Well, that does require a capable current handler with enough plate power dissipation. Fuling, I have one more question to ask. Some seasoned tube fans here in China insist on using driving tranformer to build circuit for tubes like 811a that has grid current when working in A1 mode, such as 805. Have you ever tried circuit using transformer in driving stage? If you have, could please say something about the strengths and weaknesses of either mode? Thank you.
I ask this question just because one of my friends, a tube fan, would rather accept a pair of ancient RCA-made driving transformers with frequency response from 20Hz to 15KHz than use this wide frequency band design.
I ask this question just because one of my friends, a tube fan, would rather accept a pair of ancient RCA-made driving transformers with frequency response from 20Hz to 15KHz than use this wide frequency band design.
Yup, I´ve tried transformers. I made them myself, and it all turned out to be a real mess.
There are two ways of IT-coupling these kind of circuits, either you use an ordinary transformer that can handle both the DC in the primary winding plus the grid current in the secondary; or you use an inverting IT where the grid current through the secondary partly cancels the primary current.
In short terms: The transformers has to be more or less custom made and it is way more complicated than driving A1 tubes.
If money is an issue, stick to the cathode (or mosfet) followers.
Suitable transformers won´t be cheap!
Though, if transformer coupling can be arranged properly I think it might be very good. A bit of stepdown, say 2:1 or 3:1, would provide a nice and low output impedance from the driver stage.
There are two ways of IT-coupling these kind of circuits, either you use an ordinary transformer that can handle both the DC in the primary winding plus the grid current in the secondary; or you use an inverting IT where the grid current through the secondary partly cancels the primary current.
In short terms: The transformers has to be more or less custom made and it is way more complicated than driving A1 tubes.
If money is an issue, stick to the cathode (or mosfet) followers.
Suitable transformers won´t be cheap!
Though, if transformer coupling can be arranged properly I think it might be very good. A bit of stepdown, say 2:1 or 3:1, would provide a nice and low output impedance from the driver stage.
Fulling could u care to explain how to wind the inverted IT .Is it just primary and secondary only, how is there additional sections for cancelation. What i understand is that the primary is drawing dc current clockwise say 20ma and secondary when loaded with 811a drawing 20ma anticlockwise. NET dc current flow in the core is 20-20=0. Assume perfect 20ma draw in primary and perfect grid current at 20ma. This way cancelation takes place right? But ideal case grid current in secondary would be different from 20-60ma. So all in gapped the core for 40ma to handle the normal operating points. That way the core has lower airgap higher mu, higher primary inductance right? Will need this info to design my inverted IT.
Thanks
Thanks
I´m not the right person to talk about transformers, but I can share what I´ve read about inverting IT´s.
If you pull, say, 20mA trough each winding but in opposite direction, they will cancel, yes. This works as long as the winding ratio is 1:1.
In this application we would probably want a bit of stepdown ratio in the transformer, in addition to that we probably want to pull more than 20mA trough the primary.
Now the DC fields will only cancel each other partially, so we need an airgap.
So far so good, but the real problems with inverting interstages are that they have way worse frequency response than regular IT´s, just because they are inverting.
This problem becomes worse when the turns ratio approaches 1:1, and stepdown transformers seems to work better in this mode.
I think an easier way to deal with IT´s for A2 use would be to use regular noninverting transformers with slightly bigger airgaps and cores to allow for the extra DC in the secondary.
By the way, search for IIT at the Audio Asylym, there are some discussions worth reading there.
If you pull, say, 20mA trough each winding but in opposite direction, they will cancel, yes. This works as long as the winding ratio is 1:1.
In this application we would probably want a bit of stepdown ratio in the transformer, in addition to that we probably want to pull more than 20mA trough the primary.
Now the DC fields will only cancel each other partially, so we need an airgap.
So far so good, but the real problems with inverting interstages are that they have way worse frequency response than regular IT´s, just because they are inverting.
This problem becomes worse when the turns ratio approaches 1:1, and stepdown transformers seems to work better in this mode.
I think an easier way to deal with IT´s for A2 use would be to use regular noninverting transformers with slightly bigger airgaps and cores to allow for the extra DC in the secondary.
By the way, search for IIT at the Audio Asylym, there are some discussions worth reading there.
Thanks to the reawakening of this thread, I finally gathered the willpower to start up my 811A project again.
After a few late evenings with my head buried in my junkbox and with paper, pen and calculator in my hands I´ve finally decided what topology to build and which parts to use.
Here it goes: MJE 350 CCS loaded ECC88 input stage (one tube shared between both channels) cap coupled to triode wired EL86 cathode followers which are direct coupled to the 811´s.
Rail voltages will be 160V for input/drivers and 460V for 811´s.
Main PSU: 330VAC - silicon bridge - 12uF PiO cap - choke (unknown inductance but tested and works fine) - 200uF ´lytics.
The choke will be in the ground rail due to insulation issues.
Filament PSU´s: 10VAC- bridge-1,7mH aircored choke-27000uF-1,1mH iron cored choke (plus resistor if necessary)-108000uF. Sweaty!
Driver PSU: 60VAC-voltage doubler- cap - choke - cap.
Haven´t figured this one out exactly yet.
The high voltage winding and all filament winding os on one transformer, the 60VAC comes from a separate one.
If there is room in the chassis I will add various paper i oil caps fo decoupling here and there in the PSU´s.
Speaking of chassis, it´s almost finished by now.
I´ve managed too squeeze everything onto a 25x40 cm aluminium sheet. The holes for the tubes was cut by my brother using a wire EDM . Incredible machinery, I wish I had one at home
The frame around the top plate will be made from 21x100 mm oak.
BTW those 1,7mH chokes in the fil. supplies comes from an old EDM machine that we tore apart a month or so ago. Heavy stuff, they can easily handle 10-15A.
After a few late evenings with my head buried in my junkbox and with paper, pen and calculator in my hands I´ve finally decided what topology to build and which parts to use.
Here it goes: MJE 350 CCS loaded ECC88 input stage (one tube shared between both channels) cap coupled to triode wired EL86 cathode followers which are direct coupled to the 811´s.
Rail voltages will be 160V for input/drivers and 460V for 811´s.
Main PSU: 330VAC - silicon bridge - 12uF PiO cap - choke (unknown inductance but tested and works fine) - 200uF ´lytics.
The choke will be in the ground rail due to insulation issues.
Filament PSU´s: 10VAC- bridge-1,7mH aircored choke-27000uF-1,1mH iron cored choke (plus resistor if necessary)-108000uF. Sweaty!
Driver PSU: 60VAC-voltage doubler- cap - choke - cap.
Haven´t figured this one out exactly yet.
The high voltage winding and all filament winding os on one transformer, the 60VAC comes from a separate one.
If there is room in the chassis I will add various paper i oil caps fo decoupling here and there in the PSU´s.
Speaking of chassis, it´s almost finished by now.
I´ve managed too squeeze everything onto a 25x40 cm aluminium sheet. The holes for the tubes was cut by my brother using a wire EDM . Incredible machinery, I wish I had one at home
The frame around the top plate will be made from 21x100 mm oak.
BTW those 1,7mH chokes in the fil. supplies comes from an old EDM machine that we tore apart a month or so ago. Heavy stuff, they can easily handle 10-15A.
I´ve heard that too, but the problem is hum.
High filament heater voltage and high mu increases the risk for hum that cannot be balanced away.
The common hum balance pot works fine on old tubes with 2,5 and 4V filaments and mu=3-5. Here we have 6,3V filaments and mu=160!
I don´t think it´s possible to AC heat 811´s without severe hum at the output. Every schematic I have seen employs DC heating except one, Bob Danielak´s cathode driven SV572-160 amp.
He mentions nothing about hum in the related text.
I guess there´s only one real way to find out: Try both!
I already have all the stuff needed for the filament supplys so I won´t bother trying AC, at least not at this point.
High filament heater voltage and high mu increases the risk for hum that cannot be balanced away.
The common hum balance pot works fine on old tubes with 2,5 and 4V filaments and mu=3-5. Here we have 6,3V filaments and mu=160!
I don´t think it´s possible to AC heat 811´s without severe hum at the output. Every schematic I have seen employs DC heating except one, Bob Danielak´s cathode driven SV572-160 amp.
He mentions nothing about hum in the related text.
I guess there´s only one real way to find out: Try both!
I already have all the stuff needed for the filament supplys so I won´t bother trying AC, at least not at this point.
I´ll run them at about 430-440V and 90-100mA, so somewhere between 39 and 44W.
I think 811A´s are built to do some serious work in professional and military equipment so I guess we would have to run them really hard if we´re going to burn them out in an audio circuit.
BTW, they are cheap enough to buy some spares.
I think 811A´s are built to do some serious work in professional and military equipment so I guess we would have to run them really hard if we´re going to burn them out in an audio circuit.
BTW, they are cheap enough to buy some spares.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.