Hello,
I am new to this forum and hope I am in the right place.
I am in the process of upgrading my stereo to a modular system using an active crossover. The system will be 4-way. I will need 4 amps per channel. The frequency ranges will be:
20 – 80 Hz into an 18” subwoofer
80 – 400 into a 15” woofer
400 – 6000 into a large horn with phenolic diaphragm
6000+ into a smaller high freq horn with alu diaphragm
I will use a larger ss amp for the subwoofers but want to use tubes for the rest.
I sometimes listen at high volumes, so want to have the option to get there without excess interference or clipping. My speakers are all vintage / refurbished Altecs, all with SPL’s around 100 or more. My listening room is about 20 X 30.
I have built many electronic items from scratch but have only done one tube item.
Does anyone have any recommendations on where to find info to build an 80 or more watt per channel stereo amp on monoblock amp?
I would like to build the best ones I can. Cost shouldn't be an issue (but I have seen kits on some websites that seem a little pricey - do tubes and transformers for such amps really warrant kit costs of thousands of dollars – I really don’t know). I would like to build them from scratch but would do a kit if the markup was reasonable.
My biggest concern is that if given a schematic with values only (ie 10uF, 100 Volts), I will not know what style/grades components to buy.
I would be infinitely grateful if given leads to complete this project.
Thanks in advance,
Jack
I am new to this forum and hope I am in the right place.
I am in the process of upgrading my stereo to a modular system using an active crossover. The system will be 4-way. I will need 4 amps per channel. The frequency ranges will be:
20 – 80 Hz into an 18” subwoofer
80 – 400 into a 15” woofer
400 – 6000 into a large horn with phenolic diaphragm
6000+ into a smaller high freq horn with alu diaphragm
I will use a larger ss amp for the subwoofers but want to use tubes for the rest.
I sometimes listen at high volumes, so want to have the option to get there without excess interference or clipping. My speakers are all vintage / refurbished Altecs, all with SPL’s around 100 or more. My listening room is about 20 X 30.
I have built many electronic items from scratch but have only done one tube item.
Does anyone have any recommendations on where to find info to build an 80 or more watt per channel stereo amp on monoblock amp?
I would like to build the best ones I can. Cost shouldn't be an issue (but I have seen kits on some websites that seem a little pricey - do tubes and transformers for such amps really warrant kit costs of thousands of dollars – I really don’t know). I would like to build them from scratch but would do a kit if the markup was reasonable.
My biggest concern is that if given a schematic with values only (ie 10uF, 100 Volts), I will not know what style/grades components to buy.
I would be infinitely grateful if given leads to complete this project.
Thanks in advance,
Jack
If I needed around 80w, starting from scratch and cost no object, I think I'd be looking at a parallel push-pull triode amp using triode-strapped pentodes or beam tetrodes. If you read Norman Koren's Emperor's New Amplifier you might be interested. It looks complicated, because it is! However, I've heard that the 6550 makes a very good triode and what Norman's designed makes sense (to me, at least).
As to your question about why kits are expensive, the main reasons, I suspect, are that they have to make a profit to make it worthwhile and they need to recoup their purchasing, production and admin costs, as well as their development costs. The volumes they sell are small. All this makes it impossible for them to offer a low selling price. Frankly, I'm surprised they survive at all!
As to your question about why kits are expensive, the main reasons, I suspect, are that they have to make a profit to make it worthwhile and they need to recoup their purchasing, production and admin costs, as well as their development costs. The volumes they sell are small. All this makes it impossible for them to offer a low selling price. Frankly, I'm surprised they survive at all!
Hi Jack
I am busy with final touches on a 120W tubby, but not yet at the stage where I can make a final circuit available. (It is more complex than the Emperor with a number of control circuits. If of interest, I use ECC88, EFC80, E182CC and 4 x 6L6GCs in UL.) So this reply might not be very useful to you, but perhaps a few general pointers might be of interest to you and others (gleaned from 50 years in audio), as you would have realised by now that trying to find direction in the sea of info on the web, often makes it a deeper sea.
1. So many power amp designs use the ECC83/12AX7 as an input stage - careful there. Unless you use a low output-impedance pre-amp (OK, so most are), a thing called Miller-effect (action of the capacitance between grid and anode) renders many designs below par; especially so if the input is glibly taken through a 1 meg volume control. Have that in mid-position resistance-wise and you start losing treble above 6 KHz, apart from phase shift problems with feedback. Folks shun pentode input stages; the increase in noise is academic while frequency response and distortion is superior.
2. About negative feedback (NFB): It is only bad in poor designs; don't always believe old wives' tales. Some sites e.g. Norman Koren (see previous ref.) put this in perspective should you care to read up. There is a whole lot of misunderstanding here often resulting from erroneous conclusions after listening sessions.
3. Same with UL output stages. Yes, ultra-linear is a misnomer (linearity is not superior), but it combines the best of triodes and pentodes, not merely a somewhere in-between. Space here is too limited for an extensive resume, but an example will illustrate(apologies to members reading other threads for repetition). GE quoted for pp KT88s working into a 6K.ohm load: Pentode operation: Max. output (Po) = 47W; output impedance (Re) = 20K; distortion (d) = 3.3%. For UL with 40% screen taps: Po = 50W; Re = 4.2K; d = 1.9%. Triode operation: Po = 23W; Re = 2.6 K; d = 1.1%. Triodes look best, but then d is at full output. Scaling UL down by 23/47 gives only 0.6% at triode max. output, and then you still have the rest in reserve, etc. (Some claim to "hear" that UL is inferior, but hearing is subjective.)
4. Watch out for raw-rectified d.c. on heaters. Yes, it is piously supposed to remove the possibility of induced hum by 6.3V ac, but a spiky dc because of insufficient smoothing after rectification is often a worse offender. I have never had hum problems from ac heater voltage induction in power amplifiers; in pre-amps you wire heaters in series and use proper smoothing, then it works.
5. Component-wise there is again a sofa-full of old wives' tales. Modern components are by-and-large of commendable quality, and you cannot easily go wrong. The main problem might be ripple current specs for power supply smoothing capacitors (you cannot just use any small say 47uF/450V capacitor after a rectifier) which I had some difficulty in getting down here, but dealers your way may be better informed.
Have fun!
Johan
I am busy with final touches on a 120W tubby, but not yet at the stage where I can make a final circuit available. (It is more complex than the Emperor with a number of control circuits. If of interest, I use ECC88, EFC80, E182CC and 4 x 6L6GCs in UL.) So this reply might not be very useful to you, but perhaps a few general pointers might be of interest to you and others (gleaned from 50 years in audio), as you would have realised by now that trying to find direction in the sea of info on the web, often makes it a deeper sea.
1. So many power amp designs use the ECC83/12AX7 as an input stage - careful there. Unless you use a low output-impedance pre-amp (OK, so most are), a thing called Miller-effect (action of the capacitance between grid and anode) renders many designs below par; especially so if the input is glibly taken through a 1 meg volume control. Have that in mid-position resistance-wise and you start losing treble above 6 KHz, apart from phase shift problems with feedback. Folks shun pentode input stages; the increase in noise is academic while frequency response and distortion is superior.
2. About negative feedback (NFB): It is only bad in poor designs; don't always believe old wives' tales. Some sites e.g. Norman Koren (see previous ref.) put this in perspective should you care to read up. There is a whole lot of misunderstanding here often resulting from erroneous conclusions after listening sessions.
3. Same with UL output stages. Yes, ultra-linear is a misnomer (linearity is not superior), but it combines the best of triodes and pentodes, not merely a somewhere in-between. Space here is too limited for an extensive resume, but an example will illustrate(apologies to members reading other threads for repetition). GE quoted for pp KT88s working into a 6K.ohm load: Pentode operation: Max. output (Po) = 47W; output impedance (Re) = 20K; distortion (d) = 3.3%. For UL with 40% screen taps: Po = 50W; Re = 4.2K; d = 1.9%. Triode operation: Po = 23W; Re = 2.6 K; d = 1.1%. Triodes look best, but then d is at full output. Scaling UL down by 23/47 gives only 0.6% at triode max. output, and then you still have the rest in reserve, etc. (Some claim to "hear" that UL is inferior, but hearing is subjective.)
4. Watch out for raw-rectified d.c. on heaters. Yes, it is piously supposed to remove the possibility of induced hum by 6.3V ac, but a spiky dc because of insufficient smoothing after rectification is often a worse offender. I have never had hum problems from ac heater voltage induction in power amplifiers; in pre-amps you wire heaters in series and use proper smoothing, then it works.
5. Component-wise there is again a sofa-full of old wives' tales. Modern components are by-and-large of commendable quality, and you cannot easily go wrong. The main problem might be ripple current specs for power supply smoothing capacitors (you cannot just use any small say 47uF/450V capacitor after a rectifier) which I had some difficulty in getting down here, but dealers your way may be better informed.
Have fun!
Johan
Wow..
80 watts,and ~100dB efficient horns?
You must like it *LOUD* 😱
8W per chan (PP 6V6's) with my not-that-efficent 3-way box speakers is enough to blow me out of the room..
I can rock the whole house with the ST-70.
80 watts,and ~100dB efficient horns?
You must like it *LOUD* 😱
8W per chan (PP 6V6's) with my not-that-efficent 3-way box speakers is enough to blow me out of the room..
I can rock the whole house with the ST-70.
Jack,
You may like to look at this design either to build "as is" or for some ideas. Its using KT88 but is suitable for KT88, 6550, EL34 and 6L6. With KT88 it gives 135W RMS in hard saturation and at 50W output frequency response was <20Hz to >60kHz.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=46346&highlight=
Some additional discussion on my design goals and methods here:
http://www.evatco.com.au/project_files/kt88.htm
I still only have a single mono block prototype running - too many other projects BUT the final mono block builds are next project on my list. I have tried it with Plitron VDV2100 Toroidal Output Transformers in place of the Hammond 1650T and did NOT find any appreciable improvement although this MAY have been different if running a stereo pair (i.e. there may be improvements in stereo imaging - bit hard to tell with only one mono block running)
Cheers,
Ian
You may like to look at this design either to build "as is" or for some ideas. Its using KT88 but is suitable for KT88, 6550, EL34 and 6L6. With KT88 it gives 135W RMS in hard saturation and at 50W output frequency response was <20Hz to >60kHz.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=46346&highlight=
Some additional discussion on my design goals and methods here:
http://www.evatco.com.au/project_files/kt88.htm
I still only have a single mono block prototype running - too many other projects BUT the final mono block builds are next project on my list. I have tried it with Plitron VDV2100 Toroidal Output Transformers in place of the Hammond 1650T and did NOT find any appreciable improvement although this MAY have been different if running a stereo pair (i.e. there may be improvements in stereo imaging - bit hard to tell with only one mono block running)
Cheers,
Ian
Hi Ian,
You say in your post #2 "no reply", so I will reply: Quite a good circuit and very worthwhile trying! (Also different from my approach, but it is a free world - er, well . . . And different certainly does not imply a quality difference.)
I would also not change C103 - C106; in fact, with feedback that will not stop the 20 Hz firstly, just make the previous stages work harder to make up for the attenuation. If this is a worry do it at the input. As PRR says there is not a lot down there normally but in these days of synthesizers I don't know what to expect any any more. In my own amp I have added a simple overload (clipping) indicator by way of a flashing LED to warn against roasting of expensive tweeters, etc. I keep the undesired extremities out by filters in the pre-amp.
With the varying mains we encounter here I needed to install some sort of bias stabilising network in my brute (stabilising to constant cathode current), but that would depend on where one is.
Lastly regarding R104/C100: They are in a constant current circuit - or perhaps not; the 6SN7 Ri is not that high so cheap enough to leave them in. (I was going to say with high Ri they would not be necessary.)
Otherwise well done; an emotion I am no longer capable of so often these days! (The reference to damping given is also well worth reading.)
You say in your post #2 "no reply", so I will reply: Quite a good circuit and very worthwhile trying! (Also different from my approach, but it is a free world - er, well . . . And different certainly does not imply a quality difference.)
I would also not change C103 - C106; in fact, with feedback that will not stop the 20 Hz firstly, just make the previous stages work harder to make up for the attenuation. If this is a worry do it at the input. As PRR says there is not a lot down there normally but in these days of synthesizers I don't know what to expect any any more. In my own amp I have added a simple overload (clipping) indicator by way of a flashing LED to warn against roasting of expensive tweeters, etc. I keep the undesired extremities out by filters in the pre-amp.
With the varying mains we encounter here I needed to install some sort of bias stabilising network in my brute (stabilising to constant cathode current), but that would depend on where one is.
Lastly regarding R104/C100: They are in a constant current circuit - or perhaps not; the 6SN7 Ri is not that high so cheap enough to leave them in. (I was going to say with high Ri they would not be necessary.)
Otherwise well done; an emotion I am no longer capable of so often these days! (The reference to damping given is also well worth reading.)
- Status
- Not open for further replies.