QRV-08 Build Problems/Questions

[chinook9]

I am building a QRV-08.  I am not an engineer but I have built a few kits.  The QRV-08 is my first attempt with SMD components.

BMF is helping me with this project and I thought I would document my project progress/problems here to share with others.

I am building and testing each section as I complete it. I tested the PS after the components were installed and all tested well.

I have installed the Super Regulator components and started testing.

On powering up, three of the LEDs lit but H3 did not. I decided to test/repair this rail first.

I read the schematics, tracing from Br3 to follow the signal and took measurements.

IC14 (right channel positive rail) doesn't look right. 
Output (center tab) and Vout (big one) are both 23.92 volts.
Input (right tab) is 25.42 volts (looks good)
Adj/Gnd (left tab) is 23.48 volts.

By comparison, IC7 has the following measurements.
Output (center tab) and Vout (big one) are both 16.44 volts. (looks like they could be good)
Input (right tab) is 25.42 volts  (looks good)
Adj/Gnd (left tab) is 15.18 volts. (looks like it could be good)

We are trying to determine if IC14 is bad, or are there other problems that cause these readings?  IC14 was very easy to install and looks good.

I was very careful installing the LEDs and I think I got it right, but the voltage drop across H3 is 1V.

[peranders]

About H3, is it mounted correctly? You can test this with using diode test with your voltmeter. The LED should send out a weak light. Voltage across E39?

What is your output voltage?

Can you tell me the voltages at IC11, pin 2, 3, 6 and 7?

[chinook9]

While taking the measurements, I decided to reflow the IC11 pins  The soldering job on this IC was my worst and didn't look good.  Upon powering up H3 was lit brightly. 

Voltage across E39 is 14.71V, E38 is 14.82, R39 is 14.71, R38 is 14.76

Measuring output voltage at the output block is 1 millivolt in each both right and left channel.

The voltages for IC11 are:

Pin 2 6.95V (measured at E57)
Pin 3 6.95 V (measured at E51, G47, and G45)
Pin 6 7.11V (measured at V30 pin 1)
Pin 7 13.61V (measured at E43 and G43)

The voltages for IC15 are:

Pin 2 6.98V
Pin 3 6.98 V
Pin 6 7.19V
Pin 7 13.67V

The voltages for IC2 are:

Pin 2 6.95V
Pin 3 6.95 V
Pin 6 7.11V
Pin 7 13.60V

The voltages for IC6 are:

Pin 2 6.95V
Pin 3 6.94V
Pin 6 7.12V
Pin 7 13.61V

How do these look?

Note that I have added a couple of attachments for reference, especially so that you can see what components are installed.

[peranders]

It seems to work now.

[chinook9]

I finished installing all the amplifier components.  I got the following test measurements with jumpers installed at J3 and J6 but no jumpers installed on J1,J2,J4, or J5

IC1 Right channel
Pin 2 at E29 -0.023V  (voltage across E29 0.028V)
Pin 3 at E28 -0.023V (voltage across E28 0.000V)
Pin 6 at pin 6 -4.48V

Voltage across E1 0.028V
Adjusting P2 did not change this measurement.

Offset -0.027V measured across channel output and ground


IC3 Left channel
Pin 2 at R29 0.000V (voltage across R29 0.000V)
Pin 3 at R28 0.000V (voltage across R28 0.000V)
Pin 6 at pin 6 -9.00V

Voltage across R1 0.045V
Adjusting P1 did not change this measurement.

Offset 0.000V measured across channel output and ground

All of the larger 100uF/35V capacitors were warm but not hot.
LEDs were consistently glowing brightly.

I thought I would check with you to see if the problem is clearly apparent to you, before I start checking placement of components and reflowing soldering joints.

I also researched testing of transistor Vbe but I have been unsuccessful in determining how to do it.

Any recommendations would be appreciated.

[peranders]

You should be able to trim down the input offset. Have you tested to move the jumper from J5 to J4 or vice versa?

Since you have DC servo you should obtain the offset of the used opamp plus some contributions the input bias currents of the opamp. If you use OPA134 or some other JFET opamp those errors should be small.

All in all it seems to work. Have you listened to the amp?

[chinook9]

Thank you!

I indicated that "Voltage across R1 0.045V" and "Voltage across E1 0.028V." these may actually be negative because I'm not sure I had the red/black probes placed correct across the resistors.

I have not tried with any jumpers to adjust offset.  I'll try different jumpers and see what I can do with P1 and P2.  I assume I'm trying to get to 0.000.

I am using the OPA134. Can I get the required offset from the specification sheet?

I haven't tried to listen yet.  I may try that later.

[peranders]

You should adjust the compensation current so it cancel out the input bias but the most important thing is that the DC servo output is within it's limits so you get a good regulation of the offset control loop. In other words: you should have a jumper on either J1 or J2/J4 or J5.

[chinook9]

To get to the desired jumper and trim pot setting, I decided to take a comprehensive set of measurements for each channel and each jumper
for that channel.  I took measurements for each jumper for each channel with the trim pots (P1 or P2) Fully CCW, 0900,1030,1200,1330,1500, and Fully CW.

I did this, recording each measurement, until I found measurements that were very close to 0.000VDC across E1 or R1 and DC offset of 0.000V across the outputs for the channel.

Using this method the best measurements were found using J1 jumpered and  P1 at 0900 and J4 jumpered with P2 at 0900. 

I then fine tuned P1 and P2 and ended up with Right channel +0.007V across E1 with DC offset -0.002V and Left channel 0.005V across E1 with DC offset 0.000V

I was surprised how much the voltage can change with just a small rotation of the trim pot.  On one measurement a Fully CCW to Fully CW movement of the trim pot resulted in a 1.359 change in the voltage across the resistor and a 5.60V change in DC offset.

I listened to a few of my favorite songs and I am amazed.  The clarity, detail, and imaging are the best I have heard (used HD800, AKG K812, and Oppo PM2).  My greatest surprise, however, was the dynamic range and the ease with which the amplifier drives the headphones.  I listen to a few cuts at a high volume and the amp appeared to work effortlessly and did not heat up at all... slightly warm but not hot.

P-A, you are to be congratulated on designing an amazing amplifier.   

[peranders]

https://www.neurochrome.com/sjostrom-qrv08-headphone-amp/

Here you can see some measurements made by "tomchr". We have very good results although they can be improved further but this requires a test setup for 28000 dollars (or more!)!

I'm glad that you are pleased. I would be very happy for small review and some pictures for my Facebook page. Send the text and the pictures directly to me in that case.

[chinook9]

I will be glad to send a review and pictures.  It may take a couple of weeks before I get it in an enclosure but I will do it.

I had read the Neurochrome page and considered building one of Tom's amps.  His amp is probably very good also but costs significantly more to build.

Based on Tom's page, the single concern I had about the QRV-08 was the power, but there is no problem in this respect.  The HD800 requires power and the QRV-08 pumped Pink Floyd's very low frequency organ tones flawlessly. 

Thank you again!