excellent site for beginners on the basics of tube guitar amps:
http://www.freewebs.com/valvewizard/index.html
http://www.freewebs.com/valvewizard/index.html
Ok ive studied your text. I have an Class D TA2024 over here wich needs a higher input ( ok i know the resistor mod but would be nice to design a tube pre amp for it).
Ive have a couple of 6FQ7 tubes over here.. and they seem have a low rp and mu.
rp: 6700
mu: 20
lets say we take this diagram:
An externally hosted image should be here but it was not working when we last tested it.
So we have for example:
rp: 6700
mu:20
Rk:3300
Rp:220k
Output resistor 1M
If im wright you have to calculate the parallel resistance of Rp and 1M ?
Thats about 180K.
Amplification factor would be:
mu 180k/(rp+180k+Rp (mu+1) = 14 times.
Is this right ?
But about the problems you write about... so you calculate the fequency of in this case 180K (Rp+1M) and 2uF ?
1 / 2 pi R C = 0,5 hertz ?
Dear All,
I'm new at this forum and I'm currently building an RH84 se amp:
RH 84 - Tube Audio ...... RH DESIGN
I would like to know the overall voltage gain of the circuit, because my cd player has only 1V rms output. Is it enough to drive it to the clipping point or no?
My transformer ratio is 1:25 of course. (square root of 5000/8)
Thank you very much.
I'm new at this forum and I'm currently building an RH84 se amp:
RH 84 - Tube Audio ...... RH DESIGN
I would like to know the overall voltage gain of the circuit, because my cd player has only 1V rms output. Is it enough to drive it to the clipping point or no?
My transformer ratio is 1:25 of course. (square root of 5000/8)
Thank you very much.
no. The 1M is almost negligible. Output impedance is approx. equal to the parallel of Rp and rp.
For the input, I'd use a 100K volume pot (with no extra R in series), and likely one or more LED or a Zener diode (whatever fits better the required Vgk) in place of the R//C for cathode BIAS.
Oh, anyway, for the "automatic" bias case with cathode resistor, 2200uF WRT 3.3K is way too high a value for the by-pass...
no. The 1M is almost negligible. Output impedance is approx. equal to the parallel of Rp and rp.
For the input, I'd use a 100K volume pot (with no extra R in series), and likely one or more LED or a Zener diode (whatever fits better the required Vgk) in place of the R//C for cathode BIAS.
Oh, anyway, for the "automatic" bias case with cathode resistor, 2200uF WRT 3.3K is way too high a value for the by-pass...
Aah ok... but i ment 1M with Rp to calculate the amplifie factor
Can you explain why 2200uF with 3.3K is too high ?
The Xc of the by-pass cap at the minimum frequency (would be 20Hz, but it's better to be conservative and use 10Hz instead) must be "negligible" (<= 1/10) WRT the value of the R. That is:
Xc = 1/2*pi*f*C <= R/10 @ f=10Hz
Thus:
C >= 1/2*pi*R
With R=3.3K, C ~= 47uF. You may use a somewhat bigger C to be even more conservative (say 100 or 220uF), but there is really no point on using a much larger value such as 2200uF.
Xc = 1/2*pi*f*C <= R/10 @ f=10Hz
Thus:
C >= 1/2*pi*R
With R=3.3K, C ~= 47uF. You may use a somewhat bigger C to be even more conservative (say 100 or 220uF), but there is really no point on using a much larger value such as 2200uF.
I can calculate 1/2*pi*R.. if you use 3300ohm C would be like 48uF.
But i don't get the other calculation.
The capacitive reactance (Xc)= 1/2*pi*f*c = 1/2*pi*10*47uF = 338.6 ? right ?
for 20hert example.. its would be Xc of 169.3 ?
And for 2200uF with 10herts it would be Xc 7.23
what is the idea behind these value's ?
As said, the idea is that you want Xc << R at the minimum frequency of interest (20Hz).
About your calculation, ~170 ohm is << 3.3K, so that value is ok. Of course so it is ~7 ohm, but there's no point in goin' so far. A ratio of 1:10 if enough. Using too big a capacitor only brings in more troubles, as e.g. it will have worse performance at higher freq. Oh, I'd use a much higher operating voltage cap, at least ~50V.
As for 47uF vs. 48uF, that does not really make any difference! Consider that capacitors (and particularly electrolitics) do have typ. 20% tolerance. A nominally 47uF cap. can be in fact anything between 38u and 56u, a 100uF one may range from 80u to 120u and so on!
I said 47u 'cause that's the closest standard value.
About your calculation, ~170 ohm is << 3.3K, so that value is ok. Of course so it is ~7 ohm, but there's no point in goin' so far. A ratio of 1:10 if enough. Using too big a capacitor only brings in more troubles, as e.g. it will have worse performance at higher freq. Oh, I'd use a much higher operating voltage cap, at least ~50V.
As for 47uF vs. 48uF, that does not really make any difference! Consider that capacitors (and particularly electrolitics) do have typ. 20% tolerance. A nominally 47uF cap. can be in fact anything between 38u and 56u, a 100uF one may range from 80u to 120u and so on!
I said 47u 'cause that's the closest standard value.
The picture becomes clear now ! 
undervolting or overvolting is bad for any tube...a tube rated for 6.3 can work with voltages 5.9 to 6.9 volts with out problems.......
if your filament supply is higher even under full load, you can always add a resistor in series wit the the filaments to trim off excess voltage. it is a simple appplication of ohms' law...
Heater Voltage
This is an interesting problem. Higher voltage increases the output/gain, but shortens life. Lower voltage does the opposite. However, we must remember that line voltages varied by up to 20% around the nation for a good part of the period of tube manufacturing--and tubes had to take this variance. Very seldom was the filter supply regulated.
I have done quite a bit of testing on varying voltages and my general recommendation is to keep power tubes at the rated voltage--this seems to give the best average performance. Small signal tubes are more forgiving, and so I like to run that at the rated value, or up to 5% under to save life. You can run them hot, but that will shorten life.
There are a few exceptions to the rule. 5687, for example, like to run hot as do 12B4s, at least with my testing of lots to date.
This is an interesting problem. Higher voltage increases the output/gain, but shortens life. Lower voltage does the opposite. However, we must remember that line voltages varied by up to 20% around the nation for a good part of the period of tube manufacturing--and tubes had to take this variance. Very seldom was the filter supply regulated.
I have done quite a bit of testing on varying voltages and my general recommendation is to keep power tubes at the rated voltage--this seems to give the best average performance. Small signal tubes are more forgiving, and so I like to run that at the rated value, or up to 5% under to save life. You can run them hot, but that will shorten life.
There are a few exceptions to the rule. 5687, for example, like to run hot as do 12B4s, at least with my testing of lots to date.
donerlank
Anyone who has built Fred Nachbauers Miniblok set? Or his Miniblok PP amp? If 13em7's are hard to come by in sunny SA, will 12ax7's work or maybe the new JJ ecc832/12dm7's? or perhaps 7025's which are listed as available here? If I understand Fred's schematic & write up correctly, the hi-gain half is used as a preamp and the low gain hi current half as the output. The miniblok 2 is a pp amp using 2 of these tubes, the 2nd tube high gain stage is used as a phase inverter for pp. The project appeals to me to cut my teeth with a kit, or can anyone recommend a better set project. It needs to be simple. I'm a blockhead at this stage.
Anyone who has built Fred Nachbauers Miniblok set? Or his Miniblok PP amp? If 13em7's are hard to come by in sunny SA, will 12ax7's work or maybe the new JJ ecc832/12dm7's? or perhaps 7025's which are listed as available here? If I understand Fred's schematic & write up correctly, the hi-gain half is used as a preamp and the low gain hi current half as the output. The miniblok 2 is a pp amp using 2 of these tubes, the 2nd tube high gain stage is used as a phase inverter for pp. The project appeals to me to cut my teeth with a kit, or can anyone recommend a better set project. It needs to be simple. I'm a blockhead at this stage.
This is a great idea for a thread.
This is one of the best beginner sites I've found, it walks you step by step through the design and construction of a simple single ended amp:
http://boozhoundlabs.com/howto/
I am afraid that this site is not reachabe now. I have seen it 3 days ago and it was a good site for beginners.
Thank you nthere, tried boozhound in spite of what you said (just in case) always being hopeful, I have visited Hans Niejbaum's 6C33C OTL site as well as Tom Mellow's OTL site. OTL is the way I will eventually go, as Philips OTL's were marketed under the Novosonic brand in my teenage days.Ist encounter with Hi-Fi, still mono in those days, stereo came a bit later. Novosonic's had a kick like a proverbial Mississippi mule. Could feel drums against your chest if u were listening to a percussion band. Handled classical music equally well; a pity this avenue of sound reproduction wasn't followed thru, though some were more in love the the lush, mellow Dynaco sound. Sound of transformer windings? Vere nice though. But otl's aside for the moment, I need to familiarise myself with an easy inexpensive project that works. Then...OTL's. Yum, Yum. I understand the principle of the late Fred Nachbauers topology very well, I am just not sure I can substite the 13em7's with other tubes (i.e dissimilar dual trodes) but I'll get there. Don
donerlank
Thanks,Wintermute, successfully accessed Boozehound and perhaps you will be pleased to know that the lights have started flickering for me. Isnt knowledge wonderful? Knowledge gives power, power gives enablement, enablement brings satisfaction. Dunno who said that. Perhaps it was Confucious, could even be Julius Ceasar or Shakespeare, does it matter?, or better still:Mr Tube. I will now be able to choose a tube, do some maths to determine if it will work in a given circuit, even modify the circuit to suit the tube and later on design my own circuit. I must remember to thank Boozhound also.
Thanks,Wintermute, successfully accessed Boozehound and perhaps you will be pleased to know that the lights have started flickering for me. Isnt knowledge wonderful? Knowledge gives power, power gives enablement, enablement brings satisfaction. Dunno who said that. Perhaps it was Confucious, could even be Julius Ceasar or Shakespeare, does it matter?, or better still:Mr Tube. I will now be able to choose a tube, do some maths to determine if it will work in a given circuit, even modify the circuit to suit the tube and later on design my own circuit. I must remember to thank Boozhound also.
i can buy it back from you if you can bring it to the house....
you can buy JJ 12ax7 tubes here: http://www.epektos.com/catalog/90270c5d-17be-428d-8204-30703444af81.aspx
Last edited: