AliExpress cheap low distortion analog sine generator - review and measurements

[IVX]

The limit is in my ADC E1DA cosmos

BTW, I sold almost 2000pcs and didn't see [email protected] H3 worse than -133-134db. I recommend using Cosmos ADC at 768kHz or warm it up with thermostat-board(or even a hair dryer is ok).

 

[pma]

Hi, I used normal metallized resistors selected under 0.5%, cermet trimmers Bourns, MKP and NP0 capacitors, operational amplifiers thanks to the sockets I tried many types and chose the best of which I had a sufficient amount in the reduction of soic8 to dip8. JFET original 2SK30 from Toshiba, OPA definitely were not the prescribed TL071 and NJM/NE5532. I have three boards, so I still want to make generators for 100Hz and 10kHz. Victor generator (1st gen) has a very poorly designed power supply so it makes noise and distorts more.

So, you are not posting any specific information, right? What is your point, then?

 

[IVX]

So, you are not posting any specific information, right? What is your point, then?

I think he said that Victor OSC has an oversimplified power circuit, just tl431. Archimago used batteries for the same reason. I tried Victor's early schematic, the same or similar SSS posted, H3 at 2Vrms was -150-155db, 4-5Vrms -140db, H2 was much worse, I even added the trimpot for the symmetry adjusting(R14=R15=220ohm seems too high value, SSS posted version with 33ohms). Victor says he's new OSC is better.

 

[phofman]

What exactly do you mean as "slightly". Could you define how "slightly" (exactly in Hz) in case of the measurement below.

I do not know what the actual frequency shift is. I could analyze it, if you could post (some download link) some 20 -30 minutes of straight recorded flac at full resolution, no resampling, ideally wasapi exclusive/asio/linux alsa hw:X. Thanks!

 

[pma]

I do not know what the actual frequency shift is.

Within minutes/hours of operation it is in order of 0.1Hz. During short-term operation (like tens of seconds) in 0.01Hz. The bin width used is 0.37Hz ......

The change of frequency may rather come with mechanical impact that would affect the setting of multi-turn pots.

 

[pma]

I have tested the generator with added high current output buffer, more here.

 

[pma]

This is the best result that I can get from this cheap generator + Cosmos ADC (without adjusting internal multi-turn pots). Cosmos DIP switches set to 7.6V input range.

THD = -120.7dB, SINAD = 116.5dB, SNR = 118.6dB.

 

[JohnPM]

Please what would happen if the fundamental slightly shifts within the window of samples (as long FFTs take seconds/tens of seconds of samples)?

Not much, a corresponding broadening of the peak. Same is true for a digital source, all clocks vary.

 

[pma]

Not much, a corresponding broadening of the peak. Same is true for a digital source, all clocks vary.

Indeed.

Analog generator


DAC

 

[phofman]

Not much, a corresponding broadening of the peak. Same is true for a digital source, all clocks vary.

Thank you. And what happens to the values of the peaks? IIUC broadening should mean the bin energy (value) gets spread to the adjacent bins.

 

[dualazmak]

It is my private interest and attention that the product-name printed on the PCB can be read as "Low-distortion Audio-range Osoi l lator"!
The adjective word "Osoi" (遅い) in Japanese means "late" or "slow" or "delayed", and the printed "lator" easily reminds me English word "later"; they are quite funny, almost one of the funniest and/or most ridiculous jokes, at least for me...

 

[pma]

I think is is just wrong typing, wrong characters used, related to extra low cost and country of origin. I would not try to find any special intentions. The web description denotes it properly. At $16 including shipment and VAT, I cannot have any objections. It is unimaginable to produce it, sell and ship at such price.

 

[phofman]

It is unimaginable to produce it, sell and ship at such price.

I do not see any expensive components there, or are there any? IMO it's "just" an exceptionally well designed circuit and PCB (maybe with a great deal of luck, we do not know).

 

[dualazmak]

I think is is just wrong typing, wrong characters used, related to extra low cost and country of origin. I would not try to find any special intentions. The web description denotes it properly. At $16 including shipment and VAT, I cannot have any objections. It is unimaginable to produce it, sell and ship at such price.

Maybe so, and I hope so, but we can see little pre-marketing product check and possible poor QC on the product...

 

[pma]

I do not see any expensive components there, or are there any? IMO it's "just" an exceptionally well designed circuit and PCB (maybe with a great deal of luck, we do not know).

The parts are cheap, that's right. They must be cheap to make it cheap. However, the film capacitors are not that cheap. OK, make a part list, production and technology preparation, add manufacturing and testing costs, delivery chain, and tell me how much it would cost when produced in EU, USA or Japan. Have you worked in a production company? I did, as a senior design engineer.

Maybe so, and I hope so, but we can see little pre-marketing product check and possible poor QC on the product...

Yes, but again, consider the price. Regarding QC, this is an analog product and there must be manual setting of 3 multi-turn pots to make it work properly.
It might be made nicer and sold at $100. Then I would never consider to buy and test one.

Guys, this is meant to be a technical thread. Please keep it on technical, engineering side. If you have any hints to make circuit improvements, they would be welcome.

 

[pma]

Sample measurements of Purifi 1ET400A EVAL2 (high gain 27.5dB), with this analog generator.



5W SINAD is close to @amirm measurement here.

 

[Nopiano]

At $16 including shipment and VAT, I cannot have any objections. It is unimaginable to produce it, sell and ship at such price.

I wanted to say similar. Even shipping an empty box would cost more, surely?!

 

[JohnPM]

Thank you. And what happens to the values of the peaks? IIUC broadening should mean the bin energy (value) gets spread to the adjacent bins.

Yes, but the total doesn't change. REW cares about energy within the span, not peak values.

 

[phofman]

Yes, but the total doesn't change. REW cares about energy within the span, not peak values.

Thanks. Please how is the bin span determined? If the span is larger, does it mean the total fundamental energy also includes noise from close frequencies unrelated to the fundamental? Just trying to learn the details, thanks!

 

[JohnPM]

Thanks. Please how is the bin span determined? If the span is larger, does it mean the total fundamental energy also includes noise from close frequencies unrelated to the fundamental? Just trying to learn the details, thanks!

If the Use AES17-2015 standard notch option is selected the fundamental power for THD and THD+N calculations will be the power within a one octave span around the fundamental frequency. If that option is not selected it will be the power in the main lobe of the fundamental. When the fundamental level is not well above the noise floor using the standard notch will produce a higher figure than the fundamental main lobe, in those cases it is better not to use the standard notch setting.