Hi all,
I am from the chipamp forum.
Last months I had some discrete experiences, but I cannot hear much differences between chipamps and discrete amps. I think its not worth to make a discrete (bipolar or MOSFET) class AB amp in anycase. Except a single end class A project.
I made a single end 100mW headphone amp with just a IRF510 MOSFET and saw the diffierence..
After that experience, I tried to make a bigger Class A for my speakers however I coulndt solve the thermal problems and for now solid state high power class A idea is dead for me
Anyway,
Now coming my real question.
I want to try to make a tube amplifier. It must be a single end Class A one and 20W per channel at least. However I have no idea about the tubes.
May you advice a
- Really simple,
- Adjustment free (or minimal),
- Cheap, good sounding (better than a solid state AB)
and of course with more available parts tube amplifier pls?
Thanks a lot in advance for your helps.
I am from the chipamp forum.
Last months I had some discrete experiences, but I cannot hear much differences between chipamps and discrete amps. I think its not worth to make a discrete (bipolar or MOSFET) class AB amp in anycase. Except a single end class A project.
I made a single end 100mW headphone amp with just a IRF510 MOSFET and saw the diffierence..
After that experience, I tried to make a bigger Class A for my speakers however I coulndt solve the thermal problems and for now solid state high power class A idea is dead for me
Anyway,
Now coming my real question.
I want to try to make a tube amplifier. It must be a single end Class A one and 20W per channel at least. However I have no idea about the tubes.
May you advice a
- Really simple,
- Adjustment free (or minimal),
- Cheap, good sounding (better than a solid state AB)
and of course with more available parts tube amplifier pls?
Thanks a lot in advance for your helps.
You end up with some kind of terrible tube crying for 1200V and loads of fresh air to heat up ....
To be serious - you may end up with tubes like 211, 845 or 813 that gives the required power. There's a rather unkinky Russian tube (hard to find) GU-72 that is suitable too.
OR you can go for PSE - parallelled tubes like EL34, 6550 or just a pair of good old 6L6 or 807.
I would recopmmend you the second solution as you can use rather kind stuff like transformers that isn't constructed to withstand 10kV or PSUs that gives Total Makeover a new meaning ....
To be serious - you may end up with tubes like 211, 845 or 813 that gives the required power. There's a rather unkinky Russian tube (hard to find) GU-72 that is suitable too.
OR you can go for PSE - parallelled tubes like EL34, 6550 or just a pair of good old 6L6 or 807.
I would recopmmend you the second solution as you can use rather kind stuff like transformers that isn't constructed to withstand 10kV or PSUs that gives Total Makeover a new meaning ....
Hi!
I think that 20W output power SE is often too hard for a beginner in tubes. IMHO 5W is a good starting project. IMHO output transformer is a key component in a tube amp. Therefore I could recommend that you first check what OTs you can afford (they are typically expensive...) and then consider tubes selection. Maybe Hammond OTs is a good thing to start with. There are also Russian OTs like VR20CSE (25W out, 120mA DC max, plate resistance depends on windings connection 2.5 ... 10 kOhm, loads 4 / 8 / 16 Ohm, ~$90).
Some plate resistance - output power - primary voltage amplitude - current amplitude are (95% OT efficiency assumed):
2 500Ohm - 5W - 162V - 65mA
2500 - 10 - 229 - 92
2500 - 20 - 324 - 130
5 000 Ohm - 5W - 229V - 46mA
5000 - 10 - 324 - 65
5000 - 20 - 459 - 92
Note that you need both higher power supply and anode bias current than the one needed for primary swing. It depends already on the tube how much more... Anyway, a 400V 100mA means 40W anode dissipation (to hope for 10W output power). From Soviet tubes I know I could recommend 6p45s (6n45C in Bulgarian letters, EL509 equivalent), in Bulgaria they are ~$15/pc and you would not need to parallel tubes. Well, there are much more tubes that would work in 400V 100mA, like, for example, SU GU50, LS50 clone that is several times cheaper - I tested one just yesterday and I liked the sound I got at a glance. I think you will find out that the OT is much more expensive than many tubes. You might like to have a look at the prices of one of the most famous audio DHT, 300B and OTs from Tango and Tamura... And I think that results from "whatever" transformer and tube amp are likely to be unpleasant and confusing.
When selecting OT you need to keep in mind that for triodes (I assume triode SE class A operation everywhere herein...) Rp/Ri (primary impedance to internal triode impedance) ratio needs to be higher than 3 for good distortion and damping factor performance. 6p45s/EL509 has Ri of about 300 Ohm, while GU50 has about 1 kOhm, for example - so a Rp/Ri=5 would be 1.5 kOhm vs 5 kOhm. I myself like higher Rp, but if you are chasing higher output power you might need to consider other options, too...
Best regards,
Andrey
I think that 20W output power SE is often too hard for a beginner in tubes. IMHO 5W is a good starting project. IMHO output transformer is a key component in a tube amp. Therefore I could recommend that you first check what OTs you can afford (they are typically expensive...) and then consider tubes selection. Maybe Hammond OTs is a good thing to start with. There are also Russian OTs like VR20CSE (25W out, 120mA DC max, plate resistance depends on windings connection 2.5 ... 10 kOhm, loads 4 / 8 / 16 Ohm, ~$90).
Some plate resistance - output power - primary voltage amplitude - current amplitude are (95% OT efficiency assumed):
2 500Ohm - 5W - 162V - 65mA
2500 - 10 - 229 - 92
2500 - 20 - 324 - 130
5 000 Ohm - 5W - 229V - 46mA
5000 - 10 - 324 - 65
5000 - 20 - 459 - 92
Note that you need both higher power supply and anode bias current than the one needed for primary swing. It depends already on the tube how much more... Anyway, a 400V 100mA means 40W anode dissipation (to hope for 10W output power). From Soviet tubes I know I could recommend 6p45s (6n45C in Bulgarian letters, EL509 equivalent), in Bulgaria they are ~$15/pc and you would not need to parallel tubes. Well, there are much more tubes that would work in 400V 100mA, like, for example, SU GU50, LS50 clone that is several times cheaper - I tested one just yesterday and I liked the sound I got at a glance. I think you will find out that the OT is much more expensive than many tubes. You might like to have a look at the prices of one of the most famous audio DHT, 300B and OTs from Tango and Tamura... And I think that results from "whatever" transformer and tube amp are likely to be unpleasant and confusing.
When selecting OT you need to keep in mind that for triodes (I assume triode SE class A operation everywhere herein...) Rp/Ri (primary impedance to internal triode impedance) ratio needs to be higher than 3 for good distortion and damping factor performance. 6p45s/EL509 has Ri of about 300 Ohm, while GU50 has about 1 kOhm, for example - so a Rp/Ri=5 would be 1.5 kOhm vs 5 kOhm. I myself like higher Rp, but if you are chasing higher output power you might need to consider other options, too...
Best regards,
Andrey
Patrick Turner has a 35W SE amp (4 x EL34) - in triode mode about half that. http://www.turneraudio.com.au/se35cfbmonobloc.html Not TERRIBLY complicated except for obtaining the output transformer (He makes his own).
5-6W would be MUCH easier, with a single EL34 or 6550 output tube. Search for "Tubelab simple SE".
5-6W would be MUCH easier, with a single EL34 or 6550 output tube. Search for "Tubelab simple SE".
For cheap output transformers you might like to have a look at this article. Well, it is significantly below 10W (I would expect couple of decent watts) and I have not tested it, but you could use the transformers for filament supply in case you don't like the sound...
Best regards,
Andrey
Best regards,
Andrey
Hi Tom,
It doesnt look like terribly complicated. But I am newbie and I dont think I can have a success start up with it. Its is a bit risky for me I guess.
My first target is to have a real class A amplifier, however my experiences are on solid state and I expect to have a simple and ralitively "cold" amplifier.
Thanks a lot.
Dear Andro,
I also think like you on this matter. After a short internet searching, I found some OT projects. One of them is Andrea Ciuffoli's SEPP project. Its very simple and relatively cheap one. He uses 6C33C-B x 2 as output pairs, ECC82 to drive them and as input pairs...
Anyway, Its very look like a solid state amplifier.
This is it;
However I am confused on SEPP concept! How can a SE output become a Push Pull at the same time? Whats that? Is it possible to prevent crossover distortion with that way?
And a final question; tubes in this project widely available and cheap ones as Andrea says?
Thanks...
It doesnt look like terribly complicated. But I am newbie and I dont think I can have a success start up with it. Its is a bit risky for me I guess.
My first target is to have a real class A amplifier, however my experiences are on solid state and I expect to have a simple and ralitively "cold" amplifier.
Thanks a lot.
Dear Andro,
I also think like you on this matter. After a short internet searching, I found some OT projects. One of them is Andrea Ciuffoli's SEPP project. Its very simple and relatively cheap one. He uses 6C33C-B x 2 as output pairs, ECC82 to drive them and as input pairs...
Anyway, Its very look like a solid state amplifier.
This is it;
However I am confused on SEPP concept! How can a SE output become a Push Pull at the same time? Whats that? Is it possible to prevent crossover distortion with that way?
And a final question; tubes in this project widely available and cheap ones as Andrea says?
Thanks...
Think of something like the RH84 or the RH807 , but with a 6550/KT88 as the output tube. This gives triode like performance but with pentode like power. You might expect to get about 10W of power from a KT88, but you are likely to be surprised at how far those 10W go compared to a SS amp.
This would also be one of the cheapest approaches as the RH designs place much less of a demand upon the Output transformer and so it can be cheaper. It will also produce one of the best bass responses of any SE design.
There is also the option of going PP - its easy to get your 20W this way.
Shoog
This would also be one of the cheapest approaches as the RH designs place much less of a demand upon the Output transformer and so it can be cheaper. It will also produce one of the best bass responses of any SE design.
There is also the option of going PP - its easy to get your 20W this way.
Shoog
Hi!
I did not recommend 6S33S (6C33C) for several reasons:
- I could not get a really nice sound from these myself - well, under some conditions I liked some music much, but not all...
- note that the filament is 44W per tube...
- they are hard to drive (you need to develop about 200 Vp-p on ~200pF input capacitance, and the grid resistor should be less than 100 kOhm for stable operation with higher voltage and/or plate dissipation).
Add: OTL approach is something I would not recommend - I see no normal way to ensure decent Rp/Ri value and therefore the distortion needs much more complicated schematic, NFB etc... which is not the best tubes can give us IMHO ;-) I like two stage SE designs. But the schematic is your last concern, I think - it is, IMHO, much more important to get the right concept for your case. My own opinion is that very few things can be done with tubes without output transformers - in other words, trying tubes OTL is not the real thing IMHO. However, here are couple of links with an interesting approach for OTL:
http://members.aol.com/sbench102/inv6as7.html
http://members.aol.com/sbench102/inverted.html
For the tube watts and SS watts - I didn't wanted to say that myself
Anyway, moving from SS to tube amps is often accompanied with new understanding of the output power needed...
Sorry, not much time now :-(
Best regards,
Andrey
I did not recommend 6S33S (6C33C) for several reasons:
- I could not get a really nice sound from these myself - well, under some conditions I liked some music much, but not all...
- note that the filament is 44W per tube...
- they are hard to drive (you need to develop about 200 Vp-p on ~200pF input capacitance, and the grid resistor should be less than 100 kOhm for stable operation with higher voltage and/or plate dissipation).
Add: OTL approach is something I would not recommend - I see no normal way to ensure decent Rp/Ri value and therefore the distortion needs much more complicated schematic, NFB etc... which is not the best tubes can give us IMHO ;-) I like two stage SE designs. But the schematic is your last concern, I think - it is, IMHO, much more important to get the right concept for your case. My own opinion is that very few things can be done with tubes without output transformers - in other words, trying tubes OTL is not the real thing IMHO. However, here are couple of links with an interesting approach for OTL:
http://members.aol.com/sbench102/inv6as7.html
http://members.aol.com/sbench102/inverted.html
For the tube watts and SS watts - I didn't wanted to say that myself
Anyway, moving from SS to tube amps is often accompanied with new understanding of the output power needed...Sorry, not much time now :-(
Best regards,
Andrey
Is there TUBE WATT and SEMICONDUCTOR WATT exists?
At the moment I am using a lateral MOSFET output class AB amp for my living room. Its 150W rated per channel but I need only 10 or 15W per channel, why I use that much power is just for headroom I need.
Anyway,
I have a pair of Wharfedale Diamond 9.6, theyre not "hard to drive" speakers (91dB at 1 semiconductor watt ) but I dont think just 10 watt amplifier will be enough for me!! Or am I mistaken?
At the moment I am using a lateral MOSFET output class AB amp for my living room. Its 150W rated per channel but I need only 10 or 15W per channel, why I use that much power is just for headroom I need.
Anyway,
I have a pair of Wharfedale Diamond 9.6, theyre not "hard to drive" speakers (91dB at 1 semiconductor watt
) but I dont think just 10 watt amplifier will be enough for me!! Or am I mistaken?I have a friend that built couple of single triode (per channel) amp,
6S45P (6C45n) and 6ZH52P (6*52n) - well, he's well below 1W (at least with the 6S45P one) and he needs volume regulator with his 82 dB boxes. Well, this is not for parties... ;-)
Best regards,
Andrey
6S45P (6C45n) and 6ZH52P (6*52n) - well, he's well below 1W (at least with the 6S45P one) and he needs volume regulator with his 82 dB boxes. Well, this is not for parties... ;-)
Best regards,
Andrey
Andrea Ciuffoli' amplifier is an OTL, (Output Transformer Less), this kind of amplifier has its own difficulties and I wouldn't recommend it to a beginner. Compared the cost of an OTL to a normal transformer coupled amp is difficult, although you don't need an output transformer you need a power transformer in both cases. For an OTL you will also need 2 high quality capacitors for the power supply, something in the order of at least 4700uF for 200V of very good quality in order to minimize hum and noise, these are much more expensive than the smaller ones normally used in a transformer coupled amp.
What is correct is that the tubes 6C33C are quite inexpensive, you can buy them directly from Russia for about 20USD each, this is very competetive compared to many tubes used in SE amps. One idea is therefore to build a SE amplifier using 6C33C, one tube gives 18W and you can connect them in parallell for more power, there are many schematics available on internet.
Finally the OTL shown on Andreas site is not really a SEPP, it is an inverted Futterman push-pull amplifier where both output tubes are driven from a phase splitter. For DC both output tubes are connected in series but for AC they work as they if they are connected in parallell, exactly as a push-pull amplifier, and exactly like in a push-pull amplifier this amplifier have or not have cross over distortion as a push-pull amplifier has, it is quite easy to completely remove cross over distortion in both this OTL and any push-pull amplifier, it really depends on the choice of operating point. For more text about OTL see here http://www.tubetvr.com/otl.html and here http://members.aol.com/aria3/otlpaper/otlhist.htm
Regards Hans
BTW, maybe I should mention that I really believe that a good OTL amp really is the best amplifier types possible to build but that is my own opinion and other may have other preferences.
What is correct is that the tubes 6C33C are quite inexpensive, you can buy them directly from Russia for about 20USD each, this is very competetive compared to many tubes used in SE amps. One idea is therefore to build a SE amplifier using 6C33C, one tube gives 18W and you can connect them in parallell for more power, there are many schematics available on internet.
Finally the OTL shown on Andreas site is not really a SEPP, it is an inverted Futterman push-pull amplifier where both output tubes are driven from a phase splitter. For DC both output tubes are connected in series but for AC they work as they if they are connected in parallell, exactly as a push-pull amplifier, and exactly like in a push-pull amplifier this amplifier have or not have cross over distortion as a push-pull amplifier has, it is quite easy to completely remove cross over distortion in both this OTL and any push-pull amplifier, it really depends on the choice of operating point. For more text about OTL see here http://www.tubetvr.com/otl.html and here http://members.aol.com/aria3/otlpaper/otlhist.htm
Regards Hans
BTW, maybe I should mention that I really believe that a good OTL amp really is the best amplifier types possible to build but that is my own opinion and other may have other preferences.
Hi, Hans!
I believe your opinion for the OTL is more adequate than mine. I have simply not tested anything with OTL. However, a friend of mine built several OT transistor amps ( Nemesis - like) and the results are interesting.
One more thing for beginner's OTL - there are failure modes that may damage the speakers... OT is a pretty effective protection IMHO.
6S33S (6C33C) - yeap, they are cheap (don't forget they need specific sockets...). But I suspect that investing in other tubes, like 6p45s (6n45C), 6S19P (6C19n), EC360, EL36, EL34 etc may be more wise...
Best regards,
Andrey
I believe your opinion for the OTL is more adequate than mine. I have simply not tested anything with OTL. However, a friend of mine built several OT transistor amps ( Nemesis - like) and the results are interesting.
One more thing for beginner's OTL - there are failure modes that may damage the speakers... OT is a pretty effective protection IMHO.
6S33S (6C33C) - yeap, they are cheap (don't forget they need specific sockets...). But I suspect that investing in other tubes, like 6p45s (6n45C), 6S19P (6C19n), EC360, EL36, EL34 etc may be more wise...
Best regards,
Andrey
A Watt is a Watt, regardless of technology. The difference between SS and tubes is the accessability of power capability.
SS amps clip hard. As a consequence, a reserve of power is needed as insurance against the destruction of tweeters. Many more speakers have been damaged by clipping an underpowered amp than have been damaged by too much power.
Tube amps compress, due to saturation, before hard clipping sets in. So, access to the full power capability is possible. Another factor that makes tube amps seem louder at a given power level than SS is the difference in the harmonic distortion spectrum. The 2nd harmonic predominates in many tube circuits. The 2nd harmonic reinforces the fundamental and fools the ear. The sound seems louder than it actually is.
Before a rational decision can be made about the amp. The speakers that will be used MUST be studied. What's the sensitivity of the speakers? What does the speakers' impedance curve look like? Is the curve reasonably flat or does the curve have significant dips, particularly in the power hungry bass region?
Apply Paul Joppa's 102 dB. rule. In a "typical" listening space, an amp/speaker combination should be capable of 102 dB. SPL peaks at a 1 M. distance.
Dear Ozgur,
(Again - I assume transformer SE unless specially noted)
You could assume efficiency of 20...25%. Therefore within the specified ratings of 6P45S you can get 10W per tube. Unlike SS tubes typically withstand more than the specified maximum ratings - but this is a long conversation and I would not dive a beginner in it. OK, say 10W per tube.
You could parallel tubes, yes. But you might need to measure and match them (unless you order matched pairs). I don't like paralleling and I am not really experienced in that area, but I think that with higher transconduction and amplification types oscillation issues increase when paralleling. I am afraid that could be an issue with paralleled 6P45S.
I am sure Google would give you a lot for the EL509, you could use the information for orientation. I guess you have no Bulgarian letters on your PC, so I will give it a try for you...
http://www.google.bg/search?hl=bg&q=6П45С&meta=
Best regards,
Andrey
(Again - I assume transformer SE unless specially noted)
You could assume efficiency of 20...25%. Therefore within the specified ratings of 6P45S you can get 10W per tube. Unlike SS tubes typically withstand more than the specified maximum ratings - but this is a long conversation and I would not dive a beginner in it. OK, say 10W per tube.
You could parallel tubes, yes. But you might need to measure and match them (unless you order matched pairs). I don't like paralleling and I am not really experienced in that area, but I think that with higher transconduction and amplification types oscillation issues increase when paralleling. I am afraid that could be an issue with paralleled 6P45S.
I am sure Google would give you a lot for the EL509, you could use the information for orientation. I guess you have no Bulgarian letters on your PC, so I will give it a try for you...
http://www.google.bg/search?hl=bg&q=6П45С&meta=
Best regards,
Andrey
Dear Eli,
"The speakers that will be used MUST be studied"
What do you mean? Should I make a test and plot a response (impedance) curve before use it?
Interesting... You say, I may burn my 200W RMS speakers with a 15W tube amplifier, is it? Thats really interesting, isnt it possible to compansate impedance unlinearity problems with a snubber circuit on outputz? Or is there a simple way to linearize speakers response curve?
And about the 2nd harmonic issue; As my knowledge.. the odd harmonics are more audiophile than the evens.. The evens look like horns both side of signal but odds doesnt distort the signal as evens, they make the signal as egg shape not more!
Am I right? If so, MOSFET output will have a better sound characteristics..
"The speakers that will be used MUST be studied"
What do you mean? Should I make a test and plot a response (impedance) curve before use it?
Interesting... You say, I may burn my 200W RMS speakers with a 15W tube amplifier, is it? Thats really interesting, isnt it possible to compansate impedance unlinearity problems with a snubber circuit on outputz? Or is there a simple way to linearize speakers response curve?
And about the 2nd harmonic issue; As my knowledge.. the odd harmonics are more audiophile than the evens.. The evens look like horns both side of signal but odds doesnt distort the signal as evens, they make the signal as egg shape not more!
Am I right? If so, MOSFET output will have a better sound characteristics..
No, he said that for SS amps...Dxvideo said:Dear Eli,
"The speakers that will be used MUST be studied"
What do you mean? Should I make a test and plot a response (impedance) curve before use it?
Interesting... You say, I may burn my 200W RMS speakers with a 15W tube amplifier, is it?
Best regards,
Andrey
Sorry for my poor English...
Anyway,
Dear Andrey,
Do you think a stereo 10W SE amp will enough for me? I am not a loudy listener, I am not giving parties for my friends. Its just for my own (also my wife and her sister) listening , and in a 30-35 m2 living room.
If so, I can use one EL509 and theyre widely available in ebay.
Conclusion;
- I must use output transformerred SE design as a beginner,
- I can use EL509 like available tubes as output device,
But there are still question marks exist.
- How can I find output transformers?
- Should I design my own SE amplifier (if so how?)
- Or may you give a really problemless design with that kind of tubes?
Thx again.
Anyway,
Dear Andrey,
Do you think a stereo 10W SE amp will enough for me? I am not a loudy listener, I am not giving parties for my friends. Its just for my own (also my wife and her sister) listening , and in a 30-35 m2 living room.
If so, I can use one EL509 and theyre widely available in ebay.
Conclusion;
- I must use output transformerred SE design as a beginner,
- I can use EL509 like available tubes as output device,
But there are still question marks exist.
- How can I find output transformers?
- Should I design my own SE amplifier (if so how?)
- Or may you give a really problemless design with that kind of tubes?
Thx again.
No what Eli was saying is that you need to know both the efficiency, and also the load impedance over the frequency range of your speakers. This is not just for solid state amplifiers, tube SE amplifiers tend to have high output impedance and limited output power capability and interact with the speaker system in ways that a "perfect" voltage source (read SS amp) usually won't.
You stated that your speakers are 91dB efficiency presumably at 8 ohms (4 ohms??) and 1w or sometimes 2.83Vrms (1W into 8 ohms, but 2w into 4 ohms.) The manufacturer of your speaker system can provide this information. Note in the case that if your speakers aref 4 ohms and 2.83Vrms is the value given the actual 1W efficiency is only 88dB. (Marketing games)
What you are looking for in terms of impedance over frequency is something that looks relatively flat, large impedance maxima are typical at the woofer and box resonance frequencies, and these shouldn' t be too big an issue, what you really want to avoid are very low impedance and large dips in impedance at any frequency.
A 20W DHT SE is not a good place to start, with 30+ yrs of experience I have yet to tackle one with this much power - mostly too expensive and I am not that enthusiastic about >1KV plate voltages. (I like pure single output triode designs ) Voltages are very high, representing a significant hazard, and parts are expensive. (In the past I've designed and built PP amps up to about 150Wrms.)
I second the endorsement for the tubelab simple se design - it's a very good place to start, with a single KT88 and UL operation you might get to 100dBspl or so - if you don't listen extremely loud this may be more than enough to get you where you want to go. Parallel output tubes and possible UL operation are options to get you more power. Contact George over at tubelab and see what he recommends.
Take a look at this thread for more on the tubelab simple se:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=107665
As a beginner you need something simple to start and get your feet wet. My first SE amplifier design was a 1.6W SET (single ended triode) running a pair of 94dB efficient speakers - and it got surprisingly loud in a 3.2 M x 5M room with 2.5M ceiling.
I abandoned solid state hifi in my mid 20's and have never looked back. I'd take that plunge again in a moment.
You stated that your speakers are 91dB efficiency presumably at 8 ohms (4 ohms??) and 1w or sometimes 2.83Vrms (1W into 8 ohms, but 2w into 4 ohms.) The manufacturer of your speaker system can provide this information. Note in the case that if your speakers aref 4 ohms and 2.83Vrms is the value given the actual 1W efficiency is only 88dB. (Marketing games)
What you are looking for in terms of impedance over frequency is something that looks relatively flat, large impedance maxima are typical at the woofer and box resonance frequencies, and these shouldn' t be too big an issue, what you really want to avoid are very low impedance and large dips in impedance at any frequency.
A 20W DHT SE is not a good place to start, with 30+ yrs of experience I have yet to tackle one with this much power - mostly too expensive and I am not that enthusiastic about >1KV plate voltages. (I like pure single output triode designs ) Voltages are very high, representing a significant hazard, and parts are expensive. (In the past I've designed and built PP amps up to about 150Wrms.)
I second the endorsement for the tubelab simple se design - it's a very good place to start, with a single KT88 and UL operation you might get to 100dBspl or so - if you don't listen extremely loud this may be more than enough to get you where you want to go. Parallel output tubes and possible UL operation are options to get you more power. Contact George over at tubelab and see what he recommends.
Take a look at this thread for more on the tubelab simple se:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=107665
As a beginner you need something simple to start and get your feet wet. My first SE amplifier design was a 1.6W SET (single ended triode) running a pair of 94dB efficient speakers - and it got surprisingly loud in a 3.2 M x 5M room with 2.5M ceiling.
I abandoned solid state hifi in my mid 20's and have never looked back. I'd take that plunge again in a moment.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.