buy their cheap product with the crappy software. if somehow they come accross this thread, then listen this... sell your very good hardware and very good software separately at cheap price each. if you have both very good hardware and very good software combined and sell at expensive price, then i will think twice to buy them, but if both sold at cheap price, then you are the winner.