tube amp for biamped speaker
I am building a speaker using 2 18cm 8 ohm woofers (push push) and a 1" dome tweeter (6 ohms). Scan Speak 18W and 9900. I have never built a tube amp but have heard a few.
I intend to biamp them with tubes. the idea being to use the preamp-2nd order 2.5khz crossover-power amp in one box to improve WAF.
There will be 3 souces for this system (CD, Ipod, and Cable TV audio).
I am looking at about 10W for the tweeter and about 50W for the woofers. I am not comfortable with very high DC voltages.
my concept is to use Class A Push Pull EL84 for the tweeter and Class A Push Pull EL 34 for the woofer but the EL34 would only produce 20W in class A and would have to baised to Class AB for 40W. Any other tubes that canbe used with lower DC voltages?
I also hope that each amp (Low Pass and High Pass) would have only 1 driver tube to reduce tube count.
BTW I favour push pull so that the output trannies are smaller.
Since all my soruces are High signal sources (300mV+) I hope to get by using just a single 3 way 2 pole selector and a single tube for volume control. Any ideas on this?
Anyone have a circuit for a 12db 2 way tube crossover? How many tubes will this need.
another alternate is to get a 3 box design with one box being the input selector, volume control and crossover and 2 boxes each having a 10W and 50W tube amp next to each speaker.
I am building a speaker using 2 18cm 8 ohm woofers (push push) and a 1" dome tweeter (6 ohms). Scan Speak 18W and 9900. I have never built a tube amp but have heard a few.
I intend to biamp them with tubes. the idea being to use the preamp-2nd order 2.5khz crossover-power amp in one box to improve WAF.
There will be 3 souces for this system (CD, Ipod, and Cable TV audio).
I am looking at about 10W for the tweeter and about 50W for the woofers. I am not comfortable with very high DC voltages.
my concept is to use Class A Push Pull EL84 for the tweeter and Class A Push Pull EL 34 for the woofer but the EL34 would only produce 20W in class A and would have to baised to Class AB for 40W. Any other tubes that canbe used with lower DC voltages?
I also hope that each amp (Low Pass and High Pass) would have only 1 driver tube to reduce tube count.
BTW I favour push pull so that the output trannies are smaller.
Since all my soruces are High signal sources (300mV+) I hope to get by using just a single 3 way 2 pole selector and a single tube for volume control. Any ideas on this?
Anyone have a circuit for a 12db 2 way tube crossover? How many tubes will this need.
another alternate is to get a 3 box design with one box being the input selector, volume control and crossover and 2 boxes each having a 10W and 50W tube amp next to each speaker.
Re: tube amp for biamped speaker
One should never overestimate the WAF of a tube amp design. It is often (usually, almost always) much less than we think it should be!
It sounds like you have a handle on what might be required in that department.
Your idea for a source switch is fine, I typically use a two-pole/four position rotary switch (non-shorting or break-before-make) as a source selector.
As for getting everything in one box... I suppose it could be done but I would recommend your three box design. For a preamp, high-end amp and low end amp you are likely to need three power transformers. You will of course also need four output transformers. Easier to put the preamp and crossover in one box and your power amps in separate boxes.
A couple of question though about the power requirements. How efficient are your speakers going to be, how large is your listening room and how loud to you like to listen? The reason I ask is that with 89bB/W/M speakers a couple watts for the tweeters and a dozen or so watts for the low end would be awfully loud in a small to medium sized room.
navin said:
One should never overestimate the WAF of a tube amp design. It is often (usually, almost always) much less than we think it should be!
It sounds like you have a handle on what might be required in that department.Your idea for a source switch is fine, I typically use a two-pole/four position rotary switch (non-shorting or break-before-make) as a source selector.
As for getting everything in one box... I suppose it could be done but I would recommend your three box design. For a preamp, high-end amp and low end amp you are likely to need three power transformers. You will of course also need four output transformers. Easier to put the preamp and crossover in one box and your power amps in separate boxes.
A couple of question though about the power requirements. How efficient are your speakers going to be, how large is your listening room and how loud to you like to listen? The reason I ask is that with 89bB/W/M speakers a couple watts for the tweeters and a dozen or so watts for the low end would be awfully loud in a small to medium sized room.
Hi
You could also build the crossover's into the power amp, just choose a right RC combination: put one in front of the driver stage and one between driver and output stage (assuming a two stage design, which I believe should be sufficient for a PP EL84 and/or EL34.) Another option is to build the crossover from steve bench which uses an ECC88, one per channel. This approach attains it advantages too, because you keep the crossover thing apart from the power amp thing. As for the power - you can go the traditional way connecting tubes in parallel for double power: you will double current from the PS and an output tranny with half the impedance.
Steve Bench's crossover
http://members.aol.com/sbench101/#Electronic
Erik
You could also build the crossover's into the power amp, just choose a right RC combination: put one in front of the driver stage and one between driver and output stage (assuming a two stage design, which I believe should be sufficient for a PP EL84 and/or EL34.) Another option is to build the crossover from steve bench which uses an ECC88, one per channel. This approach attains it advantages too, because you keep the crossover thing apart from the power amp thing. As for the power - you can go the traditional way connecting tubes in parallel for double power: you will double current from the PS and an output tranny with half the impedance.
Steve Bench's crossover
http://members.aol.com/sbench101/#Electronic
Erik
For the woofer amp, consider KT88s running "hot and low". You'll get about 40 W. from a PP pair set up in Class "A" pentode mode. If you can tolerate Class "B" operation when peak power is produced, the 50 W. target is in sight. Max. linearity in pentode mode requires regulated screen grid B+.
Only 2 stages looks like differential splitter/drivers to me. That's good for the loop NFB pentode mode needs. Big question, can a 12AT7 provide enough gain for the woofer amp's needs?
Marchand 24 dB./octave crossovers use 4 twin triodes per channel. A 12 dB/octave crossover would use 1 less "bottle" in each channel.
Only 2 stages looks like differential splitter/drivers to me. That's good for the loop NFB pentode mode needs. Big question, can a 12AT7 provide enough gain for the woofer amp's needs?
Marchand 24 dB./octave crossovers use 4 twin triodes per channel. A 12 dB/octave crossover would use 1 less "bottle" in each channel.
Re: Re: tube amp for biamped speaker
the tweeter is the ScanSpeak 2905-9900 the woofers are 18W8546 in push push. approximate sensitivity about 88-90db for each.
my idea was to build the power amps into the base of each speaker cabinet and only keep the preamp-crossover visible. this will mean that the pcrossover-power amp wire will be about 16 ft (5m). can a tube drive such a long length of cable?
thi sis why I was was looking at builiding 2 power amps (EL84/EL34 for example) into one box. It will mean two more transformers but the cost of tehse will be offset by the fact that the power amps will be hidden. besides the transformers will be very good ballasts for the speakers.
thanks for the link it looks nice. now should the crossover be inside the power amp or inside the preamp?
I would look at parallel tubes but i am told that keeping 4 tubes (Parallel Push Pull) biased similarly is not easy. hence i was looking at a singple pair per channel.
the woofer and tweeter have the same sensitvity. but at 2.5K (my exected crossover freq). the energy at the woofer will be 5 times that of the tweeter hence the woofer amp will need to about 5 times the tweeter amp.
thanks guys for all the help. sure would love more!
still a bit confused.
Sherman said:
the tweeter is the ScanSpeak 2905-9900 the woofers are 18W8546 in push push. approximate sensitivity about 88-90db for each.
my idea was to build the power amps into the base of each speaker cabinet and only keep the preamp-crossover visible. this will mean that the pcrossover-power amp wire will be about 16 ft (5m). can a tube drive such a long length of cable?
thi sis why I was was looking at builiding 2 power amps (EL84/EL34 for example) into one box. It will mean two more transformers but the cost of tehse will be offset by the fact that the power amps will be hidden. besides the transformers will be very good ballasts for the speakers.
ErikdeBest said:
thanks for the link it looks nice. now should the crossover be inside the power amp or inside the preamp?
I would look at parallel tubes but i am told that keeping 4 tubes (Parallel Push Pull) biased similarly is not easy. hence i was looking at a singple pair per channel.
EC8010 said:
the woofer and tweeter have the same sensitvity. but at 2.5K (my exected crossover freq). the energy at the woofer will be 5 times that of the tweeter hence the woofer amp will need to about 5 times the tweeter amp.
thanks guys for all the help. sure would love more!
still a bit confused.
Re: Re: Re: tube amp for biamped speaker
If you think about it, the energy in music at 2.4kHz (just below your crossover) is unlikely to be significantly different to that at 2.6kHz (just above). In which case, given that the tweeter is the same sensitivity as the woofers, each amplifier must deliver the same power.
What sort of music requires 5 times the power in the bass compared to the treble?
navin said:
If you think about it, the energy in music at 2.4kHz (just below your crossover) is unlikely to be significantly different to that at 2.6kHz (just above). In which case, given that the tweeter is the same sensitivity as the woofers, each amplifier must deliver the same power.
What sort of music requires 5 times the power in the bass compared to the treble?
I don't know what kind of preamp you have in mind, but because you're using sources with a high output you probably won't need any gain. And then there are the cables - long cables require a low Z-out source. No gain (well, actually something around 0,95) and low Z-out are the main characteristics of a cathode follower, the configuration used for the ECC88 in the crossover. So (in my opinion) you can use the ECC88 for both crossover and preamp buffer. Note you have to run two signals cables to each speaker (you can use balanced cable with common ground, but I think this won't be good because of crosstalk issues - has been discussed in some thread...)
As for the power you need - have you had a look at this article?
http://www.tubecad.com/articles_2001/Active_Crossovers_and_Filters/index.html
Erik
. If you decide you don't need gain in the preamp you can use the ECC88 crossover as pre-amp
to go this way you can use the crossover as
This way you can use the
Re: Re: Re: Re: tube amp for biamped speaker
I was given to understand that music has about half it's energy under 300Hz and the other half above 300Hz.
thanks for the link, I might not need gain but i will need a volume control so the preamp was only for the volume control and to drive the long cables to the power amps.
EC8010 said:
I was given to understand that music has about half it's energy under 300Hz and the other half above 300Hz.
ErikdeBest said:
thanks for the link, I might not need gain but i will need a volume control so the preamp was only for the volume control and to drive the long cables to the power amps.
Well for the bass amp you could use a pair of 6c33's (6s33s)
The tweeter amp could use it's samller brother, the 6c41c(6s41s)
These are low voltage Russian types, both extensively used in single ended designs, but there is no reason not to use them in PP.
Having very low Rp makes transformer design easier.
Re: Re: Re: Re: Re: tube amp for biamped speaker
Quite possibly, although that's not what I was arguing. My point is not so much about the power distribution of music as a consideration of what happens when one amplifier clips. I feel that if we consider our music to contain roughly the same amplitudes immediately above and below the crossover frequency, then I would prefer both amplifiers to clip at the same time - not one before the other. If the drivers are the same sensitivity, then that implies that the amplifiers ought to deliver the same power.
navin said:
Quite possibly, although that's not what I was arguing. My point is not so much about the power distribution of music as a consideration of what happens when one amplifier clips. I feel that if we consider our music to contain roughly the same amplitudes immediately above and below the crossover frequency, then I would prefer both amplifiers to clip at the same time - not one before the other. If the drivers are the same sensitivity, then that implies that the amplifiers ought to deliver the same power.
Re: Re: Re: Re: Re: Re: tube amp for biamped speaker
let me wrap my mind around this....for many years i lived with teh belief that the tweeter amp need not be as big as the woofer amp as there is less energy in the high frequencies.
let me explain what i understand so that you can correct me.
say we have a 2W (usually this is represented in volts but since we understand power in watts i am representing it in watts) music signal. 1W of this would be below 300Hz (assuming my earlier assumption is right) and 1W would be above 300hz. so in this case it would mean that both amplifiers (if indeed we are using 2) deliver the same power to clip at the same input level.
if the 2W music signal is divided at say at 2500Hz (assuming 20% is above 2500Hz and 80% is below 2500Hz) 1.6W would be below 2500Hz and 0.4W would be above 2500hz meaning that the amp delivering power below 2500Hz has to 4 times bigger so that both amps clip at the same input level.
i just found this link that might help explain what's in my mind.
http://sound.westhost.com/bi-amp.htm#power_dist
EC8010 said:
let me wrap my mind around this....for many years i lived with teh belief that the tweeter amp need not be as big as the woofer amp as there is less energy in the high frequencies.
let me explain what i understand so that you can correct me.
say we have a 2W (usually this is represented in volts but since we understand power in watts i am representing it in watts) music signal. 1W of this would be below 300Hz (assuming my earlier assumption is right) and 1W would be above 300hz. so in this case it would mean that both amplifiers (if indeed we are using 2) deliver the same power to clip at the same input level.
if the 2W music signal is divided at say at 2500Hz (assuming 20% is above 2500Hz and 80% is below 2500Hz) 1.6W would be below 2500Hz and 0.4W would be above 2500hz meaning that the amp delivering power below 2500Hz has to 4 times bigger so that both amps clip at the same input level.
i just found this link that might help explain what's in my mind.
http://sound.westhost.com/bi-amp.htm#power_dist
Aside from the difference in sonic signature between the 2 types, "pure" Class "A" operation does not expose you to crossover/notching distortion trouble. OTOH, there is more "waste" heat when Class "A" operation is employed. The maximum theoretical efficiency for Class "A" operation is 0.5, while pi/4 is max. for Class "B". "AB" operation lies between "A" and "B".
JMO, go as deeply into Class "A", consistent with O/P power requirements, as you can. It sounds better. Crossing over into Class "B" at peak power is OK. SPL will be high; so, minor sonic aberations get masked.
Hello navin, you're absolutely correct in your argument. The problem arises because musical signals are not sine waves. If you look on an oscilloscope, you will see that music contains short burst of high frequencies (perhaps a cymbal) and that although these bursts have quite high voltages, they don't last for very long. You've mentioned power and energy, but the two are quite distinct because energy is power summed over time. When music is analysed and said to have very little HF energy, these sums are taken over perhaps 5s. This is why there's no problem in using a tweeter that is rated at 8W continuous in a 100W system. Once the tweeter is protected by the crossover, the energy it receives (summed over a few seconds) causes less heating than a continuous 8W sine wave would.
The problem is that although the averaged energy is low, those sparse peaks are high voltage, and that's why you need an amplifier capable of swinging the voltage to accomodate them. If you were a Class AB transistor amplifier manufacturer, you could take advantage of this by using a treble identical amplifier to the bass amplifier but fitting a far smaller heatsink and smaller power supply (but the same voltage). This makes a big saving.
The problem is that although the averaged energy is low, those sparse peaks are high voltage, and that's why you need an amplifier capable of swinging the voltage to accomodate them. If you were a Class AB transistor amplifier manufacturer, you could take advantage of this by using a treble identical amplifier to the bass amplifier but fitting a far smaller heatsink and smaller power supply (but the same voltage). This makes a big saving.
Datasheets can be found at Frank's site . They are too large to attach here.
(Search for 6s33 and 6s41).
(Search for 6s33 and 6s41).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.