Here's something you could do with that unused half of the 6DJ8 -- reduce distortion as much as possible by using it in a mu-follower. (Not really necessary; just thought it was aesthetically pleasing.)
--
That's not far off from your idea of using an SRPP there. Mu-follower is more for reducing distortion than for max voltage swing, so seems to me to be a better fit as an input voltage amp.
--
That's not far off from your idea of using an SRPP there. Mu-follower is more for reducing distortion than for max voltage swing, so seems to me to be a better fit as an input voltage amp.
Attachments
Last edited:
Try 18k 3W for the plate resistor, 300R for the cathode resistor. Or in the cathode, use a red LED and a 1N4148 small signal diode in series. Cathodes of the diodes (banded ends) towards ground. That would put about +2.4V at the 6DJ8 cathode.
With those values and a +249V B+ there, you should see +105V on the plate and about +2.4V on the cathode. 8mA through the 6DJ8.
--
Last edited:
This looks good, and yes, very similar to the SRPP. I'll give this a shot probably.
Thanks!
Blair
Wow, I guess I should have been able to figure this out. But George Tubelab to the rescue:
"You can not run two bridges off of one transformer configured in the manner that you have them. It presents a short across the transformer through two diodes. Follow the electricity from the bottom of the winding up through the diode closest to the (bridge for negative DC) into ground, out of ground on the top (positive DC supply) bridge through the upper left diode and right back to the top of the transformer winding. This will make the transformer verry unhappy."
There it is. I still have trouble understanding it completely, but these things sink in very slowly. I'll ponder it a while...
"It presents a short across the transformer through two diodes. " That much is clear.
--
I've been doing some reading.
Miles Prower posted some info on 807 (same as 6L6 but 5-pin base and has a plate cap). Miles stated that these tubes have a harshness in the upper mids/trebles most likely caused by generation of odd order harmonics, and so require a little feedback to smooth them out. At one point he liked to use a little plate-to-grid feedback around the output tubes (a la Schade), and add a little global NFB after that, to taste.
The global NFB will be very easy, as we're tapping the output transformer secondary and feeding that signal to the LTP's grounded grid.
The "partial plate" or Schade feedback is more difficult. I'm not sure how to choose the value for the return resistor from the 6L6 plate to its grid. I also don't know if the circuit we've come up with is any good for that type of feedback. Apparently you want a high gain, high dynamic plate resistance driver tube. The 6CG7 is neither of those, at least not by tube standards. We do, however, have some gain to burn.
If anybody out there is watching:
Would this proposed circuit benefit from some Schade feedback around the output stage 6L6's, or is it only suitable for global NFB?
Would a 12AT7 (higher rp and mu) LTP work for Schade feedback in this circuit?
Thanks.
--
PS - Can anyone point me to an easy to follow set of instructions for determining the value of the plate-to-grid resistor in this type of feedback network? I do have the original Schade paper for the 6L6, which I'll read again. I noticed that Schade put a capacitor in series to block DC, but many here do not. Is the cap not necessary at all?
--
Miles Prower posted some info on 807 (same as 6L6 but 5-pin base and has a plate cap). Miles stated that these tubes have a harshness in the upper mids/trebles most likely caused by generation of odd order harmonics, and so require a little feedback to smooth them out. At one point he liked to use a little plate-to-grid feedback around the output tubes (a la Schade), and add a little global NFB after that, to taste.
The global NFB will be very easy, as we're tapping the output transformer secondary and feeding that signal to the LTP's grounded grid.
The "partial plate" or Schade feedback is more difficult. I'm not sure how to choose the value for the return resistor from the 6L6 plate to its grid. I also don't know if the circuit we've come up with is any good for that type of feedback. Apparently you want a high gain, high dynamic plate resistance driver tube. The 6CG7 is neither of those, at least not by tube standards. We do, however, have some gain to burn.
If anybody out there is watching:
Would this proposed circuit benefit from some Schade feedback around the output stage 6L6's, or is it only suitable for global NFB?
Would a 12AT7 (higher rp and mu) LTP work for Schade feedback in this circuit?
Thanks.
--
PS - Can anyone point me to an easy to follow set of instructions for determining the value of the plate-to-grid resistor in this type of feedback network? I do have the original Schade paper for the 6L6, which I'll read again. I noticed that Schade put a capacitor in series to block DC, but many here do not. Is the cap not necessary at all?
--
Hi Ron,
I'm back in the saddle, and got my 6l6 tubes. I will try to get it wired up on the drivers this weekend sometime.
To get a rough idea what my B+ is actually going to be, you or anyone looking at this thread think there is harm in applying power without a driver stage and biasing the power tubes?
Thanks,
Blair
I'm back in the saddle, and got my 6l6 tubes. I will try to get it wired up on the drivers this weekend sometime.
To get a rough idea what my B+ is actually going to be, you or anyone looking at this thread think there is harm in applying power without a driver stage and biasing the power tubes?
Thanks,
Blair
There should be no problem with powering up the output tubes, but I suggest powering up with no tubes at all and set the bias a bit negative of target.
Shut down, install the output tubes and re-apply power.
The output tubes should bias up on the cool (safe)side. You can then adjust bias for target cathode current.
Shut down, install the output tubes and re-apply power.
The output tubes should bias up on the cool (safe)side. You can then adjust bias for target cathode current.
OK,
I checked my voltages. I did not bias it because I have a hum I need to locate the source of.
Be fore that though, here are some issues.
The plates for not drivers are around 100V.
I only have 2.25v on the cathode of the 6CG7.
The G2 of the 6L6 has 271v on it while at low bias, the plates are at 266v.
Any ideas?
Thanks,
Blair
I checked my voltages. I did not bias it because I have a hum I need to locate the source of.
Be fore that though, here are some issues.
The plates for not drivers are around 100V.
I only have 2.25v on the cathode of the 6CG7.
The G2 of the 6L6 has 271v on it while at low bias, the plates are at 266v.
Any ideas?
Thanks,
Blair
Did you take those voltage measurements with tubes in, or tubes out?
Did you measure how much plate+g2 current the 6L6's were drawing? If you have something like a 10R resistor in the cathode of the 6L6, you can measure the voltage across that, and that will give you your plate+g2 current draw.
Another way to do it is to measure the DC resistance of the transformer primary windings from center tap to plate. If they measure something like 100 ohms from plate to OPT center-tap, then a drop of 7V would mean 70mA per plate.
We will want to adjust the B+ to the 6CG7 to make it so that its plate voltage is more like +130V. That +100V isn't so far off.
Do you have a working schematic for the circuit so far?
--
Did you measure how much plate+g2 current the 6L6's were drawing? If you have something like a 10R resistor in the cathode of the 6L6, you can measure the voltage across that, and that will give you your plate+g2 current draw.
Another way to do it is to measure the DC resistance of the transformer primary windings from center tap to plate. If they measure something like 100 ohms from plate to OPT center-tap, then a drop of 7V would mean 70mA per plate.
We will want to adjust the B+ to the 6CG7 to make it so that its plate voltage is more like +130V. That +100V isn't so far off.
Do you have a working schematic for the circuit so far?
--
Hi Ron,
Tubes in. Here is the schematic I worked from, but I put 39K on the plates of the 6cg7, so I might want to go back to the schematic value of 20-25K.
Here are my voltages, and yes I know the 6L6 tubes are running pretty hot. I will have to tinker with my bias circuit to drop the negative voltage.
6CG7
Plate 99.6V
Cathode 2.3V
6DJ8
Plate 107.2V
Cathode 2.5V
6L6
Pins 2,7: 6.66vAC
Pin 3: 286V
Pin 4: 290V
Pin 5: -19v
Pin 8: .778V
[/IMG]
Tubes in. Here is the schematic I worked from, but I put 39K on the plates of the 6cg7, so I might want to go back to the schematic value of 20-25K.
Here are my voltages, and yes I know the 6L6 tubes are running pretty hot. I will have to tinker with my bias circuit to drop the negative voltage.
6CG7
Plate 99.6V
Cathode 2.3V
6DJ8
Plate 107.2V
Cathode 2.5V
6L6
Pins 2,7: 6.66vAC
Pin 3: 286V
Pin 4: 290V
Pin 5: -19v
Pin 8: .778V
OK, I think putting 39k in the 6CG7 plates is causing problems. It's making the 6CG7 plate voltage really low, which makes it so that the LTP won't be able to swing much voltage. We want that stage to swing as many volts as possible, and then we can use up some of that swing for negative feedback.
If you can get the voltage at the 6CG7 cathodes to be +4V, that would make it more likely that you'll be able to swing +/- 28V, which will allow for 3dB of gNFB to add to your local plate to grid feedback on the 6L6's. You could even reduce the 6CG7 plate resistors to 15k, which is often done with 6SN7's.
Most of the voltage gain comes from the first stage, then some from the second stage. The second stage is the phase splitter/push-pull driver.
Voltages are so low in this circuit that you could easily use a 12AT7 LTP to drive the 6L6's. I know that would leave empty tube sockets, but it's the power transformer that is calling the shots here. If it's only going to provide +270V, then that's it.
The advantage of using a 12AT7 LTP is that it has higher mu and higher gm, so will balance better and provide enough volts out from a single stage. There's also the advantage of getting rid of one stage and its RC time constant. Simplicity is good.
--
If you can get the voltage at the 6CG7 cathodes to be +4V, that would make it more likely that you'll be able to swing +/- 28V, which will allow for 3dB of gNFB to add to your local plate to grid feedback on the 6L6's. You could even reduce the 6CG7 plate resistors to 15k, which is often done with 6SN7's.
Most of the voltage gain comes from the first stage, then some from the second stage. The second stage is the phase splitter/push-pull driver.
Voltages are so low in this circuit that you could easily use a 12AT7 LTP to drive the 6L6's. I know that would leave empty tube sockets, but it's the power transformer that is calling the shots here. If it's only going to provide +270V, then that's it.
The advantage of using a 12AT7 LTP is that it has higher mu and higher gm, so will balance better and provide enough volts out from a single stage. There's also the advantage of getting rid of one stage and its RC time constant. Simplicity is good.
--
Cancel that 12AT7 idea. Voltages are so low that a 12AT7 won't be able to work. A 6DJ8 might work. Maybe a 12BH7A would work. Even a 5687 or something like that. With so low a B+ to work with, you'll need to look at tubes that can work with low plate voltages. That usually means the high gm, low rp types.
Also, again, the plate-to-plate feedback around the 6L6 may not work. The driver stage may not have enough gain to properly drive the 6L6 once that has plate-to-plate feedback around it.
I would get the amp working as best you can with no feedback applied, and only then start adding feedback a little at a time, until it sounds and/or measures how you want it to.
With so little B+ voltage available for the driver stage, you are probably going to be limited to a tube type with low rp, high gm and medium-mu, that can work well with only about 100V to 150V on its plate. 6DJ8, 5687 and 12BH7A come to mind.
6CG7 really should have at least 140V on its plate for good operation. Eli D has stated that the 12AT7 really gets good with 200V on its plate. Unfortunately, you have barely more than that available from your B+.
--
I would get the amp working as best you can with no feedback applied, and only then start adding feedback a little at a time, until it sounds and/or measures how you want it to.
With so little B+ voltage available for the driver stage, you are probably going to be limited to a tube type with low rp, high gm and medium-mu, that can work well with only about 100V to 150V on its plate. 6DJ8, 5687 and 12BH7A come to mind.
6CG7 really should have at least 140V on its plate for good operation. Eli D has stated that the 12AT7 really gets good with 200V on its plate. Unfortunately, you have barely more than that available from your B+.
--
Hi Ron,
I'm sure if I drop the plate resistor values on the 6CG7 I can get the voltage up higher.
There is roughly 270v available at the filter caps for the drivers. The 6DJ8 works well with low plate voltages, so I am not as worried about that tube. If anything, I can rewire it as a mu follower which will run fine with no (or a very small) plate resistor.
Granted, these voltages are lower than what you are used to seeing, I really think it is a matter of tweaking to get it working well.
Blair
I'm sure if I drop the plate resistor values on the 6CG7 I can get the voltage up higher.
There is roughly 270v available at the filter caps for the drivers. The 6DJ8 works well with low plate voltages, so I am not as worried about that tube. If anything, I can rewire it as a mu follower which will run fine with no (or a very small) plate resistor.
Granted, these voltages are lower than what you are used to seeing, I really think it is a matter of tweaking to get it working well.
Blair