Colt45 said:SEP, not SET
Aka pentode connected. Screen regulation is a good thing here.
Note that the output impedance will be pretty high and this should be taken into consideration in pairing with loudspeakers. (Damping factor less than 1?) This thing will probably look a lot like a current source to most loudspeakers.
From Kevin
Not sure what you mean by screen regulation - a voltage regulator on the cct that feeds the screen of the tube?
Isn't the output impedance (as seen by the speakers) determined by the secondaries of the O/P trafo?
I thought that speakers wanted current rather than voltage - ala Krell amp design philosophy...
Jess
Not sure what you mean by screen regulation - a voltage regulator on the cct that feeds the screen of the tube?
Isn't the output impedance (as seen by the speakers) determined by the secondaries of the O/P trafo?
I thought that speakers wanted current rather than voltage - ala Krell amp design philosophy...
Jess
Originally posted by JesseG
Not sure what you mean by screen regulation - a voltage regulator on the cct that feeds the screen of the tube?
There is a 200V zener diode connected to the screen - so the screen supply is regulated as a result.
Isn't the output impedance (as seen by the speakers) determined by the secondaries of the O/P trafo?
Not completely, the transformer just transforms the source impedance of the tube driving it by the 1/ turns ratio^2. The individual taps just change the reflected ratio. So lets say that your transformer has a 20:1 step down ratio then this squared is 400 on say the 16 ohm tap (Primary Z 6.4K). Lets say your tube has an RP of 10K then the reflected source impedance is just 10K/400 or 25 ohms, not neglecting the dcr of the secondary winding say 4 ohms (and ignoring the dcr of the primary) this is a source Z of 29 ohms from the 16 ohm tap.. Damping factor will be about 0.55 - no longer much point to looking at it this way as the amp is providing very little damping to the driver.
I thought that speakers wanted current rather than voltage - ala Krell amp design philosophy...
Actually Krell builds excellent voltage sources, an "ideal" voltage source will source an infinite output current into any load impedance. In the real world voltage sources have current limitations that modify this behavior. A current source otoh produces a voltage that is proportional to the load impedance for a given current, whereas a voltage source produces an output voltage that is not a function of the impedance connected across it. (Like a Krell) This implies that the output voltage will vary substantially with the load impedance, most speakers are not designed to operate with such a driving source and will have an acoustical output that is a function of its source impedance and the output current rather than just the voltage applied across it.
A pentode connected amplifier output stage is usually used in conjunction with NFB to lower its source impedance to something useful with most speakers. Triodes incidentally have much lower plate resistances and so avoid this issue.
Hopefully I have not muddied the waters too much, and perhaps someone else can explain this better than I..
Not sure what you mean by screen regulation - a voltage regulator on the cct that feeds the screen of the tube?
There is a 200V zener diode connected to the screen - so the screen supply is regulated as a result.
Isn't the output impedance (as seen by the speakers) determined by the secondaries of the O/P trafo?
Not completely, the transformer just transforms the source impedance of the tube driving it by the 1/ turns ratio^2. The individual taps just change the reflected ratio. So lets say that your transformer has a 20:1 step down ratio then this squared is 400 on say the 16 ohm tap (Primary Z 6.4K). Lets say your tube has an RP of 10K then the reflected source impedance is just 10K/400 or 25 ohms, not neglecting the dcr of the secondary winding say 4 ohms (and ignoring the dcr of the primary) this is a source Z of 29 ohms from the 16 ohm tap.. Damping factor will be about 0.55 - no longer much point to looking at it this way as the amp is providing very little damping to the driver.
I thought that speakers wanted current rather than voltage - ala Krell amp design philosophy...
Actually Krell builds excellent voltage sources, an "ideal" voltage source will source an infinite output current into any load impedance. In the real world voltage sources have current limitations that modify this behavior. A current source otoh produces a voltage that is proportional to the load impedance for a given current, whereas a voltage source produces an output voltage that is not a function of the impedance connected across it. (Like a Krell) This implies that the output voltage will vary substantially with the load impedance, most speakers are not designed to operate with such a driving source and will have an acoustical output that is a function of its source impedance and the output current rather than just the voltage applied across it.
A pentode connected amplifier output stage is usually used in conjunction with NFB to lower its source impedance to something useful with most speakers. Triodes incidentally have much lower plate resistances and so avoid this issue.
Hopefully I have not muddied the waters too much, and perhaps someone else can explain this better than I..
Thanks, Kevin. I still have a hard time grokking the impedance thing. And as for damping factor, well...
I think I understand your explanation (if I didn't, it is because I am dense
)Let me see if I have it... you are saying that because of the much higher output impedance of a pentode, it has a different set of operating conditions than a triode and must be configured to either a: a - much higher ratio OPT or b - use of specific NFB to lower the impedance inherent in the pentode (I would imagine that use of both of these in appropriate amounts could solve the problems)
Can you suggest modifications to Ado's schematic that would amend the design?
Jess
6012 ---6146---EC360
Hi Folks,
On my previous msg I planned to built 6012 SE but sadly this tube asking a verry big Current 230mA/ch on 250V since I don't have any stock of big current the Power Trafo, I decided to change the project to reBuild & Improve the EC360 SE due to this "Real triode" power tube is so cheap in my local shop(only $20 each) & Now already got the final schematic with choosen components to get the pleasure sound, my SE now ready for 4x EC360(2 tubes each Ch paralel config for higher POut) & Driving my dummy Jamo Bookshelf 86dB with easy.
For Jess, sorry on my Schematic I used 2.5k primary trafo actualy that's my test trafo ex.300B/2A3 it's not recomended, You should use the higher impedance for safe operation (4K at least, see the attached text from tubelab for detail). Also the Rk you better use 30W at least.
Thanks & regards,
Ado
6146 tubelab.com
The plate voltage is not regulated, but the screen voltage is. This is exactly what you need with these tubes.
Lacking impedance data on the transformer in the original design, you are going to need an educated guess. The plate voltage of 750 is the max for this tube with this voltage you would need an 8K ohm transformer and get about 100 to 120 watts. Hammond does not make this transformer, although it could be found. If you don't need this kind of power I would use a lower plate voltage, and get a lower power output. Your tubes will last longer.
At 600 volts you would get over 90 watts with a 6K ohm transformer and 80 watts with a 7K ohm transformer. The higher impedance transformer offers slightly lower distortion to go with the lower power.
At 450 volts you will get 60+ watts with a 4K ohm transformer, and 50 or so with a 6K transformer. Same story with the distortion.
So you need to pick a desired wattage, or a close guess. The higher power requires more money in transformers, and power supply parts. Then you can pick an output transformer.
Hammond lists 3 transformers for use with a pair of 6146's. To squeeze the most power from a modest power supply, use the 1650N (4300 ohms) run it at 500 volts or less. The 1650P (6600 ohms) would offer less power in the same amp at lower distortion. It could also operate at up to 600 volts to get more power. If you have the space and the money the 1650R (5000 ohms) would provide the most power and could provide 75 watts at 500 volts.
If this was my choice, and a first amp. I would use the 1650P and set up the power supply for something between 425 and 465 volts using 500 volt capacitors. Any higher voltages get more complicated quickly. You could expect 50 watts or more from this combination.
Hammond only makes one transformer that has the power capability to adequately support both channels drivem to full power. That is the 278CX. It does not have 5 volt windings. You could use a SS rectifier, or extra transformers. The voltage might be over 500 with SS, you would need HV caps.
If you don't plan to run the amp at full volume a lot (I don't think so) you could use a smaller transformer. The other possibility is a seperate power supply for each channel.
Hi Folks,
On my previous msg I planned to built 6012 SE but sadly this tube asking a verry big Current 230mA/ch on 250V since I don't have any stock of big current the Power Trafo, I decided to change the project to reBuild & Improve the EC360 SE due to this "Real triode" power tube is so cheap in my local shop(only $20 each) & Now already got the final schematic with choosen components to get the pleasure sound, my SE now ready for 4x EC360(2 tubes each Ch paralel config for higher POut) & Driving my dummy Jamo Bookshelf 86dB with easy.
For Jess, sorry on my Schematic I used 2.5k primary trafo actualy that's my test trafo ex.300B/2A3 it's not recomended, You should use the higher impedance for safe operation (4K at least, see the attached text from tubelab for detail). Also the Rk you better use 30W at least.
Thanks & regards,
Ado
6146 tubelab.com
The plate voltage is not regulated, but the screen voltage is. This is exactly what you need with these tubes.
Lacking impedance data on the transformer in the original design, you are going to need an educated guess. The plate voltage of 750 is the max for this tube with this voltage you would need an 8K ohm transformer and get about 100 to 120 watts. Hammond does not make this transformer, although it could be found. If you don't need this kind of power I would use a lower plate voltage, and get a lower power output. Your tubes will last longer.
At 600 volts you would get over 90 watts with a 6K ohm transformer and 80 watts with a 7K ohm transformer. The higher impedance transformer offers slightly lower distortion to go with the lower power.
At 450 volts you will get 60+ watts with a 4K ohm transformer, and 50 or so with a 6K transformer. Same story with the distortion.
So you need to pick a desired wattage, or a close guess. The higher power requires more money in transformers, and power supply parts. Then you can pick an output transformer.
Hammond lists 3 transformers for use with a pair of 6146's. To squeeze the most power from a modest power supply, use the 1650N (4300 ohms) run it at 500 volts or less. The 1650P (6600 ohms) would offer less power in the same amp at lower distortion. It could also operate at up to 600 volts to get more power. If you have the space and the money the 1650R (5000 ohms) would provide the most power and could provide 75 watts at 500 volts.
If this was my choice, and a first amp. I would use the 1650P and set up the power supply for something between 425 and 465 volts using 500 volt capacitors. Any higher voltages get more complicated quickly. You could expect 50 watts or more from this combination.
Hammond only makes one transformer that has the power capability to adequately support both channels drivem to full power. That is the 278CX. It does not have 5 volt windings. You could use a SS rectifier, or extra transformers. The voltage might be over 500 with SS, you would need HV caps.
If you don't plan to run the amp at full volume a lot (I don't think so) you could use a smaller transformer. The other possibility is a seperate power supply for each channel.
6146 is a gorgious tube
Use it in PPP with a 3.2k opt, 600V on the anode, groundreferenced 345-350V regulated on g2 (50%Ultralinear) , around -80V on g1 (adjust to a Ia of 26mA) and you will get 35W of EXCEPTIONALLY CLEAN high efficience output!!!
Offcourse there is a downside, needs around 320Vp input on the grids, floating low capacitance powersupplies, heatertransformers a.s.o...
In this configuration really hard to beat in ANY RESPECT WITH ANY OTHER TUBE!
Use it in PPP with a 3.2k opt, 600V on the anode, groundreferenced 345-350V regulated on g2 (50%Ultralinear) , around -80V on g1 (adjust to a Ia of 26mA) and you will get 35W of EXCEPTIONALLY CLEAN high efficience output!!!
Offcourse there is a downside, needs around 320Vp input on the grids, floating low capacitance powersupplies, heatertransformers a.s.o...
In this configuration really hard to beat in ANY RESPECT WITH ANY OTHER TUBE!
Hi JesseG
Sorry I have no scanner, but it is simply a kind of "circlotron" with the opt in 50% ultra-linear connection by connecting g2 of the tubes to a regulated voltage-source with 340-350V. It also needs a regulated, seperately adjustable g1 "fixed bias" source of -78V +-20%. The driver I used was my "universal-driver" wich can swing almost +-400V with 1kV Ub. But used 4 tubes in cascade working into a kind of White cathode-follower with also 4 cascaded tubes to drive finally 8 paralelled 6146 per side working into a 420ohm autotransformer with a load-depending voltage stepdown of either 7:1 or 10:1 (6 layers in serial with 6x1.5 layer in serial with 6x1.5 layers in parlell for 8ohm load or 9 layers in serial with 9 layers in paralell for 4ohm. A rather expensive driver because it had a total of 4xE288CC and 4x6BX7.
This driver is certainly overkill if you use only a pair of tubes on the output and settle for higher distortion you can use a simple much simple cascade with anode-resistor loading, or even only a selfbiased 6BX7 running at about 400V on the anode and resistor loaded (but the distortion of the driver will be higher than that of the outputstage)
It all depends on the goals you set, mine was lozza power with good efficience at the lowest possible distortion, complexity or cost of the circuit where no object, neither was "easy building" or "safety" like there would in commercial projects.
Sorry I have no scanner, but it is simply a kind of "circlotron" with the opt in 50% ultra-linear connection by connecting g2 of the tubes to a regulated voltage-source with 340-350V. It also needs a regulated, seperately adjustable g1 "fixed bias" source of -78V +-20%. The driver I used was my "universal-driver" wich can swing almost +-400V with 1kV Ub. But used 4 tubes in cascade working into a kind of White cathode-follower with also 4 cascaded tubes to drive finally 8 paralelled 6146 per side working into a 420ohm autotransformer with a load-depending voltage stepdown of either 7:1 or 10:1 (6 layers in serial with 6x1.5 layer in serial with 6x1.5 layers in parlell for 8ohm load or 9 layers in serial with 9 layers in paralell for 4ohm. A rather expensive driver because it had a total of 4xE288CC and 4x6BX7.
This driver is certainly overkill if you use only a pair of tubes on the output and settle for higher distortion you can use a simple much simple cascade with anode-resistor loading, or even only a selfbiased 6BX7 running at about 400V on the anode and resistor loaded (but the distortion of the driver will be higher than that of the outputstage)
It all depends on the goals you set, mine was lozza power with good efficience at the lowest possible distortion, complexity or cost of the circuit where no object, neither was "easy building" or "safety" like there would in commercial projects.
I sure wish I knew enought about this to understand what you are saying. I think get the part about the g2 power supply.
I don't get the part about the 50% ultra-linear. I thought that ultra-linear was done with a 45% tap on the opt primary. I think I will need a diagram to understand this.
So, these tubes need a very high voltage driver stage to operate well?
Jess
hi jesse
In a circlotron half (50%) of the load is in the anodecircuit and the other 50% in the cahode circuit. Look at g2 (g2 works really like a "quasi-triodes" anode) referenced to the cathode. Can you see that this "quasi-triode works as cathode follower? Thats why this
circuit needs high drive voltage.
But you also get a cathodefollowers low distortion and low output impedance + almost double power and almost doubled efficience because the high impedance anodecurrent is added to the low impedance cathode cicuit.
Further more, the "quasi-anode" g2 needs almost no power (at comparable lower voltage) and can be easely stabilized, giving you almost the same benefits than a fully stabilized power supply.
As to the sometimes stated 43 (or what ever) % connection for UL, this is always stated for a specific tube under specific conditions. In this amplifier, the g2 voltage and load impedance is choosen to give the best results with the naturally occuring 50% connection. Not all amplifiers have the freedom of choosing the g2 voltage and therfore they have to jiggle the tap on the opt. Anyway, in this amplfier the anode does NOT have to swing below the g2 voltage (like it often does in amplifiers with Ua=Ug2). A further advantage. Did I mention the opt?
1 quarter impedance, no cross-over distortion possible, dc loaded only by Ig2 (wich gets slightly LOWER at full load).
NO?
This is the to my knowledge the best topology for a beam-tetrode pp amplifier. Some migth think that a lower voltage version could be improved by turning it into a full CF by referencing the drive voltages to the anodes. I would have done that if it would bring anything worth the higher drive voltages, but it is not worth the effort. In a OTL things would be different, I wish I had a dozen of 13E1 at hand, lol
Well, some day I´ll try those E130L with a low impedance ratio autotransformer...who knows, migth be just as good and would not require 600V on the anodes, 300V should do...
In a circlotron half (50%) of the load is in the anodecircuit and the other 50% in the cahode circuit. Look at g2 (g2 works really like a "quasi-triodes" anode) referenced to the cathode. Can you see that this "quasi-triode works as cathode follower? Thats why this
circuit needs high drive voltage.
But you also get a cathodefollowers low distortion and low output impedance + almost double power and almost doubled efficience because the high impedance anodecurrent is added to the low impedance cathode cicuit.
Further more, the "quasi-anode" g2 needs almost no power (at comparable lower voltage) and can be easely stabilized, giving you almost the same benefits than a fully stabilized power supply.
As to the sometimes stated 43 (or what ever) % connection for UL, this is always stated for a specific tube under specific conditions. In this amplifier, the g2 voltage and load impedance is choosen to give the best results with the naturally occuring 50% connection. Not all amplifiers have the freedom of choosing the g2 voltage and therfore they have to jiggle the tap on the opt. Anyway, in this amplfier the anode does NOT have to swing below the g2 voltage (like it often does in amplifiers with Ua=Ug2). A further advantage. Did I mention the opt?
1 quarter impedance, no cross-over distortion possible, dc loaded only by Ig2 (wich gets slightly LOWER at full load).
NO?
This is the to my knowledge the best topology for a beam-tetrode pp amplifier. Some migth think that a lower voltage version could be improved by turning it into a full CF by referencing the drive voltages to the anodes. I would have done that if it would bring anything worth the higher drive voltages, but it is not worth the effort. In a OTL things would be different, I wish I had a dozen of 13E1 at hand, lol
Well, some day I´ll try those E130L with a low impedance ratio autotransformer...who knows, migth be just as good and would not require 600V on the anodes, 300V should do...
Ok, this is good learning for me.
I will try drawing this up, per your description, and post it when I am done and maybe you can tell me where I ha gotten it wrong
Please don't hold your breath - I am very busy with paying work right now. But I WILL do this. I am really interested in building a 6146 amp.
ttfn
Jess
I will try drawing this up, per your description, and post it when I am done and maybe you can tell me where I ha gotten it wrong
Please don't hold your breath - I am very busy with paying work right now. But I WILL do this. I am really interested in building a 6146 amp.
ttfn
Jess
JesseG said:
Wait a minute, how cn you have both ultralinear operation (screengrids being run off of a tap on the OT) AND a regulated supply to the screen grids? Just a word of warning, I have been flying around the world and I'm in a bit of a haze, so my apologies if I've made any really stupid comments
Isaac
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.