I've got all the parts for a more or less Loftin White 2A3 build, but I realize I'm not totally clear on how a portion of the circuit works—specifically the area I've circled in red. I've been familiarizing myself with a variety of Loftin White-inspired circuits and most of them seem to have a hum pot on and cathode resistor and cap on the 2A3. This one does not and would love some insight into what is happening in this section. Perhaps it's obvious, but it's all new to me. Wanted to start with a simple circuit as my first ground up from schematic build.
After some poking around online, I've found that a 6SQ7 is a suitable replacement for the 2A6 (different heater and base). That said, I've got a couple of 6SL7 tubes on hand and would like to save some money. As far as I can tell, the circuit is only using the triode portion of the 6SQ7 and my question is can I use a 6SL7 instead (only one triode section)? They seem pretty close in spec however the amplification factor is different as well as more plate current draw and higher transconductance. As long as my power supply can handle the extra current draw how will the amplification factor play out downstream?
Thanks in advance!
After some poking around online, I've found that a 6SQ7 is a suitable replacement for the 2A6 (different heater and base). That said, I've got a couple of 6SL7 tubes on hand and would like to save some money. As far as I can tell, the circuit is only using the triode portion of the 6SQ7 and my question is can I use a 6SL7 instead (only one triode section)? They seem pretty close in spec however the amplification factor is different as well as more plate current draw and higher transconductance. As long as my power supply can handle the extra current draw how will the amplification factor play out downstream?
Thanks in advance!
This is not a simple circuit for beginers to tweak , you should build it exactly as it is . In the red square is the cathode resistor for 2A3 ( 3K 20W ) , the cap in parallel C3 and a cap to B+ C4 , probably for ripple canceling . The rest is a resistive divider R6 - R5 to the 2A3 cathode and filter C2 that is generating the supply voltage for the driver tube .
It is direct coupled so you can't just replace 2A6 with some other tube you want , the plate voltage must be 150V as it is the voltage on the 2A3 grid
It is direct coupled so you can't just replace 2A6 with some other tube you want , the plate voltage must be 150V as it is the voltage on the 2A3 grid
Last edited:
Thanks for the prompt reply.
I've gone through it pretty thoroughly and it's my understanding that the 2A6 is functionally identical to the 6SQ7 (both dual diodes with a triode). The only difference is the base and different filament voltage. The base and pinout are easy enough and my PT has the right filament secondary. Other than that, the circuit was going to remain the same, just looking for a clear explanation of a portion of the schematic. Which I understand a bit better now. Thanks for the clarification.
My question regarding the 6SQ7 vs. the 6SL7 is really because I have the 6SL7 on hand and it doesn't seem THAT far off if only using one of the triode sections of the 6SL7. More amplification factor, but perhaps I should just get the 6SQ7 and call it a day.
I've gone through it pretty thoroughly and it's my understanding that the 2A6 is functionally identical to the 6SQ7 (both dual diodes with a triode). The only difference is the base and different filament voltage. The base and pinout are easy enough and my PT has the right filament secondary. Other than that, the circuit was going to remain the same, just looking for a clear explanation of a portion of the schematic. Which I understand a bit better now. Thanks for the clarification.
My question regarding the 6SQ7 vs. the 6SL7 is really because I have the 6SL7 on hand and it doesn't seem THAT far off if only using one of the triode sections of the 6SL7. More amplification factor, but perhaps I should just get the 6SQ7 and call it a day.
Thanks @Rikaro it makes a lot more sense. I did plan on omitting the meter as all the nice ones were a little pricey.
One other thing, I'm reading the "N" on the 2.5V filament in the schematic as a center tap. Is that correct? My PT doesn't have a CT on the 2.5V secondary and I intend to use a couple of resistors for a virtual CT. Any issues with that downstream?
joneci,
2A3 tube:
max gain, u = 4.2,
plate impedance, rp = 800 Ohms
110 Ohms of un-bypassed cathode resistance: u x 110 Ohms, 4.2 x 110 = 462 Ohms.
The plate resistance is now 800 Ohms + 462 Ohms = 1262 Ohms; that is Bad, the damping factor went down, the gain went down, the output power went down.
Two 25 Ohm resistors: 12.5 Ohms x 4.2 = 52.5 Ohms
800 Ohms + 52.5 Ohms = 852.5 Ohms, much better!
At 2.5V, 50 Ohms across the filament winding only adds 50mA more current from the filament winding (if your power transformer can not stand 50mA more filament current, versus 2.5A filament current, then get a better power transformer).
Happy building and listening.
Sorry, I edited this post from a 45 tube to a 2A3 tube.
2A3 tube:
max gain, u = 4.2,
plate impedance, rp = 800 Ohms
110 Ohms of un-bypassed cathode resistance: u x 110 Ohms, 4.2 x 110 = 462 Ohms.
The plate resistance is now 800 Ohms + 462 Ohms = 1262 Ohms; that is Bad, the damping factor went down, the gain went down, the output power went down.
Two 25 Ohm resistors: 12.5 Ohms x 4.2 = 52.5 Ohms
800 Ohms + 52.5 Ohms = 852.5 Ohms, much better!
At 2.5V, 50 Ohms across the filament winding only adds 50mA more current from the filament winding (if your power transformer can not stand 50mA more filament current, versus 2.5A filament current, then get a better power transformer).
Happy building and listening.
Sorry, I edited this post from a 45 tube to a 2A3 tube.
Last edited:
Thanks for this @6A3sUMMER. Just to be clear we're talking about the resistor value for the virtual center tap, correct? The goal is to have enough damping? Would this impact the input driver tube, the 6SQ7?
My PT will output 2.5V at 5.5A so should be good. I went overkill to someday mess with DC filaments... someday.
My PT will output 2.5V at 5.5A so should be good. I went overkill to someday mess with DC filaments... someday.
Good point!
So that would mean about 0.3W less output and ~20% increased source impedance.
BTW, also the OT primary DCR adds losses and source impedance.
Slow warmup of the rectifier is essential in this circuit - if rectifier heats faster than the input tube, there will close to 300V on the 2A3 grid instead of 150. 5V4 is same as 83V but with octal base, much easier to find. Likewise powering it with driver tube missing will destroy the 2A3 as well.
Thanks @Tom Bavis. I have both a 5U4GB and a 5AR4 on hand. I plan to use the 5AR4 for the soft start and working on finalizing a design for a choke input PS with a small 2uF tuning capacitor before the choke to get the B+ to 450V DC - my PT has 420V and 480V CT secondaries. Also, considering a thermistor to slow things down further.
joneci,
I seem to have made a mistake.
But the schematic in Post #1 does not make it clear about the connections of the filaments of:
2A3
Input tube.
I had not referred to that schematic.
I believe the 2A3 self bias resistor is the 3k resistor.
A floating 2.5V AC filament supply would normally connect to the top (hot end), of that self bias resistor with a pair of resistors from the filament ends.
Those un-bypassed resistors would be in series from the filaments to the 3k bias resistor.
That is when the un-bypassed resistance causes an increase of the 2A3 plate resistance.
Any un-bypassed resistance from the filaments to the rest of the circuit (eventually to return to ground/B-, will increase the 2A3 plate resistance.
Often input tubes filament supplies also use either a center tap to ground, or a pair of resistors to form a "pseudo center tap"; but indirectly heated cathodes do not have the same effect as DHT filaments, in relation to plate impedance, rp.
I seem to have made a mistake.
But the schematic in Post #1 does not make it clear about the connections of the filaments of:
2A3
Input tube.
I had not referred to that schematic.
I believe the 2A3 self bias resistor is the 3k resistor.
A floating 2.5V AC filament supply would normally connect to the top (hot end), of that self bias resistor with a pair of resistors from the filament ends.
Those un-bypassed resistors would be in series from the filaments to the 3k bias resistor.
That is when the un-bypassed resistance causes an increase of the 2A3 plate resistance.
Any un-bypassed resistance from the filaments to the rest of the circuit (eventually to return to ground/B-, will increase the 2A3 plate resistance.
Often input tubes filament supplies also use either a center tap to ground, or a pair of resistors to form a "pseudo center tap"; but indirectly heated cathodes do not have the same effect as DHT filaments, in relation to plate impedance, rp.
I thought on this and wonder if there has ever been any "improvements" to the Loftin White typology to make it... safer. I feel like I may have come accross an article a while back but can't seem to recall where. Something about a transistor to limit the current to the 2A3 if the driver tube fails...
Thanks for the follow up... by mistake are you referring to the suggested virtual center tap resistor values for the 2.5V filament secondary? Or your read of the schematic?
joneci,
1. I was referring to the virtual center tap.
For a good example, the schematic in Post # 16 has a 50 Ohm adjustable pseudo center tap pot. It adjusts to the electrical center of the 2A3 filament to reduce the resultant hum.
But a 50 Ohm pot across the 2A3 filament is effectively 25 Ohms to the complex self bias network.
25 x 4.2 = 105 Ohms added to the 2A3 plate impedance.
800 + 105 = 905 Ohms plate impedance, rp (not too bad, a good compromise between a 25 Ohm pot, and a 100 Ohm pot).
2. Yes, it was a very late hour, and I had trouble analyzing and following the filament connections in the schematic of Post # 1.
3. I remember a friend that did a modified Loftin White 2A3 amplifier, the 2A3 had 200 or 250mA plate current during power-up warm-up.
Then, a lot of additional circuitry solved that and other DC coupled circuit problems.
Schematic in post # 16 is one interesting solution.
4. We each have our preferences.
I used to use an interstage transformer between the driver and the 2A3.
Then I used RC coupling between the driver and the 2A3.
I am not a fan of DC coupling, but that is my choice.
1. I was referring to the virtual center tap.
For a good example, the schematic in Post # 16 has a 50 Ohm adjustable pseudo center tap pot. It adjusts to the electrical center of the 2A3 filament to reduce the resultant hum.
But a 50 Ohm pot across the 2A3 filament is effectively 25 Ohms to the complex self bias network.
25 x 4.2 = 105 Ohms added to the 2A3 plate impedance.
800 + 105 = 905 Ohms plate impedance, rp (not too bad, a good compromise between a 25 Ohm pot, and a 100 Ohm pot).
2. Yes, it was a very late hour, and I had trouble analyzing and following the filament connections in the schematic of Post # 1.
3. I remember a friend that did a modified Loftin White 2A3 amplifier, the 2A3 had 200 or 250mA plate current during power-up warm-up.
Then, a lot of additional circuitry solved that and other DC coupled circuit problems.
Schematic in post # 16 is one interesting solution.
4. We each have our preferences.
I used to use an interstage transformer between the driver and the 2A3.
Then I used RC coupling between the driver and the 2A3.
I am not a fan of DC coupling, but that is my choice.
Last edited:
@directdriver thansk for this! now you've got me thinking about this approach. oye... is there an easy way to tell how much current the 6SL7 is drawing? To my untrained eye, it seems like theres not quite enough info. It would be helpful know the voltage accross the 6SL7 cathode resistor. I ask, because I roughed in a power supply - using what I have - and want to see if this is possible. Here's what I've got...
The additional circuitry gives me pause... there's something elegant about the Loftin White but I also don't want to over complicate it.
joneci,
In my opinion the only 'elegant' thing about the Loftin White is the fact that it is DC coupled.
Loftin White does not have . . .
Low frequency rolloff from either an interstage transformer, or some RC coupling parts values (RC coupling can be sub Hz if you are worried about low frequency rolloff).
Driver plate to output tube grid has zero phase shift at low frequencies; but interstage transformers do and RC couplings do (however slight of a phase shift), but so do output transformers, and Loftin White has an output transformer.
High >> 20kHz frequency rolloff from an interstage transformer.
The simplicity of a 2 stage RC coupled SE is very good; it is very hard to do anything simpler except a 1 stage amplifier (with lots of other tradeoffs).
DC coupling has many problems to solve, build a known tried and proven DC coupled circuit; otherwise when you modify and roll your own circuit and finally get it working well in all aspects, you will be an expert.
Just my opinions.
In my opinion the only 'elegant' thing about the Loftin White is the fact that it is DC coupled.
Loftin White does not have . . .
Low frequency rolloff from either an interstage transformer, or some RC coupling parts values (RC coupling can be sub Hz if you are worried about low frequency rolloff).
Driver plate to output tube grid has zero phase shift at low frequencies; but interstage transformers do and RC couplings do (however slight of a phase shift), but so do output transformers, and Loftin White has an output transformer.
High >> 20kHz frequency rolloff from an interstage transformer.
The simplicity of a 2 stage RC coupled SE is very good; it is very hard to do anything simpler except a 1 stage amplifier (with lots of other tradeoffs).
DC coupling has many problems to solve, build a known tried and proven DC coupled circuit; otherwise when you modify and roll your own circuit and finally get it working well in all aspects, you will be an expert.
Just my opinions.