My_Ref Fremen Edition - Build thread and tutorial

Nice build ;)

I suppose you like what you hear... :D

1st impression is that it sounds fast and tight, with clear separation of instruments. Excellent harmonic structure with brass etc. Maybe it's a touch lightweight, but that could be the toroids (used with several amps previously and tend to have this character), or maybe it needs to some use for the parts to settle down.

Actually this was quite a tricky build:
- the 1/4" terminals are rather close to the rectifier diodes and require sleeving and careful fitting
- access to a screw to fasten the 3886 is blocked by other components and it's not possible to use an ally bar to clamp it, for the same reason.

I will probably change it to R-cores and Rev A and a nice box if it become my favourite amp.

Just checking... for Rev A (changing from FE):
- change R3 from 0.5R to 0.33R
- delete C10 and R39

Is that right?
 
1st impression is that it sounds fast and tight, with clear separation of instruments. Excellent harmonic structure with brass etc. Maybe it's a touch lightweight, but that could be the toroids (used with several amps previously and tend to have this character), or maybe it needs to some use for the parts to settle down.

Lightweight? Weird... did you observe suggested orientation for all parts?

Also input cabling could have a role, it seems you recicled pairs from an ethernet cable, isn't it?

VA rating of transformers?


Actually this was quite a tricky build:
- the 1/4" terminals are rather close to the rectifier diodes and require sleeving and careful fitting
- access to a screw to fasten the 3886 is blocked by other components and it's not possible to use an ally bar to clamp it, for the same reason.

Never had problems with faston tabs... they're a bit tight with 35mm caps, though.

The LM3886 should be fixed by a nut over a M3 grub screw:

attachment.php


attachment.php


I will probably change it to R-cores and Rev A and a nice box if it become my favourite amp.

Rev A is an improvement, as r-cores if you can find them now that Selectronic is no longer in activity...

Just checking... for Rev A (changing from FE):
- change R3 from 0.5R to 0.33R
- delete C10 and R39

Is that right?

Correct
 

Attachments

  • IMG_2760.jpg
    IMG_2760.jpg
    279.8 KB · Views: 650
  • Socket-Screws-2_billboard.jpg
    Socket-Screws-2_billboard.jpg
    40.2 KB · Views: 665
Maybe it's a touch lightweight, but that could be the toroids (used with several amps previously and tend to have this character), or maybe it needs to some use for the parts to settle down.

Hi Dave,

Like Dario, I'm a little surprised to hear lightweight. Of course, I out of touch using an older version, but I think the same things probably affect bass (Dario, please jump in and correct me). I see that you are already using the Mundorf power supply caps. That's good and was found to help the bass in early testing.

As Dario implies, with Toroids, a higher VA rating seemed to help bass as well. I've been using 300 VA Antek toroids, which is overkill, but a bigger VA is an option if you can't find R-core.

The other thing I noticed is that it appears you are running a DC coupled input (no input capacitor blocking DC). Normally, that cap and the input impedance resistor define the bass corner frequency of the bandwidth as well as affects the sound character a bit. When you are DC coupled, I'm not sure what defines the low end of the bandwidth, but I think the characteristics of the source device output impedance become part of the equation. I hope someone smarter can explain this phenomenon.

Anyway, I hope you enjoy it, I have.

Jac
 
The other thing I noticed is that it appears you are running a DC coupled input (no input capacitor blocking DC). Normally, that cap and the input impedance resistor define the bass corner frequency of the bandwidth as well as affects the sound character a bit. When you are DC coupled, I'm not sure what defines the low end of the bandwidth, but I think the characteristics of the source device output impedance become part of the equation. I hope someone smarter can explain this phenomenon.

The low frequency bandwidth will be defined by the feedback cap C9.

I run without an input cap, and the low frequency weight is superb. I also use 300va torroids, but also some non standard mods which probably help.

i'd suggest to get used to it for a while, and let your (Dave's) amp/ears burn in.
 
The lightweight sound was not the amp!!!!
I don't need an input cap for a very good reason!

Here's what happened:
- after my usual bench tests to check stability and correct operation I put the amp in my system and took and posted the picture
- after 1/2 hour of listening Dario posted asking for my thoughts on the sound
- I made the schoolboy error of commenting immediately
- after further listening I realised something was really not right. I then looked againa the the schematic and realised the gain of the FE is significantly higher than my usual amp(s)
- I turned up the bass level on my active XO so that the <100Hz matches the main speakers - leanness instantly fixed!

Anyway, apologies for the false alarm. I will not be commenting again on SQ until the amp has clocked up some decent hours and I have done some proper comparisions.
 
Last edited:
The lightweight sound was not the amp!!!!
I don't need an input cap for a very good reason!
(...)
Anyway, apologies for the false alarm. I will not be commenting again on SQ until the amp has clocked up some decent hours and I have done some proper comparisions.

Don't worry, the important thing is that you fixed the 'problem' ;)

Keep us informed :)
 
The lightweight sound was not the amp!!!!
I don't need an input cap for a very good reason!

- I turned up the bass level on my active XO so that the <100Hz matches the main speakers - leanness instantly fixed!
A

Dave,

I'm sorry. I didn't mean to suggest that you need an input cap. I was just rambling on about things that were different from standard and possibilities that would affect bass.

And Spartacus is right about the low end cutoff frequency in a direct coupled FE amp. I knew that, but wasn't thinking clearly. I should think before typing next time.

Anyway, I'm glad you found the source of the issue. Enjoy your experiments.

Jac
 
Jac,

I also didn't mean to sound aggressive in my reply. Anyway I am confident to run without an input cap because:
1) The FE doesn't receive any signal at LF
2) I am confident my XO does not output dc (at least <2mV)
3) The DC protection will hopefully save the day if it's ever needed

Interim report (can't resist) after a few hours of listening:- Wow! This is a very, very good amp!
 
Hi
Amplifier sounds very good, but I'm not getting rid of the noise of the background.
Can not find fault with the location of the components.
Seeing that the C9 has the opposite direction than on yours on my PCB V1.7
Marking on pcb is turned around.Is that right?
I have shielded RCA cables.
Should the screen be connected to the return (outer jack.) in both ends or just at the signal source?
 
Seeing that the C9 has the opposite direction than on yours on my PCB V1.7
Marking on pcb is turned around.Is that right?
I still can't understand why you don't post pics...BTW

Don't worry about C9 orientation, different revisions could have different orientation but the circuit is always the same, the important thing is that you followed PCB's markings.

I have shielded RCA cables.
Should the screen be connected to the return (outer jack.) in both ends or just at the signal source?

It depends on cable's type, if it's a coaxial cable like Andrew already wrote yes you MUST connect both sides.

If it's a balanced one with screen used as single-ended then screen can be connected only at one side because you have the two signal wires.

Also in this case a pic would make things faster and easier.
 
...................If it's a balanced one with screen used as single-ended then screen can be connected only at one side because you have the two signal wires...............
This advice is usually wrong.
It is better to connect the screen/shield of a screened twisted pair to both chassis/enclosure at the cable's two ends.
Connecting at both ends removes most of the risk of the two enclosures being at a different voltages.
Connecting at both ends helps with attenuating the High Frequency interference.
 
Hi
Can not say I completely understand this with the ground loop
and other phenomena that cause noise in amplifiers.
Youtube and forum have many suggestions for cure,
but everything will affect the sound quality.
I think that insulators, tranformers, filters of all kinds will destroy
some of the things that make MY_Ref amps special.
The only thing I have at hand that stops audible background noise is
small caps at input, but I do not understand the reason why this happens and the sound quality is affected.
Do these caps (1uf and 4,7uf) function as a filter and for what?:confused:
I am sorry that I do not have pictures of the test setup, but I have not
working camera or mobile phone camera at this time.
Thanks for the help and advice I have received.:)
 
Last edited:
The only thing I have at hand that stops audible background noise is
small caps at input, but I do not understand the reason why this happens and the sound quality is affected.
I used to have some noise as well with my standard Gainclones without an input capacitor.
The MyRef Fremen's are much much quieter than the standard Gainclones...I wondered if that is because of the input cap as well...(I'll have to try it tonight without input caps and will get back to you)
 
Hi

Do these caps (1uf and 4,7uf) function as a filter and for what?:confused:
I am sorry that I do not have pictures of the test setup, but I have not
working camera or mobile phone camera at this time.
Thanks for the help and advice I have received.:)

Hi rg,

The input cap, specifically C13, is there for safety, to block DC which might damage your speakers. Combined with R13 (100k), C13 forms a high pass filter with a corner frequency of 1.6Hz. Ideally, with a corner frequency that low, it would have no affect on the sound of the amp. However, if a larger value cap were used, it would possibly degrade bass.

But nothing is ideal. In real life, all of the signal goes through C13, so there is an affect on sound because C13 is not perfect. This is why choosing a good C13 is important to sound quality in the FE.

Some people are confident of their system and are willing to gamble that nothing will fail and damage their amp or speakers. Plus, the FE has a speaker protection circuit. For example, Dave S is confident and runs his FE without C13. That's OK, it's his choice and he has lots of experience. For less experienced builders and the more cautious folks like me, using C13 is smart and the FE can still have excellent sound.

By the way, the reason I use C13 is that I had a DC problem. I had a Squeezebox Touch that was a digital music source for my stereo. My pre-amp is direct coupled (no blocking cap), so when the Squeezebox failed and put 1.5 VDC into my system, C13 kept everything safe.

Jac
 
Hi
Thanks for reply!
I should have had pictures.Sorry!
I am using a CD player and preamp as signal source. (Abrahansen stereo)
I use XLR cables from CD to preamp and balanced out from preamp.
At the FE-amplifier I connect pins 1 and pin 3 (ground and negative)
I do not know why this goes, but it gives the same good sound as with
RCA directly from the CD player which also has volume control and single ended outputs.
By good sound, I mean when I turn up volume so much that I do not hear the background noise of the stereo system. It is not good sound in delicate soft songs.
I use "best boom parts" in my FE amps.Cap in C13 is in place.(Hot-to Hot on PCB?)
It's when i connect an extra C13 between HOT and return input wires that this noise disappears. I can not use this for anything as it changes the sound negatively.
Like putting a blanket over speaker. What's really going on when I'm testing
with these extra caps (1uF and 4.7 uF) Are they like a filter?
As you can see, I can understand very little about the construction of electronics and how the parts work together, but I hope to learn a little when I build these Amps.
I think the noise must come from the signal source, or not?
CD-Preamp and my 2 Abrahamsen amps have no background noise.
I would like to try with an I-pad on battery to see if it gives less noise.
Have tested many versions to ground the stereo without luck.
Has failed to influence the noise in any way.
Maybe I have to buy input isolator (ISO-MAX or ART DTI transformer / isolator)?
Will noise go through these insulators too Or affect the amplifier negatively?