Ciao,
following you can find some link about a project I developped and show on Audioreview magazine in Italy.
It is a 300B pushpull amp with a tube rectifier and some particular solution on input schematic and output trafo.
There is only one cap between the input and the grid of 300B, the swing is good, about 160 vpp.
I use a fixed bias, the feedback is activable by switch, incase.
the output trafo is a double C cores, the a-a is 3.200 ohm but there are two secondaries connected in parallel with 3 ohm of impedance, not tap.
The tubes are 417A in input, the 6463 as splitter and 300B JJ.
The rectifier is a GZ34 Philips/Mullard (or similar nos) and the total current is about 240 mA, is reccomanded, if the soviet or JJ GZ34 will be used, to split the HT.
1) http://www.multitask.it/diy/intro.tif
is a introduction page of the magazine
2) http://www.multitask.it/diy/sch.1-300Bout.jpg
is a amplifier schematic
3) http://www.multitask.it/diy/sch.2-300Bpower.JPG
it is a power supply
4) http://www.multitask.it/diy/sch.3-300Bfil.JPG
it is a filaments supply, two way : ac and dc (I prefer the ac)
In ac I get about 0,8 mV at the output as noise, very low.
5) http://www.multitask.it/diy/freqtresp.jpg
these are a freq. response without (upper) and with (lower) feedback, for different loads. On 8 ohm, w/o FB the -3dB are at 80KHz
6) http://www.multitask.it/diy/ccl.JPG
this is a CCl, this is in function of the load, continuos (red line) and impulsive (black line); the line is quite vertical (ideal) to show the capacity to give current when we lower the load.
7) http://www.multitask.it/diy/parts-amp.rtf
it is a list of component, one channel
8) http://www.multitask.it/diy/APERTURA1.JPG
is a photo of one amp
Comments are welcome
Bye
Walter
following you can find some link about a project I developped and show on Audioreview magazine in Italy.
It is a 300B pushpull amp with a tube rectifier and some particular solution on input schematic and output trafo.
There is only one cap between the input and the grid of 300B, the swing is good, about 160 vpp.
I use a fixed bias, the feedback is activable by switch, incase.
the output trafo is a double C cores, the a-a is 3.200 ohm but there are two secondaries connected in parallel with 3 ohm of impedance, not tap.
The tubes are 417A in input, the 6463 as splitter and 300B JJ.
The rectifier is a GZ34 Philips/Mullard (or similar nos) and the total current is about 240 mA, is reccomanded, if the soviet or JJ GZ34 will be used, to split the HT.
1) http://www.multitask.it/diy/intro.tif
is a introduction page of the magazine
2) http://www.multitask.it/diy/sch.1-300Bout.jpg
is a amplifier schematic
3) http://www.multitask.it/diy/sch.2-300Bpower.JPG
it is a power supply
4) http://www.multitask.it/diy/sch.3-300Bfil.JPG
it is a filaments supply, two way : ac and dc (I prefer the ac)
In ac I get about 0,8 mV at the output as noise, very low.
5) http://www.multitask.it/diy/freqtresp.jpg
these are a freq. response without (upper) and with (lower) feedback, for different loads. On 8 ohm, w/o FB the -3dB are at 80KHz
6) http://www.multitask.it/diy/ccl.JPG
this is a CCl, this is in function of the load, continuos (red line) and impulsive (black line); the line is quite vertical (ideal) to show the capacity to give current when we lower the load.
7) http://www.multitask.it/diy/parts-amp.rtf
it is a list of component, one channel
8) http://www.multitask.it/diy/APERTURA1.JPG
is a photo of one amp
Comments are welcome
Bye
Walter
waltube said:
Very nice looking build.
I couldn't read the article so, I apologize if it's described there. Interesting way to separate DC and AC for the splitter. I'm not sure the sonic effect of the cap would be any less in that position than done in the normal way. Are there other reasons you chose that approach?
On your schema, the voltage at the cathode of the LTP is shown as 90V, and the voltage at the grid is 100. Enhancement mode tube?
Sheldon
Hi,
there is an error I made on writing output schematic.
The link is now:
http://www.multitask.it/diy/300Bout.jpg
The error was on Voltage on splitter cathode and 417A anode.
Sheldon, thanks for the advise.
The reason is to eliminate one cap; the 417A , for me, is a very good tube and in this configuration it helps me to connect directly the two stages.
Regarding sonic performances I am not sure that in this way I can see the heaven but technically, for me, is a good solution (not so common).
GND stand for Ground; the R16 (10 ohms on each channel) is about the ground loop.
In my configuration this break the loop of input cables.
Is not shown, but inside the chassis there is another trafo (toroid) only for filaments and -V for bias. There are 4 different 5 Vac fil. for each 300B; about ac I use a Bourns 2w 100 ohm conductive plastic (from RS) to limit the hum; it is very good (but expensive) stuff.
Regarding the output trafo, there is a peak about at 150 KHz reasoneble far from audio.
I check it on lab, ouside the amp.
Regarding feedback, without it the output impedance is 1,8 ohm; I set the FB a 6 dB, so the Zout goes at 0,9 ohm (good). I test also 12 db of FB but the sonic results is non so "brillant" as 6 db (with big woofers) and no FB with small woofers.
This due a different Zout (changing in dumping factor); it is incredible the difference.
I put, on the back, a switch where I can choose from no FB, 6dB and 12 dB; every time when we use the small speaker the no FB has the best music performances.
These amp, and few some other, are working with strong loudspeakers; the one in the photo was working with a pair of B&W 800 with a very good results on sound quality and a very good dynamics.
But I made, for this freind, a pair of 4 x 300B monos with same theory but more power (45 w rms, goes to 75w impulsive).
Bye
Walter
ps: in which way I can edit my first post to change one link?
there is an error I made on writing output schematic.
The link is now:
http://www.multitask.it/diy/300Bout.jpg
The error was on Voltage on splitter cathode and 417A anode.
Sheldon, thanks for the advise.
The reason is to eliminate one cap; the 417A , for me, is a very good tube and in this configuration it helps me to connect directly the two stages.
Regarding sonic performances I am not sure that in this way I can see the heaven but technically, for me, is a good solution (not so common).
GND stand for Ground; the R16 (10 ohms on each channel) is about the ground loop.
In my configuration this break the loop of input cables.
Is not shown, but inside the chassis there is another trafo (toroid) only for filaments and -V for bias. There are 4 different 5 Vac fil. for each 300B; about ac I use a Bourns 2w 100 ohm conductive plastic (from RS) to limit the hum; it is very good (but expensive) stuff.
Regarding the output trafo, there is a peak about at 150 KHz reasoneble far from audio.
I check it on lab, ouside the amp.
Regarding feedback, without it the output impedance is 1,8 ohm; I set the FB a 6 dB, so the Zout goes at 0,9 ohm (good). I test also 12 db of FB but the sonic results is non so "brillant" as 6 db (with big woofers) and no FB with small woofers.
This due a different Zout (changing in dumping factor); it is incredible the difference.
I put, on the back, a switch where I can choose from no FB, 6dB and 12 dB; every time when we use the small speaker the no FB has the best music performances.
These amp, and few some other, are working with strong loudspeakers; the one in the photo was working with a pair of B&W 800 with a very good results on sound quality and a very good dynamics.
But I made, for this freind, a pair of 4 x 300B monos with same theory but more power (45 w rms, goes to 75w impulsive).
Bye
Walter
ps: in which way I can edit my first post to change one link?
waltube said:
Seems like the sonic contribution of the cap will be similar in either case. But should not be a problem to use a good quality cap in either position. It is, as you say, not a common way to arrange the LTP, but does allow you to use different operating conditions for the splitter/driver. It's an interesting option. If I were building it, I'd be tempted to try both ways, just for fun.
waltube said:
The switchable feedback is a nice feature. Most box speakers are designed for an amp with high damping factor. If we are building our own speakers, we can design for any damping factor we want. The amp, enclosure, and driver, combined, control driver damping, and we can use any combination that gives the damping we want. A lower damping factor is usually helpful for open baffle bass too, also full range drivers (which are usually highly damped physically).
Sheldon
There is really nothing uncommon about that phase splitter, it is nearly identical to the one published in the old Mullard 5-20 design 50 yrs ago and I have used a very similar circuit in EL34 based PP amplifiers for over 20 yrs now. (This DC coupled splitter configuration is NOT new!) The way it is drawn makes it look different, but it's not. Balance is IMPORTANT! Trimming the plate loads carefully may pay some dividends. The other trick is to get it to swing enough voltage to drive a 300B well - which with care it will do with just a little voltage margin.
As far as the issues with frequency response that looks like an output transformer resonance outside of the audio passband and may not be an issue per se. I would possibly consider a simple first order LPF in front of the first stage of the amplifier that is -3dB at 50kHz or so to avoid exciting that resonance, but in all likelihood it is not an issue.
Stability margin as suggested should be checked, more than 60 degrees of phase shift in the region of that resonance is a concern depending on the phase angle resulting from a speaker reactive load being added.
I think you will quickly find that you do not need global negative feedback at all in this design, 15+ 300B PP KTA amplifiers sold tells me so..
As far as the issues with frequency response that looks like an output transformer resonance outside of the audio passband and may not be an issue per se. I would possibly consider a simple first order LPF in front of the first stage of the amplifier that is -3dB at 50kHz or so to avoid exciting that resonance, but in all likelihood it is not an issue.
Stability margin as suggested should be checked, more than 60 degrees of phase shift in the region of that resonance is a concern depending on the phase angle resulting from a speaker reactive load being added.
I think you will quickly find that you do not need global negative feedback at all in this design, 15+ 300B PP KTA amplifiers sold tells me so..
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.