I am using a M-Audio Fast Track Pro as source (connected to my Macbook) and that voltage is the input signal in my tube amp. I tested the gain on one of the grills of el84 and it came out almost 23. I used a 100hz sine wave mp3 and measured input at 80mV and the signal on grid was 1.80V. The soundcard is also connected to that SS amp (this is the actual SS amp+speakers Technics sc-dv280 http://exaudio7.netserwer.pl/allegro/technics/_wie%BFe/sc-dv280/sc-dv280_3.jpg ) by another output.
P.S. i tested them in paralel with same amount of input signal, at one point "60-70% of the volume knob from the soundcard" the tube amp starts distorting heavy but SS amp goes way way up and crisp clear with deep bass.
P.S. i tested them in paralel with same amount of input signal, at one point "60-70% of the volume knob from the soundcard" the tube amp starts distorting heavy but SS amp goes way way up and crisp clear with deep bass.
Last edited:
From post 23:
"Your amp will distort when the peak signal swing from the splitter is the same as the bias voltage. For example, with output tubes cathode biased at 11V, full power is developed when the splitter puts out 11V peak, or 22V peak to peak or 7.8Vrms. So you can see it easy to overdrive the amp with just one gain stage."
With 0.4V rms from your sound card and 22.5 gain, you get 9V rms at the grids of the EL84. I'd guess it's clipping (which is what you have said).
With your current speaker setup, what you could try is use your EL84 amp left and right channels to drive the mid+high of each speaker, and let the SS amp drive the bass driver of each speaker (or vice-versa). At the moment it sounds like you are driving an 8 ohm bass driver in parallel with 6 ohm mid+high with an OPT 6600:8R, which is doing things the hard way
"Your amp will distort when the peak signal swing from the splitter is the same as the bias voltage. For example, with output tubes cathode biased at 11V, full power is developed when the splitter puts out 11V peak, or 22V peak to peak or 7.8Vrms. So you can see it easy to overdrive the amp with just one gain stage."
With 0.4V rms from your sound card and 22.5 gain, you get 9V rms at the grids of the EL84. I'd guess it's clipping (which is what you have said).
With your current speaker setup, what you could try is use your EL84 amp left and right channels to drive the mid+high of each speaker, and let the SS amp drive the bass driver of each speaker (or vice-versa). At the moment it sounds like you are driving an 8 ohm bass driver in parallel with 6 ohm mid+high with an OPT 6600:8R, which is doing things the hard way
Ok, I did the preamp with common triode with local negative feedback. I attached the current scheme. After the mod it sounds terrific. The bass is so deep and everything vibrates at lower frequencies. The overall loudness is much less at maximum level, but I don't get any distorsion. At least at first tests. Please tell me if I did the feedback correct.
Zin=137k
Zout=9.2k
ra=13k
Open loop gain about -25.85
Zoa=10k
The calculated total gain of the stage is about -7
*later edit
I measured currents and voltages and it is kinda nuts:
A=324V
B=321V
C=311V
el84 - anode current 37mA for one tube and 39mA for the other. anode voltage is 311V for the 37mA one and cathode voltage is 11.3V.
preamp 6n1p-ev - anode current 5mA anode voltage 83V and cathode voltage is 3.75V
PI 6n1p-ev - anode current - 0.6mA with 281V on anode and 28V on cathode. neither is what I initially calculated. it sounds good, but I get some crackling here and there. is this ok with the applied feedback?
Zin=137k
Zout=9.2k
ra=13k
Open loop gain about -25.85
Zoa=10k
The calculated total gain of the stage is about -7
*later edit
I measured currents and voltages and it is kinda nuts:
A=324V
B=321V
C=311V
el84 - anode current 37mA for one tube and 39mA for the other. anode voltage is 311V for the 37mA one and cathode voltage is 11.3V.
preamp 6n1p-ev - anode current 5mA anode voltage 83V and cathode voltage is 3.75V
PI 6n1p-ev - anode current - 0.6mA with 281V on anode and 28V on cathode. neither is what I initially calculated. it sounds good, but I get some crackling here and there. is this ok with the applied feedback?
Attachments
Last edited:
As I was scanning thru "Valve Amplifiers" by Morgan Jones I discovered what the heck I was trying to do...it was the Bevois Valley amp 
I mean, it fits perfect for my actual scheme. Got a few problems with it thou. I am asking you to adjust or tell me if it's ok using in the same setup my actual output tranny witch is about 5.3k impedance with 8ohm for speaker. The one in the Bevois amp is a 8.8k I think. then...I don't have a choke, can I just use a 100ohm resistor instead? My power supply is 240V VAC at 400mA but I'll use a bridge rectifier not a tube so I get about 330V as well. That "285V regulator" ... I don't know what that is, but I can drop some volts with a resistor if need be for 285V. Then...my EL84's are connected to tranny by only anodes. The G2 is wired directly into the PS. Not in the transformer. Because I only have one center tap and 2 leads for primary. How do I adjust for the Bevois scheme?
And the final question would be if I can use 6n1p-ev instead of e88cc. As I saw on some page this info:
"The 6N1P has not similar with E88CC / 6DJ8 (the 6N1P require double the filament current and has Stepness in 3.8-5.1mA/V) mean they are not directly interchangeable. The S is about 4.35 ma/V , the 6DJ8/ECC88 has a S of 12.5 ma/V and a gain of 33 and a lower internal resistance. The curves of the E88CC / 6DJ8 have a better linearity too. So, we can say that these two tubes are not identical!"
So, if only different Stepness then I could redo some of the math involved to adjust.
Now to get all the correct capacitors and resistors and get to work.
I have attached the scheme of the Bevois Valley amp for reference
I mean, it fits perfect for my actual scheme. Got a few problems with it thou. I am asking you to adjust or tell me if it's ok using in the same setup my actual output tranny witch is about 5.3k impedance with 8ohm for speaker. The one in the Bevois amp is a 8.8k I think. then...I don't have a choke, can I just use a 100ohm resistor instead? My power supply is 240V VAC at 400mA but I'll use a bridge rectifier not a tube so I get about 330V as well. That "285V regulator" ... I don't know what that is, but I can drop some volts with a resistor if need be for 285V. Then...my EL84's are connected to tranny by only anodes. The G2 is wired directly into the PS. Not in the transformer. Because I only have one center tap and 2 leads for primary. How do I adjust for the Bevois scheme?
And the final question would be if I can use 6n1p-ev instead of e88cc. As I saw on some page this info:
"The 6N1P has not similar with E88CC / 6DJ8 (the 6N1P require double the filament current and has Stepness in 3.8-5.1mA/V) mean they are not directly interchangeable. The S is about 4.35 ma/V , the 6DJ8/ECC88 has a S of 12.5 ma/V and a gain of 33 and a lower internal resistance. The curves of the E88CC / 6DJ8 have a better linearity too. So, we can say that these two tubes are not identical!"
So, if only different Stepness then I could redo some of the math involved to adjust.
Now to get all the correct capacitors and resistors and get to work.
I have attached the scheme of the Bevois Valley amp for reference
I removed the copyrighted image you attached. You need to get permission before scanning and posting schematics out of contemporary books.
The Bevois (and similar circuits) represents a classic way to drive EL84; the Bevois has the advantage of being explained in great detail. The 285V regulator can be either the Maida or the two-transistor regulators that Morgan describes in the power supply chapter, along with the design equations.
Your output transformer is not optimal as far as impedance, but it can work, especially if your speakers have impedance higher than 8 ohms. Does it have ultralinear taps on the primary? If so, you're in good shape. If not, you've got some redesign to do to make everything work with a pentode output stage. If you can get by with less power (5-6W), you could also configure as triode, but there, the non-optimal impedance of your output transformer will translate to higher distortion. Still, it would be a pretty sweet little amp.
If you're going to build an ECC88 input stage, use an ECC88. They're neither rare nor expensive.
The Bevois (and similar circuits) represents a classic way to drive EL84; the Bevois has the advantage of being explained in great detail. The 285V regulator can be either the Maida or the two-transistor regulators that Morgan describes in the power supply chapter, along with the design equations.
Your output transformer is not optimal as far as impedance, but it can work, especially if your speakers have impedance higher than 8 ohms. Does it have ultralinear taps on the primary? If so, you're in good shape. If not, you've got some redesign to do to make everything work with a pentode output stage. If you can get by with less power (5-6W), you could also configure as triode, but there, the non-optimal impedance of your output transformer will translate to higher distortion. Still, it would be a pretty sweet little amp.
If you're going to build an ECC88 input stage, use an ECC88. They're neither rare nor expensive.
ups, that was stupid of me. I am sorry, I haven't thought of that. I'll pay more attention on that from now on.
I don't have ultralinear taps on my output transformer. It's custom made by a friend, so it's usual, normal 2 wire with center tap on primary (but I heard last night the deepest bass from it). If the redesign is too hard then I'll stick with what I already got but I don't know if it's ok with the measurements I get after I apply negative feedback on preamp. Seems weird, pre-amp triode gets double current and PI triode gets less than half. Is that...normal?
I don't have ultralinear taps on my output transformer. It's custom made by a friend, so it's usual, normal 2 wire with center tap on primary (but I heard last night the deepest bass from it). If the redesign is too hard then I'll stick with what I already got but I don't know if it's ok with the measurements I get after I apply negative feedback on preamp. Seems weird, pre-amp triode gets double current and PI triode gets less than half. Is that...normal?
The reason the current increased on the first stage after the FB was applied is because the cathode voltage of the PI is now flooding back to the grid of the first stage, up setting it's bias. Add a .1 mfd cap before the 1 meg FB resistor (i.e. in series with it) to prevent the dc from flooding back, and then the current draw for the first stage will then return to normal.
Dave
Dave
The 100 mfd cathode bypass caps you show on your schematic are sufficient, although many suggest that a larger value adds more definition to bass frequencies.
The 68 pf caps are specific for the output transformers that Jones used in his builds. Their value also depends on how much NFB is applied. I would build your amplifier without them, but realize that installing caps of appropriate values in those locations is a future possibility to fine tune your amplifier after it is built. To determine proper values (if needed) typically requires a 10 kHz square wave source, dummy load, and scope to assess the amplifier's HF transient stability. It also requires a decent understanding of the waveforms generated, and how different approaches to HF stabilization can best be applied.
Adding the FB coupling cap will likely eliminate the strange behavior at turn off, as the first stage will no longer be influenced by the dc voltages of the PI stage.
Dave
The 68 pf caps are specific for the output transformers that Jones used in his builds. Their value also depends on how much NFB is applied. I would build your amplifier without them, but realize that installing caps of appropriate values in those locations is a future possibility to fine tune your amplifier after it is built. To determine proper values (if needed) typically requires a 10 kHz square wave source, dummy load, and scope to assess the amplifier's HF transient stability. It also requires a decent understanding of the waveforms generated, and how different approaches to HF stabilization can best be applied.
Adding the FB coupling cap will likely eliminate the strange behavior at turn off, as the first stage will no longer be influenced by the dc voltages of the PI stage.
Dave
Thank you very much. I will further read that book so I get a better understanding of the different stages, but for now I will have to properly do all the math involved in my amp. This is a great experience for me and I learnt a lot! Thinking where I began and where I am now makes me feel very good
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.