I've been working my way through Morgan Jones' book for quite some time and have a fairly clear grasp of everything up to the point of the Bevois Valley amp design.
I follow the design of the Concertina phase splitter right up to the section entitled 'Calculation of the Cathode Bias Resistor and Feedback resistor'
At what point is the 47k Rl for the first anode chosen? I don't see any mention of it before this point so I'm assuming it's something simple that I've missed.
In the same section: where is the 190V derived from to find the anode current?
Any help would be much appreciated.
TIA
cgrums
I follow the design of the Concertina phase splitter right up to the section entitled 'Calculation of the Cathode Bias Resistor and Feedback resistor'
At what point is the 47k Rl for the first anode chosen? I don't see any mention of it before this point so I'm assuming it's something simple that I've missed.
In the same section: where is the 190V derived from to find the anode current?
Any help would be much appreciated.
TIA
cgrums
To answer your second question first, the 190V comes from observing that the ECC88 is very linear at around 90V. 285V - 90V = 195V.
As to the first question, 47k gives a nice loadline, but almost anything would do.
As to the first question, 47k gives a nice loadline, but almost anything would do.
That helped quite a bit, thanks for the info.
In the application of the feedback is the 'unmodified sensitivity' (~ 238 mVRMS - don't have the book right in front of me) referring to the sensitivity of the Concertina stage (without feedback)? If so would that 239 mVRMS be the input voltage required for an anode voltage swing of 8.6xx VRMS (required for full power output)? This seems low...am I way off?
In the application of the feedback is the 'unmodified sensitivity' (~ 238 mVRMS - don't have the book right in front of me) referring to the sensitivity of the Concertina stage (without feedback)? If so would that 239 mVRMS be the input voltage required for an anode voltage swing of 8.6xx VRMS (required for full power output)? This seems low...am I way off?
It's a combination of the gain of the input tube with the concertina. The latter is slightly less than unity.
The sensitivity numbers are as you suspect. As a check, 8.6V/0.239V = 35, about the mu of the first tube.
The sensitivity numbers are as you suspect. As a check, 8.6V/0.239V = 35, about the mu of the first tube.
Thanks SY...I'm getting there slowly.
Please forgive this next as it's probably very evident...the Concertina is composed only of the second of the two tubes shown? I misread and thought that as they were drawn as one device that both tubes comprised the Concertina. So the device shown is actually an input stage and a phase-splitter in one envelope? Is the input stage analyzed as a common cathode stage?
Please forgive this next as it's probably very evident...the Concertina is composed only of the second of the two tubes shown? I misread and thought that as they were drawn as one device that both tubes comprised the Concertina. So the device shown is actually an input stage and a phase-splitter in one envelope? Is the input stage analyzed as a common cathode stage?
Ha! Fantastic! Coming from a mechanical engineering background this is a whole different way of thinking but I think I'm slowly making sense of it all.
My current plan is to use this circuit for the basis of a birthday present for my brother. He currently has a pair of JBL 2600 bookshelfs (89 db) and we'll be looking in to making him a more efficient pair in the future.
As of right now I'm not planning a pre-amp and was going to use a rotary source selector switch as well as a pot for volume attenuation. What considerations need to be taken into account in the selection of the resistance of the pot? I know it's related to the input impedance of the input stage but am not quite sure how to put that relationship in numbers.
As of right now I'm not planning a pre-amp and was going to use a rotary source selector switch as well as a pot for volume attenuation. What considerations need to be taken into account in the selection of the resistance of the pot? I know it's related to the input impedance of the input stage but am not quite sure how to put that relationship in numbers.
100K is a value high enough that most any source can drive it.. and it's about as high as you can go without running into high-frequency rolloff. The input stage has a fairly high effective capacitance due to Miller effect, and the series portion of the pot forms an R-C network with that.
Ahh...I see. So is the R-C network formed with only the pot resistance and the miller capacitance or does the input resistance of the input stage add to that as well?
The input resistance of the input stage is so high you can ignore it for all practical purposes.
Instead of starting a new thread I'll just post this here: I've been up to my eyeballs in reading about the link between an amplifiers power rating and the efficiency of a given speaker. Having never heard/used a tube power amp before (all of my experience is with headphones) and seeing as I'm building this as a gift, is 10W PP going to do a pair of 86 dB bookshelf speakers any good? Obviously I can make a gift of a pair of more efficient speakers but I'd like the recipient to be able to use the amp with their existing setup.
Depends on whether the recipient is married. If he is, he won't be allowed to turn it up loud, so 10W will be fine. If not, you'd better hope he prefers Bach to Led Zeppelin.
^ Sounds about right.
Although I'm running 87dB/watt speakers from a 15 watt SS amp (well past it's best, too), and it goes loud enough for me. Although I don't try to get realistic concert levels.
James
P.s. Nice sig, EC8010
Although I'm running 87dB/watt speakers from a 15 watt SS amp (well past it's best, too), and it goes loud enough for me. Although I don't try to get realistic concert levels.
James
P.s. Nice sig, EC8010
It's been a bit of time since I last posted on this, I've only had tiny chunks of time here and there to work on this. I've got my chassis machined, tubes in and have begun the heater wiring.
I'm using PSUD2 to model the power supply as I've read that the 285V B+ doesn't need to be regulated in this design. I've got one RC filter of 2k and 47 uF after the 220 uF cap (that connects to the OPT).
My latest question is this: can anybody help me through the process of calculating the current draw from the 320V connection (OPT)? I'm not finding anything in the book showing this derivation directly.
Any guidance would be much appreciated.
I'm using PSUD2 to model the power supply as I've read that the 285V B+ doesn't need to be regulated in this design. I've got one RC filter of 2k and 47 uF after the 220 uF cap (that connects to the OPT).
My latest question is this: can anybody help me through the process of calculating the current draw from the 320V connection (OPT)? I'm not finding anything in the book showing this derivation directly.
Any guidance would be much appreciated.
Add together all of the currents. I know that the input stage and phase splitter currents are given, and I'm fairly sure the output stage idle current is, also. If not, it should be pretty easy to find it in the Mullard data sheets.
So According to the Mullard datasheet the EL84's run ~ 80 mA idle current per channel. I've used 160 mA for the draw of both channels as well as a second current tap of 34 mA for the input/phase splitter in my PSUD2 sim. Everything looks good so far but I'm not clear on one thing: how am I able to see ripple voltage in PSUD2? It doesn't seem to be listed as a traceable output yet I've seen numerous posts where people are able to read the ripple voltage...what am I missing? TIA!
- Status
- Not open for further replies.