• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

IEC 60318-4 Clone Coupler, is it worth it?

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
5,805
Likes
10,505
Location
Monument, CO
Two SMAs, one for each side, so loosely coupled.

I measure uV signals all the time using single-ended connections. That is how most RF/mW/mmW systems are designed...
 

IVX

Addicted to Fun and Learning
Forum Donor
Joined
Mar 3, 2019
Messages
908
Likes
2,084
Location
South of China, SHZ area
Earfonia
thank you for the taobao link, I'll buy right now the same coupler. I have an idea to play with the new 10mm planar driver, a previous version of that was used in P1 and modified in P2, so now it is kinda 3rd version of that driver. It is a lot more sensitive vs P1 but seems even less bass, so I want to find out the bass there with the clone of IEC711 318-4.
 

Skeptischism

Active Member
Joined
Sep 6, 2019
Messages
228
Likes
113
Custom SMA?

Ha!! that would be good! but no, more like the sort of thing used on test equipment and specifically for usage—and evaluation of— Fully differential Amplifiers (aka FDAs.) instrumentation amplifiers, and ADCs. there are many different types, if you care to look.

One per phase.

Remember, this is for connection to/from prototypes and evaluation of my own PCBs. not just THD at 1khz. An example, tuning out IMD humps in ESS DACs, or for tuning the networks in wide bandwidth amplifiers. very low noise, very low output impedance power supplies. etc. the sort of thing that could be affected by slightly different, or inconsistent termination between different times connecting the DUT. Each of my boards will be built to connect to it, just like so. with a consistent mounting pitch/standard.

Professional enough for engineers?

Dont really need to satisfy connection to commercial amplifiers and anyway I would knock up an adapter cable if the need arose. After testing is complete, XLR, or whatever desired connector will be added. I could also knock up a very small PCB with a PCB mount XLR, to simulate the finished design.

Also, not all approaches to notches are as simple and cheap as an OPA1612 and a handful of caps/resistors. Samuel Groner's Passive Notch using many parallel, matched MELF and many parallel thin film caps, dont come cheap. Obviously it needs to be interfaced to an amplifier for gain. you could argue its not necessary, but hey, no harm in over building in instruments IMO and we have the parts already.

Anyway mate, dont take me the wrong way re your design, its just doesnt meet my specific need. I cant stand mini jacks, its a pet hate. I wish they would go away in general :facepalm:. we can do better. i'm definitely opinionated about it, it's not personal. I particularly like your software from what ive seen, too. well done. If you do a great job on that part with your system as well, I may look past the mini jacks (or replace them ... lol). Sounds like it'll be a great tool for people and we might even get one eventually for secondary, for portable stuff and just because I have a problem with buying electronics.

OK, I think that's enough on this tangent. I'll update the thread and post my own once i've started building my DIY CIEMs
 

Attachments

  • EV1HMC199AMS8.jpeg
    EV1HMC199AMS8.jpeg
    47 KB · Views: 27
  • FDA.jpg
    FDA.jpg
    89.7 KB · Views: 24
Last edited:

Skeptischism

Active Member
Joined
Sep 6, 2019
Messages
228
Likes
113
there are also SSMA, which are smaller and would allow tighter coupling, but I'm unsure if rigid cables are available. the coupling is a trade-off vs consistency in setup when chasing harmonics around and anyway its not like XLR is particularly tightly coupled. I may end up with something different in the end if it proves to be a problem.
 

IVX

Addicted to Fun and Learning
Forum Donor
Joined
Mar 3, 2019
Messages
908
Likes
2,084
Location
South of China, SHZ area
Skeptischism, I never saw thin-film caps yet but resistors including MELF(I use that stuff in Cosmos too). Probably a very expensive thing, who knows(google doesn't about thin-film caps) )) My APU notch uses MELF and C0G caps to get -140-150db harmonics vision i.e. the same precision level as AP SYS2***. Good enough to me and 99,999% of my users, including potential ones ;) As an example, this is the result table of my Cosmos DAC output 7.5Vrms 1kHz sine after 1kHz LPF(included in the Cosmos DAC as a switchable option). AP10 means AP with forced 10V range, AP20 AP forced to 20V range to have less distortion, APU means my notch I mentioned above, harmonics #2, 3, and 5:
AP10 -144.21 -143.37 -150.16
AP20 -140.27 -134.89 -148
APU -144.27 -137.7 -152.17
And the screenshot for that, please note AP's THD+N reading after APU is -100db, the notch factor is -30db, hence the residual THD+N of Cosmos APU -130db, and THD-N -140-150db. Should I give up my APU and redesign that for bloody expensive thin-film caps aka flux-capacitors passive notch? ;) Don't worry, I'm not defensive, just kidding.

2021-07-28_17-28-34.jpg
 

Skeptischism

Active Member
Joined
Sep 6, 2019
Messages
228
Likes
113
You google fu is not great .... ;) I dont believe you googled it at all, because all I did was enter just a few words ('thin film silicon capacitors') and came up with this and the results just keep coming; pages and pages of them .... It is synonymous with silicon capacitors (the 2 are very similar and often used interchangeably; google tells me perhaps the marketing department is behind the 2 names) . It is the same process that would be used for any tight tolerance on chip capacitor (in other words, you ar already using them and dont realise it). I would expect to trim with them if required, but the bulk is ceramic NP0/C0G. they are only available in very small values

https://forum.digikey.com/t/silicon-thin-film-capacitors/12836
https://www.mouser.com/new/passive-components/capacitors/silicon-rf-capacitors-thin-film/n-5g95
https://encyclopedia2.thefreedictionary.com/thin-film+capacitor
https://smartech.gatech.edu/bitstream/handle/1853/42741/Wang_Yushu_201112_Mast.pdf
https://www.sciencedirect.com/scien...c419c946&pid=1-s2.0-0026271467900819-main.pdf
https://passive-components.eu/silicon-and-silicon-substrate-based-integrated-capacitors/

you say you arent defensive, yet it is clear you are ...
 
Last edited:

Skeptischism

Active Member
Joined
Sep 6, 2019
Messages
228
Likes
113
So either you are playing silly buggers, or ... i'm not sure. thats 2 things in a row you have tried to make me look silly with, when all you had to do was look.. I'd rather not enter a pissing contest. Its funny, because you are trying to 'punch down' and missing.

No, you shouldnt give up and do what i'm doing. Nobody would buy it. I only wish I had an AP. This is DIY and these obsessive bits are kinda personal. i'm not selling, I just fetishise high spec parts, for my own satisfaction and you should learn that you cant please everyone.
 
Last edited:

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
5,805
Likes
10,505
Location
Monument, CO
DonH56, you have no choice but if you can use the advantages of balanced inputs, which are easily implementable in the audio-band, what is the reason to go with unbalanced?

I seem to have stepped into a pissing contest that I don't really care about.

Technically, most of my current work uses differential connections (at GHz frequencies), and I have used them in the past for RF as well, it depends on the application.

Going differential requires additional circuitry and/or more expensive op-amps, adds power, may or may not improve noise and distortion (in general should cancel even harmonics, increase SNR by 3 dB, and better reject common-mode noise, but gains in practice depend upon implementation and signal levels), requires more expensive cables and connectors (plus more space for the connectors), etc. And you have to run differential end-to-end to fully realize the benefits, natch. For most folk unbalanced works well and noise and distortion are inaudible. Of course there are benefits to balanced (differential) operation but whether they are generally audible and worth the extra cost and effort depends upon the system IMO.

I have a few articles about the advantages of differential operation but understand it is not in general required for acceptable (if not outstanding) audible performance. I would definitely favor them in a production environment with so many components and so many long mic and line runs plus all the power, lighting, and other stuff in the area. For reproduction (playback) I look harder at the cost/benefits given balanced components generally come at a price premium (though that is changing somewhat). Given a choice among comparably-performing products, I would always choose a component offering (properly-implemented) balanced inputs, but unless I had a clear need for differential I/O it would not be a dominant factor in my purchase. So I suppose the key words in your statement are "if you can use the advantages" -- if I could, and particularly if I need them for noise rejection or whatever, then sure.

FWIWFM - Don
 
Last edited:

Skeptischism

Active Member
Joined
Sep 6, 2019
Messages
228
Likes
113
@DonH56 It would generally be higher noise in differential, no? distortion generally marginally better on even harmonics, but yeah unless you are dealing with long cables, not really worse or better. As you say, all about the implementation and for measurement I believe you need both (or at least its desirable). Some of the quietest amps i've ever seen and heard are single ended. That being said, at least in my case for a DAC, creating a SE output would mean a further stage and be slightly more expensive. If i'm using a balanced dac, I typically use a balanced output and a high performance FDA is about the same price as an opa1612.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
5,805
Likes
10,505
Location
Monument, CO
@DonH56 It would generally be higher noise in differential, no? distortion generally marginally better on even harmonics, but yeah unless you are dealing with long cables, not really worse or better. As you say, all about the implementation and for measurement I believe you need both (or at least its desirable). Some of the quietest amps i've ever seen and heard are single ended. That being said, at least in my case for a DAC, creating a SE output would mean a further stage and be slightly more expensive. If i'm using a balanced dac, I typically use a balanced output and a high performance FDA is about the same price as an opa1612.

Simplistically, in a differential circuit noise is uncorrelated so combines in an RSS fashion: Ntot = sqrt(N1^2 + N2^2). The signal doubles so Stot = (S1 + S2). The net result in twice the signal but only sqrt(2) the noise for a net increase in SNR of 3 dB (20*log10(sqrt(2))). The caveat is that, if you do not or can not use the extra signal level, then any SNR benefit may be lost. Assuming the noise is low enough using SE circuits, natch. Put another way, you should always gain 3 dB in SNR, but the noise floor may be higher since there are two gain "paths" and generally more circuitry. This is most often seen (IME) when people add pro gear to consumer signal chains that do not need (nor want to see) such high signal levels.

Ideally even harmonics are suppressed (fixed the typo in my original post) so net distortion should be significantly lower, but again depends upon implementation.

Some (many?) components use differential circuits internally.

If you have a balanced component already then might as well use it (I would). I do not quite get why you'd need to add a stage for a SE output unless using one side of the differential output causes too much of a performance penalty (that is often true, requiring a single-ended to differential converter at the output).

I am not against differential circuits (note not all balanced circuits are fully differential) by any means. I prefer them, but there are plenty of single-ended designs that have quite respectable performance, so it is one factor in my decision and not usually a deal-breaker. For audio, the exceptions when differential I/O is a requirement for me tends to be for live sound reinforcement and recording, studio work (which I am no longer doing), and applications like my rear subs where I wanted to avoid ground loops.
 
Last edited:

IVX

Addicted to Fun and Learning
Forum Donor
Joined
Mar 3, 2019
Messages
908
Likes
2,084
Location
South of China, SHZ area
why you'd need to add a stage for a SE output
For an economic reason only, my solutions are cost-effective by conception. Yeah, my notch uses simple and cheap C0G caps,(actually, Samuel Groner also recommends C0G for his passive notch), but I don't want to adjust 4 trimmers instead of two, and increase the parts count twice.
PS: I found ICP/CCLD definition, and will be ready tomorrow to try the coupler.
p134-f3.gif
 

Skeptischism

Active Member
Joined
Sep 6, 2019
Messages
228
Likes
113
yes, to be clear, I meant noisier for the same output level, but I see what you mean re correlation. I was thinking it would add in series and if its for the same output level this would produce higher noise.

I do not quite get why you'd need to add a stage for a SE output unless using one side of the differential output causes too much of a performance penalty (that is often true, requiring a single-ended to differential converter at the output).

As you deduce, because using one side of a balanced output is often suboptimal, meaning a BAL-SE stage is required.

Just in case Earfonia hadnt seen the thread I was referencing. Its remarkably FUD free for over there except for some occasional cable talk, but that is faily rare. Home Made IEMsHome Made IEMs. some of the guys are getting pretty impressive results. some are truly beautiful objects.
 

Attachments

  • 9559595.jpeg
    9559595.jpeg
    185.9 KB · Views: 23
  • 9614164.jpeg
    9614164.jpeg
    295.1 KB · Views: 24
  • 9670996.jpeg
    9670996.jpeg
    669 KB · Views: 23
  • 9690407.jpeg
    9690407.jpeg
    77.1 KB · Views: 24

Skeptischism

Active Member
Joined
Sep 6, 2019
Messages
228
Likes
113
even built with 'only' C0G the BOM for his notch is not cheap in time (matching) or money. thats what Chris is doing with his PCB (he bought 2, of which i'm building one, we live a few hundred km apart, so we need 2 measurement setups and I dont blame him, c0g are superb, but i'm looking for an excuse to use the silicon caps :) I actually bought them for input filtering.
 
Last edited:

Skeptischism

Active Member
Joined
Sep 6, 2019
Messages
228
Likes
113
Yep, I decided the same. it is doable to DIY everything, but choosing my battles with limited time available to me.
 
OP
Earfonia

Earfonia

Active Member
Joined
Feb 21, 2019
Messages
200
Likes
225
Location
Singapore
Earfonia
could you shortly explain how to connect that coupler to an ADC, as I see you using B&K preamp? Any info about that? Seems I'll get the coupler tomorrow, so I wanna be ready.

The coupler microphone as other condenser need bias voltage that's the B&K unit is for. Like the 48V phantom power for the common music recording condenser mic, the ICP / CCLD condenser measurement mic requires less voltage and less current, around around 22 V ± 2 V (typ. 21 V) with constant current around 3 to 4.1 mA (nom. 3.55 mA). Something you can DIY with a simple circuit. We can actually use audio interface XLR input that usually equipped with 48VDC Phantom power, step it down to 21V and to drive the ICP microphone and the mic pre-amp inside the coupler. With that you can skip the B&K unit.

I use the audio interface pin 2 (hot) and 3 (cold) to connect to the coupler RCA output. RCA center pin to XLR pin 2, and ground to XLR pin 3. The idea is to avoid ground loop with the audio interface.
 
Top Bottom