My first amplifier - 6n3c-E

[joachimlarsen]

For the last 15-20 years I have been interested in valve/tube radios, and I have a small collection. Recently I decided to build my first valve amp. I have the (very) basic knowledge about electronics and valves, but I am not an expert, and I am trying to keep my first "project" very simple - and cheap.

I made a working "test" amp with one half ECC83 and one EL84 single ended, based upon the 'RH 84 SE' I found on the Internet.
I made it from the parts I had. Some resistors didn't have the right values, and the output transformers had to high impedance. BUT IT WORKED! Off course it was just a "test", and I was surprised how easy it was to build a working amp. Well, it was not perfect, but who cares? I was fun to make.

Now I'm trying to build an upgraded version - this time with bigger output valves. I don't need more power - in fact the EL84 gave more than enough. I just want a cleaner sound with more (soft) bass and "room". I'm not sure, if its a good idea to use a large valve at low output load, instead of a smaller one at higher load.
I don't want to build PP, because I don't need the power. PP demands larger power transformers, and I rather save the money and build another amp instead. So I go the SE way.

This is my design

http://www.h783.dk/valve

First EL34 and KT66 came into my mind, but then I found four NOS Russian 6p3c-E on the big E-place. I don't know if they are good valves, but I think they are "good enough" for my first valve project.
I bought a pair of Hammond 125 ESE transformers rated at 80mA. I think the are big enough for a test run. Later I might replace them with a pair of 1628SEA's and use the 125's for a new (smaller) project (EL84 or ECL82 SE-P). That's what its all about: The NEXT project! Why haven't I thought of building amplifiers before? I enjoy it a lot. In fact I am planning to create some kind of 'vintage' amp with classic European valves like CL4 or UBL1 SE with a preamp stage made of Philips "red" valves. That would look great!

And now the questions! I modified one channel of the 'RH 84' and replaced the EL84 with the 6p3c-E, but I don't really know how to set the bias for this valve. According to the valve data I have read, the maximum plate voltage is 250 V. My bias measures 20 V above ground, and with a plate voltage of 240, the plate current is 30 mA. That's well below the recommended 73 mA, but is it TO LOW? I think the sound is clear and warm, and I actually like the sound and really don't need more volume. On the other hand, I don't know how much volume I can expect from this valve. So the questions is: Do you think I have to change the bias resistor, and how would it change the sound and/or volume?

The next question is about the driver stage: The ECC83. I am aware of the difference between 81/82 and 83. The ECC83 is a high gain preamp, but the sound is much more "direct", and I have several used Telefunken's to plug in. I find it very difficult to control the gain of this preamp. If I increase the size of the plate resistor, the amplification can get too high, and the output gets distorted. I have seen several layouts with big resistors, some of them around 10-20K Ohm, but if I do that, the sound gets to loud to feed into the output stage. Obviously I am doing something wrong - but what? Wrong ECC83 bias? How much power can I feed in to the output stage? Do you think I should "boost" the output stage instead? To be honest, I like to be well below the limit - and running both the preamp and output stage at "70%" max.

I am not sure if I am going to keep the ECC83 or replace it with a 6N7GT I have in my collection. It's the same octal socket as the output valves, but its not very important how the final amp is going to look.

This is my first real valve amp, and because I want to make it simple, I don't expect a "perfect" sound. The fundamental idea is to get a "safe" amp - not red hot glowing plates. Somewhere between 1 and 5 watts is excellent. Why not keep the EL84? Well, it's to easy.

Any comments are welcome. Its a bit difficult to explain this in English, but I certainly hope you get the basic idea of what I am trying to build.

>Joachim

[dsavitsk]

Hi Joachim,

Nice project. A couple of quick comments.

First, the load on the first tube should be several times higher than the tube's rp (plate impedance.) You can generally find the rp in the datasheet. For the ECC83, rp is between 60K and 80K depending on where you run it. So, you need a load higher than that. Also, the amplification factor of that tube is 100x which maybe too high. On the other hand, a 6sn7 or 12AU7 has an amplification factor closer to 20 and an rp of 8k or so which might be a little more to your liking. For an in between choice, perhaps a 12AT7, of 6N1p if you like the Russian tubes.

Second, if you connect the output tubes second grid to the plate with a small resistor rather than to B+, you will reduce power, and get arguably better sound -- this essentially converts it from a pentode to a triode.

[ErikdeBest]

Hi Joachim

Just to help you get a better grasp on what is going on in the circuit you may have a look at this tutorial, where the author explains the building of a SE amplifier (driver tube + output tube + trafo).

http://boozhoundlabs.com/howto/

Though Morgan Jones is mostly into PP, his book (3rd edition) also shows the development of a SE design with lots of details. Surely worth a read!

Erik

[soundbrigade]

Now that you have 4 tubes - 6P3S, you can run them in parallell and double the power.

Denmark eh?! Röde pölse med smörrebröd ... You're close to Germany from where you can get decent transformers if/when you want to upgrade.
Check Welter Electronics, As Jan Zuerst and Röhrentechnik.

[Sheldon]

Originally posted by dsavitsk
Hi Joachim,

Nice project. A couple of quick comments.

First, the load on the first tube should be several times higher than the tube's rp (plate impedance.) You can generally find the rp in the datasheet. For the ECC83, rp is between 60K and 80K depending on where you run it. So, you need a load higher than that. Also, the amplification factor of that tube is 100x which maybe too high. On the other hand, a 6sn7 or 12AU7 has an amplification factor closer to 20 and an rp of 8k or so which might be a little more to your liking. For an in between choice, perhaps a 12AT7, of 6N1p if you like the Russian tubes.

Second, if you connect the output tubes second grid to the plate with a small resistor rather than to B+, you will reduce power, and get arguably better sound -- this essentially converts it from a pentode to a triode.

This design utilizes partial feedback on the output stage - so not really pentode like. A detailed discussion here, including some discussion on the second page about using a high impedance input stage: Partial Feedback Amplifiers

Sheldon

Edit: Looking at the schematic, I notice that it does not have the feedback resistor for the RH design. In which case Dave's comments apply. In fact, the load resistor for the input stage is waaay too low, if the schematic is accurate. Joachim, you mentioned that you based this on Alec's design. Is there some reason you didn't follow through on that implementation?

[joachimlarsen]

I was trying to keep the signal level down, so I changed the load resistor (not the right way to reduce power I understand, but in a desperat attempt to do something, and it worked).
This morning before leaving for work, I made a few changes according to the answers I got on this forum (thank you!). I was lucky to find one ECC82 and replaced the plate load resistor to 22K. I also raised the bias resistor to 1.2K to lower the plate current and reduce the amplification factor. Now I have more power from the output valve, and the signal is just a little bit distorted, when I turn the volume control to full. I need to fine tune this design and forget about the ECC83 (or ECC808).

I also changed the bias resistor to the 6n3c-e to around 0.3K to increase the plate current, but I don't think the sound got louder or better at all. Now the bias is back at 0.68K. Do I risk something, if I have to low plate current?

by the way it's a great link (thanks again)
http://boozhoundlabs.com/howto/

And I really like this forum.

[joachimlarsen]

Edit: Looking at the schematic, I notice that it does not have the feedback resistor for the RH design. In which case Dave's comments apply. In fact, the load resistor for the input stage is waaay too low, if the schematic is accurate. Joachim, you mentioned that you based this on Alec's design. Is there some reason you didn't follow through on that implementation?

...sorry, I forgot something

Well, the volume was very high, because I used the high gain ECC83 (wrong tube for this design?). I didn't see the point in having a feedback resistor. I'm not sure if I do now. Do I need feedback? And why a resistor - not at capacitor?

<Joachim

[Sheldon]

Originally posted by joachimlarsen
Well, the volume was very high, because I used the high gain ECC83 (wrong tube for this design?). I didn't see the point in having a feedback resistor. I'm not sure if I do now. Do I need feedback? And why a resistor - not at capacitor?

<Joachim

If you want the amp to work as designed, yes, you need the feedback resistor. The point of the feedback resistor is to provide feedback (the clue is in the name). It will reduce distortion, and it will lower your gain. Regarding your question about a capacitor vs. a resistor: There is no either or, you need both in this circuit (again, if you want it to work as the designer intended). Look at the RH amp schematic you said you based your design on. The capacitor blocks DC from the driver to the grid. The feedback resistor goes from output plate to driver plate. Try 100k. Try half and double that value if you like, and see which sounds best, then tune some more to taste.

Joachim, you said you based this design on the RH amp. I applaud your desire to try a few new twists, but simply leaving out a critical element of the design defeats the purpose of following it in the first place. "Hmm, I don't understand what this resistor is doing, I'll just leave it out", is not an approach I'd recommend.

Now, if you want just a basic SE amp, do as Dave suggested and connect the pentode screen to the plate and leave out the feedback resistor. But that is NOT a variation on the RH design. If you want to play a little, do both and compare.

Sheldon

[Sheldon]

Originally posted by joachimlarsen
And why a resistor - not at capacitor?

<Joachim

Do you mean can't the feedback resistor be replaced with a capacitor? No, not if you want any sound to come out. The amount of feedback is a function of the ratio between the feedback resistor and the output impedance of the driver. Let's say you use an infinitely large resistor (same as leaving it out), well no feedback, right? If you use a very small resistor (piece of wire) you will have 100 percent feedback, and as you try to move the grid, the output from the plate will counteract it completely - so no change in current, no signal out.

If you place a capacitor where the feedback resistor should go, then for AC purposes you have no resistance, hence 100 percent feedback, hence no sound again.

Sheldon

[joachimlarsen]

If you place a capacitor where the feedback resistor should go, then for AC purposes you have no resistance, hence 100 percent feedback, hence no sound again.

Sheldon [/B][/QUOTE]

-

Now I understand! It's very kind of you to explain it in basic words. In fact I have just reconnected the feedback resistor, and I am surprised how different the sound is. More bass, less distortion - just a lot better.
I admit I was working backwards, because I was removing components I "didn't understand". Maybe I was working to fast to get the amp together, so I forgot to think. I'll try to follow the schematics from now on (almost).

Thank's again
-Joachim