pentodes are more like Norton's, very high mu and highish plate resistance, if you look at plate curves for pentodes, almost horizontal in most cases, and so ignored in designs, that is why pentode mu's are not to be found in many datasheets..
"μg2g1" is very useful, because in a pentode, the screen is like the plate in a triode....the pentode tube cathode currents are controlled two ways, first is by g1 and second is by g2....
if you know your "μg2g1" and assign a g2 voltage on your pentode, then you can calculate g1 voltage because of this relationship...something like, Vg1 = [Vg2 x 0.63]/"μg2g1" .....reference, RCA tube manuals...
mu is plate voltage related, while gm is cathode current related, a look at datasheets will show...
voltage gains = gm x Rl,
gm is cathode current related, so the higher the cathode current the higher the gm,
with tubes, increasing the plate supply B+ increased cathode current and therefore increased the voltage gain.....
plate voltage otoh tells you how much plate swing is available to the next stage...
it is really fun ad exciting to play with tubes...
Well - there's a quite better tutorial by Tony 🙂 🙂 🙂
My preference for pentodes (earlier question) is rather practice based, not too serious, but on balance ... (I was so tempted to say 'because they sound better' but will not tempt the wrath of a certain portion of the membership . . .)
Firstly at low signal output they have some 25% of the distortion of triodes (yes! I know it is a comparatively small contribution to the whole amplifier!) Then perhaps more serious, they have quite smaller Miller contribution (if unknown you will have to look that up). As input stage that may be important as one does not know what impedance the input signal is fed from. That, giving an unknown pole and thus phase angle at h.f., can have an effect on final NFB stability. Then I find that a pentode anode voltage is less critical that that of a triode, which makes it rather easier to set anode voltage for direct coupling to a next stage phase splitter. Finally, the screen decoupling capacitor is part of an extra shelving element at l.f., again handy for controlling l.f. NFB stability (phase angle).
Disadvantage: Because of the presence of a screen grid the noise (partition noise) is higher than for a triode, all else being equal. But I have never found that to reach troublesome levels when using a pentode as input tube for a tube power amplifier.
As said on the whole not serious factors, just all rather handy where high NFB levels are used. (Perhaps also the practical factor of getting a high amplification as early as possible in the signal chain for an overall low signal/noise factor.) Not to belabour, but that again leaves the option to use a medium-µ triode phase splitter, again Miller effect etc. etc.
(As you will gather, not unexpectedly, the design of a power amplifier rests on careful consideration/balance of a number of factors.)
My preference for pentodes (earlier question) is rather practice based, not too serious, but on balance ... (I was so tempted to say 'because they sound better' but will not tempt the wrath of a certain portion of the membership . . .)
Firstly at low signal output they have some 25% of the distortion of triodes (yes! I know it is a comparatively small contribution to the whole amplifier!) Then perhaps more serious, they have quite smaller Miller contribution (if unknown you will have to look that up). As input stage that may be important as one does not know what impedance the input signal is fed from. That, giving an unknown pole and thus phase angle at h.f., can have an effect on final NFB stability. Then I find that a pentode anode voltage is less critical that that of a triode, which makes it rather easier to set anode voltage for direct coupling to a next stage phase splitter. Finally, the screen decoupling capacitor is part of an extra shelving element at l.f., again handy for controlling l.f. NFB stability (phase angle).
Disadvantage: Because of the presence of a screen grid the noise (partition noise) is higher than for a triode, all else being equal. But I have never found that to reach troublesome levels when using a pentode as input tube for a tube power amplifier.
As said on the whole not serious factors, just all rather handy where high NFB levels are used. (Perhaps also the practical factor of getting a high amplification as early as possible in the signal chain for an overall low signal/noise factor.) Not to belabour, but that again leaves the option to use a medium-µ triode phase splitter, again Miller effect etc. etc.
(As you will gather, not unexpectedly, the design of a power amplifier rests on careful consideration/balance of a number of factors.)
Last edited:
Folks interested in small-signal pentodes would probably enjoy the Frank Blohbaum (mit der Umlaut) article in the Volume 0 _Linear Audio_. His improvement is to cascode the G2 and to return its current into the anode stream instead of simply losing it to signal ground. He used a conventional bipolar transistor for good transconductance, but it could be done with a small high-Gm vacuum triode.
All good fortune,
Chris
All good fortune,
Chris
low open loop gains, lower feedback factors and so low closed loop gains, just as with ss amps, makes for a good sounding amp....W.M. Leach
Thank you all for your responses! You all have been quite generous with your time and knowledge.
And thank you PRR for earlier clarifying a point in the datasheets that had been confusing me.
Tony, that was a very helpful explanation.
Johan, interesting points about pentodes. From casual reading I had the Triode/Pentode Miller capacitance issue reversed.
Casual reading is perhaps the cause of most of my confusion. There is quite a bit of information out there (internet) that is either wrong or woefully incomplete. I will need to get a real book by a person of real knowledge and a tube manual if I am going to proceed. I really don't want to be guilty of inadvertently contributing further to the mass of misinformation. (which is exactly what will happen if I keep posting blindly!)
And thank you PRR for earlier clarifying a point in the datasheets that had been confusing me.
Tony, that was a very helpful explanation.
Johan, interesting points about pentodes. From casual reading I had the Triode/Pentode Miller capacitance issue reversed.
Casual reading is perhaps the cause of most of my confusion. There is quite a bit of information out there (internet) that is either wrong or woefully incomplete. I will need to get a real book by a person of real knowledge and a tube manual if I am going to proceed. I really don't want to be guilty of inadvertently contributing further to the mass of misinformation. (which is exactly what will happen if I keep posting blindly!)
Last edited:
dc biasing tubes is the easy part, just do not forget plate dissipation/voltage limits, cathode current limits and you should be fine...
add to that also the heater to cathode voltage specs, they were put in datasheets for a reason...i bet that from now on you will look at data sheets from a new perspective....
small signal voltage amplification for tubes, used just a fraction of these limits, an inspection of several existing circuits reveals this.....the curves gives you choices...
add to that also the heater to cathode voltage specs, they were put in datasheets for a reason...i bet that from now on you will look at data sheets from a new perspective....
small signal voltage amplification for tubes, used just a fraction of these limits, an inspection of several existing circuits reveals this.....the curves gives you choices...
Careful here. Force! It applies both ways! In fact, what you might have read was a pentode with its comparative high Rp, being a poor source for the "Millered" load placed onto it by a following triode.
But when an input triode is used, I am also concerned by that triodes' Miller effect on the driving source's input impedance. If low most likely fine, but some feeding pre-amps may have some pretty high output impedances reacting with the first stage Miller influence. As the first stage is already in the NFB loop, that Miller capacitance might well influence h.f. NFB stability.
This does not occur commonly but must be kept in mind as a possible feedback phase angle contributor. Should not be difficult to deal with, only be aware of its potential effect. (Folks often lose sight that it is also an element in the NFB topology.0
Thanks again, Tony and Johan, for your guidance!
When I return from traveling, I will try a low or medium μ tube in the first position and see if this gets me where I want to be. If not, I can easily reduce the gain of my preamp, or perhaps run the EL34 as triodes (I could even reconnect the NFB). Lots of options.
At least the amps work correctly and sound very good, so I have a firm foundation to return to if the experiments fail!
When I return from traveling, I will try a low or medium μ tube in the first position and see if this gets me where I want to be. If not, I can easily reduce the gain of my preamp, or perhaps run the EL34 as triodes (I could even reconnect the NFB). Lots of options.
At least the amps work correctly and sound very good, so I have a firm foundation to return to if the experiments fail!
i like diy'ers who go out of their way to learn the basics of how circuits work, and not imbibe snake oils as is the case with many others...i tend to stay away from them...
Is this as simple as inserting the new tube or does anything need to be changed? 50's-era power amp input sensitivities are somewhat inconvenient these days and the noise performance increase would also be good... 🙂
Yes, exactly. 
I’m looking for more info on reducing gain by using a lower-mu tube at the phase splitter.
I’m looking for more info on reducing gain by using a lower-mu tube at the phase splitter.
i built a pp JJ6L6GC pp amp, i used the 6sn7 at first stage voltage amp and the 6SL7 ltp phase splitter, awesome combination...in a mullard 5-20 topology of course..
Utterly important!
I have been in this profession for some almost 70 years (ouch!) and how often have I encountered folks knowing a lot more than I will ever know, go awry in their designs (plus fancy tricks) because they lost sight of some of the basics!
I am not a stick-in-the-mud conservative (in my research work one often had to be the opposite!), but it is always better never to lose sight of basics - in every walk of life! (End of sermon)
[OT but how am I sorely tempted to quote the unique Mark Twain (from memory):
First get your facts right - then you can twist them as you please.
Last edited:
807’s?
Did you ever use 807’s for your push pull? I have a matching quad of 80’7’s that I’d like to use already. I have a Hammond 400-0-400@465ma/6.3v@6amps and a separate 5v@6amps(dual 5U4G’s maybe;-)) is like to keep amp simple as possible even though I’d be building stereo version. I don’t have OPT’s yet, I’m reading 10k,6.6k and I have that ampeg schematic with the 4k
Did you ever use 807’s for your push pull? I have a matching quad of 80’7’s that I’d like to use already. I have a Hammond 400-0-400@465ma/6.3v@6amps and a separate 5v@6amps(dual 5U4G’s maybe;-)) is like to keep amp simple as possible even though I’d be building stereo version. I don’t have OPT’s yet, I’m reading 10k,6.6k and I have that ampeg schematic with the 4k
Sorry to resurrect but this thread seems very relevant.
I am considering building the CB "Mullard 4-30" and also a bit concerned with the gain. Can a 5751 be used as a drop-in to the 12AX7 here?
Besides that, were there any conclusions to the discussions here with regards to using an alternative to EF86 or changing the operating points for use with an ECC81 or 12BH7A ?
I am considering building the CB "Mullard 4-30" and also a bit concerned with the gain. Can a 5751 be used as a drop-in to the 12AX7 here?
Besides that, were there any conclusions to the discussions here with regards to using an alternative to EF86 or changing the operating points for use with an ECC81 or 12BH7A ?
Hi,
I did end up building the 4-30s. Excellent amps.
I built them as designed, at first, and then tried changing a number of things.
I pretty much ended up right back with the Claus Byrith 4-30, at the end of a long road.
So, my recommendation would be to build it to the schematic first.
The EF86 in triode is excellent--supremely linear. I would say better than ECC81 or 12BH7 as a Voltage Amp. If you can find some vintage Mullards, they are quite nice.
Also, since the EF86 is a single pentode and the ECC81 and 12BH7A are dual triodes, they would have an extra triode flapping in the wind, unless you built a stereo amp (two channels, one chassis).
Strangely, the 12AX7 does a great job as Phase Splitter in this arrangement. I have yet to find better; but I also have a few nice vintage bottles to choose from, so...
Yes, the 5751 will do fine as the Phase Splitter, and will drop right in.
Regards,
John
I did end up building the 4-30s. Excellent amps.
I built them as designed, at first, and then tried changing a number of things.
I pretty much ended up right back with the Claus Byrith 4-30, at the end of a long road.
So, my recommendation would be to build it to the schematic first.
The EF86 in triode is excellent--supremely linear. I would say better than ECC81 or 12BH7 as a Voltage Amp. If you can find some vintage Mullards, they are quite nice.
Also, since the EF86 is a single pentode and the ECC81 and 12BH7A are dual triodes, they would have an extra triode flapping in the wind, unless you built a stereo amp (two channels, one chassis).
Strangely, the 12AX7 does a great job as Phase Splitter in this arrangement. I have yet to find better; but I also have a few nice vintage bottles to choose from, so...
Yes, the 5751 will do fine as the Phase Splitter, and will drop right in.
Regards,
John
Thanks John, for the information.
For clarification, I meant using the 12BH7A or 12AT7 in the PI position instead of the 12AX7, but I think you answered that regardless.
I actually have alot of NOS 12AX7's, RCA, Telefunken, Mullard ... so that works out.
This this looks like a project I'll be pursuing. I prefer PCB's and will try to recreated CB's original ones.
Did you build as monos or stereo? The recommended Lundahl PT seems like it could easily support a stereo build.
For clarification, I meant using the 12BH7A or 12AT7 in the PI position instead of the 12AX7, but I think you answered that regardless.
I actually have alot of NOS 12AX7's, RCA, Telefunken, Mullard ... so that works out.
This this looks like a project I'll be pursuing. I prefer PCB's and will try to recreated CB's original ones.
Did you build as monos or stereo? The recommended Lundahl PT seems like it could easily support a stereo build.
Last edited:
I built mine as a pair of mono amps.
Apologies in advance if I am missing the mark here but I, personally, would encourage you to consider a point-to-point build. First of all because it is more fun, and this circuit is quite simple. It can be built with just 2 or 3 terminal-strips per channel.
Also, since tubes get hot, PCBs can run into trouble with broken traces (expanding and contracting), which can be very hard to pin-point.
But etching PCBs is fun, too.
Best regards,
John
Apologies in advance if I am missing the mark here but I, personally, would encourage you to consider a point-to-point build. First of all because it is more fun, and this circuit is quite simple. It can be built with just 2 or 3 terminal-strips per channel.
Also, since tubes get hot, PCBs can run into trouble with broken traces (expanding and contracting), which can be very hard to pin-point.
But etching PCBs is fun, too.
Best regards,
John