• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

ESS THD ‘Hump’ Investigation

Zoide

Active Member
Joined
Oct 20, 2018
Messages
100
Likes
75
Tones due to static signals are described in every introductory text on delta-sigma converters. DC offsets caused significant tones in the earliest designs (low-order, single-bit). Higher-order loops, dither (noise decorrelation), and multibit designs pretty much killed them many years ago. But it seems like the DEM schemes used to improve the multibit converters has brought this old problem back. Ironically back in the mid-1990's I was designing a DAC and considered DEM. I scrapped it mainly due to power and voltage headroom issues, but did see the modulation described in the slides when I used a simple shifter. I had totally forgotten about that. I went to a PRNS (pseudo-random noise source) and that squelched the tones.

I have some old notes for a college class I gave on DS converters but when I tried to resurrect the notes a couple of years ago I failed. It was in AmiPro; MS Office had but no longer has a converter, and a comercial converter got the text but not the figures -- and it was pretty much all figures! Blah. It had a few slides on tones with pretty pictures. Including Bessel functions, not one of my most favorite things... They are also used in the derivation of SFDR for an ideal converter.
You can get AmiPro here ;)

https://winworldpc.com/product/amipro/3x
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,894
Likes
16,710
Location
Monument, CO
Thanks! I actually downloaded a version from somewhere, but to run it required massive system changes (like changing system fonts, various registry keys, etc.) so I'd essentially need a second PC. Might be worth it; I have a lot of lectures and some conference papers I presented that are all in AmiPro. I had pulled them and tweaked them so have Word versions of some, but since I only rarely needed the originals I missed it when MS dropped their AmiPro filter. :(
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,705
Location
Hampshire

Reyer

Member
Joined
Feb 5, 2019
Messages
25
Likes
32
Thanks! I actually downloaded a version from somewhere, but to run it required massive system changes (like changing system fonts, various registry keys, etc.) so I'd essentially need a second PC. Might be worth it; I have a lot of lectures and some conference papers I presented that are all in AmiPro. I had pulled them and tweaked them so have Word versions of some, but since I only rarely needed the originals I missed it when MS dropped their AmiPro filter. :(

Alternatively you could install Virtualbox to create a virtual machine using some old OS ISO. As far as I see, they're available on the same webpage.
However I'm not sure if they'll work in your case if you need an external hardware to connect to it.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,894
Likes
16,710
Location
Monument, CO
The only external HW I need are keyboard, mouse, and monitor. These are just presentations (class lectures and papers I've presented over the years). This came up about a year ago when I needed to retrieve an old SiGe HBT paper, and again when I wanted to pull my DS converter notes into a thread here. And yesterday when I read the interesting post on tones and realized I had some relevant material from my grad class lecture but no way to get to it. :(
 

exaudio

Member
Forum Donor
Joined
Jun 11, 2018
Messages
46
Likes
75
@DonH56 , IBM has made Lotus SmartSuite 99 freely available at the Internet Archive. You can download it here: https://archive.org/download/LotusSmartSuite99

It should be able to open your AmiPro .SAM files. Here's what it looks like:

lotus-amipro.png
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,894
Likes
16,710
Location
Monument, CO
@DonH56 , IBM has made Lotus SmartSuite 99 freely available at the Internet Archive. You can download it here: https://archive.org/download/LotusSmartSuite99

It should be able to open your AmiPro .SAM files. Here's what it looks like:
<elided>

Thanks, that looks promising! I know I found a bunch of versions 2-3 years ago when I really needed it. I need to research why I never loaded it. I know one of the versions would do things like overwrite system files, fonts, etc. during the install. I debated doing it, converting the files, then restoring my PC from my backup but ultimately just made new slides for the talk. I bought a highly-rated converter but it utterly failed to restore any of the images (even though most were WMF imports). I need to look into this again.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,894
Likes
16,710
Location
Monument, CO
Easy to say but I'm an analog designer and despite decades of using computers from PCs to Crays I am not really much of an expert on them. Using a VM sounds like a good idea, however. I found some of my old notes and it seems like user reviews for the IBM installs did point out some system fonts and such being overwritten. They went unnoticed by a lot of folk. There were also some cases of registry corruption but it's hard to say what really might have happened.

There are maybe a couple of dozen (of a couple of hundred) that a really care about, and probably only a few I'd really like to recover (one a paper Frank Goodenough goaded me to present; he was quite a guy!) The rest I could recreate, just lazy, and the past few years work has been so demanding I've had to drop giving lectures at the local university so it's not really critical.

Side note for Matlab users: I had some programs I wanted to hand to the students but did not know if they would run in the student version. I pinged Mathworks to ask and said I would pay for a student version but without an ID could not get one in the store and did not want to pirate one by having a student buy it for me. They said "no problem" and sent me the student version free, and updated my main licenses. Very nice! I'm sure they figured the goodwill was worth it, but when I asked another mathematical analysis program company for which I had an old version they basically said "here's a link to the latest version, which you may download at full retail price, since your version is too old to qualify for an upgrade". It was much cheaper and not too hard to convert all the code to Matlab.
 
OP
jackenhack

jackenhack

Active Member
Forum Donor
Joined
Oct 25, 2018
Messages
203
Likes
497
Location
Stockholm, Sweden
Ok, so I finally got new DAC:s to replace the dead one. Desoldered the old one and replaced it. I only had lead-free solder, and naturally, I got two short circuits as you can see in the pictures. And because I had to use lead-free solder, I had to heat it again, to insane levels, because the devil makes lead-free solder. I lifted one of the copper foils from the PCB. DAC destroyed. There's an old trick that usually works when fixing small short circuits like this, just put a lot of flux around the part and put a large dab of solder on an iron, lightly dragging the soldering iron across the short, but it didn't work, because, you guessed it, lead-free solder. Gaaah!

IMG_3003.jpg

So now I have to order a new Khadas DAC. Not too expensive, but I could do without the waiting time. Will be back...
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,705
Location
Hampshire
Ok, so I finally got new DAC:s to replace the dead one. Desoldered the old one and replaced it. I only had lead-free solder, and naturally, I got two short circuits as you can see in the pictures. And because I had to use lead-free solder, I had to heat it again, to insane levels, because the devil makes lead-free solder.
I've been using tin-bismuth solder paste lately. Its melting point is even lower than that of tin-lead. 0.4 mm QFN is no problem with this stuff. For hand soldering, especially removing bridges like in your case, a eutectic alloy is a must, and many common lead-free solders are not.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,300
Location
China
Do we have conclusions now? I have been following this thread for months.
Now i think it's really beneficial if someone can share the actual spectrum analyzer graphs at different frequency at different level and multitone td+n at different level to really dig out what is actually happening. Now it's kinda like we are only seeing half of the picture.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,894
Likes
16,710
Location
Monument, CO
The conclusion is that, if one must use lead-free solder, use a tin-bismuth alloy.

Oh, and I think it was stated early on that the hump is inaudible with musical sources.
 

Miska

Addicted to Fun and Learning
Audio Company
Joined
Feb 20, 2019
Messages
615
Likes
448
If I read the graph correctly, it occurs about 65 to 90dB below the signal, but when the signal has already been attenuated by 20 to 40dB.

20 to 40 dB is quite typical attenuation to have with a volume control through... On my S2D for example to Shure SRH1540 the setting would be -32 dB...
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,424
Likes
2,794
Location
South of China, SHZ area, - Слава Україні
THD Hump?
jackenhack, thank you for your INIT hack but have you any idea why your I2C dump has no acknowledges? Are you sure that the host send nothing more after that INIT? I asking because es9038q2m on my bench right now and I trying to set it up for SPDIF GPIO#1 input..
PS: ESS and other tiny fabless(cmedia etc) always hide application info, it kinda disease. Trivial chip can take weeks of reverse engineering! I spent less than an hour to start ak4490 DAC because all info is in the public domain.
 
OP
jackenhack

jackenhack

Active Member
Forum Donor
Joined
Oct 25, 2018
Messages
203
Likes
497
Location
Stockholm, Sweden
jackenhack, thank you for your INIT hack but have you any idea why your I2C dump has no acknowledges? Are you sure that the host send nothing more after that INIT? I asking because es9038q2m on my bench right now and I trying to set it up for SPDIF GPIO#1 input.

It is the initial setup dump. Don't know why the ESS chip didn't send ACK back. But I do know the unit worked after the dump. I even had it connected and did change the sample rate and bitrate, but no further communication to the ESS chip other than the first setup. Probably just set and forget. So there are no new changes to the THD compensation. Only the default.
It went up in smoke later when trying to change the values of the registers with a Bus Pirate. I accidentally added 3.3 volts on top of the regular 3.3 volt supply. Opps.

As soon as I get my life back on track, I'll order a new Khadas DAC and try again.
 

IVX

Major Contributor
Joined
Mar 3, 2019
Messages
1,424
Likes
2,794
Location
South of China, SHZ area, - Слава Україні
Thank you for reply. Your dump mostly equal to default and I didn't send in may setup default values, maybe this is the reason but my DAC has no output yet. Probably that IC super sensitive and was dead during my soldering(hot air 270C, and I forgot to tie pin#22 to GND for the first turn ON). Seems I have to buy typical inexpensive chifi 9039q2m PCBA and make dump too.
 
OP
jackenhack

jackenhack

Active Member
Forum Donor
Joined
Oct 25, 2018
Messages
203
Likes
497
Location
Stockholm, Sweden
Seems I have to buy typical inexpensive chifi 9039q2m PCBA and make dump too.

Yeah, I'm interested in getting what kind of changes to the THD compensation registers make to the output. According to the manual, it's there to compensate for external components. Could it cause an ESS DAC based system to measure even better? I have a feeling that most Chinese manufacturer leaves it set to the default values and call it quits. Some, like SMSL, use it to make "analogue" sounding filters for tube sound and the like, by just messing with the second and third harmonics. Topping leaves their software configuring to a German company. Maybe not enough in-house knowledge on programming the XMOS processor? Perhaps it's time to take off the cover on my SMSL SU-2 and check if they use the compensation registers other than for making specific sounds.
 
Top Bottom