Milanraf,
The first step in determining all that needs to be done to the amp is to very carefully draw out a schematic of exactly how your amplifier is wired.
Production changes or not, it is necessary to know exactly how your amp is wired.
Sorry, I know that is a lot of work, but a true schematic is really needed to proceed further.
Otherwise, we all are shooting arrows at the target in a pitch black room (after someone has turned us around about 90 or 120 degrees, give or take).
The first step in determining all that needs to be done to the amp is to very carefully draw out a schematic of exactly how your amplifier is wired.
Production changes or not, it is necessary to know exactly how your amp is wired.
Sorry, I know that is a lot of work, but a true schematic is really needed to proceed further.
Otherwise, we all are shooting arrows at the target in a pitch black room (after someone has turned us around about 90 or 120 degrees, give or take).
I've done it. Here are the results:
The ripple is still there. I no more hear the "Buzz" (less higher order harmonics) but the 100Hz is still there, just a little less.
I've checked and corrected the original schematic with the actual layout from my amp.
Here it is:
I am not sure where the EL34 screen grids go...they are connected to wires that go to the output transformer, but I don't know how they are connected inside. I drew another coil...hope it's correct...
The caps have been painted by Cary, so no labels. And my MM is dead...
Not present, as you can see.
Please see the schematics.
The CHASSIS labels refer to a star center located on the back of the ground screw terminal.
I'll try that whenever I can get out of home. Even RS Components has reduced and partially suspended parcels delivery in Lombardy due to COVID-19. This is not a great period for DIY Audio works
...and not only for that...Not even present!
I'll measure as soon as I can get a new battery for my MM
Gold Aero E34L
Thank you
One way to stop the 5AR4 arcing would be to place a filter choke between the cathode of the 'AR4 and the big 'lytic. That would get rid of the huge damaging peak currents the tube is seeing. You would need to place a small capacitance, in the 20 to 40uF range, directly from the cathode of the 5AR4 to ground, otherwise the choke input filter will give you some loss of B+ voltage. As for a choke, 3 to 5 henries or more at maybe 500mA would do fine. It looks like your chassis is big enough to accommodate such a choke. The choke will also give you additional filtering, too.
Unlike metanastis I am not suffering from arches. Nonetheless that big cap right out of the rectifier is quite unusual and dangerous. And even if it's so big, still there is too much ripple for low volume listenings.
So you mean I'd better add an inductor instead of a resistor, like others have suggested?
So you mean I'd better add an inductor instead of a resistor, like others have suggested?
I don't think it's possible to put it inside: a 3H 500mA from Hammond is quite big:
193N Hammond Manufacturing | Mouser Italia
I could put it outside, between the filter caps. It is also quite good looking.
Anyway, before just give it a try, I'd better evaluate all the options. This Hammond is not exactly cheap...
Here it is:
That 150K value doesn't look right. There's no way a 6SL7 has only a gain factor of less than 2. 150k is too high. You're better off putting the 50K pot in there and dial in the AC balance.
The schematic output transformer shows that the output tube screens are connected to a Floating primary.
I hope there is a DC path back to B+.
Perhaps each screen has about the same DCR to the center tap of the primary.
How close are the output transformers to the power transformer?
Are the output transformer coils oriented at right angles to the power transformer coils?
I bet the chassis is magnetic steel, not aluminum, right?
According to the schematic, all grounds connect to a central point (chassis). Yes?
Lots of chance for ground loops.
For example, the standby switch is connected to the chassis, and not connected (first) to the negative terminal of the 1100uF first B+.
This schematic looks pretty close to the early Cary 807 amp.
You have either a scope or sound card. What is the volts peak to peak of the 100Hz ripple?
I hope there is a DC path back to B+.
Perhaps each screen has about the same DCR to the center tap of the primary.
How close are the output transformers to the power transformer?
Are the output transformer coils oriented at right angles to the power transformer coils?
I bet the chassis is magnetic steel, not aluminum, right?
According to the schematic, all grounds connect to a central point (chassis). Yes?
Lots of chance for ground loops.
For example, the standby switch is connected to the chassis, and not connected (first) to the negative terminal of the 1100uF first B+.
This schematic looks pretty close to the early Cary 807 amp.
You have either a scope or sound card. What is the volts peak to peak of the 100Hz ripple?
Last edited:
I think it is an UL connection, mybe the OT ha 5 wires on primary
The GZ34 with big cap is not a real problem.
P.e. Quicksilver use a big ones after on GZ34 and they works for years without problem.
If you took out he 6SL7, the noise is still there. You can do this without problem.
Another thing, maybe the big cap are lost the efficency,
The GZ34 with big cap is not a real problem.
P.e. Quicksilver use a big ones after on GZ34 and they works for years without problem.
If you took out he 6SL7, the noise is still there. You can do this without problem.
Another thing, maybe the big cap are lost the efficency,
The screens extra coil I drew on the schematic is just my guess. I don't know the exact OT layout. They're enclosed in custom Cary labeled metal boxes and I don't see an easy way to take them out to see eventual labels.
I can confirm this.
This is something I will investigate as soon as I will be able to have a battery for my MM. But, since the last italian government provisions for COVID-19, it's gonna take a long time...
The OTs seem to have 5 wires on the primary, at least they appear so from inside the chassis. I cannot reach the OTs, they are enclosed in Cary labeled metal boxed.
No more tha 2 cm. But they are enclosed in massive metal boxes.
No way to check this
Correct
Please see previous posts. I've updated the schematic with more precise indication of ground points (do a browser refresh: SHIFT + reload or close-reopen).
Only the center tap of heaters coil is connected to chassis, right in the center.
All the other gounds go to the ground screw on the back panel.
I'll try that modification. But will it have effect on ripple noise?
It is difficult to do absolute amplitude measurements with a sound card but I think I can find a way to do a rough calibration with some known source. Let me try.
Feed the amp with sine wave and measure the grids of the power tubes and adjust the pot until they are at equal AC amplitude. For push-pull amp, this is essential.
Now I know why I do not use potted transformers.
You can not tell the coil orientation of them.
You have to test that with 50Hz or 60Hz applied to the power transformer; and later with the choke;
Then put a scope on an output transformer, rotate it, and measure for minimum hum from the power transformer to the output transformer.
Repeat with the choke and the output transformer.
I can see the coil orientation of all my power transformers, chokes, and output transformers.
A test setup for orientation is not needed.
I was practically born with a measurement device in my hand. 8th grade rudimentary scope. 9th grade calibrated current measurements.
Then on from there.
It seems you have several potential sources of hum, at least 3 causes of magnetic coupling, several ground loops, perhaps more items too.
It is hard to determine which is the outer layer of the onion to pull off first, in order to get to the next layer, and on, and on.
I am guessing that the resistor that shows 150k for the phase invertor voltage divider is more like 15k (makes the division closer to the gain of the stage that drives it).
I also notice that the cathode resistors of the 6SL7 are not equal. ???
If the 22k negative feedback resistor is part of that equation, then that cathode resistor would be Larger instead of smaller than the other cathode resistor, but it is not.
You can not tell the coil orientation of them.
You have to test that with 50Hz or 60Hz applied to the power transformer; and later with the choke;
Then put a scope on an output transformer, rotate it, and measure for minimum hum from the power transformer to the output transformer.
Repeat with the choke and the output transformer.
I can see the coil orientation of all my power transformers, chokes, and output transformers.
A test setup for orientation is not needed.
I was practically born with a measurement device in my hand. 8th grade rudimentary scope. 9th grade calibrated current measurements.
Then on from there.
It seems you have several potential sources of hum, at least 3 causes of magnetic coupling, several ground loops, perhaps more items too.
It is hard to determine which is the outer layer of the onion to pull off first, in order to get to the next layer, and on, and on.
I am guessing that the resistor that shows 150k for the phase invertor voltage divider is more like 15k (makes the division closer to the gain of the stage that drives it).
I also notice that the cathode resistors of the 6SL7 are not equal. ???
If the 22k negative feedback resistor is part of that equation, then that cathode resistor would be Larger instead of smaller than the other cathode resistor, but it is not.
Last edited:
By connecting the standby switch to the ground screw, instead of connecting it to the negative of the first filter cap (1100uF), all the transient current causes the ground screw to pick up the voltage from the induced current.
Instead, connect that standby switch to the negative of the first 1100uF cap.
Otherwise, you will notice that the input stages cathodes are connected to that same ground screw, and any transient voltage caused by the 1100uf charging current is applied to the cathodes of the input tubes.
Grids and Cathodes are equally sensitive to undesired voltages, including transients.
Of course, eventually the negative of the 1100uF cap has to connect to the rest of the circuit grounds. But make the center tap travel directly through the standby switch to the 1100uF cap negative first.
Instead, connect that standby switch to the negative of the first 1100uF cap.
Otherwise, you will notice that the input stages cathodes are connected to that same ground screw, and any transient voltage caused by the 1100uf charging current is applied to the cathodes of the input tubes.
Grids and Cathodes are equally sensitive to undesired voltages, including transients.
Of course, eventually the negative of the 1100uF cap has to connect to the rest of the circuit grounds. But make the center tap travel directly through the standby switch to the 1100uF cap negative first.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.