How much gain you have in each stage is determined a lot more by the circuit rather than the maximum gain the tube is capable.
That said, let say you want an 8 watts output (8Vrms on a 8 ohm speaker load). And the transformer you want to use is ~3500 ohms (21:1 ratio). The 300B output voltage needed is 8X21 = 168V. Lets say you want an input sensitivities of 1V. You will need an overall gain of 168 or 45db.
If you multiply all the tubes' max gain, you have 70x3.5x4 = 980. So in theory you have enough raw gain. You will still need to select the right circuit with proper NFB to achieve the desired power output.
That said, let say you want an 8 watts output (8Vrms on a 8 ohm speaker load). And the transformer you want to use is ~3500 ohms (21:1 ratio). The 300B output voltage needed is 8X21 = 168V. Lets say you want an input sensitivities of 1V. You will need an overall gain of 168 or 45db.
If you multiply all the tubes' max gain, you have 70x3.5x4 = 980. So in theory you have enough raw gain. You will still need to select the right circuit with proper NFB to achieve the desired power output.
Rather than attempting to guess about how to drive the 300b, you should probably start by consulting the 300b datasheet and pick one of their operating points. They list a pretty extensive set and show ample performance specifications as well.
Once you manage that, look at the specified grid voltage for that operating point. In short, you need a driver stage that can deliver twice that voltage in peak-to-peak voltage. In other words, if the grid bias is -74V, your driver stage would need to deliver a minimum of 148 volts peak-to-peak.
Designing the driver stage is more critical... as it's not just the amount of gain you need, but the output voltage swing capability and the drive current to be able to maintain a high enough slew rate to get a wide frequency response at that voltage output. Note that the 300b has a pretty high input capacitance and that will put some additional requirements on the driver stage. To help keep the overall distortion low, your driver stage should be able to drive more voltage than is required at the minimum calculation.
In most cases, it helps to have some initial design criteria, or some specifications you are looking to achieve. Once you have those, you can start doing a design of the output stage, which will help define what the driver stage needs to deliver. It also might work out that a two-stage input/driver is required to meet the design specifications you have.
Regards, KM
Once you manage that, look at the specified grid voltage for that operating point. In short, you need a driver stage that can deliver twice that voltage in peak-to-peak voltage. In other words, if the grid bias is -74V, your driver stage would need to deliver a minimum of 148 volts peak-to-peak.
Designing the driver stage is more critical... as it's not just the amount of gain you need, but the output voltage swing capability and the drive current to be able to maintain a high enough slew rate to get a wide frequency response at that voltage output. Note that the 300b has a pretty high input capacitance and that will put some additional requirements on the driver stage. To help keep the overall distortion low, your driver stage should be able to drive more voltage than is required at the minimum calculation.
In most cases, it helps to have some initial design criteria, or some specifications you are looking to achieve. Once you have those, you can start doing a design of the output stage, which will help define what the driver stage needs to deliver. It also might work out that a two-stage input/driver is required to meet the design specifications you have.
Regards, KM
6J5/6SN7 into a 45 driving a 300B will give you a sensitivity for full power of <1Vrms regardless of the output transformer, 1:1 IT or CCS load..
The single D3A will drive a 300B directly using either IT or CCS follower, and about 800mVrms sensitivity.
A D3A/45/300B is going to have way too much gain for most modern systems and sensitivity of about 250mVrms for full output.
I've built a number of amplifiers based on all of these topologies (except D3A/45/300B) and the sensitivity is generally within 1 - 2dB of these predictions.
For calculating approximate gain at minimum need to know output transformer primary and output impedances.
The single D3A will drive a 300B directly using either IT or CCS follower, and about 800mVrms sensitivity.
A D3A/45/300B is going to have way too much gain for most modern systems and sensitivity of about 250mVrms for full output.
I've built a number of amplifiers based on all of these topologies (except D3A/45/300B) and the sensitivity is generally within 1 - 2dB of these predictions.
For calculating approximate gain at minimum need to know output transformer primary and output impedances.
You have a very nice - but potentially a bit challenging - pentode with some serious transconductance in that D3a! 35mS is a boat load of gm. So why not try driving the 300B directly with that D3a in pentode mode? A lot more work to make the distortion characteristics of both tubes meet nicely, but very rewarding when you get it working!
I used an E83F to drive a 6B4G, basically a variant of a 2A3, in pentode mode, and it's one of my absolute best amplifiers. Drive requirements are quite similar for both 2A3 and 300B - both require around 150Vpp or more voltage swing, and loads of current to go with that. Easily done in a single stage with a high transconductance pentode with gain to spare.
As an extra for that pentode mode you get low input capacitance, which makes the input less sensitive for output resistance from preceding stages.
Take note, though, that high gm pentodes can oscillate when implemented improperly. They're kinda delicate little things in that respect.
I used an E83F to drive a 6B4G, basically a variant of a 2A3, in pentode mode, and it's one of my absolute best amplifiers. Drive requirements are quite similar for both 2A3 and 300B - both require around 150Vpp or more voltage swing, and loads of current to go with that. Easily done in a single stage with a high transconductance pentode with gain to spare.
As an extra for that pentode mode you get low input capacitance, which makes the input less sensitive for output resistance from preceding stages.
Take note, though, that high gm pentodes can oscillate when implemented improperly. They're kinda delicate little things in that respect.
Hi Adolf Corkscrew, I'm about to try a D3a driving 300B via an audio note interstage t/f and am aware of the hf instability so will use ferrite beads and pay attention to wiring layout. I've used E182CC via interstage t/f as driver for decades, very nice. I have Vitavox Thunderbolt speakers. Anything else I should be paying attention to? Thanks
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.