Like I said in my last message, I was wrong about the 100nF before a 470k grid resistor ... This value is OK and there is really not much difference when using a higher value one ...
Orange drops are good capacitors, in fact, everything starting with "poly" is almost as good as the paper and "paper in oil" capacitors, like the polystyrene, polypropylene, polycarbonate, polyester ... The mica capacitors are good too.
Don't use electrolytic capacitors for bypass and coupling if possible and if you have to, use a good brand like Black gate ( $$$ ), Elna Cerafine, Silmic or Elnac, Nichicon "MUSE" series KG, KZ and ES, also FW ...
But avoid using ceramic and tantalum capacitors for audio bypass and coupling, they are very bad ...
A friend of mine buy "paper in oil" capacitors from Russia on eBay, they are not very expensive and are really good !
The capacitors have a big influence on the sound quality but it is better not to use any when possible !
Alain.
I like saying Siemens, don't like that "mho" expression very much. Well, we use to say "voltage" in Spanish here when that's an English word, we should use other (I usually do!), but my friends stare at me and I have to say voltage again!
Sometimes it's no what a calculator or a simulator tell us, but what does it sound is the real thing. I have blocking distortion several times with 100n caps!
Sometimes it's no what a calculator or a simulator tell us, but what does it sound is the real thing. I have blocking distortion several times with 100n caps!
Yes, that's a very important part just to get a great tone, I always use polyester caps as coupling caps, and multilayer/multicap small ones for pF values, but I'm afraid I can't find them for high voltage operation, only 50V.
I like those little orane bugs...
Hi slideman82, I got something new for you ...
I said before the idle current is the same for cathode bias and fixed bias and it is true ...
But ... For a AB1 class push-pull amplifier, it make a very big output power and distortion differences ...
Usually, I never design AB1 push-pull amplifier stage myself, I just do like the recommendations of the tubes databooks, it is the easy and sure way ... But in your case, you already have the power and output transformers then the amplifier have to be design exactly for them ...
The class "A" amplifier have a better sound but very low power, you told me you like to make a 30W amplifier, but with your transformers, it may not be possible to go that high, the best is to design a AB1 amplifier to get the maximum power.
In class "A", a "cathode bias" can be use with no problems because the currents don't change a lot with the signal, but in a class "AB1", you are much better to use a "fixed bias" because the currents change a lot and it make the bias much higher at maximum power, then you get a very low output power ... Look by yourself :
Notice the loadline slope, it have to be like that with any other operation points because your transformer is a 3500 ohms PP ... The impedance "see" by each tube is 875 ohms, the fourth of the total impedance like for any push-pull amplifier ... The total voltage ratio is the square root of ( 3500 / 8 ) = 20,9165 : 1 ... Then the half for each tube is 10,45825 : 1 ... The square of 10,45825 x 8 = 875 ohms ...
The loadline have to cross the operation point you have choice ( under the current and dissipation limits of the tube ... ), at 300V and 25ma in that case ... To draw it up to the left or top edge of the graphic, you have to find the point where it will start, in that case, it is on the top edge at 330ma ... Then 330ma - 25ma = 305ma, from the ohm law ( E = R x I ), 875 ohms x 0,305A = 266,875V, since your operation point is at 300V, 300 - 266,875 = 33,125V ...
That is the point the load line have to start from, crossing your operation point and finish somewhere on the bottom line ... You can draw them with the little MSPaint program you have in your computer, it is very easy. You save the graphic in .BMP 16 colors first to keep just the importants colors, then in the .PNG format because it make very small files ...
Like you can see, both amplifier end stage work at the same operation points, currents and input signal ... But the "cathode bias" one have a output power of only 5,565W and the "fixed bias" have a output power of 19,883W and a much better waveshape ... Almost 4 times more power ...
The frequency response of both is not perfect, that's beacause I use a average quality output transformer for my design, with a coupling factor of only 0,99 ... I suppose your OPT is not the best of the world ... The first quality and expensive OPTs have a coupling factor of 0,999 or more and their frequency response are very flat well beyond the audible range.
The primary inductance is very important too, I use the measure you give me, 11,25H ( 5,625H per side ), the 105 ohms square resistors on my schematics are not really there in real life, they just represent the 210 ohms DC resistance of your OPT for my SIMetrix Spice simulator. You should download one, it is a fantastic useful and powerful software and it is totally free ...
The real powers of those amplifiers must be highers because the currents of my 6CA7 Spice model is lower than the one of the Tung-Sol datasheet caracteristic curves, with the best quality transformers, the power of the "fixed bias" one can reach over 21W ...
The "cathode bias" of those designs need a minimum 330,4V supply and the "fixed bias" need just a minimum 302,5V supply ... That's another reason to use the "fixed bias" because you can get a much higher power with the voltage your supply can deliver ... Of course, if your supply can deliver much more than 302,5V, you can design an amplifier with much more than 20 watts, maybe near 30 watts ...
Cheers,
Alain.
I said before the idle current is the same for cathode bias and fixed bias and it is true ...
But ... For a AB1 class push-pull amplifier, it make a very big output power and distortion differences ...
Usually, I never design AB1 push-pull amplifier stage myself, I just do like the recommendations of the tubes databooks, it is the easy and sure way ... But in your case, you already have the power and output transformers then the amplifier have to be design exactly for them ...
The class "A" amplifier have a better sound but very low power, you told me you like to make a 30W amplifier, but with your transformers, it may not be possible to go that high, the best is to design a AB1 amplifier to get the maximum power.
In class "A", a "cathode bias" can be use with no problems because the currents don't change a lot with the signal, but in a class "AB1", you are much better to use a "fixed bias" because the currents change a lot and it make the bias much higher at maximum power, then you get a very low output power ... Look by yourself :
An externally hosted image should be here but it was not working when we last tested it.
Notice the loadline slope, it have to be like that with any other operation points because your transformer is a 3500 ohms PP ... The impedance "see" by each tube is 875 ohms, the fourth of the total impedance like for any push-pull amplifier ... The total voltage ratio is the square root of ( 3500 / 8 ) = 20,9165 : 1 ... Then the half for each tube is 10,45825 : 1 ... The square of 10,45825 x 8 = 875 ohms ...
The loadline have to cross the operation point you have choice ( under the current and dissipation limits of the tube ... ), at 300V and 25ma in that case ... To draw it up to the left or top edge of the graphic, you have to find the point where it will start, in that case, it is on the top edge at 330ma ... Then 330ma - 25ma = 305ma, from the ohm law ( E = R x I ), 875 ohms x 0,305A = 266,875V, since your operation point is at 300V, 300 - 266,875 = 33,125V ...
That is the point the load line have to start from, crossing your operation point and finish somewhere on the bottom line ... You can draw them with the little MSPaint program you have in your computer, it is very easy. You save the graphic in .BMP 16 colors first to keep just the importants colors, then in the .PNG format because it make very small files ...
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Like you can see, both amplifier end stage work at the same operation points, currents and input signal ... But the "cathode bias" one have a output power of only 5,565W and the "fixed bias" have a output power of 19,883W and a much better waveshape ... Almost 4 times more power ...
The frequency response of both is not perfect, that's beacause I use a average quality output transformer for my design, with a coupling factor of only 0,99 ... I suppose your OPT is not the best of the world ... The first quality and expensive OPTs have a coupling factor of 0,999 or more and their frequency response are very flat well beyond the audible range.
The primary inductance is very important too, I use the measure you give me, 11,25H ( 5,625H per side ), the 105 ohms square resistors on my schematics are not really there in real life, they just represent the 210 ohms DC resistance of your OPT for my SIMetrix Spice simulator. You should download one, it is a fantastic useful and powerful software and it is totally free ...
The real powers of those amplifiers must be highers because the currents of my 6CA7 Spice model is lower than the one of the Tung-Sol datasheet caracteristic curves, with the best quality transformers, the power of the "fixed bias" one can reach over 21W ...
The "cathode bias" of those designs need a minimum 330,4V supply and the "fixed bias" need just a minimum 302,5V supply ... That's another reason to use the "fixed bias" because you can get a much higher power with the voltage your supply can deliver ... Of course, if your supply can deliver much more than 302,5V, you can design an amplifier with much more than 20 watts, maybe near 30 watts ...
Cheers,
Alain.
I must say this analysis you've made is great! You have spent time and brain just to help me, and I'm very grateful! Really! I thought this topic was ended until I post results of the finished amp, but I was wrong!
Well, I have to clear some thing: primary impedance is 5k, not 3.5k, don't know why they used this value, the same goes for Marshall's, datasheet says 3.5k for this voltage (the only commercial amp I've seen with 3.5k is the Vox AC50). Also, voltage with low load is 390Vdc.
Well, I have to clear some thing: primary impedance is 5k, not 3.5k, don't know why they used this value, the same goes for Marshall's, datasheet says 3.5k for this voltage (the only commercial amp I've seen with 3.5k is the Vox AC50). Also, voltage with low load is 390Vdc.
That's the way I practice the electronic theory I study since 40 years ... I like very much doing maths, amplifiers design, Spice simulations and schematics drawing. I work in the R&D departements of many electronics companys over 15 years and my job was only designing all kinds of electronic equipments with computers and softwares. But I am not very good with a solder iron and screwdrivers, that's not really my cup of tea ...
Then, if your transformer is a 5K and the supply voltage is high enough, you can have near 30W power in class AB1 with fixed bias ... But the bias will have to be adjusted carefully to not exceed the maximum 25W plate dissipation, the 8W screen grid dissipation and the 150ma cathode current of each 6CA7 at full power or they will fry ...
You can measure the current of the plates and screen grids at full power with a DC ammeter by inserting this one where are the 0 ohm R5 and R9 in this schematic and divide by 2 to get the currents in each tube. Of course, the cathode current is the plate current + the screen grid current ... You will also have to measure the voltage at the plate and the screen grid at full power with the same bias voltage ...
An externally hosted image should be here but it was not working when we last tested it.
You can take the measures at full power and 60Hz using a little power transformer at the input and a 50W 8 ohms resistor instead of the speaker. To calculate the plate and screen grid dissipation for each tube, just multiply their current by their voltage because P = V x I ...
You better to start the measures for the plates dissipation with a very high bias like - 40V and turn it down slowly until you get near the maximum dissipation, then check the screen grids dissipation. You don't have to worry too much about the cathodes currents, it will probably be lower then the 150ma maximum all the times.
Give me some news when you get to that point !
I can give you good tips for the power supply, preamplifier and phase splitter design too !
Alain.
Thanks a lot! Sure you're gonna know about it! Well, I've been messing around with solder iron, transistors and other stuff since 2005, I study EE (between 3rd and 4th year), and I build guitar and bass amplifiers for friends or so, some kind of technician, and I repaired some guitar amps. I'm a guitar player, and I'm always looking for the best tone, but, this is subjective! I will try a 3 stage 12ax7 preamp with tonestack at the end of it, before a cathodyne or a concertina phase splitter (with or without previuos triode stage, have to try), but also I want to try a LTP phase splitter too.
So, you are suggesting me to build the fixed bias version because it has better distortion figures, am I right?
So, you are suggesting me to build the fixed bias version because it has better distortion figures, am I right?
Not really, I use high efficiency speakers (100 dB of SPL) and I will use it for local gigs, and, certainly, it's enough power for the city pubs. The point is to push ir a bit, so power amp will break a bit, so I can get big tone. I've got a 60W solid state amp and never crank it when I played. So, I always wanted a 30-35W EL34 PP guitar amp, I'm a Marshall fan, but 50W is too much, and I like Eminence speakers... so...
It's more because a cathode bias is not very suitable for class "AB1", the voltage drop across the cathode resistor is very high and the current change a lot at maximum power and the bias change proportionally ... If you make the amplifier in class "A", it is but you will have a very low output power, maybe just 10 watts ...
With the 5000 ohms OPT and power transformer you already have, you can't do anything you like, you have to design the circuit for them ...
I will be back with a good and simple phase splitter idea in few days because I am very busy this week with outside work before it get too cold, winter is coming fast !
Last edited:
Today I met a guy who's been playing with tubes since he was 12, and told me 3.5k OT primary impedance for AB1 EL34 is too low, and said that EL34's needed about 6k. I know some Marshall's have this primary impedance. What's exactly the effect of increasing OT primary impedance. Iwant to know about this because Mullard datasheet specifies 3.5k.
The Mullard EL34 datasheet.
This is an interresting datasheet for your project with a 3,5K transformer. If you can get 375V from your supply, you can have 35W with a cathode bias and very low distortion. I notice they use a separate 260 ohms cathode resistor for each tube, maybe, that's why I don't get a nice waveshape in my spice simulation because I use a common resistor for both tubes ...
I will try it with those Mullard specs to see the difference and also show you my phase splitter idea when I will not be too busy ... I don't finish my outside work yet and I also have to repair my 1990 chevy motor before winter ... Construction and car mechanic are other kinds of DIY I like to do !
Alain.
I haven't read the thread, but I played (badly - my chops suck, and Im not afraid to admit it ) through a Matchless Clubman 35 that I borrowed from a friend years ago. I really liked the sound and feel of that amp, which was a cathode-biased EL34 design. It's a fairly simple amp, which I think was it's biggest strength. Coupled with a few all-tube stomp boxes it was able to do everything I wanted and it could go from a whisper to a scream.
The matchless site I linked above has some demo tracks you can listen to, and if you're interested in copying it just Google "Matchless clubman 35 schematic".
) through a Matchless Clubman 35 that I borrowed from a friend years ago. I really liked the sound and feel of that amp, which was a cathode-biased EL34 design. It's a fairly simple amp, which I think was it's biggest strength. Coupled with a few all-tube stomp boxes it was able to do everything I wanted and it could go from a whisper to a scream.The matchless site I linked above has some demo tracks you can listen to, and if you're interested in copying it just Google "Matchless clubman 35 schematic".
I haven't read the thread, but I played (badly - my chops suck, and Im not afraid to admit it
The matchless site I linked above has some demo tracks you can listen to, and if you're interested in copying it just Google "Matchless clubman 35 schematic".
That's really good info! I use well known schematics for reference all the time so I learn not to do wrong things. Matchless are really good amplifiers! BTW, this Clubman 35 has a similar preamp of what I have in mind for a friend's amp I'm going to build.
Thanks a lot!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.