|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
65.9.9.22
Folks,
This post is prompted by the review of the Tube Audio Lab WE91 300B Monoblocks posted on the SET Asylum and the subsequent response by the builder who states that:
"It definitely has a very distinct sound (interesting and old), partially thanks to the pentode driver and 2:1 plate to screen grid suppressor voltage ratio.
...
If any of you are going to build a WE91A with the 6C6, I'd suggest extra caution be taken about the value of caps used for the screen grid (suppressor) of the 6C6. To get more refined bass, I'd recommend 2uF - 2.8uF oil caps. Anything outside this range or different material of caps will give you a slight upper midrage glare that nobody wants".
Thus my question: has anyone here compared the sound of different pentode drivers, i.e. 6C6 vs. EF37 vs. 717A, etc.
I am curious to learn from your experience,
Follow Ups:
hey-Hey!!!,
The metal 6AC7 for one, type 12HL7 too, as well as the 6CL6 and a directly heated one type 7905( one of the highest gm DH tubes ).
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
I have 3 amps, all using a pentode driving a SET.
One is a WE91A clone, using WE310A.
One is a Yamamoto A08S, using 717A driving a 45.
One is a EF37A driving a PX4.
So I had a bit of experience on all 3 tubes, but I have not been able to fix the output tube and hear the difference of all 3 drivers.
There is a definite "signiture sound" of a pentode driving a SET. I would say the tone is more refine, more timbre to the instrument. But the driving ability of a pentode is very weak. Straightly limited to class A1. In this aspect, I think 45 and especially a PX4/PX25 is better suited for pentode driver as the current requirement is less than a 300B.
The problem also is that not all 3 pentodes are the same, and the driving ability of all 3 pentodes are different. I have extensively experimented with the EF37A, with different plate voltage and screen voltage. Every time I increase the plate current, the sound becomes more refine, more detail. I would suspect it was due to the fact that the driver becomes less slew rate current limited.
Currently I am using the EF37A with 4.5mA plate current and 114V screen voltage, running the EF37A near its maximum dissipation. I had previously used a 6J7, which has a lower transconductance of around 1.2mA/V, and sound less dynamic that a EF37A which has around 1.8-2mA/V transconductance.
For the WE91A, if you read Thorsten's post, you can see that he uses a positive feedback in screen to tackle the HF rolloff due to the driver's high output impedance.
So not just the driver tube, but the operating point of the tube would change the sound.
Thank you very much for the details - this is exactly what I was looking for.
Hi.
i'm also using EF37a driving 2a3 and i have to say they are SOUND Increable together but i don't think i can get full power using the EF37 as driver tube .so 300b is even harder to drive , i would not recommend either of those pentode listed as drive for 300b.
anyways what is your plate load R for the EF37? my own experiene is that the screen grid voltage have biggest impact to the sound of the whole amp .
LT
IIRC, apart form the slightly increased voltage requirements (depending on operating point), I did not think the 300B much more difficult to drive than the 2A3? Same also for the PX25 vs. 300B - less voltage required for PX25, but...
Of course, I could be way off-base.
Cheers
Hi Daniel.
to drive the 300b correctly I don't think two stage pentode front can do the job . you might get away with two stage if you only want 4 or 5 watts out the 300b but not full power . 300b is much harder to drive than the 2a3 . the bias requirements on the 300b is about 1/3 of what the 2a3 requirements for full power.
cheers
LT
The 300B may have about a third more voltage drive requrement (to get close to max output) than a 2A3, however the current requirements are about the same. I did not think it was difficult to get decent gain, and therefore voltage swing, out of pentodes; current is the challenge. I don't believe a pentode is any more appropriate for driving a 2A3 than 300B or that for this topology a 2A3 is more suited to 2-stage. I feel that pentodes are probably better for driving the 45, which has significantly lower requirements for driver current than 2A3, 300B, PX4, or PX25.Again, it is ages since I have looked at the technicalities of driving DHTs, so could be way off.
Cheers
Edits: monor clarification of points.
Edits: 02/15/09 02/15/09
Hi, I even tried the EF39 from Hytron, WWII mfg ... as sub of 6J7 ! No difference with the RCA 6J7 . The EF39 is given as semi-remote pentode but operating point remained the same. These have a metallic base and a white opaline looking glass . A very very nice construction , something looking like a Mullard EF37A. Did somebody test other semi remote toobz AKA 6K7 as there are tons laying around ? cheers Pierre
I have tried various plate resistor. From 220k to currently 47k. In my experience, for a pentode, the plate resistor only affects the gain and output impedence, not the sound (assuming the impedance is enough to drive the next stage). Thorsten once wrote that chose the lowest plate resistor that will give you enough gain. The output impedance of the pentode driver is nearly equal to the plate resistor. A 47k resistor allows me to have just enough gain to use a passive preamp. I think the amp would be -1db at 20kHz on 1W output. Mind you, that is a PX4, much easier to drive than a 300B.I totally agree with you that the screen voltage has a immense impact on the sound. Look at the plate curves, an increase in the screen voltage will push the knee of the plate curve to the left (The area which the plate curve turns 90 degrees). So the plate voltage will not swing into the distorted area of the "knees" of the plate curve.
I also think that the plate current will affect the sound, mainly the current delivering ability to drive the miller capacitance of the SET.
In my EF37A driver, I tried to maximize both parameters, which in the end were limited by the max. plate dissipation.
The 717A has a better current handling ability, but the maximum anode voltage was lower, limiting the maximum voltage swing. For a 300B, we need a +/- 70V peak to peak to drive it to max power in A1. The 717A maximum plate voltage is only 180V, where as the EF37 is 300V. On the other hand, the max. plate current of the 717A is 7.5mA where as the EF37 is 6mA.
The WE310A or 6C6 are even worse than the 717A or EF37.
My current favourite is the EF36/EF37/EF37A. Not just the sound, but the red colour coke bottle with a grid cap on the amp just look nice. Some day, I will try and find the silver or gold coated version from Mullard, just for the looks!
My next try will be on using those super pentodes like E810F, E180F and D3a. I am just a bit intimidated by their high transconductance and their tendency to oscillate.
Anybody has any experience on these "super pentodes"?
P.S. The C3m and C3g family is another choice, only that I do not have any loctal sockets. Yamamoto recently use the C3m as the driver of their new 300B amp.
Raymond
Edits: 02/14/09 02/14/09
NT
Hi Raymond.
look like we are a pentode lover . your plate load on the EF37 seem to be in a very low side , I assume you run 2 stage SE ?
47k load will not give you much gain as you already aware of the low load on the pentode.
i think 82k is about the right load with around 300v in .
what is your screen grid resistor ?
i'm on my way designing the drive for my new 300b SE so i'm picking the right drive at the moment , too many choice here to thinking about,
Cheers
LT
Hi LT,
I used a low plate resistor of 47k mainly to minimize the output impedance of the driver. Also a smaller plate resistor will give you a higher plate voltage.
My B+ is 350V and the plate voltage is 180V. I need +/-50V peak to peak anode swing to drive the PX4 to full output. The screen voltage is about 116V. So the plate will never have a lower voltage than screen. If you look at the plate curves, if the plate voltage is lower than the screen, you run into the very non linear part of the plate curves. I tried to minimize that.
The aim is to prevent the driver clipping before the DHT.
The other factor that affects the plate voltage is the plate current. In a pentode, the plate current is controlled by the screen voltage and grid voltage. So if you want a higher plate voltage, you either lower the plate current, or lower the plate load resistor. Other than that, you need to increase the B+.
The way I did it was select the plate current that would minimize the DHT slew rate limitations first. Fixed that, then calculate the lowest plate resistor that can give me enough gain. I then look at the plate curves and select the screen voltage that would give me that plate current.
I than look the the grid voltage and fixed that. Finally I calculate the B+ I needed to get the plate voltage. The plate plate voltage should be > screen voltage + negative bias voltage of the DHT. The B+ would be plate voltage + (plate current x plate resistor).
Finally check the plate dissipation and see if it is over spec.
Thorsten did a nice write up in diyaudio about how to design a good pentode driver for DHTs.
Well LT, I am not a pentode lover. My favourite output tube is still the PX4 :-)
But I have to say the sound of a pentode driving a SET just sounded right to me. It is the tone and the timbre. I think it may not be too good for rock and roll.
I am not using a screen grid resistor, but like Yamamoto, a zener chain to fix the screen voltage.
Please see my circuit in the diyaudio post.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=137447&highlight=
sorry .
I would have said pentode driving SET .
anyways I can't read your circuit due to not a menber of other forum.
can you post it here please .
LT
PS: I'm thinking of VR tube for the screen grid if room is allow .
The problem with using zener is the noise from a current starved zener. Guys over in diyaudio helped me out and I was able to kill off must noise.
Here is the schematic. Note that the PX4 is operating over the spec, slightly. I really should have get the KR version of the tube, as it can handle a plate dissipation of 20W.
I tried the PX25, it just do not sound as good as a PX4. For guys here in the asylum who are more North American, the tips is that many DHT and DH rectifiers from Europe sounded as good as WE products but much cheaper. Worth a try. You will be suprised.
Raymond
Edits: 02/14/09 02/14/09
LT,
Pete Millett uses the EF36/37A as front end for his 829B amp. I've linked the schematic and web page below.
The neat thing is that he uses a VR 0A3 tube for screen grid voltage reference.
http://www.pmillett.com/829b_amplifier.htm
Enjoy,
hey-Hey!!!,
That is another pentode input stage design. It can also be done with a cascode, and to preserve headroom, use a MOSFET on top of the triode. 5687/6H30 is a good place to start.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: