[Norway]

Hi

Would this be the correct place to share and ask questions regarding problems with Cornet 3 DIY kits?

If so I will probably have a few questions...

Rgds
HR

[Norway]

Well, a simple question wouldn't hurt anyway.

Input impedance = 47K
Output impedance = 470K

Remove R7 (47K) and input impedance should be infinite, right?

[hagtech]

Yes, the input impedance is roughly 47k at zero volt bias.  This is in addition to maybe 1uA of grid current (I hit the sweet spot in biasing where it should be negligible) and perhaps 40pF.  Output impedance is a little less than 1k above 20Hz.  At DC it is 470k. 

jh

[Norway]

OK, this might become a bit lengthy...

So I finished the Cornet and measured the main voltages throughout the board. 9V was about 8.8. 210V was about 200V, 190V was about 180V and so on. Meaning that all the "high voltage" points were about 10V under the documented values. Yellow LED's lit up and everything looked great though.

Connected it to the record player and amp and there was a lot of pops and hiss and general noise from both channels, but one channel quickly settled down while the other one continued. Did a quick measure of voltage on the output and input of both chanels and found that the right channel had wildly fluctuating voltages sometimes reading over +/- 1V. I also had several 100mV on the input of the right channel.
Removing the power from the Cornet I also measured the input impedance and found the left channel at 46.6Kohm and the right at about 38Kohm and rising up to 42Kohm. Concluded that there must be a bad soldering or component somewhere around the input of the right channel and found that I could measure a short circuit over C7R. I didn't have an equivalent capacitor, so I put the project to rest for a few days.

When I had a equivalent capacitor to C7R i control measured before de-soldering it and found that it was good...? Without replacing C7R I connected power and measured the voltage on the output and input again and found that there was still a lot more DC (fluctuating voltage) on the right channel than on the left channel, but not as wild as earlier.
So I connected the Cornet to the record player and amplifier and played a few LP's with it. Sound was normal but it was clear that the right channel still had an issue with noise, but not as bad as on first try.

New round of measurements without power connected and with tubes removed. Input impedance right channel ca 44.4ohm and left channel still 46.6Kohm. removed R7R (47K) and measured input impedance to ca 1.5Mohm and falling before settling  at about 1.34Mohm. Did the same on left channel and could not find any impedance, i.e. infinite.
Concluded that there must be a short circuit somewhere between signal input and the first 12AX7 grid but could not find any. Removed R8R as well and measured again between the grid connector on the right channel to ground, still 1.34Mohm between grid and ground. Now I am really confused since the little stub of copper between grid of the first 12AX7 tube socket and R8 couldn't possibly be shorted to anything else within the circuit that could find a path to ground.

Anyway, since left channel was good I thought I could figure it out by doing the same procedure. Remove R8L and measure and so on. To get better access I now unscrewed the ground plug and took of the backplane of the box and before starting I attached R7L again to measure input impedance and check that I had 46.6Kohms as before, but once again things have changed. I now had more or less the same as the right channel, about 44.5Kohms. Detached R7L again and found high impedance between ground and grid on left channel, although 2-300KOhm higher than on the right channel.

Now I am wondering if I have dreamt some of these measurements since things seem to change every now and then? And I also wonder if I have misunderstood the circuit diagram when I assume that grid of the first 12AX7 tube shouldn't measure impedance to ground when there is no tube in the socket and with R8 de-soldered?

B.t.w. I am using a ST micro TIP50 instead of the Fairchild one supplied because one of them were missing and I haven't received a replacement yet. But from what I can tell the spec is the same and the voltages are good throughout the board, so I don't think that is of any importance.

[Norway]

After more work at the test bench I have come to believe that the high impedance I measure between grid and ground (with no power, no tubes and no R8) just is what it is. I cannot understand it, but both channels have this high impedance now, so the most logical explanation is that I have measured the left channel wrong a 100 times before I got it right.

With that said I have now turned my attention to the most obvious difference between the channels, which is that they have different brands of the TIP50 NPN transistors. Measuring the collector voltage I can see that there is some difference and that the "good channel" has a stable 89.1V TIP50 collector voltage while the "bad channel" has a voltage at 94.4 that is quite unstable and fluctuating as much as down to 93.7V to 94.2V. Since this is close to the actual output of the Cornet I assume that this is what is causing the noisy and fluctuating signal on the right channel output.
The question then is if it is the ST Micro TIP50 that is causing this or something else in proximity of the TIP50.

Also changed the 12AX7 tubes with two backup Sovtek 12AX7WA tubes and found that even with the strange TIP50 collector voltage the noise in bad channel got much weaker. Not that I think that the Mullard tubes following the Cornet has anything to do with the noisy channel, just an observation.

Well, think I am closer to the core of the problem now. Can do a few more measurements around the TIP50 and see if I can verify what causes the unstable collector voltage.

[Norway]

Here is some measurements done today

Code: [Select]

SOVTEK 12AX7
R-Ch                   L-Ch
AU7 Gr-G2 =  93,8V     AU7 Gr-G1 =   86,7V
AU7 Gr-P2 = 185,6V     AU7 Gr-P1 =  185,9V
AU7 Gr-K2 = 101,1V     AU7 Gr-K1 =   94,7V
AU7 K2-P2 =  85,8V     AU7 K1-P1 =   92,8V
R19R      =   0,8V     R19L      =    0,8V
D2R       =   1,9V     D2L       =    1,9V
Q2R Gr-E  =   8,3V     Q2L Gr-E  =    8,3V
Q2R Gr-B  =   8,9V     Q2L Gr-B  =    8,9V
Q2R Gr-B  =  98,7V     Q2L Gr-B  =   91,8V
Q2R E-C   =  90,9V     Q2L E-C   =   84,1V

And the same for the Mullard tubes

Code: [Select]

MULLARD 12AX7
R-Ch                   L-Ch
AU7 Gr-G2 =  92,3V     AU7 Gr-G1 =   93,3V
AU7 Gr-P2 = 188,1V     AU7 Gr-P1 =  187,4V
AU7 Gr-K2 =  99,3V     AU7 Gr-K1 =  100,9V
AU7 K2-P2 =  90,1V     AU7 K1-P1 =   87,8V
R19R      =   0,8V     R19L      =    0,8V
D2R       =   1,9V     D2L       =    1,9V
Q2R Gr-E  =   8,3V     Q2L Gr-E  =    8,3V
Q2R Gr-B  =   8,9V     Q2L Gr-B  =    8,9V
Q2R Gr-B  =  96,4V     Q2L Gr-B  =   98,2V
Q2R E-C   =  88,6V     Q2L E-C   =   90,4V

As I said the noise was more or less gone with the Sovtek tubes, so I changed back to the Mullards and found that the noise was now gone there as well. Seems like things are settling down, which is good.
Assume the notable voltage differences from left to right channel with the Sovtek tubes are because the tubes are used and in a unknown state. Also the voltage differences form left to right channel around the TIP50 with both sets of tubes can be because I use one ST Micro and one Fairchild, but the difference is now so small that it's probably just coincidental differences in precision of tubes and other components. Will however change the Fairchild TIP50 to another ST Micro just to check while I am waiting to receive the Fairchild TIP50. Then again, if it seems to work just fine I will probably just keep the ST Micro's in.

One question though. The circuit diagram has a high voltage input of 210V while mine only has 200V, so the high voltage throughout the circuit in my Cornet is always about 10V lower than in the diagram. Can this have any effect on the performance of the Cornet or is it within the operational limits?

[Norway]

And by all means, if there are other people out there building Cornet's it would be interesting to see what voltages you measure at the main high voltage spots given in the circuit diagram.

[hagtech]

Some lower cost DVMs don't read very accurately.  Having the high voltage within 10% is fine.

jh

[Norway]

So, just to put my mind to rest I changed both TIP50's to ST Micro, and it didn't make any difference voltage vise. Then again, as I said the hiss and pop that was very audible in the right channel at first is now more or less gone. I can hear a little hissing if i put my ear close to the speaker when using the Mullard 12AX7, but with Sovtek and Golden Dragon there is nothing.

Anyway, reading a few other threads about noise in form of hiss and pops for the Cornet 2 I have turned my attention to the heater voltage. For the Golden Dragon tubes I measured about 6.15V and 6.05V and for the Mullards I measured 6.0V and 5.95V, and that might be a bit low. Could it be that the large plates of the Mullards has an effect here?

Although I am happy with the sound now I think I will try to lower the resistance of the heater resistors to around 8.7-8.9ohm just to see if it has any effect on the Mullard tubes. I am going to order some tubes of different brands to do some tube rolling anyway, might just as well add some resistors to the mix.

[Norway]

Incredible bass reproduction in this thing, rock solid and punchy. Definitely the best phono I have ever owned in every way, I am addicted.

Next step, build a simple reverse RIAA filter and test how it sounds with a CD player/DAC as source. Oh, and I have of course ordered a few tubes I think can compete with the tubes following the Cornet, but I think that will be tough without spending a serious amount of money.

Don't know what makes the Cornet reproduce bass the way it does, but if it could be done similarly  in a tube buffer for use with standard sources like CD players and DAC's that would be fantastic.