s1a and s1b represent differential outputs of a 5V DAC, quiescent at 2.5VDC.
So yeah, they are "balanced" as I've drawn it.
GND is also well withing the 10V range of common modes that this stage can
reject. So 2.5VDC offset can be ground referenced instead with no problem.
Each input presents a 47K AC impedance (47K+47K=94K differential) to DAC.
DC resistance to GND are slightly differrent. But common mode rejection will
hide constant DC error anyway.
If you use a single ended input, I would use S1a as the active signal, and
S1b as the ground reference. As the impedance of S1a is always true 47K
no matter what S1b is doing. The input impedance of S1b depends on S1a
being balanced opposite, else S1b input impedance drops to 31K? Just use
S1a input if single ended...
S1b already has a resistor to GND, you don't have to do anything more to
ground reference it. Any residual balanced mode behavior will only help to
break ground loop problems, so no need to "fix" S1b to local ground. It can
float a little with ground from the previous piece of upstream equipment.
Last edited:
Sorry, meant DC mode rejection in this instance, not common mode.
Does both, so I confuse myself...
Last edited:
You will also need dmtriodep.inc to run the simulation.
Generic triode at Duncan Amps. Link marked "PSpice"
As I suspected! Front end LTP don't need power filter,
DWCF common mode rejection takes care of all that.
Now we got Voltage gain! And current gain...
Almost like differential version of Aikido, no idle threat!
Note upper right corner: I have injected 2VPP of power
supply ripple just to illustrate how immune the circuit is.
You won't see any of it in the output.
Though PSRR is indeed still excellent, front end seems
to need at least 150V to work with even half the input
levels shown. Not yet sure if anything can be done to
drop that minimum figure back to the 80V? Its probably
necessary headroom for Mu to swing at LTP plates.
Generic triode at Duncan Amps. Link marked "PSpice"
As I suspected! Front end LTP don't need power filter,
DWCF common mode rejection takes care of all that.
Now we got Voltage gain! And current gain...
Almost like differential version of Aikido, no idle threat!
Note upper right corner: I have injected 2VPP of power
supply ripple just to illustrate how immune the circuit is.
You won't see any of it in the output.
Though PSRR is indeed still excellent, front end seems
to need at least 150V to work with even half the input
levels shown. Not yet sure if anything can be done to
drop that minimum figure back to the 80V? Its probably
necessary headroom for Mu to swing at LTP plates.
Attachments
Last edited:
Hi Chokesrule,
I wouldn’t worry too much about being 23 volts over or 10 volts under. Recall that you will get a voltage drop on all the series resistances in the psu, the series resistance of the choke will also decrease a tad when in use and valves have 'guidelines'.
If you are trying to measure current through the srpp stage, just measure the voltage across one of the cathode resistors and divide by the resistors value. It may even be safer to clip (before turn on) the -ve lead of the DVM to the bottom of the lower cathode resistor (i.e. where it is grounded) and use the + ve lead to measure across the resistor and then B+.
yes, good idea, but this is diyaudio and it is all about learning
. I have found psu designer to be invaluable when tweaking a power supply.many have found AC heaters to sound good, just keep the wiring twisted tight, right up to the valve pins.
Cheers,
Rich
I have just done the valve rectification with 5uag and have changed the transformers from 30 va to 100va....my valves run at 217volts dc and .014 ma and the difference in the sound is crazy.....its extreme....If you are going to build one buy the 100va transformer and the valve rectifiers it is a "must" not an "option"
Sorry to jump in here.
Richard I see you're using 6N2P in a SRPP topology no?
I've got a load of these and never used them.
I'm running a SRPP with 6N6P's at about 18mA for a headphone amp and sounds great.
What are the 6N2P's like? What B+ and cathode resistor values are you using for the 6N2P's?
p.s
I use AC heaters with a bias around 40 volts and it is dead quiet.
Richard I see you're using 6N2P in a SRPP topology no?
I've got a load of these and never used them.
I'm running a SRPP with 6N6P's at about 18mA for a headphone amp and sounds great.
What are the 6N2P's like? What B+ and cathode resistor values are you using for the 6N2P's?
p.s
I use AC heaters with a bias around 40 volts and it is dead quiet.
Last edited:
Just try it out. You might like the difference as long as you are in the linear range.
I found the bass became faster and stronger. (my wife who knows nothing about hi-fi was a testament). I tweak and ask her to come and listen.
The sound will change between 8-22mA.
The 6N6P's IMHO likes a higher current.
Are your cathode resistors bypassed on the bottom tube?
I also split the upper cathode resistor and took the output off in betweem. So 150R resistors on each. But 100R and 50R on the upper tube with the output off between these (for headphones though).
I found the bass became faster and stronger. (my wife who knows nothing about hi-fi was a testament). I tweak and ask her to come and listen.
The sound will change between 8-22mA.
The 6N6P's IMHO likes a higher current.
Are your cathode resistors bypassed on the bottom tube?
I also split the upper cathode resistor and took the output off in betweem. So 150R resistors on each. But 100R and 50R on the upper tube with the output off between these (for headphones though).
Last edited:
yes the resistors are bypassed with an 1500uf oscon.....I did not notice a lot of difference, but I am using a tripath on some proacs which is said to be not ideal, so may be when I sort all that out there will be a lot more to come......I will try using a larger current, but just want to cool off little from this rectifier/transformer upgrade....it has done wonders....I am always hyper after the first switch on I think its the worry that you have done something wrong and its going to go BANG!!!!.....One thing I am finding is it takes about 3 plays of a cd then the sound goes from mental to extreme !!!!I don't know if it takes that long for all the chokes and capacitors to sort themselves out or what...I am going to try putting a grid choke in place of the grid resistors but I might be slow to change as I an really enjoying this right now.
Why does your resistor split trick, only apply to headphones???Want me to try it on mine ??
Why does your resistor split trick, only apply to headphones???Want me to try it on mine ??
I got the idea from Broskie:
SRPP+ All-in-One & Mu Followers
For high impedance headphones it helps to reduce the load or something. My values are approx, wasn't sure on how to work out the split.
I've also bypassed with lytics with small MKP 0.01 uf caps on the cathode resistors. Just back and forth trying different combinations. That's what it's all about and all the fun.
I'm not using chokes on my PS.(Can't buy chokes here)
Just HEXFREDS with a simple RCRC circuit.
SRPP+ All-in-One & Mu Followers
For high impedance headphones it helps to reduce the load or something. My values are approx, wasn't sure on how to work out the split.
I've also bypassed with lytics with small MKP 0.01 uf caps on the cathode resistors. Just back and forth trying different combinations. That's what it's all about and all the fun.
I'm not using chokes on my PS.(Can't buy chokes here)
Just HEXFREDS with a simple RCRC circuit.
As I understand the LED (depending which colour) creates a constant voltage and most at a current of 18-20mA. Yellow about 1.85 V.
So they are only suitable for certain tubes.
I measured the voltage difference across the top cathode resistor and get about 2.75 V (2.75/150 = 18.6 mA).
Maybe someone who understands the LED concept more than me will jump in.
So they are only suitable for certain tubes.
I measured the voltage difference across the top cathode resistor and get about 2.75 V (2.75/150 = 18.6 mA).
Maybe someone who understands the LED concept more than me will jump in.
HI Brit,
The cathode resistor is about 220 ohms with a B+ of about 170V, cathode is not bypassed with a cap. The 6n2p can only run a couple of mA max. With a few mods you should be able to sub them in. They are best for current output dacs, due to their high mu. However, their other properties do limit their ability to drive low impedence and long cables. Use a 2uf cap on the output with a 1M output to ground resistor.
I've got a 6c45P output in the pipeline, but not complete yet.
Chokesrule,
Yeah, valve rectifiers are the way to go. The 5u4G is vastly overspecified for the job, the 5Y3 and 5W4 are also nice, 5W4's are cheap, but beware that the input cap size is limited to 4uF. Glad to see that you are enjoying it.
Cheers,
Rich
The cathode resistor is about 220 ohms with a B+ of about 170V, cathode is not bypassed with a cap. The 6n2p can only run a couple of mA max. With a few mods you should be able to sub them in. They are best for current output dacs, due to their high mu. However, their other properties do limit their ability to drive low impedence and long cables. Use a 2uf cap on the output with a 1M output to ground resistor.
I've got a 6c45P output in the pipeline, but not complete yet.
Chokesrule,
Yeah, valve rectifiers are the way to go. The 5u4G is vastly overspecified for the job, the 5Y3 and 5W4 are also nice, 5W4's are cheap, but beware that the input cap size is limited to 4uF. Glad to see that you are enjoying it.
Cheers,
Rich
Last edited:
Thxs Rich. Not really suited to my headphone amp then. Got a box of them and no use for them
I'm just trying the 6N1P to compare with 6N6P. Running the 6n1P with 5 mA (think maybe this should be higher) and 6N6P with 18mA.
First impressions are that 6N1P is more neutral and clear. 6N6P is a bright tube with lots of life if you like that. Sometimes very bright depending on the music. Good for jazz maybe.
I'm just trying the 6N1P to compare with 6N6P. Running the 6n1P with 5 mA (think maybe this should be higher) and 6N6P with 18mA.
First impressions are that 6N1P is more neutral and clear. 6N6P is a bright tube with lots of life if you like that. Sometimes very bright depending on the music. Good for jazz maybe.
Last edited:
Thxs Rich. Not really suited to my headphone amp then. Got a box of them and no use for them
I'm just trying the 6N1P to compare with 6N6P. Running the 6n1P with 5 mA (think maybe this should be higher) and 6N6P with 18mA.
First impressions are that 6N1P is more neutral and clear. 6N6P is a bright tube with lots of life if you like that. Sometimes very bright depending on the music. Good for jazz maybe.
Interesting to see how you get on with that, there is also the 6N23P to try. As I have no russian 6N1P or 6N23P I may buy some next...
Currently my 6N6P SRPP stage is on 8.5mA but anything that works in the sweet spot at less power would be good, especially as I recently upped my GU50 idle current from 55mA to 60mA, losing yet another 10mA from the pot.
I can confirm now however that the GU50 in pentode mode driven by an SRPP sounds extremely fine, with an air of space and dynamics previously never heard in my amp!
As far as using LEDs on an SRPP lower is concerned however I think you'd need to work the top resistor to match every time, for instance my 6N6p has a 8.5V cathode voltage (via a 1k resistor top and bottom) which balances the top and bottom tubes to an equal midpoint - even if I unplug them both and try 6N23p tubes instead. With a Led you are fixing the expected operating point - but you get more gain, albeit with less local feedback.
However I deliberately wanted a big cathode voltage on mine too as my SRPP needs to swing a couple of hundred volts - not the same application as a headphone amp!
This morning I was tolling with the idea of using some SV83 pentodes in a headphone amp. (came across a 300b headphone amp and thought of the 4 pentodes I have).
I wonder if I could use sv83 strapped as triodes in srpp mode, but that would require 4 tubes for 2 channels.
Or maybe the triode, 6N6P, grounded cathode input, and a SV83 pentode as output tube as this is a cheap linear tube and would give me plenty of power to drive headphones in SE mode instead of PP.
So many possibilities and so little time
Valid point there. Maybe a LM317 as a CCS would be easier to balance with the top tube.
I also have some 6080 which are great as a CF in the headphone amp due to their low resistance but I don't have the huge heater current capability on hand. 3 Amps per heater!!!!!!!!!
I wonder if I could use sv83 strapped as triodes in srpp mode, but that would require 4 tubes for 2 channels.
Or maybe the triode, 6N6P, grounded cathode input, and a SV83 pentode as output tube as this is a cheap linear tube and would give me plenty of power to drive headphones in SE mode instead of PP.
So many possibilities and so little time
Valid point there. Maybe a LM317 as a CCS would be easier to balance with the top tube.
I also have some 6080 which are great as a CF in the headphone amp due to their low resistance but I don't have the huge heater current capability on hand. 3 Amps per heater!!!!!!!!!
Last edited:
Well I just tried a CCS LM317 on the bottom tube cathode and not sure what happened.
Was horrible. All airy, distorted and no bass.
I connected both cathodes to 1 LM317.
Not sure if I needed to double the current so I tried both. Both the same.
68R resistor set at 18mA.
then 34R set at 36mA.
Do I need a bypass cap?
Was horrible. All airy, distorted and no bass.
I connected both cathodes to 1 LM317.
Not sure if I needed to double the current so I tried both. Both the same.
68R resistor set at 18mA.
then 34R set at 36mA.
Do I need a bypass cap?
Last edited:
hi, my mark2 lampi is running nice and I have just been messing with this Duncan psu designer program. I know my tubes are running a 0.014amps so in the designer I set the load as constant current 14ma and set up my psu. The program was not impressed at all!!!!!!!
From the rectifier I have CLCLCRC that is 1uf::40h700ohm::47uf::40/700::470uf::4.2ohm resistor::470uf
the tubes are 5uac so I need 1uf for the first cap
the rest of the caps ???because i had them on hand
the psu designer says "current sink causes voltage to sink below zero for more than 5 mains cycles"
I dont know what it means but is it bad?????? I can play around changing capacitor values (not the chokes/resistor) to eventually find something like 1uf::L::10uf::L::47uf::R::470uf does set off any program warnings....Should I change to this psu???? Is my current set up bad??
I have also been reading in Morgan Jones, that chokes should be snubbed with a pair of 220nf capacitors ....on just before choke/one after and both snubbers also attached to ground...does any one recomend this????
From the rectifier I have CLCLCRC that is 1uf::40h700ohm::47uf::40/700::470uf::4.2ohm resistor::470uf
the tubes are 5uac so I need 1uf for the first cap
the rest of the caps ???because i had them on hand
the psu designer says "current sink causes voltage to sink below zero for more than 5 mains cycles"
I dont know what it means but is it bad?????? I can play around changing capacitor values (not the chokes/resistor) to eventually find something like 1uf::L::10uf::L::47uf::R::470uf does set off any program warnings....Should I change to this psu???? Is my current set up bad??
I have also been reading in Morgan Jones, that chokes should be snubbed with a pair of 220nf capacitors ....on just before choke/one after and both snubbers also attached to ground...does any one recomend this????
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.