I only found out about E1DA and its products - better said Igor's or IVX here on forum - yesterday.
I wanted to ask him regarding specifics of his units and whether they are also at least potentially usable for my intended use by sending him a private message.
It turned out I had to re-verify my email, renew the account etc due to not posting for a while. After that, I realized I count as a new member, unable to send PM until posting some times - and that at least some of my conversations from the past vanished into thin air. Like those that actually did help finding help that would otherwise be most likely impossible to obtain.
So, I would like ask IVX to send me PM how we can get in touch - as I am sure he would not mind answering the questions few/some/many other members might/may find a red flag to the bull.
Tired of that for life and do not want to enter yet another "round".
I wanted to ask him regarding specifics of his units and whether they are also at least potentially usable for my intended use by sending him a private message.
It turned out I had to re-verify my email, renew the account etc due to not posting for a while. After that, I realized I count as a new member, unable to send PM until posting some times - and that at least some of my conversations from the past vanished into thin air. Like those that actually did help finding help that would otherwise be most likely impossible to obtain.
So, I would like ask IVX to send me PM how we can get in touch - as I am sure he would not mind answering the questions few/some/many other members might/may find a red flag to the bull.
Tired of that for life and do not want to enter yet another "round".
Ok but I believe it is a good idea to keep the conversation here opened way to answer to similar questions once. So, I will repeat my answer about the DSD format for Cosmos ADC. Technically, Cosmos ADC has all for that but ES9822 has no DSD modulator i.e. that chip can convert PCM to DSD, and Cosmos ADC USB bridge chip may send that to your PC as DSD but who may need that surrogate DSD? Marketing people? 
Hint: if somebody is interested to see mostly H3 as low as possible, I can recommend replacing RT6863S opamp for OPA2156. H3 will be -160db@1kHz or lower, however, H2 goes up for about 10db(-143-145db). I have to say that about 10% of RT6863S also shows H3 at -160db but usually -153db+/-3db.
About the same with OPA1656. OPA1612, if you are lucky, may reduce H3 for about 3db, and keep H2 the same i.e. -150-155db but some of OPA1612 may increase H2 up to -145db.
It is clear that the issue is unsymmetrical input impedance but I can not sacrifice a residual noise for that as well to keep thd+N cool. So, I was pushed to find the opamp with max symmetrical input stage and Ok loop gain for the H3 performance simultaneously.
About the same with OPA1656. OPA1612, if you are lucky, may reduce H3 for about 3db, and keep H2 the same i.e. -150-155db but some of OPA1612 may increase H2 up to -145db.
It is clear that the issue is unsymmetrical input impedance but I can not sacrifice a residual noise for that as well to keep thd+N cool. So, I was pushed to find the opamp with max symmetrical input stage and Ok loop gain for the H3 performance simultaneously.
I receiving strange questions regarding APU's Notch, and really don't understand what to reply. Probably, not all users really understand how the device works and so on. After all, I decided to draw a simplified diagram of that, I hope it helps.
Attachments
Last edited:
Hi...
Been following this thread, and also did some experiementation with my own diy passive 1KHz notch.
I was having a hard time myself understand howto use it in conjounction with REW and a (mic.)calibration file, but i made it at last
Along the way, i asked some quistion's regarding the FS (FullScale) and the [Manuel fundamental] setting in REW.
The quistion's are answered here by the owner of REW, John: Link!
Well, i also did read this Link!, where Archimago created a calibration file, but differently than the way i "learned" to do it.
Anyway i see neither here, or on Achimago's blog, the use of [Manuel fundamental]
[Manuel fundamental] is where you set the "input VRMS" before the notch and Mic. calibration file, so that REW can calculate the correct THD + freind's numbers.
Okay then, I think @IVX that there should be some explanation and maybe howto's to set it all up and to use it correctly.
There are a lot of armateurs like me around, donno if i'am tho only one, which didn't get it but i doubt
Another thing is, when looking at your'e nice simplified diagram (which is a very good thing to show IMO), i don't understand why the add of +20dB and buffer at the end of the chain, this does not sum up with 30dB in my head.
Please explain further, I (or proberly others) need more indeep explanation.
Rgds; Jesper.
Been following this thread, and also did some experiementation with my own diy passive 1KHz notch.
I was having a hard time myself understand howto use it in conjounction with REW and a (mic.)calibration file, but i made it at last
Along the way, i asked some quistion's regarding the FS (FullScale) and the [Manuel fundamental] setting in REW.
The quistion's are answered here by the owner of REW, John: Link!
Well, i also did read this Link!, where Archimago created a calibration file, but differently than the way i "learned" to do it.
Anyway i see neither here, or on Achimago's blog, the use of [Manuel fundamental]
[Manuel fundamental] is where you set the "input VRMS" before the notch and Mic. calibration file, so that REW can calculate the correct THD + freind's numbers.
Okay then, I think @IVX that there should be some explanation and maybe howto's to set it all up and to use it correctly.
There are a lot of armateurs like me around, donno if i'am tho only one, which didn't get it but i doubt
Another thing is, when looking at your'e nice simplified diagram (which is a very good thing to show IMO), i don't understand why the add of +20dB and buffer at the end of the chain, this does not sum up with 30dB in my head.
Please explain further, I (or proberly others) need more indeep explanation.
Rgds; Jesper.
lykkedk, APU has an active Notch filter 1/10kHz switchable but the ratio of notching is -30db for both frequencies. You can simply subtract that number from your REW reading to get the actual THD+N. Of course, you can get better precision if build the calibration file as Archimago did but not necessary. For example, I made that file just by my hands.
12 70 0
18 35 0
20 33 0
22 31 0
25 30 0
30 30 0
40 30 0
60 30 0
100 30 0
150 30 0
250 30 0
500 30 0
700 30 0
800 30 0
900 25 0
950 20 0
995 0 0
1000 0 0
1005 0 0
1100 25 0
1200 30 0
1300 30 0
1500 30 0
2000 30 0
2500 30 0
5000 30 0
10000 30 0
11000 30 0
15000 30 0
19000 32 0
20000 33 0
21000 34 0
24000 36 0
35000 70 0
60000 80 0
100000 100 0
12 70 0
18 35 0
20 33 0
22 31 0
25 30 0
30 30 0
40 30 0
60 30 0
100 30 0
150 30 0
250 30 0
500 30 0
700 30 0
800 30 0
900 25 0
950 20 0
995 0 0
1000 0 0
1005 0 0
1100 25 0
1200 30 0
1300 30 0
1500 30 0
2000 30 0
2500 30 0
5000 30 0
10000 30 0
11000 30 0
15000 30 0
19000 32 0
20000 33 0
21000 34 0
24000 36 0
35000 70 0
60000 80 0
100000 100 0
@IVX, good explanation about the active notch giving attenuation of 30dB, makes perfectly sense. Thank's!
But it's still not clear to me, that when using a calibration file - are you sure that REW will calculate the correct number's according to the 30dB attenuation?
If i'am not using the manuel fundamental box when using my 1KHz notch in REW, the THD numbers are way off.
Sry. for my ignorance, i know you are lightyears ahead of my knowledge IVX...
Would it be possible to show the use with REW and a calibration file here perhaps?
Jesper.
But it's still not clear to me, that when using a calibration file - are you sure that REW will calculate the correct number's according to the 30dB attenuation?
If i'am not using the manuel fundamental box when using my 1KHz notch in REW, the THD numbers are way off.
Sry. for my ignorance, i know you are lightyears ahead of my knowledge IVX...
Would it be possible to show the use with REW and a calibration file here perhaps?
Jesper.
Why not sure, just simply subtracting 30db from the THD+N result is match AP measurement +/-.5db or so.
Yes i know that you are right here, when compared to AP it's also verified @IVX
Today i took some measurement's with my H2 generator (Nelson Pass's). Setting the H2 around 1.1% (~ -40dB)
I took the measurement's first directly into my Behringer soundcard, and afterwards with my 1KHz into the chain
The number's are more or less equal with and without notch, witch they should be when noisefloor and THD's are way above the soundcards own internal one's
Direct :
Notch in the chain (with mic. calibration file loaded & manuel fundamental set) :
And last; Notch in the chain (with mic. calibration file loaded & manuel fundamental NOT set) :
With calibration file loaded in REW, and manuel fundamental NOT set, i can subtract my approx. -37dB my notch gives me, this way i got the right numbers
This dosen't however not work if calibration file is not loaded, then the number's will be more than 10dB off.
I accept, that the Cosmic APU is just plain -30dB for getting the correct number's, but i don't get why it's possible without calibration file...
Well i see that REW is not used often here, and i guess that other program's due this differently. - Maybe it's just this which trick's me.
I'am just trying to learn here
Thank's...
Jesper.