http://www.diyaudio.com/forums/tubes-valves/10336-my-6sl7-2a3-se-design.html
Let me first start by saying I just read the above thread....informative yet sad with all the bickering......
I recently sourced all the parts for the Hagtech Clarion 2A3 and now I am standing at the precipice and having second thoughts. I am thinking I want an integrated with enough gain to use without a preamp.
I recently began to review the schematics readily available via a Google search and have come up with several other options...I just have a hard time choosing #1, and also I have to make some fairly significant adaptations to the power supply to make the parts I have on hand work with these other schematics.
My other thought is to just build a design from scratch with the help of a few of you brainiacs here on the forum.
I have a Hammond 700VCT 200mA and an Edcor 750VCT 200mA transformers. I have a pair of James 6112HS OPTs as well. I have a bunch of different tubes (6SL7, 6SN7, 5687, 6922, 71, 27, 56, 12AX7, 12AU7, 6SJ7, and a few others that might work I can't remember now.
On the post above there was some consideration of using an ECC99 as a driver tube but never a schematic. I don't have these tubes but they are cheap and easy to obtain.
I have seriously considered the Time Bandit, JEL 2A3dx, and Sashiudo 12AX7 designs. I am now questioning the viability of the 6SL7 and 12AX7 as sole drivers of the 2A3 though without a SRPP or other configuration.
I geuss where I am going with this is 1st, is there another good 2 stage design out there I could easily adapt to my parts? And second, I am open to the idea of a "from scratch" design with assistance from this forum.
I will be driving a pair of Proac 2.5 Response (91-92dB) and will be building some Sachiko horns this month.
Last consideration and probably the easiest given I have all the parts for the Hagtech.....what about just building the Hagtech and adding a stage before the 5687 (the Hagtech takes 4V to drive to clipping) http://www.hagtech.com/pdf/clarionarticle.pdf
Maybe adding another 5687 "preamp" on the chassis, something like the "Simple 5687 Preamp" at Diyparadise.com. What would I need to be aware of in doing this? Is there an easy one tube circuit I could simply add in front of the 5687, I am not sure of the setbacks in trying this. Seems that If I need a preamp to begin with to drive the Hagtech then I should be able to just build it on to the amp. diyparadise - Simple 5687-based preamp
I look forward to the responses! BTW, this will be my first DHT so I want to keep it as simple as possible.
Thanks,
Jeff
Let me first start by saying I just read the above thread....informative yet sad with all the bickering......
I recently sourced all the parts for the Hagtech Clarion 2A3 and now I am standing at the precipice and having second thoughts. I am thinking I want an integrated with enough gain to use without a preamp.
I recently began to review the schematics readily available via a Google search and have come up with several other options...I just have a hard time choosing #1, and also I have to make some fairly significant adaptations to the power supply to make the parts I have on hand work with these other schematics.
My other thought is to just build a design from scratch with the help of a few of you brainiacs here on the forum.
I have a Hammond 700VCT 200mA and an Edcor 750VCT 200mA transformers. I have a pair of James 6112HS OPTs as well. I have a bunch of different tubes (6SL7, 6SN7, 5687, 6922, 71, 27, 56, 12AX7, 12AU7, 6SJ7, and a few others that might work I can't remember now.
On the post above there was some consideration of using an ECC99 as a driver tube but never a schematic. I don't have these tubes but they are cheap and easy to obtain.
I have seriously considered the Time Bandit, JEL 2A3dx, and Sashiudo 12AX7 designs. I am now questioning the viability of the 6SL7 and 12AX7 as sole drivers of the 2A3 though without a SRPP or other configuration.
I geuss where I am going with this is 1st, is there another good 2 stage design out there I could easily adapt to my parts? And second, I am open to the idea of a "from scratch" design with assistance from this forum.
I will be driving a pair of Proac 2.5 Response (91-92dB) and will be building some Sachiko horns this month.
Last consideration and probably the easiest given I have all the parts for the Hagtech.....what about just building the Hagtech and adding a stage before the 5687 (the Hagtech takes 4V to drive to clipping) http://www.hagtech.com/pdf/clarionarticle.pdf
Maybe adding another 5687 "preamp" on the chassis, something like the "Simple 5687 Preamp" at Diyparadise.com. What would I need to be aware of in doing this? Is there an easy one tube circuit I could simply add in front of the 5687, I am not sure of the setbacks in trying this. Seems that If I need a preamp to begin with to drive the Hagtech then I should be able to just build it on to the amp. diyparadise - Simple 5687-based preamp
I look forward to the responses! BTW, this will be my first DHT so I want to keep it as simple as possible.
Thanks,
Jeff
Jeff,
A common cathode 6SL7 section lacks the "stones" to drive a DHT's Miller capacitance.
I think you hit on a good idea in mentioning SRPP. A 6SL7 SRPP driving an IRFBC20 source follower DC coupled to the 2A3's grid will give you plenty of gain and also allow a mild Class "A2" regime.
Check Tubelab's "Power Drive" out for the "skinny" on getting the DC coupling done.
BTW, because of the 6SL7's high Miller capacitance, hold the pot. at the I/P down to 50 KOhms.
A common cathode 6SL7 section lacks the "stones" to drive a DHT's Miller capacitance.
I think you hit on a good idea in mentioning SRPP. A 6SL7 SRPP driving an IRFBC20 source follower DC coupled to the 2A3's grid will give you plenty of gain and also allow a mild Class "A2" regime.
Check Tubelab's "Power Drive" out for the "skinny" on getting the DC coupling done.BTW, because of the 6SL7's high Miller capacitance, hold the pot. at the I/P down to 50 KOhms.
Last edited:
Try a single D3A (triode mode) per tube (or maybe 5842 or 6c4s-pi) and make things easier for yourself. This kind of driver has enough current and gain to drive a 2A3. Be sure to use a smaller potentiometer (or grid leak), . say 10 - 50 K to avoid treble roll off.
Examples at AudioRoundTable.com: Group Build => 300B SE Project, Part 9 - E180F, E280F, D3a pentodes - triode connected (you could use a triode load instead of a CCS, for starters)
Examples at AudioRoundTable.com: Group Build => 300B SE Project, Part 9 - E180F, E280F, D3a pentodes - triode connected (you could use a triode load instead of a CCS, for starters)
Last edited:
plate CCS
I am wondering, with what ever I decide to go with, could I use this: K & K Audio - Lundahl Transformers, audio DIY kits and more as a plate load CCS? I ask b/c I actually own 8 of them, I was planning on replacing the cathode resistors on a couple of other amps I built with these little devices.
Eli,
I appreciate the help but the Power Drive is a bit over the top for this project. I want to keep it as simple as possible for my very first DHT SE amp.
The Time Bandit 2A3 amp uses the 6SL7 in an SRPP...I am considering this amp with some modifications in the power supply to be able to use my over rated transformer (700VCT vs 600VCT called for).
My only concern is now that I have read that thread I referred to I am concerned about using the 6SL7. Does the SRPP overcome all the issues that make the 6SL7 as a cathode follower a poor choice?
What about an 12AX7 in SRPP? I have a 50-60 of these tubes, some very good ones. I only have one pair of well matched 6SL7-GTB Sylvanias. I was hoping to use them in the next project (JEL 300B dx).
I know there seems to be alot of movement towards the more exotic ideas like the power drive and stuff but for now I am still in the transition from "knowing to understanding" and I want to keep things simple, I will have, God willing, plenty of time to move on towards the more complex exotic circuits previously described.
I am wondering about this Time Bandit 2A3 amp.....any thoughts on this amp design. Think it will give me descent performance as a first time DHT SE amp?
Also, is there a way to possibly use these CCS I have mentioned on the Time Bandit to improve it's performance, be it on the plate of a driver or the cathode somewhere else? I was planning on using them in lieu of the cathode resistor on my KT88SE I recently built.
If I were to use the other driver tubes Zobsky has recommended then I will have to basically create a circuit from scratch which, I am not opposed to doing, I will certainly need the help though. Zobsky, are you aware of a 2A3 amp that uses these tubes, one that is already developed with a track record? The D3A seems pretty easy to obtain, there are several on eBay for about $13. Why wire it in triode? Doesn't that actually reduce some of the positive characteristics I am trying to obtain with this tube? Triode is certainly easier to wire. Just wondering.
I understand part of the problem with driving tubes like the 2A3 is the Miller capacitance of the tube and needing enough current from the driver (as well as about 90v swing) to overcome it. Again, what about using another stage in front of the already designed Clarion's 5687. From what I understand the 5687 has the current needed just lacks the gain. The Hagtech Clarion also does not bypass the 5687 cathode resistor, how much gain improvement would I get in doing this simple addition?
Thanks again,
Jeff
I am wondering, with what ever I decide to go with, could I use this: K & K Audio - Lundahl Transformers, audio DIY kits and more as a plate load CCS? I ask b/c I actually own 8 of them, I was planning on replacing the cathode resistors on a couple of other amps I built with these little devices.
Eli,
I appreciate the help but the Power Drive is a bit over the top for this project. I want to keep it as simple as possible for my very first DHT SE amp.
The Time Bandit 2A3 amp uses the 6SL7 in an SRPP...I am considering this amp with some modifications in the power supply to be able to use my over rated transformer (700VCT vs 600VCT called for).
My only concern is now that I have read that thread I referred to I am concerned about using the 6SL7. Does the SRPP overcome all the issues that make the 6SL7 as a cathode follower a poor choice?
What about an 12AX7 in SRPP? I have a 50-60 of these tubes, some very good ones. I only have one pair of well matched 6SL7-GTB Sylvanias. I was hoping to use them in the next project (JEL 300B dx).
I know there seems to be alot of movement towards the more exotic ideas like the power drive and stuff but for now I am still in the transition from "knowing to understanding" and I want to keep things simple, I will have, God willing, plenty of time to move on towards the more complex exotic circuits previously described.
I am wondering about this Time Bandit 2A3 amp.....any thoughts on this amp design. Think it will give me descent performance as a first time DHT SE amp?
Also, is there a way to possibly use these CCS I have mentioned on the Time Bandit to improve it's performance, be it on the plate of a driver or the cathode somewhere else? I was planning on using them in lieu of the cathode resistor on my KT88SE I recently built.
If I were to use the other driver tubes Zobsky has recommended then I will have to basically create a circuit from scratch which, I am not opposed to doing, I will certainly need the help though. Zobsky, are you aware of a 2A3 amp that uses these tubes, one that is already developed with a track record? The D3A seems pretty easy to obtain, there are several on eBay for about $13. Why wire it in triode? Doesn't that actually reduce some of the positive characteristics I am trying to obtain with this tube? Triode is certainly easier to wire. Just wondering.
I understand part of the problem with driving tubes like the 2A3 is the Miller capacitance of the tube and needing enough current from the driver (as well as about 90v swing) to overcome it. Again, what about using another stage in front of the already designed Clarion's 5687. From what I understand the 5687 has the current needed just lacks the gain. The Hagtech Clarion also does not bypass the 5687 cathode resistor, how much gain improvement would I get in doing this simple addition?
Thanks again,
Jeff
Run the D3A in pentode if you want. You'll have a whole different animal, something similar to the pentode driven "Thorsten Legacy" amp http://www.fortunecity.com/rivendell/xentar/1179/projects/legacy/Legacy.html. You will definitely have enough gain with this mode.Sound will be more dynamic and less euphonic. Try both
I was merely being true to your initial goal i.e. a simple triode DHT amp
12AX7 aren't the best tube for audio, and a DHT amp is only as good as it's weakest link.
It shouldn't be any problem to deduce a simple triode or pentode operating point for the D3A. In fact , this can be a very good learning experience. The tube has low Rp, high gain (and transconductance too, which is why a lower grid leak is needed). Look at the spec sheets, I thought they specified some approximate operating points. You would mostly just have to redesign the input stage, and not worry about the power stage
Do I know of any amps using this as a driver. Not SE, personally, but wrt push pull, I use it as the sole driver in my 2A3 push pull interstage amp, so I can say it works well. Here's another SE circuit, (PSE, no less )New 2A3 , though I would NOT use a 470K grid leak.
I'd suggest KISS and try a plate resistor first before attempting a CCS.
Hope that helped
Last edited:
Yes, helped very much. It shouldn't be to hard to build a input stage I can switch from triode to pentode. I was just curious about this D3A tube, I am not familiar with it.
I was looking at the "New 2A3" I like the simplicity of the design. What would it take to change it to a single output tube instead of a parallel tube? Just remove one of the tube circuits? Would I need to change the B+ or cathode value any if I sans one of the 2 2A3 and just make it a simple SE single 2A3 output stage? What about the grid leak with a single 2A3 stage, what would be the best value?
I could easily adapt all the parts I have to build this design sith a single output tube, no problem. PSUD2 makes this so easy.
I really like the idea of a 2 stage D3a/2A3, I would really like some help carrying this idea through to implementation, can you guys help me come up with a final schematic with the voltages needed. I can take care of the power supply myself.
Thanks alot guys,
Jeff
Jeff,
If those CCS kits are the cascode variant, they will be fine as anode loads. The single FET design, like the IXYS 10M45S, is better than a resistor, but ...
Current sinks under cathodes are used in combination with cathode followers and differential gain blocks. A constant current device does not replace a cathode bias resistor.
Yes, the 6SL7 is a wimp. It's an excellent voltage amplifier and there the tale ends. Low gm, along with a low IB operating point, makes the 6SL7 a poor cathode follower. That's why I suggested a "Power Drive" type setup. A 6SL7 SRPP has the "stones" to drive a power MOSFET's gate. The FET's "cojones" are beyond question.
An alternative 6SL7 configuration is DC coupling a ZVN0545A source follower to a CCS loaded common cathode section's anode. Even the wimpy 6SL7 can drive the tiny capacitances of the little FET. With FET ID around 3 mA., 2A3 CMiller is a non-issue.
If those CCS kits are the cascode variant, they will be fine as anode loads. The single FET design, like the IXYS 10M45S, is better than a resistor, but ...
Current sinks under cathodes are used in combination with cathode followers and differential gain blocks. A constant current device does not replace a cathode bias resistor.
Yes, the 6SL7 is a wimp. It's an excellent voltage amplifier and there the tale ends. Low gm, along with a low IB operating point, makes the 6SL7 a poor cathode follower. That's why I suggested a "Power Drive" type setup. A 6SL7 SRPP has the "stones" to drive a power MOSFET's gate. The FET's "cojones" are beyond question.
An alternative 6SL7 configuration is DC coupling a ZVN0545A source follower to a CCS loaded common cathode section's anode. Even the wimpy 6SL7 can drive the tiny capacitances of the little FET. With FET ID around 3 mA., 2A3 CMiller is a non-issue.
Magnaquest Parafeed 2a3 amp idea
I found this: MagneQuest Transformers: DIY Schematics w/ MagneQuest Iron
Basically a two stage 5687 based preamp/driver for the 2A3
Compare to this: http://www.hagtech.com/pdf/clarionarticle.pdf
Since I have ALL of the parts to build the Hagtech Clarion, why not try something like putting an extra gain stage with a 5687 in front of the 5687 tube on the Clarion. Clearly the operating point of the 5687 on the clarion is different than the second 5687 on the Lessard schematic.
Can I basically take the front end of the Lessard and RC couple it to the Clarion?
Jeff
I found this: MagneQuest Transformers: DIY Schematics w/ MagneQuest Iron
Basically a two stage 5687 based preamp/driver for the 2A3
Compare to this: http://www.hagtech.com/pdf/clarionarticle.pdf
Since I have ALL of the parts to build the Hagtech Clarion, why not try something like putting an extra gain stage with a 5687 in front of the 5687 tube on the Clarion. Clearly the operating point of the 5687 on the clarion is different than the second 5687 on the Lessard schematic.
Can I basically take the front end of the Lessard and RC couple it to the Clarion?
Jeff
From the spec sheet at http://www.mif.pg.gda.pl/homepages/frank/sheets/128/d/D3a.pdf , triode mode notes suggest an operating point of 160V B+ and a cathode resistor of 470 ohm. This can vary depending on what you are looking to achieve.
If you're unfamiliar with the calculations involved, read this diyparadise - How to "Screw Around" Your Tube Load Line
Re. adapting the PSE circuit to SE operation, think of it as unparalleling the circuit. WRT the schematic I mentioned, I don't think his 3K output transformer was optimal (it was probably the lowest impedance he could find). 2.5K to 3 K is about right for a single 2A3.
From NJ7P Tube Database Search , the classic class 2A3 A operating points are pasted below. Some like to run their tubes a bit hotter , for more power, others cooler for longer life and,as they say, a less strained sound.
Characteristics and Typical Operation
Class A Amplifier
Plate Voltage ................................. 250 V
Grid No. 1 Voltage ............................ -45 V
Amplification Factor .......................... 4.2
Plate Resistance (approx) ..................... 800 Ω
Transconductance .............................. 5250 µ
Plate Current ................................. 60 mA
Load Resistance ............................... 2.5K Ω
Power Output (approx) ......................... 3.5 W
Characteristics and Typical Operation
Class A Amplifier
Plate Voltage ................................. 250 V
Grid No. 1 Voltage Derived from
Cathode Bias Resistor ....................... 750 Ω
Amplification Factor .......................... 4.2
Plate Resistance (approx) ..................... 800 Ω
Transconductance .............................. 5250 µ
Plate Current ................................. 60 mA
Load Resistance ............................... 2.5K Ω
Power Output (approx) ......................... 3.5 W
Re. the grid leak of the 2A3, it sets the input impedance of the power stage. The rule of thumb is to have this value equal to 10 or more times the previous stage's output impedance. The more the better, to a point (where miller capacitance losses start rolling off the high frequency). I'm sorry I dont have the calculations handy. For example , if D3A plate resistance is 2K, and you use a 5K plate resistor, the output impedance (parallel these two ) = 1.4K ohm .
This has to drive a the power stage. Theoretically, by my rule of thumb, we should be able to get by with 14KOhm. But remember, the coupling cap and the power stage grid leak form a high pass filter. Because of this, it makes sense to bump up the grid leak somewhat. Bottom line, start with something like 0.22uF, 470K, and alter from there. Hopefully, I havent forgotten something in my hasty typing.
Also, re the power drive, it is a good way to go, but not necessary as a first step. Consider leaving some space in your chassis if you want to add this in later. It is probably vital if your amp is very low power as in my old 12b4A 1.5 W amp where I ended up with blocking distortion (but thats another story). Bottom line, you should be fine with a basic 2A3 if your speakers are fairly efficient (mid to high 90s)
If you're unfamiliar with the calculations involved, read this diyparadise - How to "Screw Around" Your Tube Load Line
Re. adapting the PSE circuit to SE operation, think of it as unparalleling the circuit. WRT the schematic I mentioned, I don't think his 3K output transformer was optimal (it was probably the lowest impedance he could find). 2.5K to 3 K is about right for a single 2A3.
From NJ7P Tube Database Search , the classic class 2A3 A operating points are pasted below. Some like to run their tubes a bit hotter , for more power, others cooler for longer life and,as they say, a less strained sound.
Characteristics and Typical Operation
Class A Amplifier
Plate Voltage ................................. 250 V
Grid No. 1 Voltage ............................ -45 V
Amplification Factor .......................... 4.2
Plate Resistance (approx) ..................... 800 Ω
Transconductance .............................. 5250 µ
An externally hosted image should be here but it was not working when we last tested it.
Plate Current ................................. 60 mA
Load Resistance ............................... 2.5K Ω
Power Output (approx) ......................... 3.5 W
Characteristics and Typical Operation
Class A Amplifier
Plate Voltage ................................. 250 V
Grid No. 1 Voltage Derived from
Cathode Bias Resistor ....................... 750 Ω
Amplification Factor .......................... 4.2
Plate Resistance (approx) ..................... 800 Ω
Transconductance .............................. 5250 µ
An externally hosted image should be here but it was not working when we last tested it.
Plate Current ................................. 60 mA
Load Resistance ............................... 2.5K Ω
Power Output (approx) ......................... 3.5 W
Re. the grid leak of the 2A3, it sets the input impedance of the power stage. The rule of thumb is to have this value equal to 10 or more times the previous stage's output impedance. The more the better, to a point (where miller capacitance losses start rolling off the high frequency). I'm sorry I dont have the calculations handy. For example , if D3A plate resistance is 2K, and you use a 5K plate resistor, the output impedance (parallel these two ) = 1.4K ohm .
This has to drive a the power stage. Theoretically, by my rule of thumb, we should be able to get by with 14KOhm. But remember, the coupling cap and the power stage grid leak form a high pass filter. Because of this, it makes sense to bump up the grid leak somewhat. Bottom line, start with something like 0.22uF, 470K, and alter from there. Hopefully, I havent forgotten something in my hasty typing.
Also, re the power drive, it is a good way to go, but not necessary as a first step. Consider leaving some space in your chassis if you want to add this in later. It is probably vital if your amp is very low power as in my old 12b4A 1.5 W amp where I ended up with blocking distortion (but thats another story). Bottom line, you should be fine with a basic 2A3 if your speakers are fairly efficient (mid to high 90s)
Last edited:
This is very true for the potmeter, but in case of a grid leak, there's no problem, because the grid leak resister is parallel to the output impedance (which should be smaller than 10 Kohm) of the source.
Problem with the potentiometer is that at low volumes, a large part of the resistance is in series with the grid of the tube. And the grid-anode capacitance is amplified by mu of the d3a in triode, which is 70. This comes to almost 20pF, which in combination with the 100k series-resister forms a low pass filter in the audio band.
But other than that, the d3a in triode is a very fine driver. With a css on top, it is capable of very large V swing, with very low distortion.
Look Here: High Gm driver pentodes
You can see that the d3a is a real winner: 0.04 % D at 25v pp, and 265v max swing at 5%. Plus the low output impedance....
Well, I just bid on 4 D3a tubes on eBay, just in case I go this route.
I would like to keep everything from the Hagtech clarion the same for the output stage (i.e., I have already sourced the resistors, caps, etc..). I guess what I need to know is can I use the output stage from the Clarion Hagtech as is and just substitute this D3a circuit for the ontended 5687 tube? The B+ for the driver (5687) is at 250v already, it wouldn't take but a resistor in the power supply to drop this to around 160v for triode operation of the D3a. I guess what I need to know is what shoud I set the plate CCS mA to (I plan to use one, I have the K&K audio CCS on hand), should I use a resistor of any kind on the grid of the D3a, what about a grid leak for the D3a (need one?), also, the Hagtech uses a 33K to set the bias for the 2A3, does this need any modification?
For the triode strapping, do I need to place a resistor from g2 to the plate? Also, is g3 already connected to the cathode internally or is it not? (the sheet I faound looks like g3 is left to pin 8 and not connected internally).
Should I just start with a 470 cathode resistor and go from there? Bypass the cathode resistor?
I know it is alot of questions, but I don't want to try to figure this out on my own, I am not that knowledgable yet.
Thanks guys,
Jeff
I would like to keep everything from the Hagtech clarion the same for the output stage (i.e., I have already sourced the resistors, caps, etc..). I guess what I need to know is can I use the output stage from the Clarion Hagtech as is and just substitute this D3a circuit for the ontended 5687 tube? The B+ for the driver (5687) is at 250v already, it wouldn't take but a resistor in the power supply to drop this to around 160v for triode operation of the D3a. I guess what I need to know is what shoud I set the plate CCS mA to (I plan to use one, I have the K&K audio CCS on hand), should I use a resistor of any kind on the grid of the D3a, what about a grid leak for the D3a (need one?), also, the Hagtech uses a 33K to set the bias for the 2A3, does this need any modification?
For the triode strapping, do I need to place a resistor from g2 to the plate? Also, is g3 already connected to the cathode internally or is it not? (the sheet I faound looks like g3 is left to pin 8 and not connected internally).
Should I just start with a 470 cathode resistor and go from there? Bypass the cathode resistor?
I know it is alot of questions, but I don't want to try to figure this out on my own, I am not that knowledgable yet.
Thanks guys,
Jeff
I was reviewing all you said about the D3a tube and looked through the "guide line" sheets. You mentioned the grid leak being about 10X or more of the output impedance. The original design used a 33K, by your estimate I shold use at least a 28K....what is the drawback of using a larger than needed grid leak? Would keeping the 33K be acceptable?
From everything I have read and what you have said this
seems pretty straightforward.
The datasheet shows a plate current of 24mA with a voltage of 160. If I were to use a plate load resistor instead of the CCS, what value would I need to use?
Jeff
From everything I have read and what you have said this
seems pretty straightforward.
The datasheet shows a plate current of 24mA with a voltage of 160. If I were to use a plate load resistor instead of the CCS, what value would I need to use?
Jeff
I went back and reviewed the "how to screw around with your load lines" and it suggess the way they picked the plate load resistor was to use a multiple of x5 the internal resisance of the tube. For the D3a it is 1.9K so, I guess a 10K plate load resistor would fit the bill. The value x5 would be 9500, I am assuming the rounding up to 10k is of little significance, correct?
What about a grid leak on the grid of the D3a?
Correction, the current through the D3a is 22mA
Jeff
What about a grid leak on the grid of the D3a?
Correction, the current through the D3a is 22mA
Jeff
No, you dont need any resistors other than the setting resistors for the ccs (probably a potentiometer) Just hook one end of the ccs to the 250v and the other to the anode.
The current should be about 20mA.
I would use a 1k grid stopper on the d3a. If you use a potentiometer (47k or less); thats you grid leak resistor. If you use it as power amp (without potmeter) with a separate pre amp, use a 100k resistor from grid to ground as grid leak.
On my d3a's I didnt use additional resistors for wiring it as triode, just connected all the wires. Worked perfect, without instability or oscillating. And you have to connect them all; g2 to anode and g3 to kathode.
The cathode resistor should be about 75 ohm; 82 ohms should do the trick.
If you look at your d3a triode curves, you see that the 20mA line, and the 160v line crosses the -1.5v curve. So the bias needs to be about 1.5 v. 1.5v / 20 ma= 75 ohms.
You can bypass the cathoderesistor; I did not. If you do it, output impedance is lower, and distortion a little higher. (in my case i didnt need the extra low output impedance, so I went for the lowest possible distortion. Plus i didnt want an extra capacitor)
The cathode resistor acts as local feedback, thats why the distortion is lower if you dont bypass it. If you dont bypass it, it also gets multiplied by mu; and added to the output impedance. ( 70x82 ohms=5740 ohms extra)
The current should be about 20mA.
I would use a 1k grid stopper on the d3a. If you use a potentiometer (47k or less); thats you grid leak resistor. If you use it as power amp (without potmeter) with a separate pre amp, use a 100k resistor from grid to ground as grid leak.
On my d3a's I didnt use additional resistors for wiring it as triode, just connected all the wires. Worked perfect, without instability or oscillating. And you have to connect them all; g2 to anode and g3 to kathode.
The cathode resistor should be about 75 ohm; 82 ohms should do the trick.
If you look at your d3a triode curves, you see that the 20mA line, and the 160v line crosses the -1.5v curve. So the bias needs to be about 1.5 v. 1.5v / 20 ma= 75 ohms.
You can bypass the cathoderesistor; I did not. If you do it, output impedance is lower, and distortion a little higher. (in my case i didnt need the extra low output impedance, so I went for the lowest possible distortion. Plus i didnt want an extra capacitor)
The cathode resistor acts as local feedback, thats why the distortion is lower if you dont bypass it. If you dont bypass it, it also gets multiplied by mu; and added to the output impedance. ( 70x82 ohms=5740 ohms extra)
posting like hell....
Sorry,
I get carried away...I wish I had all day to devote to learning this stuff. This is such a great resource but sometimes I forget that it's just other people who are so kind to volunteer all this info......
So, what is the 470R cathode resistor value on the spec sheet represent?
Any idea of the gain I will get with this set up?
I am really excitited to find that I am going to be able to achieve what I wanted with minimal changes to my circuit. It will be very simple to change this back to the 5687 too if I don't like the results.
I am wondering, for the original design using the 5687 tube.....if I were to keep that tube and just build it like the designer did, what would you recommend setting the CCS at for the plate load on the 5687 and sans the 10K plate load resistor? The data sheet shows 12mA at 250v with grid at -12.5. When I draw a straight horizontal line from 0v to 250v across the 12ma horizon the characteristics appear very linear. When you draw the line from 250v/0ma to 0v 12mA it doesn't look near as good. If I am understanding a plate load using a CCS this is the difference in the linearity. Am I correct in this assumption? If so, I can sure see whay a CCS as a plate supply is much better thatn a resistor.
thanks,
Jeff
i'm sorry my previous post was an answer to 3 posts ago; your posting like hell
Sorry,
I get carried away...I wish I had all day to devote to learning this stuff. This is such a great resource but sometimes I forget that it's just other people who are so kind to volunteer all this info......
So, what is the 470R cathode resistor value on the spec sheet represent?
Any idea of the gain I will get with this set up?
I am really excitited to find that I am going to be able to achieve what I wanted with minimal changes to my circuit. It will be very simple to change this back to the 5687 too if I don't like the results.
I am wondering, for the original design using the 5687 tube.....if I were to keep that tube and just build it like the designer did, what would you recommend setting the CCS at for the plate load on the 5687 and sans the 10K plate load resistor? The data sheet shows 12mA at 250v with grid at -12.5. When I draw a straight horizontal line from 0v to 250v across the 12ma horizon the characteristics appear very linear. When you draw the line from 250v/0ma to 0v 12mA it doesn't look near as good. If I am understanding a plate load using a CCS this is the difference in the linearity. Am I correct in this assumption? If so, I can sure see whay a CCS as a plate supply is much better thatn a resistor.
thanks,
Jeff
Looks like I screwed up, the mA at zero volts should be about 26mA not 12mA as I stated earlier. This is what diyparadise's article says at least. Still the linearity is better than the line drawn from the 0mA/250v to 26mA/0v points. So, in my previous statement I said I thought I should set the CCS to 12mA....should it be set to this, it appears to be at the midpoint of the swing zone, where a grid of -6v would normally be chosen.
Jeff
Jeff
Yes, the horizontal load line kicks serious butt. However, there's more to the story. A CCS as the anode load presents a very high dynamic AC impedance. Therefore, stage gain can closely approach the full μ of the triode. If a good CCS is used and DC coupling to a tube's control grid or a FET's gate is employed, the fraction of the full μ is so large that it appears to be 100%.
Correct. The relatively non linear approach using a plate resistor is mostly euphonic even order harmonics. Using a CCS or inductive load on a triode's plate presents such a large impedance to the tube that the load line is near horizontal, pivoted into a linear region
FWIW, this is also why pentodes make decent CCS, but thats another story
Last edited:
Eli,
So, what you are saying is by adding a good CCS to the plate supply this will in fact not only increase linearity but will also increase the gain? Well then, I wonder if I just put the CCS on the original 5687 stage of the Clarion if that would provide the gain I am looking for (about 2v sens)? Boy, that would simplify things a bit.
I am using these K&K Audio CCS by default at this time, simply b/c I bought several to try out and they were very cheap ($8). I will try these out and see how they sound. I would like to know of the "best" CCS that is out there. I know there is going to be some differing opinions of the "Best" but I am interested in your idea of this. I plan to start implementing these types of things in future projects so I always want to know what is the best way to do things, even if I choose not to do so, I don't like being ignorant.
BTW, I am turning 41 in 2 days. Until 2 yrs ago I didn't even know what a soldering iron looked like. I am late into this hobby by many of ya'lls standards and still have a lifetime of learning ahead of me. If it wasn't for these forums I would still be clueless. I just wanted to thank all of you for the time you have invested in these forums and answering so many questions. Sometimes I think I couldn't have paid for a better education. Thanks again for all of you guys help, it is very much appreciated. I look forward to the day when I can offer the advice many of you do. I also look forward to the day that I can be perusing a tube manual and see something interesting about a tube and from scratch just put together a working design like most of you can. I have a long ways to go but I am getting there thanks to ya'll.
My favorite part of all this is actually building the chassis and amps out of wood and metal by hand. I love the artistic and techical fusion this hobby provides. I wish I could figure out a way to make a living at this, I would take a pay cut and leave medicine for good if I could. This hobby keeps me sane I think at times. When I have had the SH^&##% of SH^&## days at work I just long to return to my garage and tinker on these amps.....isn't this the best hobby ever?
Well, at this point I am seriously trying to decide between swapping the front end of this amp with the Lessard vs. the D3a vs. just leaving it be and maybe add a CCS on the plate and possibly a cathode bypass cap to the original design. Nice thing is I will leave enough space underneath and up top to make all sorts of adjustments down the road. I have built the Chassis 20" wide by about 10" deep and 3.5" tall. I can't wait to post some pics of the chassis when I have it glued together, this is one of the better ones I have put together, very cool looking Spanish Modena and white Tiger Maple...I am very proud of it so far.
Thanks again!
Jeff
So, what you are saying is by adding a good CCS to the plate supply this will in fact not only increase linearity but will also increase the gain? Well then, I wonder if I just put the CCS on the original 5687 stage of the Clarion if that would provide the gain I am looking for (about 2v sens)? Boy, that would simplify things a bit.
I am using these K&K Audio CCS by default at this time, simply b/c I bought several to try out and they were very cheap ($8). I will try these out and see how they sound. I would like to know of the "best" CCS that is out there. I know there is going to be some differing opinions of the "Best" but I am interested in your idea of this. I plan to start implementing these types of things in future projects so I always want to know what is the best way to do things, even if I choose not to do so, I don't like being ignorant.
BTW, I am turning 41 in 2 days. Until 2 yrs ago I didn't even know what a soldering iron looked like. I am late into this hobby by many of ya'lls standards and still have a lifetime of learning ahead of me. If it wasn't for these forums I would still be clueless. I just wanted to thank all of you for the time you have invested in these forums and answering so many questions. Sometimes I think I couldn't have paid for a better education. Thanks again for all of you guys help, it is very much appreciated. I look forward to the day when I can offer the advice many of you do. I also look forward to the day that I can be perusing a tube manual and see something interesting about a tube and from scratch just put together a working design like most of you can. I have a long ways to go but I am getting there thanks to ya'll.
My favorite part of all this is actually building the chassis and amps out of wood and metal by hand. I love the artistic and techical fusion this hobby provides. I wish I could figure out a way to make a living at this, I would take a pay cut and leave medicine for good if I could. This hobby keeps me sane I think at times. When I have had the SH^&##% of SH^&## days at work I just long to return to my garage and tinker on these amps.....isn't this the best hobby ever?
Well, at this point I am seriously trying to decide between swapping the front end of this amp with the Lessard vs. the D3a vs. just leaving it be and maybe add a CCS on the plate and possibly a cathode bypass cap to the original design. Nice thing is I will leave enough space underneath and up top to make all sorts of adjustments down the road. I have built the Chassis 20" wide by about 10" deep and 3.5" tall. I can't wait to post some pics of the chassis when I have it glued together, this is one of the better ones I have put together, very cool looking Spanish Modena and white Tiger Maple...I am very proud of it so far.
Thanks again!
Jeff
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.