Newbie here. I just put together my first 300B amp. Very excited that it turns on and makes sound instead of smoke! The satisfaction of building something that allows me to sit down an enjoy music is undescribably awesome!
As for the sound, it is quite good, i think (or is it just the "face only a mother could love" syndrome?) there are some things that it does well, like the mid and the warmth, detail and realism is very good and SETy. However, I also noticed it is a bit shouty on the high, sometimes to the point of piercing. The bass is quite loose, there is bass, but not well defined.
Is there something I have done wrong? I didn't use a specific schematic, but instead put together bits and pieces from all the info I found here and elsewhere.
Here are things I am not sure if i did correctly.
435V B+ comes from an outboard power supply that I scavenge from another amp. It is 5u4g rectified + CLC with 33uF / 5H / 1000+uF. That need an umbilical cord so I added a 22uF Obbligato oil cap on the amp chassis. I cannot change the power transformer (~735VCT) so 430V is all I can get. That seems a bit low as the 300B cathode is at 65V so I only ended up with 370V plate. Is that low? I am still struggling to understand how to calculate the output power but I do get slightly more than 9V swing at the output.
The 6C45P bias point is a bit hot but to get the amount of swing to drive the 300B I decided to push it a little. I read that 160V peak to peak is recommended. But then, with only 65V bias voltage, am I driving it to A2? Is 65V bias for 300B too low?
For the cathode bypass caps I used what I had on hand. rough calculation the cutoff should be enough. Are they sufficient? also, the 300B bypass cap do not cover the 200R pot, is that bad?
The de-coupling cap is a 0.22uF Jensen film cap. the cutoff against the 270k seems a bit high. I have read some page that recommend 10/(2*pi*R*f) instead of 1/(2*pi*R*f). Not sure if that is affecting the bass.
C5 is a 47uF Mundorf Mylitic and I bypassed it with a 6.8uF Obbligato oil. Is that too large value here? Or should I use 22uF Obbligato instead as I bought a bunch from a sale. How to determine what is the correct size here?
All the resistors are cheap generic 2W ones I got from one of the assortment kits on Amazon. I plan to get some Takman REX to replace them now that I know the values are "OK".
The 6c45p is a Sovtek. The 300B is a KR VV302B that I have lying around. I also tried a WE300B but it seems weaker. The OPT is Monolith Magnetic SA9.
My speakers are ~90dB/W. It can play decently loud with the preamp almost maxed out. Some music I can hear clipping but I think that's expected for that speaker. Though more headroom would be nice.
more questions on the issue I have with the sound.
Where do I look to figure out why the highs are so shouty?
Where do I look to figure out how to get tighter bass? Right now it is quite mushy. The bass come quick, but not well defined.
I read that having the 6C45P driving 300B might have some high roll off. It seems to be OK or am I not hearing the right thing?
I am thinking of the following changes as potential next steps.
- Fixed bias - Maybe having another 65V or so plate voltage will give more headroom?
- interstage - how much improvement will I see from that?
- maybe choke loaded? or CCS?
- separate B+ supply for the 6c45p - is this worth it? i might squeeze a few more volts for the 300B B+ and also some separation for the 2 stages
- penthode driver
Ultimately I want to make this amp sound very good.
sorry, lots of questions.
As for the sound, it is quite good, i think (or is it just the "face only a mother could love" syndrome?) there are some things that it does well, like the mid and the warmth, detail and realism is very good and SETy. However, I also noticed it is a bit shouty on the high, sometimes to the point of piercing. The bass is quite loose, there is bass, but not well defined.
Is there something I have done wrong? I didn't use a specific schematic, but instead put together bits and pieces from all the info I found here and elsewhere.
Here are things I am not sure if i did correctly.
435V B+ comes from an outboard power supply that I scavenge from another amp. It is 5u4g rectified + CLC with 33uF / 5H / 1000+uF. That need an umbilical cord so I added a 22uF Obbligato oil cap on the amp chassis. I cannot change the power transformer (~735VCT) so 430V is all I can get. That seems a bit low as the 300B cathode is at 65V so I only ended up with 370V plate. Is that low? I am still struggling to understand how to calculate the output power but I do get slightly more than 9V swing at the output.
The 6C45P bias point is a bit hot but to get the amount of swing to drive the 300B I decided to push it a little. I read that 160V peak to peak is recommended. But then, with only 65V bias voltage, am I driving it to A2? Is 65V bias for 300B too low?
For the cathode bypass caps I used what I had on hand. rough calculation the cutoff should be enough. Are they sufficient? also, the 300B bypass cap do not cover the 200R pot, is that bad?
The de-coupling cap is a 0.22uF Jensen film cap. the cutoff against the 270k seems a bit high. I have read some page that recommend 10/(2*pi*R*f) instead of 1/(2*pi*R*f). Not sure if that is affecting the bass.
C5 is a 47uF Mundorf Mylitic and I bypassed it with a 6.8uF Obbligato oil. Is that too large value here? Or should I use 22uF Obbligato instead as I bought a bunch from a sale. How to determine what is the correct size here?
All the resistors are cheap generic 2W ones I got from one of the assortment kits on Amazon. I plan to get some Takman REX to replace them now that I know the values are "OK".
The 6c45p is a Sovtek. The 300B is a KR VV302B that I have lying around. I also tried a WE300B but it seems weaker. The OPT is Monolith Magnetic SA9.
My speakers are ~90dB/W. It can play decently loud with the preamp almost maxed out. Some music I can hear clipping but I think that's expected for that speaker. Though more headroom would be nice.
more questions on the issue I have with the sound.
Where do I look to figure out why the highs are so shouty?
Where do I look to figure out how to get tighter bass? Right now it is quite mushy. The bass come quick, but not well defined.
I read that having the 6C45P driving 300B might have some high roll off. It seems to be OK or am I not hearing the right thing?
I am thinking of the following changes as potential next steps.
- Fixed bias - Maybe having another 65V or so plate voltage will give more headroom?
- interstage - how much improvement will I see from that?
- maybe choke loaded? or CCS?
- separate B+ supply for the 6c45p - is this worth it? i might squeeze a few more volts for the 300B B+ and also some separation for the 2 stages
- penthode driver
Ultimately I want to make this amp sound very good.
sorry, lots of questions.
Last edited:
The 6C45P is a high Gm tube, a type that tends to oscillate quite easily at very high frequencies. This can manifest as the "steely" or shouty sound that you have been observing. Increase the driver tube's grid stopper to a 1kOhm carbon comp resistor and solder the body of the resistor as close to the socket pin as possible.
As for the loose bass, you might want to connect to the 4 Ohm tap if you haven't done so already.
Your other thoughts about separate supplies for input and power sections are good and should be implemented if you are inclined to further experiment as you've said. Another thing you could try, is to replace the last cap (1000uF) of the outboard PSU with a much smaller, say ~100uF oil cap. Use the same oil cap for the cathode bypass of the 300B.
As for the loose bass, you might want to connect to the 4 Ohm tap if you haven't done so already.
Your other thoughts about separate supplies for input and power sections are good and should be implemented if you are inclined to further experiment as you've said. Another thing you could try, is to replace the last cap (1000uF) of the outboard PSU with a much smaller, say ~100uF oil cap. Use the same oil cap for the cathode bypass of the 300B.
Shall we do some proper calculation?
With 1 VRMS input signal the 6C45 will provide the 300B grid with some 35 VRMS (it is RC coupled so gain is not like mu!). 35 VRMS is ~ 100V pk-pk, and you would need ~ 150V pk-pk to drive a standard 300B in class A1. So you need ~ 1,5 VRMS at the grid of the 6C45; that's a bit more than what digital based sources provide on average.
But maybe you have some sort of line stage in front of the amplifier? An interstage transformer has several advantages:
- gain is like mu, so 1 VRMS input signal enough to drive the output stage;
- no risk of blocking distortion driving the 300B in class A2.
By the way, SE 300B amps begin to "shine" with 95+ dB loudspeakers...
With 1 VRMS input signal the 6C45 will provide the 300B grid with some 35 VRMS (it is RC coupled so gain is not like mu!). 35 VRMS is ~ 100V pk-pk, and you would need ~ 150V pk-pk to drive a standard 300B in class A1. So you need ~ 1,5 VRMS at the grid of the 6C45; that's a bit more than what digital based sources provide on average.
But maybe you have some sort of line stage in front of the amplifier? An interstage transformer has several advantages:
- gain is like mu, so 1 VRMS input signal enough to drive the output stage;
- no risk of blocking distortion driving the 300B in class A2.
By the way, SE 300B amps begin to "shine" with 95+ dB loudspeakers...
I hear you! but at the moment... the speakers I have will have to do. Space constraints..
I have the 330 Ohm solder directly to the socket pin. I will try the 1k tonight.
my speakers are 4 Ohm and they are connected to the 4 Ohm tap already.
as for the cap suggestion. can you explain why? the 1000uF is a pair of 2000uF (300V I think) electrolytics in series and bypassed with a 33uF motor run oil cap. Is going smaller capacitance better? I always been told the more the better.
do you have any suggestion for oil cap? i have some more Obbligato but they are rather large to get them to 100uF.
so EL34/6L6 is sharp? another thought I had was to make a 3 stage with a 6L6 driving the 300B (kinda copying the Wavac EC300B design)
Shall we do some proper calculation?
With 1 VRMS input signal the 6C45 will provide the 300B grid with some 35 VRMS (it is RC coupled so gain is not like mu!).
35 VRMS is ~ 100V pk-pk, and you would need ~ 150V pk-pk to drive a standard 300B in class A1.
So you need ~ 1,5 VRMS at the grid of the 6C45; that's a bit more than what digital based sources provide on average.
But maybe you have some sort of line stage in front of the amplifier?
An interstage transformer has several advantages:
- gain is like mu, so 1 VRMS input signal enough to drive the output stage;
- no risk of blocking distortion driving the 300B in class A2.
By the way, SE 300B amps begin to "shine" with 95+ dB loudspeakers...
I have a preamp which is almost maxed out at the input POT. It is at 4-5 o'clock! poor thing never worked that hard. But I think that gave the 1.5Vrms need to drive the 6C45P to drive the 300B.
How does CCS sound compared to interstage? CCS also give gain=mu with a flat load line. I intend to try that next as CCS is way cheaper to experiment than interstage.
"it depends"
The interstage transformer, as all signal transformers, has a limited frequency range falling off at both ends. Extending one bites the other.
The insterstage transformer allows A2 operation.
The interstage transformer introduces no blocking distortion as opposed to a capacitor.
I don't know how to describe "how it sounds" in any useful way. Personally I don't use coupling capacitors as they all come with a signature I don't like.
The interstage transformer, as all signal transformers, has a limited frequency range falling off at both ends. Extending one bites the other.
The insterstage transformer allows A2 operation.
The interstage transformer introduces no blocking distortion as opposed to a capacitor.
I don't know how to describe "how it sounds" in any useful way. Personally I don't use coupling capacitors as they all come with a signature I don't like.
some manufacturers like Monolith claims a really wide frequency range, from single digit to 200+kHz. is that trustable? others like the Japanese ones tends to be in the 20Hz up.
How would you describe the sound of interstage compared to coupling caps?
the supply for the 6C45P is at 400V right now. Eventually I will modifying the amp to go interstage and fixed bias, but with 400V and -70V or so at the grid that's 470V DC between the windings.
I think the non-bifilar ones can take that DC voltage, but the bifilar ones tends to be lower at 375VDC.
is that the distinction? but in general ppl seem to recommend bifilar in terms of sound.
How would you describe the sound of interstage compared to coupling caps?
the supply for the 6C45P is at 400V right now. Eventually I will modifying the amp to go interstage and fixed bias, but with 400V and -70V or so at the grid that's 470V DC between the windings.
I think the non-bifilar ones can take that DC voltage, but the bifilar ones tends to be lower at 375VDC.
is that the distinction? but in general ppl seem to recommend bifilar in terms of sound.
Cascode CCS and LED based bias on the cathode is my current recipe. Works quite well.
I'd be surprised if you can get more than 80Vpp out of a 6C45 which is good for a 45 or 2A3 but short of what's needed for a 300B. The D3A in triode connection will do >180Vpp comfortably on 150V plate voltage with CCS (250 - 300V supply needed) or choke loading.
The quality of the cathode bypass capacitor on the 300B is pretty critical. I use fixed bias with typically about -70 to -80V on grid depending on the plate supply voltage - eliminates that bypass capacitor.
The D3A, 6C45 and other high transconductance types are very prone to oscillation and need careful layout. (short wiring, solid grounding at RF, ferrite beads, etc.)
I'd be surprised if you can get more than 80Vpp out of a 6C45 which is good for a 45 or 2A3 but short of what's needed for a 300B. The D3A in triode connection will do >180Vpp comfortably on 150V plate voltage with CCS (250 - 300V supply needed) or choke loading.
The quality of the cathode bypass capacitor on the 300B is pretty critical. I use fixed bias with typically about -70 to -80V on grid depending on the plate supply voltage - eliminates that bypass capacitor.
The D3A, 6C45 and other high transconductance types are very prone to oscillation and need careful layout. (short wiring, solid grounding at RF, ferrite beads, etc.)
I use MM amorphous core interstages in some of my amplifiers, they work very well. With triode connected D3A, 180V supply is more than sufficient.
Long before Damir had tested a lot of drivers for the 300B SET in another forum. These are his notes on the 6C45Pi driver:
AudioRoundTable.com: Group Build >> 300B SE Project, Part 7 - Common cathode 6C45Pi triode driver
You would also like to read about other drivers, especially C3g
Of course, you can fully believe what Damir wrote-
Best diying
AudioRoundTable.com: Group Build >> 300B SE Project, Part 7 - Common cathode 6C45Pi triode driver
You would also like to read about other drivers, especially C3g
Of course, you can fully believe what Damir wrote-
Best diying
Last edited:
Your current operating point for the driver tube as calculated from the schematic you provided puts the plate to cathode voltage at around 165V or at about 168V plate-to-ground. IOW, there's 232 Volts dropped through the load resistor. The IT being a coil, presents a high impedance at AC while having very low DCR it drops only a handful of volts across its primary.
If you were to use an interstage in place of the RC coupling scheme while using the same operating point, you will need to drop the B+ voltage entering the IT down to around 168V ground referenced. That's because the IT presents a load that's close to a horizontal line and is also able to swing twice the B+ voltage. So, in this scenario the IT primary sees 168V while the secondary is grounded (assuming you retain the cathode bias scheme) and you will be well within the IT's voltage limits.
If you go fixed bias, then the -70V is added to the 168V for a total of ~240V across the windings - still well within margin.
PS. to drop to 168V from the B+ of 435V you will need a dropping resistor (R5) of 22.9KOhm. This is based on the calculated current through the tube of 11.8mA. Of course your measurement of this current will better determine the resistor to be used. Keep in mind that this tube exhibits high variability in its parameters between different samples, so unless these tubes are well matched, you might find that your measurements between the two drivers might differ quite a bit.
Last edited:
It's generally easier to achieve success with more parts doing less work than with less parts doing more work.
I was thinking it would be easier to achieve success with a 3-stage design than with a 2-stage.
Perhaps 6J5 input tube driving a smaller low-mu triode like 4P1L, 12B4, or triode-strapped EL86, driving the 300B?
If the 2nd stage had a plate choke and was DC-coupled to the 300B, that could help minimize blocking and grid current effects, etc.
Just an idea...
I was thinking it would be easier to achieve success with a 3-stage design than with a 2-stage.
Perhaps 6J5 input tube driving a smaller low-mu triode like 4P1L, 12B4, or triode-strapped EL86, driving the 300B?
If the 2nd stage had a plate choke and was DC-coupled to the 300B, that could help minimize blocking and grid current effects, etc.
Just an idea...
With those 300Bs THE very first thing I'd do is go to a soft start dc filament supply like Rod Coleman's kits After that, Bob's your uncle.
Before you do anything else try replacing the 22uF Obbligatos with some standard 47uF electrolytics in series. Or even some cheap Solens--anything but those Obbligatos. I tried Obbligatos and found them unbearable in the highs. I was swapping cables, preamp, tubes, everything. Almost tore the whole project apart before I figured out they were the problem. Remove the 6.8uF Obbligato bypass cap as well.
For the external power supply, I'd lose the 1000uF. 100 is plenty. Heck, 47 is plenty. You might improve the bass that way. Also, are you really using a 200 ohm balance pot on the 300Bs? That's too high. 25-50 ohms is plenty.
In the end you might want to try a different driver tube. The Loesch Legacy 300B using an EL84 pentode driver would be an easy swap and it's very nice.
Nice looking amps, BTW!
For the external power supply, I'd lose the 1000uF. 100 is plenty. Heck, 47 is plenty. You might improve the bass that way. Also, are you really using a 200 ohm balance pot on the 300Bs? That's too high. 25-50 ohms is plenty.
In the end you might want to try a different driver tube. The Loesch Legacy 300B using an EL84 pentode driver would be an easy swap and it's very nice.
Nice looking amps, BTW!
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.