Group Buy: Mauro Penasa MyRef Rev C kit

[hemanth.hg]

Hi,

Finally I fully assembled the amplifier, when I powered it on, Im facing the following issues

Channel 1: When I switch it on, first it makes the relay triggering sound and then after a couple of seconds it makes screeching sound intermittently, I checked for any loose connections, and did re-soldering, but the problem still persists. When I pause the source there is no screeching, but when I start again, it starts screeching intermittently.

Channel 2: When I switch it on, there is no relay triggering sound, but when I switch it off, exactly after 6 seconds I hear the relay triggered on and triggered off( two beeps) and I observed the LM3886 getting warm.

Im using a transformer Primary:0-230 Volt Secondary: 24-0-24 V/300 VA. I checked the transformer by connecting to the AC main and it is reading 23.5 and 23.6 respectively on the secondary.

Please guide me on how to debug. what could be the problem?

I have attached the snaps.

Thanks in advance.

 

[Audiodoc]

I had the same problem that you have on the second board.

Troubleshooting:

!. Don't attach any speakers yet.
2. Take a multimeter. Red pin to the speaker out (near output resistor, white box type) and black to ground (besides the relay on which you have connected the red wire. that is ground Do one thing the blue and yellow should be 24v each attach a spade connector in the holes between the two nichicon capacitors second ground so the pin beside the relay is a free ground for measurements. Check my pictures a few pages back). MM on DC output. Check the DC offset. It will be high that is why the protection circuit comes into play and the relay is not clicking. Remove the LM318 and try again.

Use the LED as an indicator. Now when you turn it on the relay will click normally.

Problem will be: A low value of C32 / bad LM318 / Dry joint around 318.

3. For the screeching channel. Check DC offset and report.

Don't worry there will be a minor problem but attach bigger pictures or send some pics to my mail.

Check some pics here:

http://www.head-fi.org/gallery/album/view/id/131889/user_id/261754//page/40

 

[hemanth.hg]

Hi Audiodoc, Thanks a lot, I will try out the steps you have provided.


Thanks,
Hemanth.

 

[Audiodoc]

You are welcome. And in DIY I hardly have done any project that worked the first time. There is always some thing to learn from our mistakes.

I have noticed one difference in your boards. You have C7 in place which is not there in Rev E. I had one more correction the C32 in the picture is actually C34 (checked it) it should be 47-100pF for stability of Rev E. Lower is better for sound but poorer for stability.

In Rev E C10 is 27pF. Your board also doesn't have a C23 besides pin 1 of LM318 (10nF in Rev E while 47nF in Rev C). In Rev E R42 is shorted with a wire bridge while C32 is 330pF (150pF in Rev C).

Kindly send few larger pictures and which version you are building.

Gist as posted by linuxguru: Rev C C10 = 22pF , C32 = 150pF , C34 = 10pF , R42 = 47 ohm / Rev E C10 = 27pF silver mica, C32 = 150 to 330pF FKP2, C 34 = 47 - 100pF silver mica.

 

[Audiodoc]

Tell me the values of the capacitors from what i can see the C10 is 22pF C 34 is 27pF. What is your C7. You should remove it and add a capacitor at C23.

 

[hemanth.hg]

Thanks again for the guidance.

The one I'm using is Rev C, version 1.4.

I checked the DC offset as per your suggestion, the values are :

Channel 1(screaching noise): 1.2V
Channel 2(delayed relay no sound): 31.6V

I even swapped the LM 318, but that didn't solve the issue.

 

[Audiodoc]

These are too high. As already described the stability of LM318 is dependent on the capacitors (compensation). Try removing C34 as in the picture I sent you and try the DC offset. I can guide you with my working rev E but then you will need some components but beware the rev is very finicky.

 

[hemanth.hg]

I removed the C34 from one of the channel(no sound one), it reduced the DC offset from 31.6 to 27.8. The problem still persists.

 

[Audiodoc]

You will have to get few capacitors from linuxguru to be placed at c34 and possibly c23 for stability. You should also check the DC between pins 4 (black multimeter pin) and 7 (red multimeter pin).It should read 2 x 12V = 24 volts.

Check http://www.ti.com/lit/ds/symlink/lm118-n.pdf

 

[hemanth.hg]

I checked the DC between the Pins 4 and 7, it shows 24.3V. I think it is fine.
As per one of the post on diyaudio, a member suggested to check the DC between the pins 9 and 10 on the LM3886, when I verified it, it showed 0.6mV, not sure if it fine.

Any ideas on what does the above values suggest?

 

[Audiodoc]

Yes that is fine equals +-12V on which the 318 runs. Did you purchase a cheap tl071 opamp.

For the 3886 will check its datasheet will mail you the same.

 

[linuxguru]

Good suggestion - replace the the LM318 with a DIP-8 TL071. If it works with a TL071, it's a stability issue that may be solved with an LM318 replacement. If it does not work with a TL071, the problem probably lies elsewhere.

Note that you need a 8 ohm load, or SPiKe protection on the LM3886 may keep activating.

 

[exljp]

I am almost thru with populating one board, got all parts. I need to power it up. I have a doubt about the power connection to the R-core transformer. It has got four wire leads in the primary winding. See the attached picture. There is black, white and red together and a Green one. So the black,white and red to be connected together to the positive and the green to the negative of the Ac outlet?

 

[Audiodoc]

Connect the 0V black and 230V red wire to the mains insulating the other two. This will be the primary.

 

[hemanth.hg]

Yesterday purchased TL071 and LM318 (two each). Tried various combination of of these on the two boards

Channel 1 (the channel with screeching sound): I tried new LM318, it didn't change anything. I used TL071, now the music is coming without screeching, but when I increased the volume of my ipod touch >50%, it starts screeching, so screeching is missing only for low volumes.

Channel 2(the channel without delayed triggering and no sound): I tried TL071, no changes. I used new LM318, now the relay started triggering properly. When I connected the speaker, then noticed the same old behavior, then switched on the board without connecting to speaker which initiated the trigger and then connected the speaker, now the music comes from the speaker, the sound level is low and it is not stable, the heat generated from the LM3886 is too much, the heat sink is getting too hot.

How can I proceed from here?

Thanks.

 

[linuxguru]

Channel 1 (the channel with screeching sound): I tried new LM318, it didn't change anything. I used TL071, now the music is coming without screeching, but when I increased the volume of my ipod touch >50%, it starts screeching, so screeching is missing only for low volumes.

It looks like SPike protection is kicking in - check the DC resistance of the speakers. It should be at least 6 ohms for an 8-ohm rated speaker.

Channel 2(the channel without delayed triggering and no sound): I tried TL071, no changes. I used new LM318, now the relay started triggering properly. When I connected the speaker, then noticed the same old behavior, then switched on the board without connecting to speaker which initiated the trigger and then connected the speaker, now the music comes from the speaker, the sound level is low and it is not stable, the heat generated from the LM3886 is too much, the heat sink is getting too hot.

This is instability, aka HF oscillations. On both channels, check all the components and compare against the BoM/layout diagram. Pay particular attention to similar-looking components like the Wimas and the small MFRs. 47R looks a lot like 470k, which looks like 47k; 22k looks like 12k, etc.

 

[Audiodoc]

My My Ref Rev E also has the same problem with my 6 ohm wharfedale 10.1s even though I have a secondary voltage of 22V on my toroid. The spike protection makes it impossible to play at high volumes however the sound is ok at normal listening levels so I don't want to replace the toroid with a 20-0-20 one. My ref needs 8ohm speakers. You can try putting two speakers in series and try channel 1 to check if it is a problem with the speaker resistance.

Channel 2 has some minor assembly error check components then reflow soldering.

 

[quad]

Also, during testing, use cheap local speakers.
My pair of 8 ohms test drivers has lasted me years of testing different
amps, and cost me under Rs 100 a pair.

 

[doors666]

Is this an issue with rev e, on diyaudio, people (incl russ white) have reported using 4 ohm speakers with rev c with no issues, even mauro penasa has stated that it should not be a problem.

 

[linuxguru]

Is this an issue with rev e, on diyaudio, people (incl russ white) have reported using 4 ohm speakers with rev c with no issues, even mauro penasa has stated that it should not be a problem.

If the voltage rails are low enough (less than +/- 28V) and the heatsinks are large enough, you can run 4 ohm speakers on either Rev C or E. Rev E with C34=100pF is as stable as Rev C.

I would however recommend staying with 8 ohm loads and no higher than +/- 32V rails for optimal sonics.