diyAudio (
-   Analogue Source (
-   -   Chipomatic balanced input RIAA (

ClaudeG 3rd February 2019 08:58 AM

Hi Salas,

That's a very nice feedback report and a very exciting message, many thanks!

It is also very promising regarding the possible performance and musicality that little board can achieve, congrats!

Following with non restricted enthousiasm


ClaudeG 3rd February 2019 09:25 AM

Hi again,

IMHO, your choice is nice re OP-rolling and on top most DIYer should be able to implement and additional sub-board to play between double and simple Op-amps, that is provided there is some space to do so around the board.

A few questions, if you don't mind...

1) If considering MM use only, with (my) 47k/100uF requirement re load (that's roughly what I need considering cable capacitance), is there a better choice thanAD8429?

2) Both stages are fed directly by the same PS, and I do understand that is also part of the seeked simplicity, but do you believe there is a benefit to feed the stages separately or at least to have some kind of additional buffer / regulation between them?

3) I understand OP rolling is suitable at IC2 level only with FET input types, right? 16 years ago I was very happy with the sonic qualities AD825 (single, 10x times noisier than your AD823)... But perhaps the world has moved on since, LOL, what do you think? I still use these in my DAC, polarised in Class A through... 2SK170 transisitors. Wasn't an overkill nor expensive back then, but I know these are nowadays quite rare and TBH Class A was possibly achievable only with resistors ;-)

Thanks for all this again


Salas 3rd February 2019 10:14 AM

1) The 1nV AD8429 is great for MM and very well behaving in general.

2) The SMD test board is already dual mono in chip regs. Dual mono usually brings a small but noticeable gain in quality. This UBiB shared shunt reg did not leave much to hope for though, forceful and meaty although resolute and smooth. When it will be better wired up with shorter thicker cables it will maybe have the chance to up a further notch. Remains to be seen what happens when doubled up also.

3) AD823 is dual and not quieter than AD825 or OPA134. Its for second stage so the intrinsic noise of IC2 is mostly covered by the first stage's. When you got a socket nothing to lose with rolling. If it works well and sounds well you keep it. If not, you move to another adequate JFET type candidate.

ClaudeG 3rd February 2019 12:06 PM

Many thanks Salas, clear as ever !

Re point 2, I was perhaps misguided by the sonic differences UBiB 1.3 vs SMPS... Reading you again, I do now understand you had an agricultural version without any regulation on your "test board", whereas the "proper PCB" version does already include an UBiB on its board AND THOUGH the need for 2 external Salas UBiB 1.3 is not required, a simple classical linear power supply (tranny + bridge+ caps) could do the trick? And have similar results as yours with UBiB 1.3?

Is my reading correct?

Again many thanks... and defo interested!


Salas 3rd February 2019 12:18 PM

The other style, the SMD board, includes LT3045 LT3094 SMD LDO dual mono linear regs. The big green board with the Mosfets to its side is the Ultrabib shunt linear regulator that was used soon after the SMPS stuff did not satisfy. The prototype green matrix boards agricultural build has no onboard regulators of any type. Just input DC connectors.

vgeorge 3rd February 2019 02:04 PM

Maybe some LME49860 in place of OPA2134 worth trying?

Salas 3rd February 2019 03:06 PM

I don't know about offset and noise when substituting there due to their different inputs technology. Still with its lowish input bias current (10nA) may worth a pop into the pcb to test results.

vgeorge 3rd February 2019 03:07 PM


ClaudeG 3rd February 2019 05:11 PM

Thanks Salas, that's what I finaly understood LOL

ClaudeG 7th February 2019 06:58 PM

Hi Salas,

As said I am very interested in this project and, modestly many years ago, I gave it some thoughts myself although by far not with your knowledge nor pushing it that far.

Therefore, pardon my simple questions, my aim is just to understand better what you are doing (which is for sure state of the art) and of course not questioning your choices or the design.

I have found my old notes again, have forgotten (LOL!) why I made a few choices back then, but anyway this has highlight a few points and I was wondering if you could enlighten me…

1) In my draft, I went for an input load of 47k and 100uF as that’s ideal for MY cartridge given my cable capacity. I see you went for 48k (not sure that makes a lot of difference ?) but no capacitive load ( ?)… is that because one can add it easily or because some calculation is needed to take into account the following chip (that seem to have a very input capacity though) ?

2) I had no way to simulate my ideas, but though back then why are people using so high resistors in the passive filter section, whereas these must generate noise and resonably low values not bring the usual associated problems ? . My choice was also pratical and based on non SMD devices (never soldered one, that will be a first, although I would love a Chipomatic with classical parts better, bare the few ICs) and also quality parts re capacity. To make it short, R1, R2, C1 and C2 where in my project respectively 7.32k, 1.07k, 0.1uF and 0.3uF (combination of 0.1uF Vishay MKP 1837 1% tol, or Wima MKP 10 as I had good experience with these). So troughly 7 times lower R values and admittely higher C values to get it right, but I had no space issues. Were my R values a problem ?

3) I guess there must be quality low tolerance SMD resistors for so few watts, but I don’t know if there are quality SMD caps that can match the above ones, that I do rate highly in terms of sonic.

4) I wanted also to get more gain for my 6.5mV Goldring, to bring the total output to a bit higher than 1V, closer to CD / line level. That is beacuse I have a passive preamp… I like it simple with reasonable gain stages. Of course due to headroom, I must confess I went for 34 dB in the first stage, so quite close to your 36dB, but I went for 32dB in the second stage… whereas my understanding is you limit yourself to 20dB. Why so ? Simulation ? Noise seems still secondary in this stage whith these still acceptable values and headroom / stability still OK… or am I mistaken ? At least it seems easy to play with R5/R6 to adjust this…

5) One of the things that stopped me back then was the need for a proper PCB board, a multilayer with a full groundplane. Of course this can still be done manualy with foils, but can I assume you thought about a proper separate ground plane in your PCB ? I guess that is one of the benefits going small and SMD, as otherwise traditional parts on a traditional perf board as your prototype could be easier to build.

6) I like your Offset correction and PS ideas a lot. I had no DC control and re PS was thinking batteries. Now, with your project, I believe some kind of CRC LPS could be well suited. Some tranny + rectifier in a box, 2m long cables to have some R and some quality caps bypassed by small MKP ones near your board, before regulation, in the main box. Dooes that make sense… or overkill ?

7) I had some standard PS decoupling caps near the OP-amps, 22uF+0.1 MKP… is that still a good idea with a regulated PS and have you made some provision of some kind for this ? Some amps seem to be sensitive in that regard…

Sorry for the many questions in one go. I also undestand our goals might have been slightly different as indeed I wasn’t concerned by cost or space issues, so no real compromise there on my side… but also no knowledge of modern parts you are using and clear SMD avoidance :-)

Many thanks for all your work and looking forward for the final PCB !


All times are GMT. The time now is 02:15 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 17.65%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio