Just reading over these posts, I like the concept of screen drive, but how do Horiz. Sweeps like the 6CB5 do in straight forward pentode or UL with lower voltage as the oroginal poster suggests?
I see references that state that these type tubes can be run in triode at higher than the stated max screen voltage. So, how would one of these tubes do in UL at say 350V B+?
I've got an old ST-70 that I just received on consignment from a guy with a fried power transformer. I have a Bogen power transformer that will easily supply the current demands for the filaments, and two 4.3K UL outputs that are good. Would there be an issue with the 6CB5 in in UL at 350V B+? If not, how would it do in pentode operation with 350V B+ and 150-200V on the screens?
I'd rather play with the 6CB5 in something like that first vs. dropping a quad into a chassis with $1k worth of iron only to cook something.
Thanks!
I see references that state that these type tubes can be run in triode at higher than the stated max screen voltage. So, how would one of these tubes do in UL at say 350V B+?
I've got an old ST-70 that I just received on consignment from a guy with a fried power transformer. I have a Bogen power transformer that will easily supply the current demands for the filaments, and two 4.3K UL outputs that are good. Would there be an issue with the 6CB5 in in UL at 350V B+? If not, how would it do in pentode operation with 350V B+ and 150-200V on the screens?
I'd rather play with the 6CB5 in something like that first vs. dropping a quad into a chassis with $1k worth of iron only to cook something.
Thanks!
In the case of the 1570s,the original triode connected 6w6s could not produce enough power for effectively driving 811s or 572s at full range because you need 18 watts of effective drive power and this is why the E34L was chosen although I use 5881s even tho there is no gain in this stage,I love the sonic presentation more.
Notice the independent power supply with the 6X5 completing the other part of the bridge just for the drive stages.
Notice the independent power supply with the 6X5 completing the other part of the bridge just for the drive stages.
This is a great thread, and as I read, there are lots of folks like me that are looking for a more straightforward design using these sweeps without insane schematics, etc.
So, I inquired with Edcor regarding building a transformer like the old Acro TO-350. I drew this out for them, and they said they could do a design like this for $98 /ea plus a $70 design fee. This would be really helpful for doubler arrangements to run lower screen voltages. This pricing is for the 100W model.
Blair
So, I inquired with Edcor regarding building a transformer like the old Acro TO-350. I drew this out for them, and they said they could do a design like this for $98 /ea plus a $70 design fee. This would be really helpful for doubler arrangements to run lower screen voltages. This pricing is for the 100W model.
Blair
In typical UL applications where the G2 DC-voltage is same as the anode voltage, 43 % UL tapping is usually optimal but I think that 43 % UL tapping is too high for sweep tubes,
which have essentially higher G2-transconductance and lower DC voltage than typical audio output tubes like 6L6 and EL34.
What G2-voltage you plan to have with 6CB5 ?
which have essentially higher G2-transconductance and lower DC voltage than typical audio output tubes like 6L6 and EL34.
What G2-voltage you plan to have with 6CB5 ?
Hi,
Look at the 6146B schemaic in this link:
http://www.clarisonus.com/Archives/Trans/Acro55.pdf
A transformer like the one I posted will let you tie the outer windings together and feed them 4-600V and use the center tap to the screens with a lower voltage all in the same output transformer. It's not UL. Well, they claim it is, but it has separate supplies for plate and screen?
Blair
Look at the 6146B schemaic in this link:
http://www.clarisonus.com/Archives/Trans/Acro55.pdf
A transformer like the one I posted will let you tie the outer windings together and feed them 4-600V and use the center tap to the screens with a lower voltage all in the same output transformer. It's not UL. Well, they claim it is, but it has separate supplies for plate and screen?
Blair
Last edited:
This is a great thread, and as I read, there are lots of folks like me that are looking for a more straightforward design using these sweeps without insane schematics, etc.
So, I inquired with Edcor regarding building a transformer like the old Acro TO-350. I drew this out for them, and they said they could do a design like this for $98 /ea plus a $70 design fee. This would be really helpful for doubler arrangements to run lower screen voltages. This pricing is for the 100W model.
Blair
If enough people are interested in this, it might make a group buy. I'd be interested.
They were a bit reluctant to entertain the idea until I drew them the image I posted. Looking at the Acrosound TO-350, it does look as if we would want it to be a 6-6.6K A-A transformer since the plates would only see a portion of the winding resistance.
I do, however think this could be a huge potential for those of us looking to build nice amps with sweeps without buying a ton of iron. Edcor will typically make the product available after the design fee is paid and executed.
Blair
I do, however think this could be a huge potential for those of us looking to build nice amps with sweeps without buying a ton of iron. Edcor will typically make the product available after the design fee is paid and executed.
Blair
Splitting the primary at the UL tap will not work. The original Acro transformer has a plate winding that has the full impedance (6600 ohms?) which gets the full plate voltage and a second seperate winding that is 43% of the plate impedance that can be fed from a lower voltage screen supply.
If you feed the plate portion of the partial primaries from one power supply which is at an AC ground (bypassed with a cap) the plates will not see the full winding, and thus will be working into a very low impedance.
I dug out my old breadboard that was used for all the screen drive experiments, and the 6LW6 triode amplifier. I am in the process of restoring it to life.
I now realize that it has been almost 4 years since it has seen power (where does all the time go) and a few parts are missing (where do all the parts go, are they hiding in a corner of the closet for fear of meeting mister power supply????).
I plan to pick up where I left it 4 years ago which involved the ability to drive the screen and control grids independantly with variable DC, and AC driver of either polarity, and now since I have the Edcors with UL taps, variable dc referenced UL.
I am not sure how much time I will have to devote to it before my upcoming road trip, but maybe I will flip the switch and fry some parts this weekend!!!!
If you feed the plate portion of the partial primaries from one power supply which is at an AC ground (bypassed with a cap) the plates will not see the full winding, and thus will be working into a very low impedance.
I dug out my old breadboard that was used for all the screen drive experiments, and the 6LW6 triode amplifier. I am in the process of restoring it to life.
I now realize that it has been almost 4 years since it has seen power (where does all the time go) and a few parts are missing (where do all the parts go, are they hiding in a corner of the closet for fear of meeting mister power supply????).
I plan to pick up where I left it 4 years ago which involved the ability to drive the screen and control grids independantly with variable DC, and AC driver of either polarity, and now since I have the Edcors with UL taps, variable dc referenced UL.
I am not sure how much time I will have to devote to it before my upcoming road trip, but maybe I will flip the switch and fry some parts this weekend!!!!
Right,
But if the split is at the 43% marker on each winding, wouldn't the plates being fed see roughly 3.75K A-A load which is 57% of 6.6K?
Blair
But if the split is at the 43% marker on each winding, wouldn't the plates being fed see roughly 3.75K A-A load which is 57% of 6.6K?
Blair
The impedance ratio is the square of the turns ratio, so each plate on a regular 6600 ohm OPT sees 1650 ohms. If my blonde arithmetic is right they will see 937.5 ohms each when fed by this arangement, which now that I looked at it is more than the 825 ohms each they see with my 3300 ohm OPT.
I think that Edcor would need to use a larger wire size for the plate windings since they will see nearly 1 amp of peak current, but could use a tiny wire for the screen winding since they will only see a few mA.
I think that Edcor would need to use a larger wire size for the plate windings since they will see nearly 1 amp of peak current, but could use a tiny wire for the screen winding since they will only see a few mA.
Well,
That does complicate things a bit. Not trying to oversimplify things, but it looked like the TO-350 had split the windings at their UL taps. I wonder if we shifted the UL point to 15-20% if it would help. This ain't my forte. I just like to build stuff, but if we could get Edcor to manufacture something like this at a reasonable cost, it seems like it would be worth it.
Not sure about that 1A current rating though.
Blair
That does complicate things a bit. Not trying to oversimplify things, but it looked like the TO-350 had split the windings at their UL taps. I wonder if we shifted the UL point to 15-20% if it would help. This ain't my forte. I just like to build stuff, but if we could get Edcor to manufacture something like this at a reasonable cost, it seems like it would be worth it.
Not sure about that 1A current rating though.
Blair
Sorry, just re-read this, and you are more concerned about the current handling than the load now. Hmm, I wonder if they can give me primary current handling data.
I would think, and others have pointed out, that the usual 43% number is too much for a sweep tube. I believe that 43% was taken as the "magic point" for an EL34, and other tubes may work better with a different ratio. A sweep tube has a more sensitive screen grid than the typical audio tube.
1A (actually .75 A in my amp) is the upper limit on peak current that these designs generate. A sweep tube, especially in screen drive can saturate down to a few volts, putting the entire power supply into one half of the OPT(650 volts into 825 ohms). This is only during transient peaks, but it is real. The same wire used in their 3300 ohm OPT should be fine since that sees the same current. The average current is much lower.
There is a rather crude way to check some of this stuff out before comitting any money to OPT design. Wire two existing OPT's together. Tie the secondaries in parallel, and feed one primary with the plates, and the second primary with the screens. A seperate power supply for each OPT can be used. The coupling will not be as good as a single OPT, but the theory can be tested.
I'm sure it would be the same wire they use on their standard 100W models.
It may not be tomorrow🙂, but I would invest $170 into something like that if it were for the greater good of the audio guys.
I do a lot if dumb things like this🙂 LOL!
Blair
It may not be tomorrow🙂, but I would invest $170 into something like that if it were for the greater good of the audio guys.
I do a lot if dumb things like this🙂 LOL!
Blair
Re the "sweet" spot for the UL tapping or ratio that's best for sweep tubes, every simulation I've run on "born to sweep" tubes (as opposed to adapted existing tubes) has convinced me that something in the order of 25% is far closer to ideal than 40 or more%. As George points out, these beasts are blessed with a large gm for G2, one can deduce 4 or more ma/volt which is on the par with G1 gm for a lot of tubes!
The more I've thought about this, the more inclined I'm becoming to implement G1 drive with a MOSFET circuit to allow me to both set the idle G2 voltage and experiment with the ratio, and use the 6CB5. The only thing lost, I think, to an actual tap is that the screen current no longer contributes to the load, but then we are back to the situation with a straight pentode anyway!
What do y'all think?
Rene
The more I've thought about this, the more inclined I'm becoming to implement G1 drive with a MOSFET circuit to allow me to both set the idle G2 voltage and experiment with the ratio, and use the 6CB5. The only thing lost, I think, to an actual tap is that the screen current no longer contributes to the load, but then we are back to the situation with a straight pentode anyway!
What do y'all think?
Rene
That's the thing about that Acro design. It's called UL, but it has two separate B+, so it works more like pentode?
Anyway, I will still do something with the 6CB5 tubes since I have right now🙂
However, it was suggested that with my 1.5K output iron, I look at the 4D32. Looks very promising! 350v SG max, 50W PD, 125W into 3K at reasonable voltages. It is an RF tetrode, but also has audio frequency data.
I got very exited about this and started looking for PP designs, but all I managed to find were folks that said they were working on a design using them. Anyone know of a link or way to find a PP 4D32 amp?
Since its a tetrode, I will likely start a new thread for information also.
Thanks!
Blair
Anyway, I will still do something with the 6CB5 tubes since I have right now🙂
However, it was suggested that with my 1.5K output iron, I look at the 4D32. Looks very promising! 350v SG max, 50W PD, 125W into 3K at reasonable voltages. It is an RF tetrode, but also has audio frequency data.
I got very exited about this and started looking for PP designs, but all I managed to find were folks that said they were working on a design using them. Anyone know of a link or way to find a PP 4D32 amp?
Since its a tetrode, I will likely start a new thread for information also.
Thanks!
Blair
In my experiments with running sweep tubes at plate voltages above 500 volts and screen voltages under 200 volts, the screen current is just a few mA so there isn't much power to contribute.
That's where I'm going. I was just starting down this path when life's interruptions derailed most of my experimenting for nearly 5 years. I still have a hectic work schedule, and a 1 to 2 year countdown to retirement, but there should be some time for some tube melting in the next few months.
Note:
G1 on a sweep tube doesn't need to go positive. The driver board that I will be using for these experiments has mosfet drive, but this board was designed for the 6L6GC in AB2 thread. Most of my more recent high powered sweep tube experiments were done with Pete's driver boards which do not use mosfets. 525 watts from 4 tubes without mosfets is OK isn't it????
Mosfet drive for G1 does eliminate the overload recovery issues, and as you suggested, allows total freedom of adjustment to put any signal on any grid, with any DC offset. I plan to mount 4 BIG Fuji fets per channel, 2 per tube, on a big heatsink.
I have to make a new amp design, and I will investigate G1, screen, and dual drive for this design. Soon I must choose an output tube so the chassis work can resume, and it will not be the 6CB5, but I have a bunch, so I can test a few and see what they do.
I found the box full of parts that I packed away when this amp was shelved, and set out to mount the mosfets on the heatsinks, but the mosfets were not in the box......bummer. I spent most of the night looking for them without success.
I have 6 of them, and they are on the short list to hook up to this breadboard. The datasheet calls for AB2 G1 drive, so mosfet drivers are needed. This is the reason I haven't tried them yet. I have been using Petes driver boards which have no mosfets.
- Status
- Not open for further replies.