Hi, I recently built the lampizator srpp output stage for my voltage out DAC. It is great! It runs of a 110v 30 ma secondary from a toroid.Then I added a 40H 30ma 700ohm choke, WOW! Hence the user name. I am about to build a markII. This time I want to bias the anode with 200v dc. I have a toroid with two 160volt 30ma secondary windings. I am wanting to give each anode its own winding. I have a few questions I am hoping some one could take pity on me and help me out.
1) The anode should recieve 10ma. But what in the circuit, determines the current the anode receives?Is it the anode itself? May be the 200ohm resistor between the upper cathode and lower anode of the double triode?The 200ohm resistor linking the lower cathode to ground?I ask because I am worried if I up the bias voltage i might up the current?
2) I have calculated if I use a 160 volt ac secondary, rectified to 223 volt dc and have a clclc filter to the anode. A further 1000ohms must be added to bring the voltage down to 200 volts AS there are 1400 ohms of resistance in the two chokes of the clclc filter. Is this right???? and where is the best place to add the resistor ???
3) I dont undersand why the lampizator uses only 0.25watt resistors in the anode circuit. Obviously it works :I am listening to music right now from them.If he wants to use 10ma and uses a 150 volts dc, then surely the resistors should be 150 multiplied by 0.01 which equals 1.5 watt??In a similar vein if I up the voltage to 200, i will need 2 watt, i.e 3 watt (to be safe) resistors?I have obviously mis understood something.
I realise there are some very clever people out there and I apologise if my questions are inane, I am just learning.
1) The anode should recieve 10ma. But what in the circuit, determines the current the anode receives?Is it the anode itself? May be the 200ohm resistor between the upper cathode and lower anode of the double triode?The 200ohm resistor linking the lower cathode to ground?I ask because I am worried if I up the bias voltage i might up the current?
2) I have calculated if I use a 160 volt ac secondary, rectified to 223 volt dc and have a clclc filter to the anode. A further 1000ohms must be added to bring the voltage down to 200 volts AS there are 1400 ohms of resistance in the two chokes of the clclc filter. Is this right???? and where is the best place to add the resistor ???
3) I dont undersand why the lampizator uses only 0.25watt resistors in the anode circuit. Obviously it works :I am listening to music right now from them.If he wants to use 10ma and uses a 150 volts dc, then surely the resistors should be 150 multiplied by 0.01 which equals 1.5 watt??In a similar vein if I up the voltage to 200, i will need 2 watt, i.e 3 watt (to be safe) resistors?I have obviously mis understood something.
I realise there are some very clever people out there and I apologise if my questions are inane, I am just learning.
Hi Chokes rule,
Don’t worry I’m still learning as well
1) It’s the cathode resistor that determines the plate current. With SRPP the voltage is divided between both tubes, i.e. if you have 200V on the top tube’s anode you will have 100V on the bottom tube. To calculate your cathode resistor value, just choose your operating point using the curves and divide the grid voltage by the current at that point. The current runs through both tubes. At 10mA sounds like you’re using the 6n6p.
2) I would actually download power supply designer from the Duncan’s amps website and use it to model your power supply. Use a ccs as the tube load and set it to 10mA
Software
you can add an RC section to the CLCL if needed, so CLCLRC
3) You can also calculate power dissipation by P = (I^2) R. Current squared x resistance. So for 10ma (0.01 x 0.01) x 200 ohms = 0.02W.
I’m using the 6n2p at the moment. Using an EI core transformer acts as a bit of a filter, since they have a lower frequency bandwidth compared to a torroid, but that is just my preference. Also I like valve rectifiers, you may like to give a low current one a try. The 6BW4 is cheap and easy to implement with a 6.3V heater.
The lampizator circuit is a good start but there are better sounding valve output stages out there.
Rich
Don’t worry I’m still learning as well
1) It’s the cathode resistor that determines the plate current. With SRPP the voltage is divided between both tubes, i.e. if you have 200V on the top tube’s anode you will have 100V on the bottom tube. To calculate your cathode resistor value, just choose your operating point using the curves and divide the grid voltage by the current at that point. The current runs through both tubes. At 10mA sounds like you’re using the 6n6p.
2) I would actually download power supply designer from the Duncan’s amps website and use it to model your power supply. Use a ccs as the tube load and set it to 10mA
Software
you can add an RC section to the CLCL if needed, so CLCLRC
3) You can also calculate power dissipation by P = (I^2) R. Current squared x resistance. So for 10ma (0.01 x 0.01) x 200 ohms = 0.02W.
I’m using the 6n2p at the moment. Using an EI core transformer acts as a bit of a filter, since they have a lower frequency bandwidth compared to a torroid, but that is just my preference. Also I like valve rectifiers, you may like to give a low current one a try. The 6BW4 is cheap and easy to implement with a 6.3V heater.
The lampizator circuit is a good start but there are better sounding valve output stages out there.
Rich
Hi Richard,
Any chance you could point me at better valve circuits than the Lampizator ones? I have built a couple of his for current DACs and been very happy. Other circuits I have seen to date are so similar you would only call them variants of the same SRPP e.g. Thosten Loesch. Tho I would say that the Lamizator circuits have the simplest power supply design.
Cheers
Any chance you could point me at better valve circuits than the Lampizator ones? I have built a couple of his for current DACs and been very happy. Other circuits I have seen to date are so similar you would only call them variants of the same SRPP e.g. Thosten Loesch. Tho I would say that the Lamizator circuits have the simplest power supply design.
Cheers
It helps to start with, what do you want the circuit to do? As a line-level amplifier, an SRPP is not a particularly good topology. It's more suited for driving highly capacitive loads to large voltage swings and trades off linearity to do so. And this circuit isn't even a good SRPP.
So, what are you trying to do? What's the desired swing and gain? What's the load impedance to be driven? Do you want low distortion?
So, what are you trying to do? What's the desired swing and gain? What's the load impedance to be driven? Do you want low distortion?
OK, so what you really want is an I-to-V converter. An SRPP is definitely out.
For most linear and high dynamic range operation, we're talking about a high bandwidth feedback circuit, with the input taken to a virtual ground. There are probably a few circuits out there, findable with a bit of searching.
Another possibility is to terminate the DAC output with a small (say, 10R) resistor, then build a very quiet 40dB voltage amp and a buffer. That won't be as good as using a proper virtual ground, but it will be a hell of a lot easier to build!
I'd avoid trick circuits and use a plain vanilla common-cathode voltage amplifier, coupled to a cathode follower output.
If you absolutely need to have something more complex (more tubes! more tubes!), the cathode follower could be traded in for a White follower or one of the Tektronix-style buffer circuits successfully used by Allen Wright in his preamps (Vacuum State - High End Hifi Equipment).
For a voltage DAC, all you need is a cathode follower.
For most linear and high dynamic range operation, we're talking about a high bandwidth feedback circuit, with the input taken to a virtual ground. There are probably a few circuits out there, findable with a bit of searching.
Another possibility is to terminate the DAC output with a small (say, 10R) resistor, then build a very quiet 40dB voltage amp and a buffer. That won't be as good as using a proper virtual ground, but it will be a hell of a lot easier to build!
I'd avoid trick circuits and use a plain vanilla common-cathode voltage amplifier, coupled to a cathode follower output. If you absolutely need to have something more complex (more tubes! more tubes!), the cathode follower could be traded in for a White follower or one of the Tektronix-style buffer circuits successfully used by Allen Wright in his preamps (Vacuum State - High End Hifi Equipment).
For a voltage DAC, all you need is a cathode follower.
thanks richard
Hi, Richard, NOW!!! I get why he only uses .25 wat resistors lol!!!However I think I have been making the wrong assumptions about this circuit.
1) when we say "we are running the anode at 200v" do we mean upper or lower? If it is lower, my transformer is not man enough and I have made an expensive boo boo. However, this srpp is so dynamic at 110volts secondary I am sure 400 volts to the upper anode would kill my speakers!!!!!
2) In a similar vein, when you say cathode resistor , do you mean the one linking the top cathode to the lower anode OR!!! the resistor linking the lower cathode to ground. I am asssuming that they are always the same value anyway, but I need to know for details. (and this might now be a wrong assumption).
3) Grid voltage.....The dac I am playing with is a voltage out CS4398. When we say grid bias voltage. Is this determined by the output voltage of the dac chip??? I down loaded the 4398 datasheet and tried to find a value, hoping it was -2V . However I can see no value.....If this is not the case , how do we set the grid voltage, so it works with the curve chart i have down loaded for the valve????
Yes you are right 6n6p....and I am begging you all not to mention any other circuits as you can surely see I am maxed out trying to get this working properly. If it is not the best, it is certainly the best by leagues that I have ever heard. I am hoping what I can learn from all of this I can go on to other circuits later but I am surely overloaded now and talk of other circuits will give give me a stroke. Thanks for your time
Hi, Richard, NOW!!! I get why he only uses .25 wat resistors lol!!!However I think I have been making the wrong assumptions about this circuit.
1) when we say "we are running the anode at 200v" do we mean upper or lower? If it is lower, my transformer is not man enough and I have made an expensive boo boo. However, this srpp is so dynamic at 110volts secondary I am sure 400 volts to the upper anode would kill my speakers!!!!!
2) In a similar vein, when you say cathode resistor , do you mean the one linking the top cathode to the lower anode OR!!! the resistor linking the lower cathode to ground. I am asssuming that they are always the same value anyway, but I need to know for details. (and this might now be a wrong assumption).
3) Grid voltage.....The dac I am playing with is a voltage out CS4398. When we say grid bias voltage. Is this determined by the output voltage of the dac chip??? I down loaded the 4398 datasheet and tried to find a value, hoping it was -2V . However I can see no value.....If this is not the case , how do we set the grid voltage, so it works with the curve chart i have down loaded for the valve????
Yes you are right 6n6p....and I am begging you all not to mention any other circuits as you can surely see I am maxed out trying to get this working properly. If it is not the best, it is certainly the best by leagues that I have ever heard. I am hoping what I can learn from all of this I can go on to other circuits later but I am surely overloaded now and talk of other circuits will give give me a stroke. Thanks for your time
valve rectification
Sorry, I forgot to mention. My transformer has the secondaries as 0 then 160v. I again understand nothing of valve rectification, but would love to try...but I have just been reading up, would I not have to have a transformer with a 160-0-160 output , the 0 going to ground and the two 160s going to the anodes???? I might just get another transformer built as I really want to make a good go of this the first lampi srpp has really got me hooked and I want to make a real proper job of my mark 2.
Sorry, I forgot to mention. My transformer has the secondaries as 0 then 160v. I again understand nothing of valve rectification, but would love to try...but I have just been reading up, would I not have to have a transformer with a 160-0-160 output , the 0 going to ground and the two 160s going to the anodes???? I might just get another transformer built as I really want to make a good go of this the first lampi srpp has really got me hooked and I want to make a real proper job of my mark 2.
I would be greatly appreciative if somebody could point me at a "fast quiet gain stage"! Would this fill the bill - the output stage taht is?
DAC End 2 - the AD1865N-K with single ended vacuum output stage
DAC End 2 - the AD1865N-K with single ended vacuum output stage
Last edited:
Another possibility is to terminate the DAC output with a small (say, 10R) resistor, then build a very quiet 40dB voltage amp and a buffer. That won't be as good as using a proper virtual ground, but it will be a hell of a lot easier to build!
Sy, what do you think about this: http://www.diyaudio.com/forums/tubes-valves/102352-6v6-line-preamp-3.html#post1225140
What if we could run something like 6v6 tube, or EL84 , as a line preamp, even without a cathode follower, it should have an acceptable output impedance, suitable enough for most power amps. Plain and simple. Should sound better than SRPP. Agree?
Thanks
Vix
What are you trying to do? If this is for a voltage output DAC, you don't need or want voltage gain. A cathode follower (or one of its variants) is a far better choice. 6V6/EL84 are not particularly good choices for CFs- you want transconductance to be high, so something smaller and cheaper like an ECC88 will greatly outperform the misapplied power tubes.
If you're trying to amplify the signal from a small resistor shunt at the output of a current-mode DAC, you need a high gain low noise tube, which 6V6/EL84 certainly are not. My first choice for that would be D3a, but there are a lot of other good choices.
If you're trying to amplify the signal from a small resistor shunt at the output of a current-mode DAC, you need a high gain low noise tube, which 6V6/EL84 certainly are not. My first choice for that would be D3a, but there are a lot of other good choices.
ChrisMmm,
As SY points out one should first decide what they want to do. In the case of an i/v dac output you may need a bit of gain if using the 1541, hence SY’s suggestion of a voltage gain and buffer. Also when converting the current to voltage with a resistor, you will need to keep it small as the distortion increases with higher resistance values. Lukas uses 100 ohms which is rather high, I have used 10-15 ohms with a 1:2 step-up transformer with good results. SRPP benefits a lot from a good power supply. Audionote use a tube regulated psu. You may like to search for the ECL85 regulator circuit published by Frank, if you wanted to give this route a try.
As for other circuits check out Lesha’s dac, he uses the 6c45p tube. I have a heard a 1541 with this tube output and it does sound very good. Thorsten also posted an ECC88 output stage here, which can use either choke or ccs load, you may have to do a search for it. It uses one triode of the tube per channel. You can also try the Aikido circuit as an output buffer.
These circuits are a good starting point and easier than a virtual ground and, of which I’m not sure of. SY is the man to you help you with that.
Lesha's DAC
Chokesrule,
1) If you’re running a B+ of 200 then in your SRPP you’re running each triode at 100V, remember that B+ is divided among the 2 triodes. So if you use 400V, you will have 200 on each triode.
2) The cathode resistor is the resistor attached to the cathode of the tube, sets the current draw and biases the tube. In SRPP, there are technically 2 cathode resistors of the same value. On the bottom valve you will notice the cathode resistor grounded.
3) I’m sure there are ways to use the dac output to bias the tube, but in SRPP the cathode resistor does this.
With your transformer, which does not have a centre tapped secondary, you can use a valve rectifier but adding 2 diodes from the plate to ground. Essentially a full wave bridge where 2 diodes are valve and 2 solid state. If you have 2 0-160 secondaries you can connect them in series for 160-0-160 but make sure you do so, so that they are in correct phase by following the transformer’s datasheet. You may like to just keep it as it is for now and upgrade to a centre tapped EI or C-core transformer that has some filament taps for the rectifier etc.
Cheers,
Rich
As SY points out one should first decide what they want to do. In the case of an i/v dac output you may need a bit of gain if using the 1541, hence SY’s suggestion of a voltage gain and buffer. Also when converting the current to voltage with a resistor, you will need to keep it small as the distortion increases with higher resistance values. Lukas uses 100 ohms which is rather high, I have used 10-15 ohms with a 1:2 step-up transformer with good results. SRPP benefits a lot from a good power supply. Audionote use a tube regulated psu. You may like to search for the ECL85 regulator circuit published by Frank, if you wanted to give this route a try.
As for other circuits check out Lesha’s dac, he uses the 6c45p tube. I have a heard a 1541 with this tube output and it does sound very good. Thorsten also posted an ECC88 output stage here, which can use either choke or ccs load, you may have to do a search for it. It uses one triode of the tube per channel. You can also try the Aikido circuit as an output buffer.
These circuits are a good starting point and easier than a virtual ground and, of which I’m not sure of. SY is the man to you help you with that.
Lesha's DAC
Chokesrule,
1) If you’re running a B+ of 200 then in your SRPP you’re running each triode at 100V, remember that B+ is divided among the 2 triodes. So if you use 400V, you will have 200 on each triode.
2) The cathode resistor is the resistor attached to the cathode of the tube, sets the current draw and biases the tube. In SRPP, there are technically 2 cathode resistors of the same value. On the bottom valve you will notice the cathode resistor grounded.
3) I’m sure there are ways to use the dac output to bias the tube, but in SRPP the cathode resistor does this.
With your transformer, which does not have a centre tapped secondary, you can use a valve rectifier but adding 2 diodes from the plate to ground. Essentially a full wave bridge where 2 diodes are valve and 2 solid state. If you have 2 0-160 secondaries you can connect them in series for 160-0-160 but make sure you do so, so that they are in correct phase by following the transformer’s datasheet. You may like to just keep it as it is for now and upgrade to a centre tapped EI or C-core transformer that has some filament taps for the rectifier etc.
Cheers,
Rich
Tired of "so similar" SRPP variants? Try something different.
Search this forum for "Differential White Cathode Follower".
Regulates own voltage amongst other obviously impossible
claims. What, two triodes doing three things at once???
Whoever come up with this lie is an obvious madman.
If you intend to run the sim in LTSpice, you will also need dmtriodep.inc
you can find the required model at Duncan Amps under the link "PSpice"
Anyways, how to explain???
First, lets look as-if a cascode with combined Mu gain of 30x30=900.
From S1a (inverting output of DAC): The node inbetween R3 and R4
tries to maintain a steady DC voltage. Virtual ground for AC purpose.
This forces top plate to hold an inverted copy of the inverted DAC.
Inverted twice = non-inverted. This is also a freebie power supply
regulation, believe it or nutz... The inverting input sees 47K to the
virtual ground node, an important detail for later.
Now R1 and R2 average S1b non-inverting DAC output with the
re-verted S1a copy at the top plate and feed into grid of V1b.
The cathode of V1b simply, or maybe not so simply, follows...
There is a high impedance across R1 and R2, because duplicate
copies of the same AC voltage on both ends. R5 assures a 47K
impedance to ground and gives entire input bridge a controlled
leak to real GND...
Interestingly, any error in V1b's cathode follower action results
in a 30x amplified error at the top plate. Half this amount gets
through R4 and drives V1a as-if a White Cathode Follower. The
plate of V1a is assiting to correct the error at V1b's cathode.
All this junk is happening at once, and none of it stomps its own
toes. PSRR and CMRR both excellent. So's linearity. Go figure...
Everything works together with bizarre synergy.
2.5VDC at the DAC outputs is totally optional. It can reject 10V
common mode either way, so ground referenced input is fine too.
Any power supply B+ from +80V to +180V should work, The top
plate will hold a regulated +50V or thereabouts, with a copy of
the output superimposed...
you can find the required model at Duncan Amps under the link "PSpice"
Anyways, how to explain???
First, lets look as-if a cascode with combined Mu gain of 30x30=900.
From S1a (inverting output of DAC): The node inbetween R3 and R4
tries to maintain a steady DC voltage. Virtual ground for AC purpose.
This forces top plate to hold an inverted copy of the inverted DAC.
Inverted twice = non-inverted. This is also a freebie power supply
regulation, believe it or nutz... The inverting input sees 47K to the
virtual ground node, an important detail for later.
Now R1 and R2 average S1b non-inverting DAC output with the
re-verted S1a copy at the top plate and feed into grid of V1b.
The cathode of V1b simply, or maybe not so simply, follows...
There is a high impedance across R1 and R2, because duplicate
copies of the same AC voltage on both ends. R5 assures a 47K
impedance to ground and gives entire input bridge a controlled
leak to real GND...
Interestingly, any error in V1b's cathode follower action results
in a 30x amplified error at the top plate. Half this amount gets
through R4 and drives V1a as-if a White Cathode Follower. The
plate of V1a is assiting to correct the error at V1b's cathode.
All this junk is happening at once, and none of it stomps its own
toes. PSRR and CMRR both excellent. So's linearity. Go figure...
Everything works together with bizarre synergy.
2.5VDC at the DAC outputs is totally optional. It can reject 10V
common mode either way, so ground referenced input is fine too.
Any power supply B+ from +80V to +180V should work, The top
plate will hold a regulated +50V or thereabouts, with a copy of
the output superimposed...
Attachments
Last edited:
hey guys,
i'm also very interested in building a dac. im currently looking at twistedpear's buffalo2: The Buffalo II Digital-to-Analog Converter. i was wondering if i could, instead of using their i/v stage, build the dac in front of vacuum state's RTP3 phono stage? Schematics
i'm also very interested in building a dac. im currently looking at twistedpear's buffalo2: The Buffalo II Digital-to-Analog Converter. i was wondering if i could, instead of using their i/v stage, build the dac in front of vacuum state's RTP3 phono stage? Schematics
Not with tubes, no. With opamps, yes.
Sure, but you'll have to eliminate the EQ and adjust the gain.
thanks for all this help rich
So Rich, this is probably this last time I need to pester you. Would you just look at my sums here and see if my reckoning is correct. I realise I can always check from actual measurements from the tubes, but I am not as yet that condfident messing with these voltages.
I am not computer literate enough to insert the 6n6p data sheet, so I apologise you are going to have to download it.
1)I look at the curves, i am interested in the -2 grid voltage curve. This is because I know from research the cs4397 voltage out dac gives -2.5 volts output signal.
2) because my psu is rigged to give 200volts, I am intersted in the 100 volt grid line as the two anodes share the voltage
3) I find the curve and grid line intersect at 20ma. This will be the draw from the transformer.
4) I now divide -2volts by 0.02ma . This gives me 100 ohms, the value of the two resistors needed for the two cathodes of the double triode tube.
5 ) I know need to work out the power of the cathode resistors. So it is .02 squared multiplied by 100ohms. The answer being 0.04 wats, meaning should really use a .25 or .5 wat resistor for safety.
6) Now assuming all the maths is right, here goes the questions.
7) I see Lukas uses a resister to ground after the input decoupling cap attached to the lower grid. How do I work out the suitable ohms??????What is its function?
8) I look at my transformer/chokes data sheets, and see that the windings are good for 30ma? So now I am wondering is it safe to run this circuit with these windings. If I am going to use 20ma, is 30ma enough head room?
9) I obviously look at the curves again and pick 75 volts as this give me a 10ma draw. It still is in a straight line region of the curve. I can using the software you sugested get the psu to give 150volts. Will it be poorer sound???This region although linear is near the curve bit of the chart.
10) This method of calculations is a way of trying any tube in srpp?
11) Thanks so much for helping me with this Alan
So Rich, this is probably this last time I need to pester you. Would you just look at my sums here and see if my reckoning is correct. I realise I can always check from actual measurements from the tubes, but I am not as yet that condfident messing with these voltages.
I am not computer literate enough to insert the 6n6p data sheet, so I apologise you are going to have to download it.
1)I look at the curves, i am interested in the -2 grid voltage curve. This is because I know from research the cs4397 voltage out dac gives -2.5 volts output signal.
2) because my psu is rigged to give 200volts, I am intersted in the 100 volt grid line as the two anodes share the voltage
3) I find the curve and grid line intersect at 20ma. This will be the draw from the transformer.
4) I now divide -2volts by 0.02ma . This gives me 100 ohms, the value of the two resistors needed for the two cathodes of the double triode tube.
5 ) I know need to work out the power of the cathode resistors. So it is .02 squared multiplied by 100ohms. The answer being 0.04 wats, meaning should really use a .25 or .5 wat resistor for safety.
6) Now assuming all the maths is right, here goes the questions.
7) I see Lukas uses a resister to ground after the input decoupling cap attached to the lower grid. How do I work out the suitable ohms??????What is its function?
8) I look at my transformer/chokes data sheets, and see that the windings are good for 30ma? So now I am wondering is it safe to run this circuit with these windings. If I am going to use 20ma, is 30ma enough head room?
9) I obviously look at the curves again and pick 75 volts as this give me a 10ma draw. It still is in a straight line region of the curve. I can using the software you sugested get the psu to give 150volts. Will it be poorer sound???This region although linear is near the curve bit of the chart.
10) This method of calculations is a way of trying any tube in srpp?
11) Thanks so much for helping me with this Alan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.