I have been building small amplifiers for a while, and I'm interested in other people's opinons of the difference in sound between the 6BM8 and the EL84.
I have several single ended and PP 6bm8 and EL84 amps. They all sound decent.
Of the 6bm8 amps, I have two motorolas. One was a 3 channel PP amp with 6bm8 high channels and an EL84 PP bass channel... The other was also 3 channels.. it had single ended high channels with 6bm8s and a PP bass channel which also used 6BM8s. The bass channels were pretty much useless to me in both amps, so I removed them and removed the high pass caps in each and replaced them with .02 caps which let the full sound through.
Of the EL84 amps, I have two zenith single ended amps. One of those was another 3 channel (which I have posted about before..) It derived the bass channel from the two single ended outputs... the other was a pretty standard SE stereo amplifier.
Of all these amps, the motorola SE amp (after I removed the bass channel) sounds the best. Does this make sense? Of all the amps, it has the smallest lowest quality looking output trannies, no NFB, and it has outstanding response with just about any speakers and it's loud.
My real question here, is how do these two tubes compare? Are the specs similar? or about the same? or completely different...
I'm very curious about this.
These were all integrated amplifiers, and I think I am going to build new chassis for each of them this winter and turn them into regular power amplifiers, becuase I have some pretty nice preamps.
I have several single ended and PP 6bm8 and EL84 amps. They all sound decent.
Of the 6bm8 amps, I have two motorolas. One was a 3 channel PP amp with 6bm8 high channels and an EL84 PP bass channel... The other was also 3 channels.. it had single ended high channels with 6bm8s and a PP bass channel which also used 6BM8s. The bass channels were pretty much useless to me in both amps, so I removed them and removed the high pass caps in each and replaced them with .02 caps which let the full sound through.
Of the EL84 amps, I have two zenith single ended amps. One of those was another 3 channel (which I have posted about before..) It derived the bass channel from the two single ended outputs... the other was a pretty standard SE stereo amplifier.
Of all these amps, the motorola SE amp (after I removed the bass channel) sounds the best. Does this make sense? Of all the amps, it has the smallest lowest quality looking output trannies, no NFB, and it has outstanding response with just about any speakers and it's loud.
My real question here, is how do these two tubes compare? Are the specs similar? or about the same? or completely different...
I'm very curious about this.
These were all integrated amplifiers, and I think I am going to build new chassis for each of them this winter and turn them into regular power amplifiers, becuase I have some pretty nice preamps.
Hi,
No, the ECL82/6BM8/PCL82/16A8 pentode section is quite different to EL84/6BQ5 in many regards. The closest thing to EL84/6BQ5 you can find in a noval based compound valve certainly is the pentode section of ECL86/6GW8/PCL86/14GW8, which can be viewed as a plate Pd derated (9W vs. 12W) EL84/6BQ5.
Tom
akimball442 said:
No, the ECL82/6BM8/PCL82/16A8 pentode section is quite different to EL84/6BQ5 in many regards. The closest thing to EL84/6BQ5 you can find in a noval based compound valve certainly is the pentode section of ECL86/6GW8/PCL86/14GW8, which can be viewed as a plate Pd derated (9W vs. 12W) EL84/6BQ5.
Tom
I have used the ECL82 in a few projects. I also had a Leak Stereo 20 which used the EL84.
My experience of the Leak was that it has a very rich smooth sound, which I understand is common to most EL84 amps.
I have made a headphone amp using the ECL82 and also have used it to drive interstage transformers. I found that it is tremendously detailed, but a little zingy and bright. In the end I had to introduce a small amount of loop feedback into both to tame the top end. I suspect that its the triode stage of the ECL82 which is the real weak link, but I have no way of been certain.
If building from scratch I would guess that a better result could be achieved with the EL84. The ECL86 seems to be a superior brother of the ECL82.
ECL82 need handling with care.
Shoog
My experience of the Leak was that it has a very rich smooth sound, which I understand is common to most EL84 amps.
I have made a headphone amp using the ECL82 and also have used it to drive interstage transformers. I found that it is tremendously detailed, but a little zingy and bright. In the end I had to introduce a small amount of loop feedback into both to tame the top end. I suspect that its the triode stage of the ECL82 which is the real weak link, but I have no way of been certain.
If building from scratch I would guess that a better result could be achieved with the EL84. The ECL86 seems to be a superior brother of the ECL82.
ECL82 need handling with care.
Shoog
Hi Shoog,
It is very similar to one section of 6SL7, and there is nothing wrong with that
I agree with both. BTW, the triode section within the ECL86 is one section of ECC83, except being a little bit plate Pd derated (this is only due to the massive heat dissipation of the power pentode nearby).
Tom
It is very similar to one section of 6SL7, and there is nothing wrong with that
I agree with both. BTW, the triode section within the ECL86 is one section of ECC83, except being a little bit plate Pd derated (this is only due to the massive heat dissipation of the power pentode nearby).
Tom
To low for what. It has a gm of 70 which I usually find way to high for normal preamp duties. Coupled with its very low plate current, and a useful bias range around -1v, and its a pain to use in modern situations.
Might be good for guitar amp use, which I intend to try.
Looking at the Triode curves, theirs no way it can be considered a 12AT7, it miles off.
The 6SL7 is a much better fit, but not perfect. Its also got the right gm.
Shoog
Shoog,
You're mixing mu and gm up. Depending on operating conditions, the mu of the 'T7 has been quoted as 60 and 70. The mu of the 6SL7 is quoted at 70. The 6BM8's triode also has a mu of 70. The gm of the 6SL7 is quoted at 1.6 mA./V. The gm of the 6BM8's triode is quoted at 2.5 mA./V. The difference is significant.
Your point about curves is valid. Splitting the difference seems reasonable. The character of the 6BM8's triode lies between the 'SL7 and 'T7.
The few 6BM8 circuits I've perused use 2 bottles for PP "finals" and common cathode/"concertina" splitter.
You're mixing mu and gm up. Depending on operating conditions, the mu of the 'T7 has been quoted as 60 and 70. The mu of the 6SL7 is quoted at 70. The 6BM8's triode also has a mu of 70. The gm of the 6SL7 is quoted at 1.6 mA./V. The gm of the 6BM8's triode is quoted at 2.5 mA./V. The difference is significant.
Your point about curves is valid. Splitting the difference seems reasonable. The character of the 6BM8's triode lies between the 'SL7 and 'T7.
The few 6BM8 circuits I've perused use 2 bottles for PP "finals" and common cathode/"concertina" splitter.
Good point.
I'am thinking of redesigning my 6080 amp to do away with the ECL82. It performs really well at moderate levels, but sounds strained in the top end when driven hard. I think I'am getting overload of the ECL82 triode before I achieve full output. I might either keep the pentode section as the interstage transformer driver and put a 5687 infront.
Seems a little waistful so I might just drop the ECL82 and replace it with PL84's and a 5687. Only one extra valve and should perform a lot better overall. Won't need a hell of a lot of modification, but I might struggle to find room under the hood for the neccisary changes.
Shoog
The ECL82 pentode used in p-p; UL 40% 8K A-A gives brilliant low thd performance with a good output tranny. A cue to performance is the 1KHz square wave shown in pic. Considering this tube was destined in small sized TV sets, it's done well to survive against the EL84/34 stuff.
With fixed bias; 270V B+ the AB power can be stretched close to 10W and no-one should underestimate the simplicity of fantastic listening power and simplistic design. The ECL82 does require a driver stage to deliver min 15+15V rms to the grids for 7W o/p.
The original Mullard paraphase splitter driver coupled to the o/p stage with this design begins to founder to distortion at 10KHz, so I abandoned any attempt to improve it, ..Instead opted for an optimised concertina. (I have a habit of changing tubes and I can't stand tweaking drift for min thd as in an LTP).
Hint: working at 270V B+; an ECC88 used as concertina with top/bottom resistors of 33K fed by an 82V front end stage will give symmetrical clipping on each o/p phase capable of 35Vrms signal each side.
Morgan Jones didn't mention in his books with the concertina phasesplitter that the optimised voltage drop across each top/bottom resistor and the tube should be v. roughly 1/3 on each. !!
It is well known that the original Mullard 10W amp requires 2x 10V rms to EL84 grids.
Implying the ECL82 pent has roughly half the gm of an EL84. This should make the ECL82 pentode as-tame-as-a-cat.
I believe a publisher/vendor has copyright details on circuit schematic details of Mullard designs. SY; is this correct ?
richy
With fixed bias; 270V B+ the AB power can be stretched close to 10W and no-one should underestimate the simplicity of fantastic listening power and simplistic design. The ECL82 does require a driver stage to deliver min 15+15V rms to the grids for 7W o/p.
The original Mullard paraphase splitter driver coupled to the o/p stage with this design begins to founder to distortion at 10KHz, so I abandoned any attempt to improve it, ..Instead opted for an optimised concertina. (I have a habit of changing tubes and I can't stand tweaking drift for min thd as in an LTP).
Hint: working at 270V B+; an ECC88 used as concertina with top/bottom resistors of 33K fed by an 82V front end stage will give symmetrical clipping on each o/p phase capable of 35Vrms signal each side.
Morgan Jones didn't mention in his books with the concertina phasesplitter that the optimised voltage drop across each top/bottom resistor and the tube should be v. roughly 1/3 on each. !!
It is well known that the original Mullard 10W amp requires 2x 10V rms to EL84 grids.
Implying the ECL82 pent has roughly half the gm of an EL84. This should make the ECL82 pentode as-tame-as-a-cat.
I believe a publisher/vendor has copyright details on circuit schematic details of Mullard designs. SY; is this correct ?
richy
Attachments
I have just been talking someone through a fully differential design with input transformer and input stage grid leak bias using the ECL82, it also uses plate to grid feedback with the triode voltage supplied through the feedback resistor. He is happy with the test rig and wants to build the full amp. I wanted to see how it performed as a speaker amp rather than a headphone amp (as designed originally). On my sensitive speakers it sounds as good as my class A 25W 807 amp - so i am happy with the result.
My only concern with the ECL82 is the triode section which I feel is easily overloaded at the grid and needs careful handling if not to sound harsh. This is why I used a step down input transformer in my design. I used microphone transformers with a voltage ratio of 9:1+1. This works for me in my application as a PP headphone amp - but something like a 3:1+1 would be more appropriate for most people. Heavily loaded the triode has a gain of 70 which is massively excessive for almost all designs not using unreasonable amounts of cathode feedback.
Shoog
My only concern with the ECL82 is the triode section which I feel is easily overloaded at the grid and needs careful handling if not to sound harsh. This is why I used a step down input transformer in my design. I used microphone transformers with a voltage ratio of 9:1+1. This works for me in my application as a PP headphone amp - but something like a 3:1+1 would be more appropriate for most people. Heavily loaded the triode has a gain of 70 which is massively excessive for almost all designs not using unreasonable amounts of cathode feedback.
Shoog
This what I tried
I simply took the design of Shoog's headphone amp and with his help I fiddled with some component values to get 30 mA through each ecl pentode. I used small (4 x 5 cm) OPT's formerly used by the Stoetkit jr.
The differences with the Shoog headphone amp are that I used 390R cathode resistors (to be honest I didn't have 330R in the house) and that I left out the common 680R resistor in the tail. I also used 2 x 2200 uF/50 volt under the cathodes because that value was available in my closet.
It worked out that by changing the resistor to G2 you can settle the Ia.
Changing it to about 7K gives you about 35 mA through each tube.
You always hear good stories about ecl82PP amps and I must say that I am not disappointed. I also have a EL84 amp with Lundahl input transformer => 2SK130 FET => STC style (Japan) el84 and very good vintage OPT's. But I haven't had the time to make a comparison. I suppose they are in the same league with a somewhat more "lush" sound of the el84 and more precision and speed in the ecl82. But that is speculation on this moment. Remember that both amps do not deliver the same amount of power.
Since yesterday I only have a mono version of the ecl82 with garter bias so it is a bit early to make comments. Perhaps the design can be made better, I will post seperate to get some comments from experts. I lack the knowledge and I only have 1 not very dependable cheap multimeter.
The sound at this point is very good and loud enough for normal listening level on my 88dB 3-way speaker.
I simply took the design of Shoog's headphone amp and with his help I fiddled with some component values to get 30 mA through each ecl pentode. I used small (4 x 5 cm) OPT's formerly used by the Stoetkit jr.
The differences with the Shoog headphone amp are that I used 390R cathode resistors (to be honest I didn't have 330R in the house) and that I left out the common 680R resistor in the tail. I also used 2 x 2200 uF/50 volt under the cathodes because that value was available in my closet.
It worked out that by changing the resistor to G2 you can settle the Ia.
Changing it to about 7K gives you about 35 mA through each tube.
You always hear good stories about ecl82PP amps and I must say that I am not disappointed. I also have a EL84 amp with Lundahl input transformer => 2SK130 FET => STC style (Japan) el84 and very good vintage OPT's. But I haven't had the time to make a comparison. I suppose they are in the same league with a somewhat more "lush" sound of the el84 and more precision and speed in the ecl82. But that is speculation on this moment. Remember that both amps do not deliver the same amount of power.
Since yesterday I only have a mono version of the ecl82 with garter bias so it is a bit early to make comments. Perhaps the design can be made better, I will post seperate to get some comments from experts. I lack the knowledge and I only have 1 not very dependable cheap multimeter.
The sound at this point is very good and loud enough for normal listening level on my 88dB 3-way speaker.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.