iFi Audio Model "micro iPhono2": Creating Schematic - help needed

[tiefbassuebertr]

iFi Audio Model "micro iPhono2" (iphono-2): Creating Schematic - help needed

Currently I am creating a schematic diagram of this MM/MC RIAA head amp:
micro – iPhono2
iFi iPhono 2 | The Ear
Review ifi iPhono 1 & 2 (und eine kleine Warnung) - Phonovorstufen und Ubertrager - Analogue Audio Association

For all used IC's I don't find out the right type number and thus I don't find the associated datasheets (except U3/U4: HC4066A in the TSSOP-14 outline = MC74HC4066A):

U1: EAA 5SA (SOT23-5, go to last image), unknown, maybe MAX828
U2: OV 4627A - maybe quadversion of Burr Brown's OPA 627 made for iFi-Audio (TSSOP-14, go to first image)
AMR/iFi audio's Content - Page 2 - Computer Audiophile
U5: OQ = H05 or 0Q = HO5 (SOT23-6, go to second image) - unknown.

Who can find out this for me ?
Thank you very much.

This URL's don't provide the wanted information:
http://chip.tomsk.ru/chip/chipdoc.nsf/vc1!readform&view=smd&cat=J&start=1&count=500
http://www.dl7avf.info/charts/smdcode/cj.html#TOC
http://www.marsport.org.uk/smd/mainframe.htm
E96 codes SMD resistors
https://learn.sparkfun.com/tutorials/resistors/decoding-resistor-markings
Transistors:
http://www.dl7avf.info/charts/smdcode/c2d.html
Q1 12T = 2N7002
Q2 2X = 2N4401
Q3 3FW = BC857B/BC557B
Q4 2XZ = ?? maybe 2N4401
Q5 38W = PMBFJ108/J108, Made in China - go to page 2 of 9 under
https://www.nxp.com/docs/en/data-sheet/PMBFJ108_109_110.pdf?
Q6 1BW = BC847B/BC547B
Q7 12T = 2N7002
Q8 12T = 2N7002
Q9 12T = 2N7002
Q10 2X = 2N4401
Q11 NU
Q12 12T = 2N7002
Q13 2X = 2N4401
Q14 12T = 2N7002
Q15 38W = PMBFJ108
Q16 6g = BC818-40/BC338-40
Q17 6g = BC818-40/BC338-40

Diodes
D1 WL = Zener 15V
D8 T4 = 1N4148
D7 SV = Zener 3V6
D9,10 WB = Zener 6V2
D5/D6: J11 923B = Zener, 8V2
https://www.onsemi.com/pub/Collateral/1SMB5913BT3-D.PDF
basics for iFi’s revolutionary ‘TubeState’ design
https://ifi-audio.com/portfolio-view/pro-idsd/
https://ifi-audio.com/portfolio-view/x-xdsd/

 

[billshurv]

You want help reverse engineering a commercial product?

 

[tiefbassuebertr]

You want help reverse engineering a commercial product?

Most important thing for me on a commercial product is in general the wish, to understand the theory of operation in detail. Unfortunately this is impossible without a schematic diagram. OTOH - most exist schematic diagrams from today's available audio devices wasn't release, because this diagrams are confideltial in most cases.

But in this case only the information concerning the used semiconductors U1 and U5 from the internal SMPS regulator with SMD marking code "EAA 5SA" and "OQ = H05" or "0Q = HO5" is important for me (I am already owner of this RIAA preamp device).

 

[tiefbassuebertr]

Unfortunately no support from the experts here - I have find out this:

U1 (Device Code "EAA": Onsemi MAX828 (DC-DC Converter)
https://www.onsemi.com/pub/Collateral/MAX828-D.PDF
datasheet MAX828EUK

U5 (Device Code "OQ", Outline TSOP-6): Richtek R7731GE (PWM Flyback Controller)
https://www.richtek.com/assets/product_file/R7731/DS7731-03QV.pdf
http://www.s-manuals.com/pdf/datasheet/r/7/r7731_richtek.pdf
R7731 - Burst Triple Mode PWM Flyback Controller | Richtek Technology

Under
http://www.s-manuals.com/pdf/datasheet/r/7/r7731_richtek.pdf
I read this:

For device code information, contact our sales representative directly
or through a Richtek distributor located in your area.

But I have found this:
SMD Marking Code (Marking Information, Marking Diagram, Device Code) - go to page C-7 under
https://www.rom.by/files/Richtek_Marking_Code.PDF

Different package naming for the same outline: SOT23-5 = SC-74a = SOT753 = TSOP-5
https://www.nxp.com/docs/en/package-information/SOT753.pdf

The actually supply voltage for the input stages and RIAA stages are create by internal SMPS DC/DC concerter.

This provides the great addvantage, that both the quality and technology (both 50/60Hz or SMPS) of/in the outdoor wall power plug haven't any influence to the sonic character. Even a high proportion of disturbances in the power grid haven't any influence through this approach.
OTOH - the quality of the internal SMPS defines the highest achievable quality in sound.

Maybe the disconnect of the internal DC/DC converter and connect with a shunt regulated power supply version like that one under
The simplistic Salas low voltage shunt regulator
is the right way to enhance additional the sound quality.

 

[nRm]

I watched your first link:

micro – iPhono2

and got very intrigued about this:

"To DC Infinity and beyond. The all new 'DirectDrive® Servoless DC Infinity' circuit has no capacitor and no DC-Servo meaning no noise and no distortion."

Is that for real? No cap, neither DC servo? How is that done? Through insane matching?

 

[PRR]

> Is that for real?

It is not even wrong. Just nice-sounding words. There's no reason a cap or servo should add noise or distortion.

It may be a finely engineered box. But their words-writer was smoking something else.

 

[nRm]

Anyway if there exists an autozeroing topology without caps neiter servo, it is interesting.

tiefbassuebertr let us know, please.

 

[tiefbassuebertr]

I think, a creating of schematic is impossible, because the used marking codes on the transistors are custom made, at least partly.
According the marking codes under
Marking smd
at the MC input stage operates a small signal N-Ch MOSFET (NXP 2N7001/12t) and a small signal NPN BjT (ONSEMI 2N4401/2x) in parallel mode. I doubt very strongly, whether this is possible.
The use of custom made marking for smd devices is a really good prevent for unauthorized clones resp. copies.

 

[nRm]

Can you tell us something about the suppossed autozeroing topology?

 

[rayma]

at the MC input stage operates a small signal N-Ch MOSFET (NXP 2N7001/12t)
and a small signal NPN BjT (ONSEMI 2N4401/2x) in parallel mode.

They're more likely to be in cascode, not in parallel.

 

[tiefbassuebertr]

They're more likely to be in cascode, not in parallel.

Two very important mistakes have occurred:

1) The SMD code for 2N7002 was read incorrectly. At Q1, 7, 8, 9, 12 and 14 there is instead of a "1" (place for first character) an exclamation mark or another special character, maybe for the country of manufacturing. The correct marking code is consequently 2T for 2N4403 / MMBT4403 and not 12T for 2N7002 MOSFET - go to
http://www.onsemi.com/pub/Collateral/MMBT4403-D.PDF

2) At the input MC stage only the pins for base and colector are connected in parallel. Emitter of PNP-BjT goes to the pos. rail and the other of NPN-BjT to the neg. rail.
Creating schematic is a little easier now.

 

[tiefbassuebertr]

Schematic Diagram iFi Micro iPhono 2V1.3

Please note, sign errors are probably still some available.
But there is definitely on the bottom of the iPhono2 at least one missprint at the switch No 3:
The actual function of IEC (Subsonic) and eRIAA (enhanced RIAA) is reversed according the presently connections of PCB-wires.
But I can even not understand the real eRIAA position in the circuit.

The voltage inverter for creating the neg. rail comes later.

Most interesting detail is the so called "Pure Class A Tube Stage" behind the integrated operational amplifier - this approach is not to be confused with a push-pull emitter follower. What is the usual term therefore in colloquial English ?

Which task do these parts have?
1) R81 (from output left to output right)
2) R2-18-19 (3 pcs. 51R in serial mode from 15VDC-input to GND - runs very hot; not showed in the diagrams)

 

[rayma]

Please note, sign errors are probably still some available.

The connections to U2's output buffer are wrong. The NPN collector goes to +V,
and both bases go to R63 and then to the U2 output. Probably the pinout you
used for the NPN is wrong.

 

[tiefbassuebertr]

The connections to U2's output buffer are wrong. The NPN collector goes to +V,
and both bases go to R63 and then to the U2 output.
Probably the pinout you used for the NPN is wrong.

This connections I had expected.
but the pinout under
http://www.onsemi.com/pub/Collateral/MMBT4401-D.PDF
and
http://www.onsemi.com/pub/Collateral/MMBT4403-D.PDF
took me to the circuit shown.
Next days I will check this again.

 

[rayma]

This connections I had expected. but the pinout took me to the circuit shown.

There's no way that can work.

 

[tiefbassuebertr]

There's no way that can work.

Yes, and according the measured voltages arround Q17/16 and Q14/9 I assume, Q17/16 isn't a transistor.
At first glance, only the connection sequence needs to be changed - i. e.. swap base and collector of Q17/16.
But that too can not be right for the following reasons:

1) Voltages at both emitters are 0V, at each base of Q14/9 are -0V58 (-580mV) and at each base of Q17/16 also -0V58 ant not +0V58.
2) Such a deviation of the pin-out sequence from the datasheet isn't conceivable for me

This means for me, Q17/16 is a three terminal adjustable current source, like this under
http://www.nteinc.com/specs/7200to7299/pdf/nte7237.pdf
or
http://www.ti.com/lit/ds/snvs746d/snvs746d.pdf
but in the same outline and smd/smt marking code than the BC818/40.
In the attached pictures you can try also to read the marking code - unfortunately much more difficult than the marking code of Q14/9.

Long story short - the question is, which component is Q17/16 in the SOT-23 outline ?
Keep in mind - schematic was created concerning Q17/16 according this datasheet:
http://www.onsemi.com/pub/Collateral/BC818-40LT1-D.PDF

 

[mkc]

Ok, my guess. Thorsten is both clever and pratical and like historical circuits.

As far as I can read Q16 reads 6G, which I believe is MMBF4393. Q9/2T is MMBT4403. I believe Thorsten has used the buffer circuit which I know from the Kaneda Active Crossover. Have a look at the follower in the link below.

http://www.diyaudio.com/forums/multi-way/113111-sounding-active-crossovers.html#post1367784

I can't say if R63/29 in you drawing is only going to the base of Qr og if both the base and gate of Q16 is tied together after R63/29

Mogens

 

[tiefbassuebertr]

Ok, my guess. Thorsten is both clever and pratical and like historical circuits.

As far as I can read Q16 reads 6G, which I believe is MMBF4393. Q9/2T is MMBT4403. I believe Thorsten has used the buffer circuit which I know from the Kaneda Active Crossover. Have a look at the follower in the link below.

http://www.diyaudio.com/forums/multi-way/113111-sounding-active-crossovers.html#post1367784

I can't say if R63/29 in you drawing is only going to the base of Qr og if both the base and gate of Q16 is tied together after R63/29

Mogens

Thank you very much - I think, your estimation is right.
http://www.onsemi.com/pub/Collateral/PN4393-D.pdf
active crossover Kaneda (image 1)
Filtre électronique triphonique de type Kanéda (JC.Gaertner)

 

[mkc]

Great! Also, I think I would have drawn the MC stage a little different. I would move R22/R38/R55/R37 and C44/C24/C52/C39, from the PSU schematic to the MC schematic. The MC stage is similar to the one QUAD used and is typically drawn with these components together with the rest of the curcuit.

The capacitor and resistor from base to emitter pusle me a bit. The resistors might be an option for trimming the difference in base current of the two transistors.

Mogens

 

[davidsrsb]

2N4401 & 2N4403 always makes me suspect a W.M.Leach based MC front end