7 years ago I decided to have an OTL SE for my headphone.
Since I had a Grado SR60(32 Ohm) I decided for this schematic:
http://utenti.lycos.it/pcclubfermo/OTLa.jpg
I buyed all components but I never finished the ampi.
Now I decided to use the components to try an OTL for 8 Ohm speakers, improving the project.
First of all I tryed this schematic:
http://utenti.lycos.it/pcclubfermo/OTLb.jpg
As you see I left the driver(e88cc) as it was and I added a 6080 in output stage. I know that this configuration give about 1W, but it is enough for me. My idea is to add one more 6080 in output stage, but first of all I wanted to know how the ampli sound.
The problem is that the sound(with 98db loudspeakers - 8 ohm) seems to be little distorted and with poor high freq.
Where I got wrong?
3 x 6080 need improved driver?
Or Cathode need lower res. value than 160 ohm(for example 120 ohm)?
I know 3 x 6080 in this configuration has an output of about 20 ohm, but many people say that it's enough to drive 8 ohm speakers.
Help needed! Thanks!
Since I had a Grado SR60(32 Ohm) I decided for this schematic:
http://utenti.lycos.it/pcclubfermo/OTLa.jpg
I buyed all components but I never finished the ampi.
Now I decided to use the components to try an OTL for 8 Ohm speakers, improving the project.
First of all I tryed this schematic:
http://utenti.lycos.it/pcclubfermo/OTLb.jpg
As you see I left the driver(e88cc) as it was and I added a 6080 in output stage. I know that this configuration give about 1W, but it is enough for me. My idea is to add one more 6080 in output stage, but first of all I wanted to know how the ampli sound.
The problem is that the sound(with 98db loudspeakers - 8 ohm) seems to be little distorted and with poor high freq.
Where I got wrong?
3 x 6080 need improved driver?
Or Cathode need lower res. value than 160 ohm(for example 120 ohm)?
I know 3 x 6080 in this configuration has an output of about 20 ohm, but many people say that it's enough to drive 8 ohm speakers.
Help needed! Thanks!
Hi,
In order to produce significant output you need to ensure that the current through the tubes and the cathode resistor is much higher. 1A standing current will give you a theoretical max output power of 4W as the peak output current can never exceed 1A and output power is Ipk^2/2*Rl. 2A will give you 4 times as much output power or 16W, the real output power will be lower as it is impossible to use all current swing.
A problem will be power dissipation as the circuit will dissipate 170V * the standíng current so you will dissipate 170W to get 4W output power.
It seems that with 40V over the cathode resistor and 170V B+ a 6080 will draw 160mA which will be 1A for 6 sections in Parallell, 160mA is a bit higher than allowed but should probably be OK, anode dissipation will be ~21 per section which is too high but it will work at least for a while. The cathode resistor should be 40Ohm and will dissipate 40W continous.
I don't recommend to build an amplifier like this but the example above illustrate the problems with building a class A OTL and this is the reason why all OTLs on the market are working in class AB except for the low power SET OTL offered by Trancendant which gives 1.5W
Regards Hans
In order to produce significant output you need to ensure that the current through the tubes and the cathode resistor is much higher. 1A standing current will give you a theoretical max output power of 4W as the peak output current can never exceed 1A and output power is Ipk^2/2*Rl. 2A will give you 4 times as much output power or 16W, the real output power will be lower as it is impossible to use all current swing.
A problem will be power dissipation as the circuit will dissipate 170V * the standíng current so you will dissipate 170W to get 4W output power.
It seems that with 40V over the cathode resistor and 170V B+ a 6080 will draw 160mA which will be 1A for 6 sections in Parallell, 160mA is a bit higher than allowed but should probably be OK, anode dissipation will be ~21 per section which is too high but it will work at least for a while. The cathode resistor should be 40Ohm and will dissipate 40W continous.
I don't recommend to build an amplifier like this but the example above illustrate the problems with building a class A OTL and this is the reason why all OTLs on the market are working in class AB except for the low power SET OTL offered by Trancendant which gives 1.5W
Regards Hans
First of all thanks for the answer.
The reason I am trying with this kind of amp is simple:
I have the components and I am very curious about the type sound it can produce. 😉
And also I have many things to learn about tubes, and I hope to starting here in this nice forum with your competent and kind help.
Now, if I understand well, you say that the problem is in the output section(not in the ecc88 SRPP driver, that was thought for only 4 sections of 6080).
So I have "only" to try to change the value of R in the cathodes of 6080s.
For example starting using an 80 ohm res, or lower(you said 40 Ohm, but I am afraid to go so "out" of plate dissipation of 6080s).
I prefer starting lowering a little, for first time.
It's correct, or I miss something?
P.s. How I can calculate actual real anode dissipation per section(with my 160 Ohm cathode resistor)?
Thanks again.
The reason I am trying with this kind of amp is simple:
I have the components and I am very curious about the type sound it can produce. 😉
And also I have many things to learn about tubes, and I hope to starting here in this nice forum with your competent and kind help.
Now, if I understand well, you say that the problem is in the output section(not in the ecc88 SRPP driver, that was thought for only 4 sections of 6080).
So I have "only" to try to change the value of R in the cathodes of 6080s.
For example starting using an 80 ohm res, or lower(you said 40 Ohm, but I am afraid to go so "out" of plate dissipation of 6080s).
I prefer starting lowering a little, for first time.
It's correct, or I miss something?
P.s. How I can calculate actual real anode dissipation per section(with my 160 Ohm cathode resistor)?
Thanks again.
Btw, I tryed to change 160 Ohm res with a 80 Ohm res.
It seems to sound louder.
I notice, once again, poor high freq(it seems to be like a cut on high frequencies, muffled sound) and little rough sound(like a very little but advertible distorsion).
I don't know why I have so poor freq. res. and this sound.
😕
It seems to sound louder.
I notice, once again, poor high freq(it seems to be like a cut on high frequencies, muffled sound) and little rough sound(like a very little but advertible distorsion).
I don't know why I have so poor freq. res. and this sound.
😕
Hi,
Using 80 ohms will give you lower output power as the current will be lower, to find the current with any resistance you need to look at tube curves like these http://www.triodeel.com/6080_p4.gif and find the point where the grid voltage is R/I where R is the cathode resistance and where I is the current you want to have for all tubes in parallell. (Example -40V grid voltage gives a current of approx 160mA for a anode voltage of 130V so the resistor should be 40/(6*.16) = 40ohm).
To drive cathode followers like this is easy as the input capacitance is very low but you need to provide enough voltage, I would imagine that a 4W version will need something like 80V Pk-Pk which should be possible from the ECC88 (which BTW is coupled like a mu follower and not as a SRPP)
Regards Hans
Using 80 ohms will give you lower output power as the current will be lower, to find the current with any resistance you need to look at tube curves like these http://www.triodeel.com/6080_p4.gif and find the point where the grid voltage is R/I where R is the cathode resistance and where I is the current you want to have for all tubes in parallell. (Example -40V grid voltage gives a current of approx 160mA for a anode voltage of 130V so the resistor should be 40/(6*.16) = 40ohm).
To drive cathode followers like this is easy as the input capacitance is very low but you need to provide enough voltage, I would imagine that a 4W version will need something like 80V Pk-Pk which should be possible from the ECC88 (which BTW is coupled like a mu follower and not as a SRPP)
Regards Hans
Thanks again for answer.
So using 80 ohm cathode res give lower than 160 Ohm Res?
I am confused...
I made in this way:
I checked the voltage across the cathode res. With 160 Ohm res I found 48V. With 80 Ohm res I found again 48V.
My conclusions were:
With 160 Ohm res I have 0,3A across this res. Each section of 6080 draws 0,05A.(or not?)
With 80 Ohm res I have 0,6A across this res. Each section of 6080 draws 0,1A.(or not?).
So with 80 Ohm res I reached the max. value for 6080.
Why you say I have lower power with 80 res than 160 res?
And, also why my sound changes a little or nothing with this change?
Again, why I have poor high freq in both cases?
Excuse me for make you loose your time answering me, but at this point I am very curious to know more about the sound of this schematic!
Thanks again.
So using 80 ohm cathode res give lower than 160 Ohm Res?
I am confused...
I made in this way:
I checked the voltage across the cathode res. With 160 Ohm res I found 48V. With 80 Ohm res I found again 48V.
My conclusions were:
With 160 Ohm res I have 0,3A across this res. Each section of 6080 draws 0,05A.(or not?)
With 80 Ohm res I have 0,6A across this res. Each section of 6080 draws 0,1A.(or not?).
So with 80 Ohm res I reached the max. value for 6080.
Why you say I have lower power with 80 res than 160 res?
And, also why my sound changes a little or nothing with this change?
Again, why I have poor high freq in both cases?
Excuse me for make you loose your time answering me, but at this point I am very curious to know more about the sound of this schematic!
Thanks again.
Hi,
When I wrote lower power with 80ohm I compared with my original recommendation of 40ohm, 80 ohm will give lower current and therefore lower max output power than with 40 ohm.
The loss of high frequencies is probably due to the cap of 500pF to ground in the mu follower, I don't understand what it is doing there, it seems to a deliberate attempt to limit high frequencies, remove this cap. (I didn't notice this cap from the beginning)
If you have the possibility please measure with an oscilloscope on the grid of 6080 to see if the mu follower give enough voltage and expected wide frequency response.
BTW, I encourage the idea to test a simple OTL circuit, when you discover the superiour sound of this compared to practically anything else I believe you probably want to build something more practical like this http://www.tubetvr.com/otl.html
Regards Hans
When I wrote lower power with 80ohm I compared with my original recommendation of 40ohm, 80 ohm will give lower current and therefore lower max output power than with 40 ohm.
The loss of high frequencies is probably due to the cap of 500pF to ground in the mu follower, I don't understand what it is doing there, it seems to a deliberate attempt to limit high frequencies, remove this cap. (I didn't notice this cap from the beginning)
If you have the possibility please measure with an oscilloscope on the grid of 6080 to see if the mu follower give enough voltage and expected wide frequency response.
BTW, I encourage the idea to test a simple OTL circuit, when you discover the superiour sound of this compared to practically anything else I believe you probably want to build something more practical like this http://www.tubetvr.com/otl.html
Regards Hans
Hi,
I am glad to here you encourage OTL design!
Yes, my idea is to discover the sound of OTL amps, and I expect to find good news in them(I ever used classical triode- single ended amps in my system). I know single ended OTLs have some limitations. The big output cap in one of them.
Do you think I can improve the sound bypassing it with a 5uF - paper/oil?
Next step is to try OTL amps like your nice web-site show.
BTW now I need to adjust this single ended OTL, just to know how OTL can sound.
I haven't an oscilloscope, so I cant meausre.
I removed the 500pF cap, but I have no improvment at all.
My sound is "darky" and seems to be little distorted, with bad dynamics impression and no "crystal" at all.
I don't know if the drive stage(mu follower) is capable to deliver the correct Voltage, but I noticed that the sound is bad also at very very low volumes.
I have to find where I am wrong.
The V+ come from a power supply with a tot. of 5000uF(5 caps of 1000uF), I think it's enogh.
Pheraps the problem come from mu follower, but I don't know how to change it, since I made a search in the web and I found nothing serious about mu follower design(formulas).
SO I don't know what I have to change to increase gain in the circuit without destroying E88CC, and also I can't control if the circuit is well designed for the purpose.
Thanks again for answers and for the link to your nice web-site.
🙂
I am glad to here you encourage OTL design!
Yes, my idea is to discover the sound of OTL amps, and I expect to find good news in them(I ever used classical triode- single ended amps in my system). I know single ended OTLs have some limitations. The big output cap in one of them.
Do you think I can improve the sound bypassing it with a 5uF - paper/oil?
Next step is to try OTL amps like your nice web-site show.
BTW now I need to adjust this single ended OTL, just to know how OTL can sound.
I haven't an oscilloscope, so I cant meausre.
I removed the 500pF cap, but I have no improvment at all.
My sound is "darky" and seems to be little distorted, with bad dynamics impression and no "crystal" at all.
I don't know if the drive stage(mu follower) is capable to deliver the correct Voltage, but I noticed that the sound is bad also at very very low volumes.
I have to find where I am wrong.
The V+ come from a power supply with a tot. of 5000uF(5 caps of 1000uF), I think it's enogh.
Pheraps the problem come from mu follower, but I don't know how to change it, since I made a search in the web and I found nothing serious about mu follower design(formulas).
SO I don't know what I have to change to increase gain in the circuit without destroying E88CC, and also I can't control if the circuit is well designed for the purpose.
Thanks again for answers and for the link to your nice web-site.
🙂
Remember the 6080's max dissipation of 13watts, so dont go over 100mA per triode section with this psu voltage. Although these tubes are super tough, I've run them much harder, by accident, and they are still running fine.
I agree with Hans in regards to how awesome OTL amps are. Something must be wrong since I've never had high frequency problems with OTLs.
Since the sound is slightly distorted, perhaps the tubes are faulty? Try another E88CC. Or maybe a resistor is the wrong value? Is the problem in both channels?
I know this circuit works ok, it doesnt take much to alter the circuit you now have to this, try it...
You really need an oscilloscope of some sort, audiotester is a great tool using your soundcard, it cost a little but well worth it. Just remember the soundcard has low input impedance,(will load any circuit you measure), and doesnt handle high voltages. A tube buffer is a good project too...
I agree with Hans in regards to how awesome OTL amps are. Something must be wrong since I've never had high frequency problems with OTLs.
Since the sound is slightly distorted, perhaps the tubes are faulty? Try another E88CC. Or maybe a resistor is the wrong value? Is the problem in both channels?
I know this circuit works ok, it doesnt take much to alter the circuit you now have to this, try it...
You really need an oscilloscope of some sort, audiotester is a great tool using your soundcard, it cost a little but well worth it. Just remember the soundcard has low input impedance,(will load any circuit you measure), and doesnt handle high voltages. A tube buffer is a good project too...
Attachments
Another thing about these power triodes, 6080 and 6AS7, is they have poor matching. One triode will perhaps have 30mA thru it and the other 60mA with same bias and voltages. So you really should have cathode resistors on each leg. I use a whopping 10ohm resistance, and even without any loop feedback the amp has full control. (My speakers have two 8" bass drivers in series, so the amp sees 16ohms in the bass, which may explain this.)
Thank you very much SemperFi. I'll check again the circuit and the E88CC. The problem is I haven't others E88CC to try, only ECC88 and I am afraid Voltage exceed the specification of ECC88.
Thanks again.
P.S. I saw your schematic. I think cathode R 280 Ohm may be high. 200 Ohm give more current. 😉
Thanks again.
P.S. I saw your schematic. I think cathode R 280 Ohm may be high. 200 Ohm give more current. 😉
Checked the circuit and tryed adding more capacitors in power supply, changing E88CC(tesla) with E88CC amperex, and first R from 470k to 220k.
Sound, high freq and dynamics are still not pleasant.
Sound, high freq and dynamics are still not pleasant.
Ok, 3/4 of the problem solved.
It seems that the amply doesn't like filaments of E88CC and 6080s connected in the same windings.
Since I have another transfo 6.3V I just separated them and now ampli sound much better(it gained many high freq, and lost some roughy) but still not lively enough.
I realized only a channel(only to try the schematic and it's sound) so mu follower is in a single tube. I know that it's not a good idea(like for the SRPP).
Maybe in a stereo version, with the high sections of mu follower in a tube and the lowers in another tube, the sound will improve.
😎
It seems that the amply doesn't like filaments of E88CC and 6080s connected in the same windings.
Since I have another transfo 6.3V I just separated them and now ampli sound much better(it gained many high freq, and lost some roughy) but still not lively enough.
I realized only a channel(only to try the schematic and it's sound) so mu follower is in a single tube. I know that it's not a good idea(like for the SRPP).
Maybe in a stereo version, with the high sections of mu follower in a tube and the lowers in another tube, the sound will improve.
😎
Glad to hear it is almost fixed. I always bias the heater voltage at some positive voltage since sometimes I get strange noises when not. Connecting one side of the heater voltage to the 6080's cathode will give a perfect bias of about 50volts. This should eliminate any problems from the heater.
The E88CC shares the voltage between them, so they are fine with this voltage. Note that the circuit I posted the E88CCs are not SRPP but so called 'aikido', which I think sounds best of the two.
The 280 ohm resistor is if only using two triodes, since you use 6, the value could be 70 ohms.
Oh, you asked in the first post how to calculate anode dissipation. The voltage over the tube multiplied with the current equals dissipation. So if you have 48 volts at the cathodes, the voltage over the tube is 170-48=122V. Next assume each triode shares the total current equally, they dont but itll be close enuff. If using 100ohm resistor you have 480mA total current, divide by 6 triodes and thats 80mA per triode. Dissipation is 122V*80mA=9.8watts per triode.
The E88CC shares the voltage between them, so they are fine with this voltage. Note that the circuit I posted the E88CCs are not SRPP but so called 'aikido', which I think sounds best of the two.
The 280 ohm resistor is if only using two triodes, since you use 6, the value could be 70 ohms.
Oh, you asked in the first post how to calculate anode dissipation. The voltage over the tube multiplied with the current equals dissipation. So if you have 48 volts at the cathodes, the voltage over the tube is 170-48=122V. Next assume each triode shares the total current equally, they dont but itll be close enuff. If using 100ohm resistor you have 480mA total current, divide by 6 triodes and thats 80mA per triode. Dissipation is 122V*80mA=9.8watts per triode.
Ok, thank you very much.
Now I am studying improvments. The sound is good but not exceptional.
I think to do this:
1)bypass the big output cap(3000 uF) with a paper-oil 2uF or more(it's a good idea or not?)
2)increase the gain of e88CC mu-follower(but I can't find good formulas in the web for mu-followers, so I don't know how to do without destrying E88CC)
3)add 1 more 6080 and decrease the value of cathode res proportionally.
4)try a tube rectifier supply for 6080. I like very much the improvment in sound that tube rectifiers give.
I think to use something like this:
EL3C
Tantalum Anodes and xenon gas filled full wave tube rectifier.
2.5A max... 😀
Now I am studying improvments. The sound is good but not exceptional.
I think to do this:
1)bypass the big output cap(3000 uF) with a paper-oil 2uF or more(it's a good idea or not?)
2)increase the gain of e88CC mu-follower(but I can't find good formulas in the web for mu-followers, so I don't know how to do without destrying E88CC)
3)add 1 more 6080 and decrease the value of cathode res proportionally.
4)try a tube rectifier supply for 6080. I like very much the improvment in sound that tube rectifiers give.
I think to use something like this:
EL3C
Tantalum Anodes and xenon gas filled full wave tube rectifier.
2.5A max... 😀
I made some experiments.
I saw that the line stage of "triodino" 6SN7 uses the same values in the mu-follower as my otl schematic but the V+(which is 335V instead of my 260V with E88CC)
triodino
Assuming that 6CG7 is similar to 6SN7 (and pin to pin compatible with E88CC) I just had to upper the V+ to 300V and tryed this tube in my OTL.
I noticed I had lower gain but much better sound(more "crystal" and dynamic speed, and also more "complex")
Someone know why 6CG7 sound better than E88CC in this situation?
And it's "electrically correct" this experiment with 6CG7?
Finally, what I have to change in the mu-follower if I want to try a high gain tube like 6SL7?
I saw that the line stage of "triodino" 6SN7 uses the same values in the mu-follower as my otl schematic but the V+(which is 335V instead of my 260V with E88CC)
triodino
Assuming that 6CG7 is similar to 6SN7 (and pin to pin compatible with E88CC) I just had to upper the V+ to 300V and tryed this tube in my OTL.
I noticed I had lower gain but much better sound(more "crystal" and dynamic speed, and also more "complex")
Someone know why 6CG7 sound better than E88CC in this situation?
And it's "electrically correct" this experiment with 6CG7?
Finally, what I have to change in the mu-follower if I want to try a high gain tube like 6SL7?
- Status
- Not open for further replies.