Trasformatore & Capasitores
Yes! the master of high end audio says so, and I agree with him but... those old timers are nice devices and can be used when you will make decent nice sounding audio aparatus for the people sorry to say that those old timers are very expensive today.
I remember from the past that one of the best sounding "hifi" was a product from Danish Arena, a neat design a hifi stereo receiver equipped with two single ended transistor amplifiers with transformers as output devices.
A minimum part design, I like it very much as it is friendly to our environment with components that can be recycled with high profitability.
So... I am, as I told earlier, ambivalent and bit confused about this person from UK who urgent and uneffected present this old basic idea of design.
Kamskoma
Yes! the master of high end audio says so, and I agree with him but... those old timers are nice devices and can be used when you will make decent nice sounding audio aparatus for the people sorry to say that those old timers are very expensive today.
I remember from the past that one of the best sounding "hifi" was a product from Danish Arena, a neat design a hifi stereo receiver equipped with two single ended transistor amplifiers with transformers as output devices.
A minimum part design, I like it very much as it is friendly to our environment with components that can be recycled with high profitability.
So... I am, as I told earlier, ambivalent and bit confused about this person from UK who urgent and uneffected present this old basic idea of design.
Kamskoma
Susan-Parker said:
Driven by a 2x250W amp, these sound just wonderful. With a 2x40W pentode amp, the amp just can't keep up at realistic volumes, however.
You might want to develop a hybrid amplifier. Dynamic Precision A1 uses a class AB amplifier to drive the ground of a class A amplifier. If I understood the operation of your amplifier correctly, you could probably use a similar approach here if more power is required.
The A1 spec is 2x800W into 2 ohms / 2x250W into 8 ohms, a slew rate of 250V/us, damping factor of 1000 from DC to 20KHz, 105dB SNR, 100dB+ CMRR and 120dB+ channel separation. At 2x400W into 4 ohms, it has 0.001% THD and 0.0003% IMD.
Have you by any chance tried implementing this amplifier with high current triodes? It would be interesting to compare the performance of the FET version with one based on e.g. the 6C33C-B.
Re: Zeus Portable Headphone Amp
Hey, you're doing this on your spare time, same as me. You don't owe me anything. 😀
Looks very promising.
Currently, I'm looking to drive some Grados and a pair of 501. The former is 32 ohms and gives me 98dB/1mV, so the 2SK389 should do nicely there (4mV exceeds cinema levels). The latter is 120 ohms and I figure about 3V to reach cinema level, so the MGF1302 is more useful there.
Lundahl has some nice transformers both for input and output.
The main advantage of using the 2SK389 is that it is inherently closely matched (max 20mV difference in Vgs), low noise and low gate charge. Also, it can be operated at low voltage with low-ish current.
The MGF1302 is more beefy, and even lower noise and gate charge, but its ideal bias point is with 60ohm Zdc primaries and two dry cells in series.
By using either of these two, you can pare things down to an input transformer, an output transformer, one transistor pair and two gate stopper resistors per channel. This since zero bias on the gate gives a "perfect" bias current given a somewhat careful choice of primary Zdc.
Susan-Parker said:
Hey, you're doing this on your spare time, same as me. You don't owe me anything. 😀
Looks very promising.
Currently, I'm looking to drive some Grados and a pair of 501. The former is 32 ohms and gives me 98dB/1mV, so the 2SK389 should do nicely there (4mV exceeds cinema levels). The latter is 120 ohms and I figure about 3V to reach cinema level, so the MGF1302 is more useful there.
Lundahl has some nice transformers both for input and output.
The main advantage of using the 2SK389 is that it is inherently closely matched (max 20mV difference in Vgs), low noise and low gate charge. Also, it can be operated at low voltage with low-ish current.
The MGF1302 is more beefy, and even lower noise and gate charge, but its ideal bias point is with 60ohm Zdc primaries and two dry cells in series.
By using either of these two, you can pare things down to an input transformer, an output transformer, one transistor pair and two gate stopper resistors per channel. This since zero bias on the gate gives a "perfect" bias current given a somewhat careful choice of primary Zdc.
Zeus and magneplanars
Hi Angel,
Zeus in 2:1 mode with a 48 volt power supply can drive 115 watts into 4 ohms with a max 1.0% THD at around 35 watts (for higher powers the THD reduces - see table below), and 85 watts into 8 ohms.
Unfortunately this complicates the circuit and moves away from something that anybody with basic skills can have a go at building.
Also the point of the amplifier is not to have any global feedback loops.
Others have built serious power output versions of my amplifier, getting 500 or so watts. For example see earlier posts on this thread by Ampman.
Personally I am really only interested in powers of the order of 25 to 50 watts max, so I haven't pursued the higher powers.
For comparison purposes one needs to look at valve amplifiers, not solid state. At low levels (with 4:1 into 8 ohms) I can get below 0.01% THD without any NFB loops (not counting the followers).
N.B. The zeus amp has over 150 dB SNR.
I have looked at a 6C33C circuit (see earlier posts on this thread) but don't have the personal resources to invest in that at the moment.
I believe that the performance and sound quality could be very similar to the mosfet version, with the exception of a higher noise floor. Power output would be quite limited, probably about 8 watts or so.
It would be quite possible to prototype a 6C33C amp using standard mains power toroid transformers. And to be honest even I would concede that the nice warm glow of a pair of 6C33Cs would be more attractive to look at than a black anodised heatsink.
Best wishes,
Susan.
Table:
Zeus 75, 2:1 Transformer Configuration.
Vsupply 48 volts (off load).
Vbias 1.6 amps (800 mA per mosfet).
Load 4 ohms.
Vac THD% Watts PSU-Amps
21.43 0.8750 114.8 4.89
20.14 0.8200 101.4 4.58
18.99 0.8300 90.2 4.33
18.12 0.8670 82.1 4.16
16.97 0.9030 72.0 3.91
16.05 0.9310 64.4 3.71
15.18 0.9560 57.6 3.53
14.34 0.9770 51.4 3.36
13.57 0.9920 46.0 3.20
12.82 0.9990 41.1 3.05
12.14 1.0040 36.8 2.91
11.47 1.0040 32.9 2.78
10.85 0.9970 29.4 2.66
10.27 0.9850 26.4 2.55
9.71 0.9630 23.6 2.44
9.19 0.9270 21.1 2.33
8.69 0.8820 18.9 2.24
8.22 0.8280 16.9 2.16
7.78 0.7660 15.1 2.08
7.35 0.6950 13.5 2.00
6.96 0.6210 12.1 1.94
6.59 0.5430 10.9 1.88
6.23 0.4680 9.70 1.83
5.90 0.3980 8.70 1.79
5.57 0.3380 7.76 1.75
5.27 0.2865 6.94 1.72
4.99 0.2438 6.23 1.70
4.71 0.2071 5.55 1.68
4.45 0.1770 4.95 1.66
4.20 0.1523 4.41 1.65
3.97 0.1415 3.94 1.64
3.76 0.1276 3.53 1.65
3.55 0.1109 3.15 1.64
3.36 0.0970 2.82 1.65
3.17 0.0845 2.52 1.64
2.99 0.0738 2.24 1.65
2.83 0.0655 2.00 1.64
2.67 0.0577 1.78 1.65
2.52 0.0516 1.59 1.64
2.38 0.0460 1.41 1.65
2.25 0.0413 1.26 1.64
2.12 0.0371 1.13 1.65
2.00 0.0337 1.00 1.64
1.89 0.0305 0.89 1.65
1.78 0.0278 0.80 1.64
1.68 0.0255 0.71 1.65
1.59 0.0239 0.63 1.64
1.50 0.0223 0.56 1.65
1.42 0.0210 0.50 1.64
1.34 0.0199 0.45 1.65
1.26 0.0193 0.40 1.64
1.19 0.0185 0.36 1.65
1.13 0.0184 0.32 1.64
1.06 0.0183 0.28 1.65
1.00 0.0184 0.25 1.64
0.95 0.0183 0.22 1.65
0.90 0.0188 0.20 1.64
0.84 0.0190 0.18 1.64
====
Hi Angel,
angel said:
Zeus in 2:1 mode with a 48 volt power supply can drive 115 watts into 4 ohms with a max 1.0% THD at around 35 watts (for higher powers the THD reduces - see table below), and 85 watts into 8 ohms.
Unfortunately this complicates the circuit and moves away from something that anybody with basic skills can have a go at building.
Also the point of the amplifier is not to have any global feedback loops.
Others have built serious power output versions of my amplifier, getting 500 or so watts. For example see earlier posts on this thread by Ampman.
Personally I am really only interested in powers of the order of 25 to 50 watts max, so I haven't pursued the higher powers.
For comparison purposes one needs to look at valve amplifiers, not solid state. At low levels (with 4:1 into 8 ohms) I can get below 0.01% THD without any NFB loops (not counting the followers).
N.B. The zeus amp has over 150 dB SNR.
I have looked at a 6C33C circuit (see earlier posts on this thread) but don't have the personal resources to invest in that at the moment.
I believe that the performance and sound quality could be very similar to the mosfet version, with the exception of a higher noise floor. Power output would be quite limited, probably about 8 watts or so.
It would be quite possible to prototype a 6C33C amp using standard mains power toroid transformers. And to be honest even I would concede that the nice warm glow of a pair of 6C33Cs would be more attractive to look at than a black anodised heatsink.
Best wishes,
Susan.
Table:
Zeus 75, 2:1 Transformer Configuration.
Vsupply 48 volts (off load).
Vbias 1.6 amps (800 mA per mosfet).
Load 4 ohms.
Vac THD% Watts PSU-Amps
21.43 0.8750 114.8 4.89
20.14 0.8200 101.4 4.58
18.99 0.8300 90.2 4.33
18.12 0.8670 82.1 4.16
16.97 0.9030 72.0 3.91
16.05 0.9310 64.4 3.71
15.18 0.9560 57.6 3.53
14.34 0.9770 51.4 3.36
13.57 0.9920 46.0 3.20
12.82 0.9990 41.1 3.05
12.14 1.0040 36.8 2.91
11.47 1.0040 32.9 2.78
10.85 0.9970 29.4 2.66
10.27 0.9850 26.4 2.55
9.71 0.9630 23.6 2.44
9.19 0.9270 21.1 2.33
8.69 0.8820 18.9 2.24
8.22 0.8280 16.9 2.16
7.78 0.7660 15.1 2.08
7.35 0.6950 13.5 2.00
6.96 0.6210 12.1 1.94
6.59 0.5430 10.9 1.88
6.23 0.4680 9.70 1.83
5.90 0.3980 8.70 1.79
5.57 0.3380 7.76 1.75
5.27 0.2865 6.94 1.72
4.99 0.2438 6.23 1.70
4.71 0.2071 5.55 1.68
4.45 0.1770 4.95 1.66
4.20 0.1523 4.41 1.65
3.97 0.1415 3.94 1.64
3.76 0.1276 3.53 1.65
3.55 0.1109 3.15 1.64
3.36 0.0970 2.82 1.65
3.17 0.0845 2.52 1.64
2.99 0.0738 2.24 1.65
2.83 0.0655 2.00 1.64
2.67 0.0577 1.78 1.65
2.52 0.0516 1.59 1.64
2.38 0.0460 1.41 1.65
2.25 0.0413 1.26 1.64
2.12 0.0371 1.13 1.65
2.00 0.0337 1.00 1.64
1.89 0.0305 0.89 1.65
1.78 0.0278 0.80 1.64
1.68 0.0255 0.71 1.65
1.59 0.0239 0.63 1.64
1.50 0.0223 0.56 1.65
1.42 0.0210 0.50 1.64
1.34 0.0199 0.45 1.65
1.26 0.0193 0.40 1.64
1.19 0.0185 0.36 1.65
1.13 0.0184 0.32 1.64
1.06 0.0183 0.28 1.65
1.00 0.0184 0.25 1.64
0.95 0.0183 0.22 1.65
0.90 0.0188 0.20 1.64
0.84 0.0190 0.18 1.64
====
Hi Susan,
Its unusual that the distortion stops rising at 10V output.
Do you understand why this may be ?
mike...🙂
Its unusual that the distortion stops rising at 10V output.
Do you understand why this may be ?
mike...🙂
Hi Mike,
It's undoubtedly a combination of factors, which I need to work out in more detail. Mosfet load lines figure here.
8 ohm load, but lower 500 mA per mosfet bias for this graph.
Distortion approximately halves going from 4 to 8 ohms.
Best wishes,
Susan.
mikelm said:
It's undoubtedly a combination of factors, which I need to work out in more detail. Mosfet load lines figure here.
8 ohm load, but lower 500 mA per mosfet bias for this graph.
Distortion approximately halves going from 4 to 8 ohms.
Best wishes,
Susan.
Re: Zeus and magneplanars
This should be enough for most people, especially since biamping is becoming rather common. With a reasonably efficient driver, that should achieve cinema levels (105dB peak).
Seems familiar
One might consider a 6N30P spud amp for evaluation. The costs should be lower, and they should allow for 4W or more. Impedance will be about twice as high, though.
Why the 8W limit? These devices are capable of dissipating 60W continous, and a lot more in peaks. Given an appropriate stepdown ratio, one would think they'd deliver quite a decent output.
To me, a major part of the attraction of a depletion mode device (which JFETs, GaAsFETs and valves all qualify as), is that it can simplify this already simple circuit further.
This makes the circuit even easier to build with reliable results. One presumes subjective sound quality might be positively affected as well, given the tendency of our ears to cope better with simple distortion patterns than complex ones.
I intend to test this circuit soon, using 2SK389's driving headphones with Lundahl transformers.
Just need to figure out the best way to realize volume attenuation (logarithmic or equal-loudness) and Ohman crossfeed without compromising the quality of the circuit. Did I say simple? 
Susan-Parker said:
This should be enough for most people, especially since biamping is becoming rather common. With a reasonably efficient driver, that should achieve cinema levels (105dB peak).
Seems familiar
One might consider a 6N30P spud amp for evaluation. The costs should be lower, and they should allow for 4W or more. Impedance will be about twice as high, though.
Why the 8W limit? These devices are capable of dissipating 60W continous, and a lot more in peaks. Given an appropriate stepdown ratio, one would think they'd deliver quite a decent output.
To me, a major part of the attraction of a depletion mode device (which JFETs, GaAsFETs and valves all qualify as), is that it can simplify this already simple circuit further.
This makes the circuit even easier to build with reliable results. One presumes subjective sound quality might be positively affected as well, given the tendency of our ears to cope better with simple distortion patterns than complex ones.
I intend to test this circuit soon, using 2SK389's driving headphones with Lundahl transformers.
Just need to figure out the best way to realize volume attenuation (logarithmic or equal-loudness) and Ohman crossfeed without compromising the quality of the circuit. Did I say simple? 
Hi Susan,
What lovely distortion characteristics you have illustrated.
At the class-A level it could be mostly low harmonic, with higher (worse sounding) harmonics falling in amplitude with increasing order. (Negligible crossover distortion)
At the class-Ab level all harmonics might start increasing as the crossover from dual to single transconduction becomes faster.
At the class-aB level where the low order crossover harmonics might no longer increase linearly with respect to maximum output amplitude, a further increasing output might then sharpen the crossover and lift the higher harmonic content.
I guess the shape of the curve is dependent upon the harmonic weighting within the calculation, and only a deeper harmonic analysis would shed more light, with separate curves for say 2nd, 3rd .. 7th..etc.
__________________________________________________________________
Angel's idea of going 'hollow' is interesting, but whatever thermal efficiency you already have will need to be divided by 3 or 4 due to cathode heating requirements, the necessity for anode voltage headroom and grid voltage variation with cathode current.
Also the input transformer ratio would become disadvantageous, so I could not see this working without separate phase splitter and buffered push-pull voltage amplifier circuitry, or input transformer amplifier stage, neither option being simple, both requiring capacitor coupling and no longer being naturally balanced to maintain symmetrical drive with aged tubes without some form of multistage nfb.
Cheers ............. Graham.
What lovely distortion characteristics you have illustrated.
At the class-A level it could be mostly low harmonic, with higher (worse sounding) harmonics falling in amplitude with increasing order. (Negligible crossover distortion)
At the class-Ab level all harmonics might start increasing as the crossover from dual to single transconduction becomes faster.
At the class-aB level where the low order crossover harmonics might no longer increase linearly with respect to maximum output amplitude, a further increasing output might then sharpen the crossover and lift the higher harmonic content.
I guess the shape of the curve is dependent upon the harmonic weighting within the calculation, and only a deeper harmonic analysis would shed more light, with separate curves for say 2nd, 3rd .. 7th..etc.
__________________________________________________________________
Angel's idea of going 'hollow' is interesting, but whatever thermal efficiency you already have will need to be divided by 3 or 4 due to cathode heating requirements, the necessity for anode voltage headroom and grid voltage variation with cathode current.
Also the input transformer ratio would become disadvantageous, so I could not see this working without separate phase splitter and buffered push-pull voltage amplifier circuitry, or input transformer amplifier stage, neither option being simple, both requiring capacitor coupling and no longer being naturally balanced to maintain symmetrical drive with aged tubes without some form of multistage nfb.
Cheers ............. Graham.
Re: Zeus and magneplanars
Hi Angel
In our small living room this is more than adequate.
These look to be a bit small for a power amp stage?
8 watts is dependent on the voltage swing on the input transformer, as this is where the gain is. 1:20 step up ratio is getting to the limits of a single stage amplification stage.
The 6C33C can drive 600 ohms, and with a low(ish) anode voltage. Thus the output transformer can still have (for tube amplification) a relatively low step down ratio it impedance match to say a nominal 6 ohms load.
I was originally looking at the 6C33Cs for the preamp-line driver.
Sounds good. I haven't used JFETs enough to be able to really comment meaningfully on this, but look forward to hearing your progress.
BW,
Susan.
Hi Angel
angel said:
In our small living room this is more than adequate.
These look to be a bit small for a power amp stage?
8 watts is dependent on the voltage swing on the input transformer, as this is where the gain is. 1:20 step up ratio is getting to the limits of a single stage amplification stage.
The 6C33C can drive 600 ohms, and with a low(ish) anode voltage. Thus the output transformer can still have (for tube amplification) a relatively low step down ratio it impedance match to say a nominal 6 ohms load.
I was originally looking at the 6C33Cs for the preamp-line driver.
To me, a major part of the attraction of a depletion mode device (which JFETs, GaAsFETs and valves all qualify as), is that it can simplify this already simple circuit further.
This makes the circuit even easier to build with reliable results. One presumes subjective sound quality might be positively affected as well, given the tendency of our ears to cope better with simple distortion patterns than complex ones.
I intend to test this circuit soon, using 2SK389's driving headphones with Lundahl transformers.
Sounds good. I haven't used JFETs enough to be able to really comment meaningfully on this, but look forward to hearing your progress.
BW,
Susan.
Hi Graham,
Yes 🙂 Although looking at the figures they seem higher than they should be so I need to check my test setup.
The flattening out is also apparent in the earlier tests I did with lower bias levels.
I will set up and get some figures, but will have to rework my test setup to be able to use the FFTs and PC card to get the levels of the fundament and it's harmonics. The the moment I am limited to 6 Vac max input.
The choice of 6C33C was to overcome these problems as it is low impedance 600 ohms and doesn't need a kilovolt on the anode 🙂
The are also suitable for backup winter heating.
Otherwise I see a two stage 6C33C as being quite practical. Still a much more advanced project than the mosfet based amplifier.
Best wishes,
Susan.
Graham Maynard said:
Yes 🙂 Although looking at the figures they seem higher than they should be so I need to check my test setup.
The flattening out is also apparent in the earlier tests I did with lower bias levels.
I will set up and get some figures, but will have to rework my test setup to be able to use the FFTs and PC card to get the levels of the fundament and it's harmonics. The the moment I am limited to 6 Vac max input.
The choice of 6C33C was to overcome these problems as it is low impedance 600 ohms and doesn't need a kilovolt on the anode 🙂
The are also suitable for backup winter heating.
Otherwise I see a two stage 6C33C as being quite practical. Still a much more advanced project than the mosfet based amplifier.
Best wishes,
Susan.
Re: Re: Zeus and magneplanars
Hi, Susan.
Actually, the 6C33C-B is not 600ohm. It has a plate resistance of about 80ohm, and a gain of about 4, which leaves the ideal resistance for matching a follower to the load at about 20ohm. Essentially, use 2x20ohm primaries, and 2x4ohm secondaries, and you should get 60W or more output.
Yes, they do, don't they... But those anodes can take a beating. You could go for the 6C45P too, which can take 10W per triode, but it's not a double triode, so matching and aging becomes more of an issue. As I understand, both of them are quite content to operate with red spots on the anode, and the 6C45P is happy with a glowing red anode.
In short, 6N30P is more powerful than it looks, and can dissipate about 4W per half. It too is designed for series pass regulation, I think.
True. However, with many CD players giving off 3Vrms, and 2x34Vrms being needed to reach 60Wrms (?), this should not be a problem, provided you use 2x20:2x4 stepdown on the output.
It would probably do a nice job of it. I'd recommend a dual triode like 6N1P, 6N30P or 6AS7, though.
Sincerely.
Hi, Susan.
Actually, the 6C33C-B is not 600ohm. It has a plate resistance of about 80ohm, and a gain of about 4, which leaves the ideal resistance for matching a follower to the load at about 20ohm. Essentially, use 2x20ohm primaries, and 2x4ohm secondaries, and you should get 60W or more output.
Susan-Parker said:
Yes, they do, don't they... But those anodes can take a beating. You could go for the 6C45P too, which can take 10W per triode, but it's not a double triode, so matching and aging becomes more of an issue. As I understand, both of them are quite content to operate with red spots on the anode, and the 6C45P is happy with a glowing red anode.
In short, 6N30P is more powerful than it looks, and can dissipate about 4W per half. It too is designed for series pass regulation, I think.
True. However, with many CD players giving off 3Vrms, and 2x34Vrms being needed to reach 60Wrms (?), this should not be a problem, provided you use 2x20:2x4 stepdown on the output.
It would probably do a nice job of it. I'd recommend a dual triode like 6N1P, 6N30P or 6AS7, though.
Sincerely.
Hi Pavel,
Distortion level overall is reduced.
I will do a series with different bias and different supply levels.
I get different figures from the HP8903B and the FFT based analysis, with the HP figures generally being up to two times larger i.e. showing more distortion. These current measurments are with the HP.
When I am making FFT measurments you can see that my graphs are now normally from 45 Hz to 96 kHz, which includes the mains induced noise 50 Hz and the computer switching stuff in the 35 plus Khz. If I display (and measure) only from 1 Khz to 20 kHz the figures are much improved.
THD+N at 2 watts into 8 ohms should be below 0.01% (FFT measured).
These figures made at the same time but only up to 30 watts are a lot better, for only an extra 250 mA bias per mosfet.
STW34NB20 mosfets
750 mA bias, 2:1 into 8 ohms.
Vac THD% Watts
15.54 0.3430 30.2
14.98 0.3090 28.1
14.43 0.2680 26.0
13.87 0.2330 24.0
13.30 0.2010 22.1
12.69 0.1700 20.1
12.11 0.1430 18.3
11.54 0.1190 16.6
10.96 0.0990 15.0
10.38 0.0830 13.5
9.80 0.0690 12.0
9.21 0.0580 10.6
8.62 0.0480 9.3
8.04 0.0390 8.1
7.45 0.0330 6.9
6.88 0.0270 5.9
6.29 0.0220 4.9
5.70 0.0180 4.1
5.11 0.0153 3.3
4.53 0.0127 2.6
3.95 0.0106 2.0
3.35 0.0085 1.4
2.76 0.0077 1.0
2.18 0.0076 0.6
1.59 0.0091 0.3
1.00 0.0138 0.1
I hope to have the time over the weekend to do more measurments.
BW,
Susan.
PMA said:
Distortion level overall is reduced.
I will do a series with different bias and different supply levels.
I get different figures from the HP8903B and the FFT based analysis, with the HP figures generally being up to two times larger i.e. showing more distortion. These current measurments are with the HP.
When I am making FFT measurments you can see that my graphs are now normally from 45 Hz to 96 kHz, which includes the mains induced noise 50 Hz and the computer switching stuff in the 35 plus Khz. If I display (and measure) only from 1 Khz to 20 kHz the figures are much improved.
THD+N at 2 watts into 8 ohms should be below 0.01% (FFT measured).
These figures made at the same time but only up to 30 watts are a lot better, for only an extra 250 mA bias per mosfet.
STW34NB20 mosfets
750 mA bias, 2:1 into 8 ohms.
Vac THD% Watts
15.54 0.3430 30.2
14.98 0.3090 28.1
14.43 0.2680 26.0
13.87 0.2330 24.0
13.30 0.2010 22.1
12.69 0.1700 20.1
12.11 0.1430 18.3
11.54 0.1190 16.6
10.96 0.0990 15.0
10.38 0.0830 13.5
9.80 0.0690 12.0
9.21 0.0580 10.6
8.62 0.0480 9.3
8.04 0.0390 8.1
7.45 0.0330 6.9
6.88 0.0270 5.9
6.29 0.0220 4.9
5.70 0.0180 4.1
5.11 0.0153 3.3
4.53 0.0127 2.6
3.95 0.0106 2.0
3.35 0.0085 1.4
2.76 0.0077 1.0
2.18 0.0076 0.6
1.59 0.0091 0.3
1.00 0.0138 0.1
I hope to have the time over the weekend to do more measurments.
BW,
Susan.
Re: Trasformatore & Capasitores
Hi Kamskoma,
Which is why we do the DIYaudio so we can afford to have the sound we aspire to - but perhaps without the chrome.
A transistor version of this circuit?
This is basic 1920's established amplification.
And this push-pull version is common by the end of the 1920's.
The triodes are amplifiers (i.e. a gain stage) and the transformers are impedance matching.
Mine is different from the commonly used configuration in that I use the input transformer for voltage amplification, and the mosfets are followers, not amplifiers (they actually have a gain of a little under unity).
Best wishes,
Susan.
Hi Kamskoma,
kamskoma said:
Which is why we do the DIYaudio so we can afford to have the sound we aspire to - but perhaps without the chrome.
A transistor version of this circuit?
This is basic 1920's established amplification.
And this push-pull version is common by the end of the 1920's.
The triodes are amplifiers (i.e. a gain stage) and the transformers are impedance matching.
Mine is different from the commonly used configuration in that I use the input transformer for voltage amplification, and the mosfets are followers, not amplifiers (they actually have a gain of a little under unity).
Best wishes,
Susan.
Hi Susan,
I've been following this thread from the start and have just spent the evening rereading it. Thanks for all your hard work and timely responses to the forum members - it's great to see such input!
At long last I'm now able to start building a pair of 75W monoblocks. I'd like to make the best version possible. You've mentioned mu-metal laminations for the input transformer in a few posts (67,219,224,256,262 and 443). Did you ever get a chance to try these out?
James
I've been following this thread from the start and have just spent the evening rereading it. Thanks for all your hard work and timely responses to the forum members - it's great to see such input!
At long last I'm now able to start building a pair of 75W monoblocks. I'd like to make the best version possible. You've mentioned mu-metal laminations for the input transformer in a few posts (67,219,224,256,262 and 443). Did you ever get a chance to try these out?
James
Hi Susan,
can I ask you, when do you will stop this project ( or when do you will come ) with " end results " ? The same results ( ideas ) published eighty years ago ( sure, without measuring of distortion at HP 😉 ) , any progres ? For what do you hope ?
can I ask you, when do you will stop this project ( or when do you will come ) with " end results " ? The same results ( ideas ) published eighty years ago ( sure, without measuring of distortion at HP 😉 ) , any progres ? For what do you hope ?
Upupa Epops said:
Hi Upupa Epops,
"can I ask you, when do you will stop this" negativity? diyAudio is full of new ideas, old ideas and re-invented ones. I see many people participating in forums devoted to tubes/valves. Do you hassle every one of those threads as well?
James
Hi James,
Thank you for your interest 🙂
Yes, it has been very encouraging.
Yes, I tried 100% Mumetal in the input laminations.
Some improvement in the 200 Hz to 1 kHz band, but considerably more saturation in the low bass at higher levels.
I would consider 50/50 mix of Mumetal/M6 to be the best compromise. We are talking about something like only a 25% improvement over this small range, but every little helps. It is the fundamentals for most instruments after all.
Best wishes,
Susan.
Thank you for your interest 🙂
nemestra said:
Yes, it has been very encouraging.
Yes, I tried 100% Mumetal in the input laminations.
Some improvement in the 200 Hz to 1 kHz band, but considerably more saturation in the low bass at higher levels.
I would consider 50/50 mix of Mumetal/M6 to be the best compromise. We are talking about something like only a 25% improvement over this small range, but every little helps. It is the fundamentals for most instruments after all.
Best wishes,
Susan.
