I'm trying out some driver tubes for my 300b SE amp. Currently listening to the EL84 in triode which is sounding much better than I expected. Setup is a LL1660/18mA used as a plate choke with 0.2uF Russian FT-2 teflon caps. My first try out has 14.5mA current, -5.2 bias, 155v a-k. using a 360R cathode resistor unbypassed. I'm sure this can be improved with some further experiments. Take the current up to 18mA and look at different a-k voltages. I can also try SIC diode bias.
Present tubes are Sovtek EL84 which is probably same as 6P14P. So I'm now wondering if I'd get any better sound out of a SV83 or 6P15P or other close cousins. I'm also intending to try the EL33.
I'd appreciate plenty of useful feedback on this!
Present tubes are Sovtek EL84 which is probably same as 6P14P. So I'm now wondering if I'd get any better sound out of a SV83 or 6P15P or other close cousins. I'm also intending to try the EL33.
I'd appreciate plenty of useful feedback on this!
Attachments
Member
Joined 2009
Paid Member
I’ve used the 6P15P in a spud amp, triode wired, and found it to be smoother than the 6E5P so I would use this tube again. Regardless, it depends on so many factors that I would not be wanting to choose for YOUR circuit without breadboarding it with your iron etc.
... you know, of course, that whichever is the more expensive, more troublesome to use and the most limited in terms of specs must be the best sounding one and if it’s rare enough that nobody else can find a good NOS copy with a certain shaped hole in the plate construction then it would be legendary in it’s sound 😀
... you know, of course, that whichever is the more expensive, more troublesome to use and the most limited in terms of specs must be the best sounding one and if it’s rare enough that nobody else can find a good NOS copy with a certain shaped hole in the plate construction then it would be legendary in it’s sound 😀
Andy,
Klausmobile had extensive testing of 6P15 (SV83) and compared triode curves with EL84 triode curves
View last sentence (and curves from tester) on his site ... "EL84 (Saratov) curve are identical to 6P15P (SV83)"
Tube Tester Files - 6П15П - 6Pi15Pi (SV83)
In pentode mode, Thorsten Loesch said a long time ago that the SV83 sounds a little better.
For triode mode, I dont know.
Best DIYing
Klausmobile had extensive testing of 6P15 (SV83) and compared triode curves with EL84 triode curves
View last sentence (and curves from tester) on his site ... "EL84 (Saratov) curve are identical to 6P15P (SV83)"
Tube Tester Files - 6П15П - 6Pi15Pi (SV83)
In pentode mode, Thorsten Loesch said a long time ago that the SV83 sounds a little better.
For triode mode, I dont know.
Best DIYing
Last edited:
Thanks for that! Smoother than the 6e5P is good. I tried a 6e6P and wasn't blown away. I also prefer not to use high transconductance tubes because of the risks of oscillation. 6P15P is definitely on my agenda.
I have Klaus' curves. I do believe the 6P14P and 6P15P are not identical - not sure where I read it but I think the internal construction is slightly different.
I have Klaus' curves. I do believe the 6P14P and 6P15P are not identical - not sure where I read it but I think the internal construction is slightly different.
The key difference between 6Π14Π and 6Π15Π lies in screen grid toughness. The screen grid in the 6Π15Π is fragile. Note the 150 V. limit. When triode wiring the 6Π15Π, tie g2 to the plate with a 1 Kohm resistor.
Sockets correctly wired for the 6Π15Π accept EL84s without incident. The usual remark about EB suffixed specimens being superior to their unsuffixed bretheren.
Sockets correctly wired for the 6Π15Π accept EL84s without incident. The usual remark about EB suffixed specimens being superior to their unsuffixed bretheren.
I believe you are using the 18 mA air gapped 1660's two A primaries and the two B primaries in series.
That is 100H.
At 20Hz the inductive reactance is 12.5k.
The EL84 triode wired u is about 20, and Gm is about 10k uMhos.
12.5k/u. 1/10k uMho
The cathode impedance is 12.5k/20 + 100 Ohms = 725 Ohms.
At 20Hz, the cathode impedance of 725 Ohms interacts with the 325 Ohm self bias resistor.
That will cause the low frequency response to roll off there (about - 3dB @ 20Hz, about - 1dB @ 40 Hz).
Some people do not use a Bypass cap across the self bias resistor.
Now you know why some people do use a Bypass cap across the self bias resistor.
Low frequency roll off, or much less low frequency roll off.
In the case of any inductive load that has lots of distributed capacitance across the windings, the same effect of frequency roll off happens at high frequencies.
Bypass or not (test the frequency extremes, and decide).
I have always wanted to try using an EL84 triode wired as the driver tube for a 2A3.
When I get around to building my next 2A3 amp, perhaps I will use an EL84 triode wired driver.
That is 100H.
At 20Hz the inductive reactance is 12.5k.
The EL84 triode wired u is about 20, and Gm is about 10k uMhos.
12.5k/u. 1/10k uMho
The cathode impedance is 12.5k/20 + 100 Ohms = 725 Ohms.
At 20Hz, the cathode impedance of 725 Ohms interacts with the 325 Ohm self bias resistor.
That will cause the low frequency response to roll off there (about - 3dB @ 20Hz, about - 1dB @ 40 Hz).
Some people do not use a Bypass cap across the self bias resistor.
Now you know why some people do use a Bypass cap across the self bias resistor.
Low frequency roll off, or much less low frequency roll off.
In the case of any inductive load that has lots of distributed capacitance across the windings, the same effect of frequency roll off happens at high frequencies.
Bypass or not (test the frequency extremes, and decide).
I have always wanted to try using an EL84 triode wired as the driver tube for a 2A3.
When I get around to building my next 2A3 amp, perhaps I will use an EL84 triode wired driver.
Last edited:
I'm using C+C windings. I haven't wired up all the coils in series - I could add A and B windings. I had no luck doing this with a Hammond 126C. Don't know if it works with the LL1660. Should be 100H. I can live with -1db at 40hz since my speakers already have a bass rolloff since I'm using Alpair 10Ms in infinite baffle. I'll see if I can reduce the size of the cathode resistor. At this a-k it should be 270R instead of 360R for 18mA current.
Someone in another forum said the EL84 had some kind of high frequency peak which gives the impression of extra clarity. I know nothing of this. But the sound does indeed have a lot of clarity to it.
Someone in another forum said the EL84 had some kind of high frequency peak which gives the impression of extra clarity. I know nothing of this. But the sound does indeed have a lot of clarity to it.
Triode wired EL84s and close relations are nice. Triode wired 6V6 family tubes are nicer. 😉 It has been said to do better than a triode wired 6V6, in this power range, a #45 is necessary.
I'd second Eli's statement.
One of the best 300B amps I've heard was something like 6J5 DC coupled to trioded 6V6 IT coupled to 300B.
My personal love for 6V6 family detracts from the objectivity a bit though 🙂
One of the best 300B amps I've heard was something like 6J5 DC coupled to trioded 6V6 IT coupled to 300B.
My personal love for 6V6 family detracts from the objectivity a bit though 🙂
I was confused. The C windings are the high impedance side.
The only thing to wonder about is the unterminated A and B windings. They will resonate.
I did not check to see if Lundahl has an 18mA 100H plate choke; if so, then there are no unterminated windings.
A high frequency peak from an EL84 that is triode wired probably means some other part or parts of the circuit are peaky, not the tube.
Or, the "high frequency peak sound", may just mean that there is no high frequency roll off.
Lot of recordings have had producers that peaked the high frequency response.
The only thing to wonder about is the unterminated A and B windings. They will resonate.
I did not check to see if Lundahl has an 18mA 100H plate choke; if so, then there are no unterminated windings.
A high frequency peak from an EL84 that is triode wired probably means some other part or parts of the circuit are peaky, not the tube.
Or, the "high frequency peak sound", may just mean that there is no high frequency roll off.
Lot of recordings have had producers that peaked the high frequency response.
Thanks for that schematic - I notice an unbypassed 110R cathode resistor. Evidently someone else who hates cathode bypasses.
I've started going through Salas' long thread on a 6V6 line stage to see what I can pick up there. He clearly prefers it to 6SN7 and a few other choices, which asks the question why so many designs using 6SN7 into 300b exist. The gain is very similar so that isn't an advantage. I'm going to try out the octal varieties as well, starting with EL33.
I'm getting some very low level hum, which I didn't expect - not sure where that is coming from in my system. Poindexter says that some 6V6 he tried were hummy, citing RCA and Bendix.
I've started going through Salas' long thread on a 6V6 line stage to see what I can pick up there. He clearly prefers it to 6SN7 and a few other choices, which asks the question why so many designs using 6SN7 into 300b exist. The gain is very similar so that isn't an advantage. I'm going to try out the octal varieties as well, starting with EL33.
I'm getting some very low level hum, which I didn't expect - not sure where that is coming from in my system. Poindexter says that some 6V6 he tried were hummy, citing RCA and Bendix.
I suggested to my friend to try 2+2 RedLed in parallel for EL84 cathode, but he doesn't want it. Because the amplifier sounds really beautiful.
My opinion, EL84-2A3 direct copuled amplifier is the best idea today for tube amplifier.
If I didn’t have my two SETs, I would make that amp today, for sure.
BTW, today's story about 6V6 reminds me a lot of the old story about 6SN7 .... Just my personal opinion.
I have RCA metal case 6V6, tried them, nothing special. Siemens EL84 is better, much better.
My opinion, EL84-2A3 direct copuled amplifier is the best idea today for tube amplifier.
If I didn’t have my two SETs, I would make that amp today, for sure.
BTW, today's story about 6V6 reminds me a lot of the old story about 6SN7 .... Just my personal opinion.
I have RCA metal case 6V6, tried them, nothing special. Siemens EL84 is better, much better.
Last edited:
Salas measures and listens to things carefully, and I am not surprised he prefers 6V6 in a driver position - I do too.
Now that you have an EL84 in your driver socket, please take the opportunity to wire in in pentode, and compare. The big opportunity is that you won't need the opamp gain-stage, when you do that. A good starting point is the Thorsten "Legacy" (Va=250V; Vg2=150V RC filtered; Ik=ca. 12mA).
Simple, and surprisingly good - in my setup, the articulation of voices was much better than 6SN7 (2-stages).
6BW7 in the same socket was better still, or you can still get Mullard EF80 (or UF80). These are all *designed* to run at Ik ca. 12-15mA, and are more linear as a result.
Pentode stages suffer if the driver's anode supply is not up to snuff, though.
Hello Rod
Thanks for the suggestions. I did struggle with a pentode connection of 4P1L but didn't manage to get any good results out of it. I'm quite happy with my NE5534 input board so I'm going to stay with triode. In my system triode connected EL84 is very nearly loud enough so I don't need much more gain, in fact the x5.5 gain from the NE5534s is far too much, though my limited understanding of op-amps suggests they are more stable at over x5 gain. With this op-amp stage I can play around with other triode choices like the 46 which I have working.
I have a few alternative valves on my radar and have wired up a pair of octal sockets to try out those. There are a lot of metal ones I could try, though I don't really want to buy a load of valves speculatively on eBay. I have some 6P6S, a Russian equivalent of 6V6, so I'll try those. I may even end up liking one of these indirectly heated valves as much as the 46, which I never thought possible. They are not worlds apart.
Thanks for the suggestions. I did struggle with a pentode connection of 4P1L but didn't manage to get any good results out of it. I'm quite happy with my NE5534 input board so I'm going to stay with triode. In my system triode connected EL84 is very nearly loud enough so I don't need much more gain, in fact the x5.5 gain from the NE5534s is far too much, though my limited understanding of op-amps suggests they are more stable at over x5 gain. With this op-amp stage I can play around with other triode choices like the 46 which I have working.
I have a few alternative valves on my radar and have wired up a pair of octal sockets to try out those. There are a lot of metal ones I could try, though I don't really want to buy a load of valves speculatively on eBay. I have some 6P6S, a Russian equivalent of 6V6, so I'll try those. I may even end up liking one of these indirectly heated valves as much as the 46, which I never thought possible. They are not worlds apart.
Andy,
the last few days I used to test the replacement for El84.
I found it, it's just an E80L.:
This is my personal model for the E80L:
.SUBCKT E80L_P 1 4 2 3 ; (E74GD) 25.3.10.
+ PARAMS: MU= 23.5 EX= 1.263 KG1= 563.0 KP=149.19 KG2= 1400
+ KVB= 16.1 VCT= 0.00 RGI=1k
+ CCG=11P CPG1=0.1P CCP=7P
RE1 7 0 1G
RE2 8 3 1G
E1 7 0 VALUE= ; E1 BREAKS UP LONG EQUATION FOR G1.
+{V(4,3)/KP*LOG(1+EXP((1/MU+V(2,3)/V(4,3))*KP))}
G1 1 3 VALUE={(PWR(V(7),EX)+PWRS(V(7),EX))/KG1*ATAN(V(1,3)/KVB)}
G2 8 3 VALUE={(PWR(V(7),EX)+PWRS(V(7),EX))/KG2*(2.5708-ATAN(V(1,3)/KVB))}
E2 8 4 VALUE={0}
RCP 1 3 1G ; FOR CONVERGENCE
C1 2 3 {CCG} ; CATHODE-GRID 1
C2 1 2 {CPG1} ; GRID 1-PLATE
C3 1 3 {CCP} ; CATHODE-PLATE
R1 2 5 {RGI} ; FOR GRID CURRENT
D3 5 3 DX ; FOR GRID CURRENT
.MODEL DX D(IS=1N RS=1 CJO=10PF TT=1N)
.ENDS
the last few days I used to test the replacement for El84.
I found it, it's just an E80L.:
This is my personal model for the E80L:
.SUBCKT E80L_P 1 4 2 3 ; (E74GD) 25.3.10.
+ PARAMS: MU= 23.5 EX= 1.263 KG1= 563.0 KP=149.19 KG2= 1400
+ KVB= 16.1 VCT= 0.00 RGI=1k
+ CCG=11P CPG1=0.1P CCP=7P
RE1 7 0 1G
RE2 8 3 1G
E1 7 0 VALUE= ; E1 BREAKS UP LONG EQUATION FOR G1.
+{V(4,3)/KP*LOG(1+EXP((1/MU+V(2,3)/V(4,3))*KP))}
G1 1 3 VALUE={(PWR(V(7),EX)+PWRS(V(7),EX))/KG1*ATAN(V(1,3)/KVB)}
G2 8 3 VALUE={(PWR(V(7),EX)+PWRS(V(7),EX))/KG2*(2.5708-ATAN(V(1,3)/KVB))}
E2 8 4 VALUE={0}
RCP 1 3 1G ; FOR CONVERGENCE
C1 2 3 {CCG} ; CATHODE-GRID 1
C2 1 2 {CPG1} ; GRID 1-PLATE
C3 1 3 {CCP} ; CATHODE-PLATE
R1 2 5 {RGI} ; FOR GRID CURRENT
D3 5 3 DX ; FOR GRID CURRENT
.MODEL DX D(IS=1N RS=1 CJO=10PF TT=1N)
.ENDS
Hi Andy,
Last 15 years I tried a dozed tubes as driver in my home made 300b se. 6V6, El84, 2A3, 12BY7, E180F, E810F etc. As Rod mentioned I got best sound from EF80 triode strapped.
Plate voltage 180 volt at 10ma. 300B plate voltage 357 volt at 50ma. Crystal clear and lush sound-nirvana at last. By the way, I use 3A5 pre with 15k:600 as parafeed output transformer. Zero hum so far. Please try it I am sure you will like it.
Regards
Last 15 years I tried a dozed tubes as driver in my home made 300b se. 6V6, El84, 2A3, 12BY7, E180F, E810F etc. As Rod mentioned I got best sound from EF80 triode strapped.
Plate voltage 180 volt at 10ma. 300B plate voltage 357 volt at 50ma. Crystal clear and lush sound-nirvana at last. By the way, I use 3A5 pre with 15k:600 as parafeed output transformer. Zero hum so far. Please try it I am sure you will like it.
Regards
The schematic in Post # 11 . . .
Very nice circuit!
But . . .
It grounds one end of the 6.3V filament of the EL84.
Another way is to use two 100 Ohm resistors, 100 Ohms from each end of the 6.3V to ground (virtual center tap).
That means there is no more than 3.15V from any part of the filament to the EL84 cathode.
Any filament to cathode leakage might result in hum.
(I know, select low leakage / low hum tubes).
A bypass cap across the self bias resistor will reduce the filament to cathode leakage hum.
But of course, the bypass cap will also increase the gain of that stage.
At least use some kind of plastic cap, not an electrolytic cap there. More expensive, but better sound.
And some do not like bypass caps at all . . .
Very nice circuit!
But . . .
It grounds one end of the 6.3V filament of the EL84.
Another way is to use two 100 Ohm resistors, 100 Ohms from each end of the 6.3V to ground (virtual center tap).
That means there is no more than 3.15V from any part of the filament to the EL84 cathode.
Any filament to cathode leakage might result in hum.
(I know, select low leakage / low hum tubes).
A bypass cap across the self bias resistor will reduce the filament to cathode leakage hum.
But of course, the bypass cap will also increase the gain of that stage.
At least use some kind of plastic cap, not an electrolytic cap there. More expensive, but better sound.
And some do not like bypass caps at all . . .