Group Buy: Mauro Penasa MyRef Rev C kit

[Audiodoc]

Recheck solder joints. Recheck if the lm318 is correctly seated. Recheck if the lm318 is getting positive and negative voltage. Recheck dry solder joints at the grounding sites.

Do the relays click. Check the dc offset before the relay. Try a TL072 dual op amp instead of the lm318.

For increasing and decreasing voltages of the amp power supply for different loads I use a manual step up step down transformer to regulate mains.

 

[Audiodoc]

Pics and wiring diagram to check for ground loops being a cause of hum. Biasing resistors will get hot and should be soldered with a gap between the resistor and board.

A 20 0 20 is preffered if one needs to play for both 4 ohm and 8 ohm speakers.

 

[jls001]

Thanks for the pointers, Audiodoc.

I had a bit of an accident while trying to measure the supply voltage to the LM3886 chip. I was very carefully measuring the V+ and V- (4 and 5) wrt Ground pin (7) and accidentally shorted pins 4 and 5. The chip is a goner. The light and sound show was thankfully short, but heart rending all the same. I've found a shop that sells the TF version that I destroyed. Hopefully it should be an authentic piece.

I do have a few TL072 opamps amongst my spares. Will try as suggested. Got spare LM318s too just in case. Which other opamps work at this position? I must be having a pair of OPA2132 (dual) and some LM series opamps too. Or is the compensation tailored for LM318 only?

Also, both relays do click when power is switched on.

I have already bought 1 kOhm resistors of 6W rating and will swap out.

Will reflow solders though I don't see any obvious flaw when I inspected the populated boards before fitting them into the cabinet.

Here's hoping that I haven't damaged any other components

 

[jls001]

To add: I also bought a few polystyrene caps to try out at the input. I want to stay capless throughout but I think it's better to be safe at this point.

 

[Audiodoc]

Most opamps will work but the lm318 is recommended for the best sonics. You should try to avoid measuring voltages at the chip to prevent short circuits. Genuine lm3886 can be purchased at diyaudiokart.

Polystyrene caps more than 2.2 microfarad is good for blocking DC.

Do the speaker protection relays click.

 

[Audiodoc]

Do a run outside the cabinet prevent any short circuits.

 

[jls001]

First I'm quickly making a hanging lamp to light up my work space better. It was lack of adequate lighting that did me in. My (cheap) LED table lamp has given up its ghost.

Let me replace the faulty chipamp, add input caps, ground the input shield to chassis, tidy up the wiring where possible, and I'll report back after powering it up again.

 

[jls001]

OK, chip swapped out, 2.2 uF caps inserted, using TL072 in place of LM318, all power supply clips checked for proper insertion and contact, input wires rechecked, speaker wires checked, minimising loop as far as it can go. 6W rated 1 kOhm resistors used. Shield grounded to chassis at RCA input socket and shield left floating at board end.

Powered on - one channel relay clicks. The other channel does not click.

Clicking channel:
AC 1 and 2: both 26.1 VAC, measured with reference to center tap.
TL072 pin 4 to pin 7 supply voltage: 7.8VDC (doesn't look normal).
LM3886 pin 5 to 7: 35.6 VDC (this is roughly the expected regulated DC voltage, so it seems fine).
DC offset at speaker terminal with input shorted: -0.502V (way too high).
LM 3886 chip gets a bit warm to the touch. The two newly inserted 1kohm/6W resistors are warmer than the chip. TL072 chip is not warm. I touched few other components too but felt no warmness on any.

The non-clicking channel (where 3886 was replaced):
AC 1 and 2: both 26.1 VAC
TL072 pin 4 to 7: 24.9 VDC (shouldn't the opamp also get the full regulated DC voltage of about 35V?)
LM3886 pin 5 to 7: -60.6 VDC (this is nearly twice the regulated voltage. Shouldn't V+ referenced to ground pin 7 be about +35V? Does this indicate a short somewhere or some blown component?)
The LM3886 chip is only very slightly warmer than ambient, the two 1kohm resistors are warmer than the LM3886 chip, but much cooler than the other channel. TL072 chip is at ambient temperature.
DC offset: between -18 to -20 mV.

No smoke, no spark

I didn't connect to speakers as one channel would not work any way as the relays are not activated.

Any idea how to proceed from here? The channel that didn't click is the one where the 3886 was replaced.

I didn't dare to measure pin 4 to 7 on the chipamps as the tip of my multimeter's probe is not thin enough to not short to pin 5.

I had fitted the transformer and associated wiring in the cabinet so it is inconvenient to power it up on the bench.

 

[doors666]

if memory serves me right, 12v zeners were used to power the opamps. so the first one isnt getting the correct voltage.
What are the unregulated dc rail values.

 

[Audiodoc]

No the opamps voltage is lower 12 0 -12 or 15 0 -15 volts.

Both channels have a problem. Can you please post some pictures. Something seems to be wrong in the regulated supply of amp 1 and unregulated supply of amp 2.

 

[Audiodoc]

The op amp voltage is to be measured between pin 4 negative and pin 8 positive which should be 2 x 12 v ie 24 volts.

 

[avesbilal]

Opamp rail voltage is +/-12 in Myref X2 I think it should be same in Myref rev C also.

 

[jls001]

Pics here:

1. Top view

2. View of the secondary AC distribution tag block:

3. View 1 of the clicking channel:

4. View 2 of clicking channel:

5. View 3 of same channel:

6. View 4:

7. View 5:

8. View 6 of 6:

 

[jls001]

@doors666: I'll refer to the circuit diagram and trace the voltage after the power supply caps.

@avesbilal: thanks for the pointer. Now I know what voltage to expect for the opamps.

@audiodoc: I was measuring the opamp supply voltage across the -ve input pin 4 and positive input pin 7. Now I'm confused: should it be pin 7 or 8? I referred to the pinout diagram of the LM318 and it says pin 4 and 7, respectively for negative and positive rails. I understand that the voltage across the V- and V+ pins should be twice Vcc, so about 24V.

Also, can someone kindly indicate what DC supply voltages to expect at pins 4 and 5 of the LM3886? I have given my observations for pin 5 (positive rail) to pin 7 (ground) for both channels.

Thanks for your kind supports :thumbsup:

I was referring to the pinout here: LM318 datasheet, Pinout ,application circuits Precision High-Speed Operational Amplifier(15MHZ OP AMP)

 

[Audiodoc]

You are correct it is 24 v between pin 4 and 7 for 318 and 4 and 8 for tl072.

I opened my amp for some measurements.

Input AC 230V transformer output 24.5 0 24.5v

DC approx 35V 0 -35V

Between pin 4 and 7 24 volts for the opamp

DC at output of two channels 5.6 and 5.7 milli volts with input open.

Somethings wrong in the supply section, dry solder, small shot circuit, unclean board, or non functional equipment.

The first time I powered my myref I had a problem of the relays clicking continuously and I am pretty good at soldering 20 plus years. The same did not happen with the tlo74s. I reflow ed everything with a lot of flux. Cleaned with isopropyl alcohol and checked components in the compensation section to achieve stability with 318.

In you case it is the supply section. The protection relay should click once and dc offsets should be below 100 mv and led should light up.

 

[Audiodoc]

Are your 10000 nippon chemicons original. I have half a dozen fakes with me. Can you replace temporarily with a better capacitor.

Solder some led to each board.

 

[jls001]

Did some measurements based on the schematic by Russ White at Post #3 here.

On the channel with the clicking relay, observations are:

1. R1 (1 KOhm) facing -ve terminal of C3 (10,000 uF/63V) to ground: -35.0 VDC
2. R1 facing the ZD1 zener: -3.4VDC (must be -12VDC as per the schematic)
3. R4 (1 KOhm) facing +ve terminal of C8 (10,000 uF/63V) to ground: +35.1 VDC
4. R4 facing ZD2 zener: +4.2VDC (must be +12V as per schematic)
5. ZD1 cathode side to ground: -3.4 VDC, same as 2 above since they are same point on the circuit (must be -12V per schematic)
6. ZD2 anode side to ground: +4.2 VDC, same as 4 above since they are same point on the circuit (must be +12V per schematic)

The voltages dropped by the zeners look wrong. It's too much. Could the zeners be bad?
The same measurements are repeatable at pin 4 (-3.4V) and 7 (+4.2V) of TL072. BTW, pin 8 measures +3.2VDC.

Regarding +/-35V, it approximately matches calculation (secondary 25.8VAC * 1.414 -2 = 34.48). So I think the rectifier diodes and power rail caps are doing their duty, but the zeners are dropping way too much voltage. Nearly 9V is lost.

Meanwhile, the other channel has refused to power up though it was showing readings earlier in the evening. The AC secondary supplies are reaching the board, but similar measurements post rectification as above are all zero. There is no output from the rectification section. Could be the MURs or the two power caps. Or does the LM3886 have some protection trigger that gets triggered under high DC or short?

The power caps are Nippon-Chemicon KMH received as part of the premium kit from Linuxguru so I don't doubt their authenticity.

 

[jls001]

Are your 10000 nippon chemicons original. I have half a dozen fakes with me. Can you replace temporarily with a better capacitor.

Solder some led to each board.

I might have some 10K caps from another project that never took off. I will have to dig deep into my pile

Any specific colour recommended for LED?

One board is completely reflowed. The other board got interrupted in the middle so I'll reflow and clean with alcohol. No harm trying. This board is temperamental, given me lots of the blues. It better sound good

 

[Audiodoc]

You can use the tlo71 or ne5534 single opamp to test not the tlo72 I posted by mistake.

May be the zener are bad. The cap will be authentic if linuxguru has supplied them.

The rectified voltages are fine.

 

[Audiodoc]

Troubleshoot the first board. The second definitely has something wrong. You may have to desolder the lm3886 and then check if the rectifiers are fine. They are hard to kill.

Any led will do I use green and red on each board. This post may help you.

http://www.hifivision.com/diy/11495-group-buy-mauro-penasa-myref-rev-c-kit-31.html