the zeners...
So we are in bussines! you can improve a lot the noise perfomance of the regulator...
As zeners are good voltages sources the capacitor across the zeners are not very good in decoupling it and garanty low noise.
Solution...from the top of the string put a resistor (say 1K)...then conect it to the decoupling capacitor...and at this point of union of the resistor and the capacitor conect the gate of the MOSFET...
Now you have a "virtual batery operation".
Hasta!!
Jorge
So we are in bussines! you can improve a lot the noise perfomance of the regulator...
As zeners are good voltages sources the capacitor across the zeners are not very good in decoupling it and garanty low noise.
Solution...from the top of the string put a resistor (say 1K)...then conect it to the decoupling capacitor...and at this point of union of the resistor and the capacitor conect the gate of the MOSFET...
Now you have a "virtual batery operation".
Hasta!!
Jorge
Jorge,
I do have this type of connection, let me explain:
Since we have two regulators with only one voltage reference (zener's) I have the connection to the gate of the mosfet's through a 12K R, one for each mosfet coming from the reference, and each gate is decoupled with 100u X 2 lytics plus a Wima 0.47.
Maybe the addition of a 0.01 cap would help?
I do have this type of connection, let me explain:
Since we have two regulators with only one voltage reference (zener's) I have the connection to the gate of the mosfet's through a 12K R, one for each mosfet coming from the reference, and each gate is decoupled with 100u X 2 lytics plus a Wima 0.47.
Maybe the addition of a 0.01 cap would help?
zeners...
I think that your regulator is allwrigt...
Any resistor that have good stability... low noise and can take the dissipation!...Choose the brand!!!
I do not recomend...because when they become hot they made funy noises as the end caps are not soldered to the nikel wire...they are good only for real power jobs...
Buenas!
Jorge
PS : the resistor is dissipating 1,62W...so 3W do the job.
I think that your regulator is allwrigt...
Any resistor that have good stability... low noise and can take the dissipation!...Choose the brand!!!
I do not recomend...because when they become hot they made funy noises as the end caps are not soldered to the nikel wire...they are good only for real power jobs...
Buenas!
Jorge
PS : the resistor is dissipating 1,62W...so 3W do the job.
Round up
Now that I have some 80 hours of listening to the Line Amp I want to sum up my impresions.
First let me say that after about 4 days of testing the LA on my system I finished a speaker project with new Fostex FE164 full range that I use wide open which are very reveling transducers.
Of course this needed break-in so I thought that after 20 or so hours this were done and started to "evaluate" the LA. My comments regarding to the slight graininess on the upper freq were actualy related to the speakers and not to the LA since the these speakers need way more than 20 hours to brake-in.
Now I can report that the slight graininess was due to the speakers and not to the LA. Other roughness I may hear are due to the CD player as most of us are familiar with. In my case I use and old Onkyo Integra DX-706 which BTW has been greatly modify during its existance of more than 12 years but by no means it can compeat to later technologies CDP.
To my knowledge, this is a very nice implemented LA and by a far degree better than all other LA I have built during the last 30 years, including SS, hybrids and Vacuums including 12AX7 and 6SL7 families and 6922 (6DJ8).
The 6922 is an excellent tube (in my book) but not to the task of driving dificult loads such as an Aleph or Zen, both of which I use, and have drive these with the 6922. In this application the tube is somewhat rough and does not control well low freq (loose) and high freq are week as is the mid bass which I find a very important ingredient to a good sounding system. It's my impresion though that the 6922 would be a very good candidate to drive vacuum amplifiers with 100K, or more, input impedance -which I have not tested even though I have some beefy PP monoblocks around.
Now coming back to the 417A (5842) as LA is my opinion that all the shortcomings of the 6922 (for this application- low impedance load) are addressed in a very nice way. High frequency is smooth and very reveling (and powerfull), I have found some CD what I thought were boring to be full of life and in no way boring. Bass and mid bass very tight and full of detail. And to say of the most important range, mids, are full of life and not lay back nor up front either, just the perspective intended by the console technician at the studio. The aural presentetion is wide and deep pleasing and musical.
All of this has been achived with 2 cents plate load resistors, yes they have magnetic caps. I also tested some low wattage carbon composition resistors on this position and the aural presentation still came even better (with maybe less detail though) but had to remove them after about 3 hours for heat noises coming from them (they were 1/2 w each x2, way below of what is needed).
Let me add also that the ladder attenuators were implemented with resistors i had at hand and very few of them are non magnetic and some of them are even carbon film. So changing this parts will definitely improve the sound.
Output audio transformer:
And best for last. This LA was conceived to be an Euridice circuit at the end so the plate resistors and the output coupling cap will be replaced by an output audio transformer which the primary will serve as plate load (as usual) and the secondary will be the output element that will replace the capacitor, so a no capacitor in series with the signal LA.
The use of an OPT will lower still more the output impedance and provide a X 10 increase on drive power!!!
This will also address a shortcomings of the stage (as preamp) as having excesive gain. The OPT will lower the gain to a factor of around 4 or even more depending on what you may use. To deal with this in the meanwhile I'm using low values of resistances as shunt elements on the attenuators. So the signal is very close to ground not a desirable thing IMO.
All in all the 417A is a powerfull triode and well deserved praise by others and now me and with no doubt it will make also an excellent voltage driver as an input stage (or even driver of the power tube in some applications) of a power amplifier.
Some detail of this implementation:
I have biased the stage at 18 mA using a +B at 180V a 5K plate reistor which gives around 93V at the plate using a 30 Ohms resistor on the cathode. Cathode voltage around 0.5+.
The PSU is all SS as are the regulators (one each channel) using only one zener voltage reference adecuated bypassed. From the regulators there is a 5H inductor connectig to the +B cap in this case I used a 330u 400V Nichicon 105°C lytic GK type that I had with no bypassing (I may try bypassing in the future and see if like it).
The rectification is by Fast recovery shotky (?) diodes in a full wave configuration using snubbers in each device (same as in the heater ac/dc). Every atemp (known to me) was made to lower any parasitics from SS devices through all the PSU.
Series regulators are MOSFET IRF840LC (low capacitance 800V devices)
The heater voltage is regulated with LM317T in a texbook variable configuration, one for each tube. One can easily dial in the requiered 6.3V.
I want to thaks Thorsten Loesch for the posted implementation he did of the Euridice preamp that served me as a starting point for my implementation. Many details of the power supply were taken directly form here. Though I added some "improvements".
(just to feel that is my design!). See his web site.
And most important Thanks to Ciro Marzio, Cristiano Jelasi & Luca Chiomenti which are the original designers of the Euridice preamp.
Anyone interest on my implementation feel free to contact me for details. I may even prepare some schematics and help around on their implementation.
Now that I have some 80 hours of listening to the Line Amp I want to sum up my impresions.
First let me say that after about 4 days of testing the LA on my system I finished a speaker project with new Fostex FE164 full range that I use wide open which are very reveling transducers.
Of course this needed break-in so I thought that after 20 or so hours this were done and started to "evaluate" the LA. My comments regarding to the slight graininess on the upper freq were actualy related to the speakers and not to the LA since the these speakers need way more than 20 hours to brake-in.
Now I can report that the slight graininess was due to the speakers and not to the LA. Other roughness I may hear are due to the CD player as most of us are familiar with. In my case I use and old Onkyo Integra DX-706 which BTW has been greatly modify during its existance of more than 12 years but by no means it can compeat to later technologies CDP.
To my knowledge, this is a very nice implemented LA and by a far degree better than all other LA I have built during the last 30 years, including SS, hybrids and Vacuums including 12AX7 and 6SL7 families and 6922 (6DJ8).
The 6922 is an excellent tube (in my book) but not to the task of driving dificult loads such as an Aleph or Zen, both of which I use, and have drive these with the 6922. In this application the tube is somewhat rough and does not control well low freq (loose) and high freq are week as is the mid bass which I find a very important ingredient to a good sounding system. It's my impresion though that the 6922 would be a very good candidate to drive vacuum amplifiers with 100K, or more, input impedance -which I have not tested even though I have some beefy PP monoblocks around.
Now coming back to the 417A (5842) as LA is my opinion that all the shortcomings of the 6922 (for this application- low impedance load) are addressed in a very nice way. High frequency is smooth and very reveling (and powerfull), I have found some CD what I thought were boring to be full of life and in no way boring. Bass and mid bass very tight and full of detail. And to say of the most important range, mids, are full of life and not lay back nor up front either, just the perspective intended by the console technician at the studio. The aural presentetion is wide and deep pleasing and musical.
All of this has been achived with 2 cents plate load resistors, yes they have magnetic caps. I also tested some low wattage carbon composition resistors on this position and the aural presentation still came even better (with maybe less detail though) but had to remove them after about 3 hours for heat noises coming from them (they were 1/2 w each x2, way below of what is needed).
Let me add also that the ladder attenuators were implemented with resistors i had at hand and very few of them are non magnetic and some of them are even carbon film. So changing this parts will definitely improve the sound.
Output audio transformer:
And best for last. This LA was conceived to be an Euridice circuit at the end so the plate resistors and the output coupling cap will be replaced by an output audio transformer which the primary will serve as plate load (as usual) and the secondary will be the output element that will replace the capacitor, so a no capacitor in series with the signal LA.
The use of an OPT will lower still more the output impedance and provide a X 10 increase on drive power!!!
This will also address a shortcomings of the stage (as preamp) as having excesive gain. The OPT will lower the gain to a factor of around 4 or even more depending on what you may use. To deal with this in the meanwhile I'm using low values of resistances as shunt elements on the attenuators. So the signal is very close to ground not a desirable thing IMO.
All in all the 417A is a powerfull triode and well deserved praise by others and now me and with no doubt it will make also an excellent voltage driver as an input stage (or even driver of the power tube in some applications) of a power amplifier.
Some detail of this implementation:
I have biased the stage at 18 mA using a +B at 180V a 5K plate reistor which gives around 93V at the plate using a 30 Ohms resistor on the cathode. Cathode voltage around 0.5+.
The PSU is all SS as are the regulators (one each channel) using only one zener voltage reference adecuated bypassed. From the regulators there is a 5H inductor connectig to the +B cap in this case I used a 330u 400V Nichicon 105°C lytic GK type that I had with no bypassing (I may try bypassing in the future and see if like it).
The rectification is by Fast recovery shotky (?) diodes in a full wave configuration using snubbers in each device (same as in the heater ac/dc). Every atemp (known to me) was made to lower any parasitics from SS devices through all the PSU.
Series regulators are MOSFET IRF840LC (low capacitance 800V devices)
The heater voltage is regulated with LM317T in a texbook variable configuration, one for each tube. One can easily dial in the requiered 6.3V.
I want to thaks Thorsten Loesch for the posted implementation he did of the Euridice preamp that served me as a starting point for my implementation. Many details of the power supply were taken directly form here. Though I added some "improvements".
(just to feel that is my design!). See his web site.
And most important Thanks to Ciro Marzio, Cristiano Jelasi & Luca Chiomenti which are the original designers of the Euridice preamp.
Anyone interest on my implementation feel free to contact me for details. I may even prepare some schematics and help around on their implementation.
Re: Round up
try shunt reg with one OA2 per channel ;
I have WOT preamp with EC8010 and OA2 per channel-nothing is better than that (at least for me);any other type of supply is worse;
remember that in SE supply is in signal path
apassgear said:Now that I have some 80 hours of listening to the Line Amp I want to sum up my impresions.
First let me say that after about 4 days of testing the LA on my system I finished a speaker project with new Fostex FE164 full range that I use wide open which are very reveling transducers.
Of course this needed break-in so I thought that after 20 or so hours this were done and started to "evaluate" the LA. My comments regarding to the slight graininess on the upper freq were actualy related to the speakers and not to the LA since the these speakers need way more than 20 hours to brake-in.
Now I can report that the slight graininess was due to the speakers and not to the LA. Other roughness I may hear are due to the CD player as most of us are familiar with. In my case I use and old Onkyo Integra DX-706 which BTW has been greatly modify during its existance of more than 12 years but by no means it can compeat to later technologies CDP.
To my knowledge, this is a very nice implemented LA and by a far degree better than all other LA I have built during the last 30 years, including SS, hybrids and Vacuums including 12AX7 and 6SL7 families and 6922 (6DJ8).
The 6922 is an excellent tube (in my book) but not to the task of driving dificult loads such as an Aleph or Zen, both of which I use, and have drive these with the 6922. In this application the tube is somewhat rough and does not control well low freq (loose) and high freq are week as is the mid bass which I find a very important ingredient to a good sounding system. It's my impresion though that the 6922 would be a very good candidate to drive vacuum amplifiers with 100K, or more, input impedance -which I have not tested even though I have some beefy PP monoblocks around.
Now coming back to the 417A (5842) as LA is my opinion that all the shortcomings of the 6922 (for this application- low impedance load) are addressed in a very nice way. High frequency is smooth and very reveling (and powerfull), I have found some CD what I thought were boring to be full of life and in no way boring. Bass and mid bass very tight and full of detail. And to say of the most important range, mids, are full of life and not lay back nor up front either, just the perspective intended by the console technician at the studio. The aural presentetion is wide and deep pleasing and musical.
All of this has been achived with 2 cents plate load resistors, yes they have magnetic caps. I also tested some low wattage carbon composition resistors on this position and the aural presentation still came even better (with maybe less detail though) but had to remove them after about 3 hours for heat noises coming from them (they were 1/2 w each x2, way below of what is needed).
Let me add also that the ladder attenuators were implemented with resistors i had at hand and very few of them are non magnetic and some of them are even carbon film. So changing this parts will definitely improve the sound.
Output audio transformer:
And best for last. This LA was conceived to be an Euridice circuit at the end so the plate resistors and the output coupling cap will be replaced by an output audio transformer which the primary will serve as plate load (as usual) and the secondary will be the output element that will replace the capacitor, so a no capacitor in series with the signal LA.
The use of an OPT will lower still more the output impedance and provide a X 10 increase on drive power!!!
This will also address a shortcomings of the stage (as preamp) as having excesive gain. The OPT will lower the gain to a factor of around 4 or even more depending on what you may use. To deal with this in the meanwhile I'm using low values of resistances as shunt elements on the attenuators. So the signal is very close to ground not a desirable thing IMO.
All in all the 417A is a powerfull triode and well deserved praise by others and now me and with no doubt it will make also an excellent voltage driver as an input stage (or even driver of the power tube in some applications) of a power amplifier.
Some detail of this implementation:
I have biased the stage at 18 mA using a +B at 180V a 5K plate reistor which gives around 93V at the plate using a 30 Ohms resistor on the cathode. Cathode voltage around 0.5+.
The PSU is all SS as are the regulators (one each channel) using only one zener voltage reference adecuated bypassed. From the regulators there is a 5H inductor connectig to the +B cap in this case I used a 330u 400V Nichicon 105°C lytic GK type that I had with no bypassing (I may try bypassing in the future and see if like it).
The rectification is by Fast recovery shotky (?) diodes in a full wave configuration using snubbers in each device (same as in the heater ac/dc). Every atemp (known to me) was made to lower any parasitics from SS devices through all the PSU.
Series regulators are MOSFET IRF840LC (low capacitance 800V devices)
The heater voltage is regulated with LM317T in a texbook variable configuration, one for each tube. One can easily dial in the requiered 6.3V.
I want to thaks Thorsten Loesch for the posted implementation he did of the Euridice preamp that served me as a starting point for my implementation. Many details of the power supply were taken directly form here. Though I added some "improvements".
(just to feel that is my design!). See his web site.
And most important Thanks to Ciro Marzio, Cristiano Jelasi & Luca Chiomenti which are the original designers of the Euridice preamp.
Anyone interest on my implementation feel free to contact me for details. I may even prepare some schematics and help around on their implementation.
try shunt reg with one OA2 per channel ;
I have WOT preamp with EC8010 and OA2 per channel-nothing is better than that (at least for me);any other type of supply is worse;
remember that in SE supply is in signal path
Hi folks,
I have to agree with everyone that the 417a with output transformer makes a very nice line amp.
I have been running mine with Bartolucci outputs, 180v CLCLC power supply, and 220ohm cathode resistor bypassed with big cerafine electrolytics.
I just modified the power supply to use 0A2 shunt regulators and think the sound improves quite a bit, but I need to do a few things. First I need to lower the cathode resistor to get a bit more current through the tube. Second, I need to drop the capacitance after the 0A2. Right now I have 100uf polypro caps which are left over from the original version. Most of the time these aren't an issue but occasionally the 0A2 / 100uf combo seems to turn into a relaxation oscillator. The 0A2 datasheet suggest no more than 0.1uf, so I think I definitely need to dig back in there.
Since I have more gain than I need in the system, I'm also tempted to unbypass the cathode resistor. Anyone try this?
---Gary
I have to agree with everyone that the 417a with output transformer makes a very nice line amp.
I have been running mine with Bartolucci outputs, 180v CLCLC power supply, and 220ohm cathode resistor bypassed with big cerafine electrolytics.
I just modified the power supply to use 0A2 shunt regulators and think the sound improves quite a bit, but I need to do a few things. First I need to lower the cathode resistor to get a bit more current through the tube. Second, I need to drop the capacitance after the 0A2. Right now I have 100uf polypro caps which are left over from the original version. Most of the time these aren't an issue but occasionally the 0A2 / 100uf combo seems to turn into a relaxation oscillator. The 0A2 datasheet suggest no more than 0.1uf, so I think I definitely need to dig back in there.
Since I have more gain than I need in the system, I'm also tempted to unbypass the cathode resistor. Anyone try this?
---Gary
Attachments
GaryB said:
Gary,
Very nice Euridice you got, and with Bartolucci tranies!!!
Guess nobody needs a cathode bypass with this stage, I would cerainly give it a try. Myself not using one, but of course I'm not using a OT.
choky's fix bias could be another thing to try, I've done this with a 6922 to a somewhat mix results, never came to realy make a comparison of its benefits.
Anyone can point me to a suitable OA2 reg link for this application???
Some test I have done on the grid stopper:
Originally I had two gridstopper one to each side of the frame with no oscillation problems and a grid leak R of some 150 or 200K.
I removed both Gridstoppers and Grid leak R and replaced only with a wire. For grid ref to ground I had only the very low value R at the attenuator. There were no oscillations and the sound was very much improved in the transparency department but the balance was tilted to a not so nice Hi freq weight, but maybe this will not happen with a OT. On the scope there was nothing to indicate this change of balance.
Later I replaced the wire with ony one gridstopper, the same Dale 220 R, non magnetic metal film 1/2 W. Lost some magic though and have not recover the balance but cleaner than using two R in //.
Seems that this resistor really makes a difference.
What are you using as gridstopper???
Originally I had two gridstopper one to each side of the frame with no oscillation problems and a grid leak R of some 150 or 200K.
I removed both Gridstoppers and Grid leak R and replaced only with a wire. For grid ref to ground I had only the very low value R at the attenuator. There were no oscillations and the sound was very much improved in the transparency department but the balance was tilted to a not so nice Hi freq weight, but maybe this will not happen with a OT. On the scope there was nothing to indicate this change of balance.
Later I replaced the wire with ony one gridstopper, the same Dale 220 R, non magnetic metal film 1/2 W. Lost some magic though and have not recover the balance but cleaner than using two R in //.
Seems that this resistor really makes a difference.
What are you using as gridstopper???
nope ;(apassgear said:
tried toobs:
7DJ8 in para(lell
, 6SN7 in paralell,EL84 triode,6V6 triode ,12BH7 paralell,10Yparafeed,etc,but I don't have any 417 to play with.I chase stash of 7788 almost a year to play,but I didn't catch that for now........ hehe.
who knows what will be next; in last time I play little with xformers-after my CD player DA just into one resistor and 160:78K iron.
conclusion-more iron in chain,more life in music.
who cares for silly terms like harmonics and distortion
Hi,
Regarding taming oscillations in the 417a, there was a lot of discussion on this on joelist a few years back. Based on that discussion I tried the following, which seems to work pretty well. I use only one grid connection to the 417a. The tube socket pin for the other grid connection is cut off, presumbably to lesson its tendency to act as an antenna. To tame oscillations, I use a ferrite bead on the wire going to the remaining grid pin. This seems to work well with a nice smooth sound and none of the high frequency harshness that tends to indicate oscillations.
Apassgear, I suspect that the balance problem you note without grid stoppers is some high frequency oscillation even if you can't see it on the scope. I'd try ferrite beads and see if that helps.
About using NiCds in the cathode of the tube, I actually tried this a couple of years back when I first built this preamp and I wasn't at all happy with the sound. For some reason it didn't work well for me.
And finally, getting back to the 0A2 issue, my power supply (and 0A2) is in a different box from the preamp, so I worry that the option of no capacitor is not a good one. I think I'll try a small russian military teflon cap and see how that does. I can compare 0.056uf and 0.01uf caps.
---Gary
Regarding taming oscillations in the 417a, there was a lot of discussion on this on joelist a few years back. Based on that discussion I tried the following, which seems to work pretty well. I use only one grid connection to the 417a. The tube socket pin for the other grid connection is cut off, presumbably to lesson its tendency to act as an antenna. To tame oscillations, I use a ferrite bead on the wire going to the remaining grid pin. This seems to work well with a nice smooth sound and none of the high frequency harshness that tends to indicate oscillations.
Apassgear, I suspect that the balance problem you note without grid stoppers is some high frequency oscillation even if you can't see it on the scope. I'd try ferrite beads and see if that helps.
About using NiCds in the cathode of the tube, I actually tried this a couple of years back when I first built this preamp and I wasn't at all happy with the sound. For some reason it didn't work well for me.
And finally, getting back to the 0A2 issue, my power supply (and 0A2) is in a different box from the preamp, so I worry that the option of no capacitor is not a good one. I think I'll try a small russian military teflon cap and see how that does. I can compare 0.056uf and 0.01uf caps.
---Gary
OSCILLATIONS AND GRIDSTOPPERS.
Hi,
Indeed, and it's even recommended to put a small cap (good Q filmcap is fine) across the voltage reference tube.
This is especially true when it is used on its own.
The cap will snubb any noise and stabilise the tubes' operation.
Of all the Stabivolt tubes the 90C1 is one of my favourites together with the 5651A.
Re: gridstoppers, if you don't need them, don't use them.
Contrary to what you may read they kill detail and dynamic range suffers too.
If you need to put one in, always solder them as close to the grid as possible and use the smallest possible value.
Listen to the difference and I think you'll agree.
Cheers,
Hi,
Indeed, and it's even recommended to put a small cap (good Q filmcap is fine) across the voltage reference tube.
This is especially true when it is used on its own.
The cap will snubb any noise and stabilise the tubes' operation.
Of all the Stabivolt tubes the 90C1 is one of my favourites together with the 5651A.
Re: gridstoppers, if you don't need them, don't use them.
Contrary to what you may read they kill detail and dynamic range suffers too.
If you need to put one in, always solder them as close to the grid as possible and use the smallest possible value.
Listen to the difference and I think you'll agree.
Cheers,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.