F6 Illustrated Build Guide

thanks for all the replies. the jericho horn is a super efficient design isn't it? my speakers are "only" rated at 87db.

i'm currently also thinking about a new pre amp and i really like the doug self pre and the b1 by nelson pass. two very different approaches, i know. obviously the b1 only makes sense if no additional gain is needed.
 
Simple stoopid stepped attenuator into My F6 driving Tannoys which are ~94db.
Even with a Simple diy sk170 Phono.. Loudness is a non issue as in I've never been brave enough to max the volume.
Let alone able to stay in my listening room at anything past 2/3rds rotation on the pot.
 
I wonder if some of you clever and experienced people here can chime in with some suggestions. I recently started playing around with some FFT measurements using REW, since i had some audio interfaces and enough curiousity (and helpful suggestions from xrk971) to do so. And since the F6 was on the top shelf and easy to access this was my first measurement victim.

So i did some measurements (just one channel for now) and found quite some 50hz and harmonic peaks the FFT spectrum at approx 1W into 4ohm (my speakers are 4 ohm so that seemed like a good idea).
attachment.php

Am i right to think that the presence of all the harmonics above 100 hz suggest mains hum and not ground loop?

Ok, the distortion for the amp itself fits nicely with the stated specs (0.05%), but i am curious how i might bring the mains hum down, and was hoping some of you had some suggestions. I am also unsure about the ripply stuff above 1k, i just tested the output of the interface i used to generate the signal and that seems very clean plugged straight back into its line input.

First of all, here is the FFT spectrum with nothing attached to the amp inputs:
attachment.php


Now, i should say the hum does not bother me for my normal listening, it is only audible when you put your ear right up to the speaker. And i guess -100 dbfs is not a terrible level for mains hum, even if its not great. So no panic, but i figure i might as well use this as a learning opportunity to see if there is anything i can do to improve this.

A few details about the build (and picture) here: F6 build from a newbie, with some (hopefully) helpful pointers for fellow newbies

And an added question to this (which may be completely unrelated and irrelevant to the subject but i am curious): The transformer is a 400VA from Toroidy, audio grade, and has a "0" wire in addition to the two primaries and two sets of secondaries. I connected this to the star gnd point together with the safety earth from the IEC inlet and also. But i since noticed that many transformers do not have this "0" wire, and i wonder if i did the right (and safe) thing with it. So any input on this also welcome!

Best,
Silas
 

Attachments

  • 2v, 4 ohm.png
    2v, 4 ohm.png
    53.4 KB · Views: 684
  • no input2.png
    no input2.png
    51 KB · Views: 669
Last edited:
Ok tried the snubbers with 47nF+10ohm with and without additional 100 nF across secondaries. No difference. I guess im stuck at this point since i dont own an oscilloscope so really cannot do any measurements (math or no math) to find the right values.

Edit: just read through the quasimodo notes, and realized that according to Mr. Johnson, the Cs in the Cs+Rs should be larger than the parallel Cx - i will try and flip things around tomorrow and report back.
 
From what I see the noise is 80db down from the fundamental which is plenty good enough in my world. The 120hz is a harmonic of the 60hz signal. Since you do sit with your ear next to your speaker I would not do anything. It will take more effort, time and money to improve on what you see on a graph that you cannot really hear from your listening position. Nothing actually to gain. Nice looking signal from the fundamental on with a touch of 2nd and 3rd harmonic. Should sound really good.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
@silasmellor

Start from simple things like do the input signal's wiring in shielded coax if its not. First see it best dressed away from any AC power wiring. Its not only the AC inlet primary side. The secondary exit points and wires can be throwing quite a field no matter if the toroid has some form of magnetic screening for its most part. If rotated and the harmonic noise changes in FFT then there is a gotcha.

On the other hand if its simply due to high PSU ripple, just try measure it on the rails with a DMM in ACV mode. A CRC or a CLC or a capacitance multiplier could solve it.

Then again if its mainly due to some low intensity ground loop in the build itself or in its interfacing with the rest of the system... it will take enough conceptual guessing and experimental investigation.
 
@2picodumbs,
Yeah I suppose more filtering probably wouldn't hurt!

@wdecho
You are absolutely right, this buzz is totally irrelevant to the listening experience, and since upgrading from a naim nap 155xs I couldn’t be happier – it sounds fantastic! The only reason I am chasing it is to try and understand the issue and learn something new :)

@Salas
Thank you for those suggestions. Tonight I spent some time rearranging the AC side, mainly to improve safety in case of bad luck and loose screws, but I also reouted the inlet AC line further away from the business end, without any visible change in the FFT. The input wires are mogami coax, routed as far from the AC side as possible (4U deluxe case) – essentially like shown in 6L6’s illustrations. I actually also rotated the toroidal a bit (about 30 degrees) during this process, but I don’t see any obvious change. I don’t know if that is enough of a rotation to warrant anything though.
My DMM measuring AC on the outputs with no signal is varies between 0-1 mV, I don’t think it is sensitive enough to pick up anything major. Though 2pD also suggested more filtering, and I am on the low side of the recommended specs with 4x15mF/rail, I lean towards keeping that in mind for my next build and being happy with what I have now :)

Thanks for the feedback and suggestions, and for making diyaudio such an awesome place!
 
Last edited: