Shilded signal path, add 4k7 grid stopper and added ferrite beans directly on the 5687 soket...really better. There is still some noise, but I need to put my ears on the tweeter to listen to it. Layout isn't good I think...well, this a test board...I'll try to make something better.
Anyway...still hear noise if I touch the heatsink. Again...lower now, but it still buzz.
Why should I get rid of 1K res? What can they do?
( listen...this is not a dissent, take it as a student question. I'm here to learn as much as possible
)
Polenta TARAGNA...mhmmm good. This is a traditional dish from Valtellina and Alta Val Brembana. Folk cooking in Italy produced some of the best dishes you can eat. I like "nouvelle cousine" and elaborate cooking, but I think that some traditional dishes are simply perfect! Nothing to modify, you have only to find the better ingredients you can.. Well I'm really luky...here there is no problem to find all you need
Next time you have to come in Veneto and try "polenta e baccalà"...ahhhhh
And, if you like cheese, maybe you'll be in love with "polenta and tosella"...
http://www.formaggio-asiago.com/prodotti/tosella_ing.html
You cook it with a spoon of cream and butter...better thing is to dip the polenta in the tosella juice
Sorry for this long OT...
Mark
Anyway...still hear noise if I touch the heatsink. Again...lower now, but it still buzz.
Why should I get rid of 1K res? What can they do?
( listen...this is not a dissent, take it as a student question. I'm here to learn as much as possible
)Polenta TARAGNA...mhmmm good. This is a traditional dish from Valtellina and Alta Val Brembana. Folk cooking in Italy produced some of the best dishes you can eat. I like "nouvelle cousine" and elaborate cooking, but I think that some traditional dishes are simply perfect! Nothing to modify, you have only to find the better ingredients you can.. Well I'm really luky...here there is no problem to find all you need
Next time you have to come in Veneto and try "polenta e baccalà"...ahhhhh
And, if you like cheese, maybe you'll be in love with "polenta and tosella"...
http://www.formaggio-asiago.com/prodotti/tosella_ing.html
You cook it with a spoon of cream and butter...better thing is to dip the polenta in the tosella juice
Sorry for this long OT...
Mark
Don't ground the heatsinks since that will increase the capacitance from a high Z point to ground. I suggested getting rid of the 1K resistors just because I didn't use them and my circuit is stable.
If you get INCREASED noise when touching the heatsinks, that's OK. It's when the noise decreases that you have a problem. There's still no substitute for a nice 100MHz+ oscilloscope.
Let me be brutally honest- the French do "fussy food" much, much better than the Italians. But, also brutally honest, the Italians do country food better than anyone in the world, including the French. My best friend and cooking partner joined me for my birthday last week; we had meals on successive nights in Lyon (at a VERY fancy place with real artwork for food) and Neviglie (near Alba, at a simple farmhouse/agriturismo). Each was a perfect example of its type, and I would not have rated one as intrinsically better than the other, just different. And both, in their own ways, absolutely perfect. Sort of like the difference between electromagnetic point-source and electrostatic dipole speakers.
If you get INCREASED noise when touching the heatsinks, that's OK. It's when the noise decreases that you have a problem. There's still no substitute for a nice 100MHz+ oscilloscope.
Let me be brutally honest- the French do "fussy food" much, much better than the Italians. But, also brutally honest, the Italians do country food better than anyone in the world, including the French. My best friend and cooking partner joined me for my birthday last week; we had meals on successive nights in Lyon (at a VERY fancy place with real artwork for food) and Neviglie (near Alba, at a simple farmhouse/agriturismo). Each was a perfect example of its type, and I would not have rated one as intrinsically better than the other, just different. And both, in their own ways, absolutely perfect. Sort of like the difference between electromagnetic point-source and electrostatic dipole speakers.
SY said:
Ok, thankyou for the advice
SY said:
Yes, I know...I've booked it as degee gift
SY said:
First of all...happy birthday...I'm late I know
I agree with you, brutality isn't needed
What I think is that Frenchs research the flavour and use food for this purpose. Think at the number of different sauces they use...think at the number of aromatic cheeses ( uff I'm hungry...again). Italians purpose is to glorify the taste of the food. Sort of like the difference between baroque cathedral and a zen garden. Obviusly this is a semplification, but I think, in the most cases, this can sum up the different spirit in cooking.
Anyway, as you already said, "both, in their own ways, absolutely perfect"
Mark
PS. electronic and food...what an amazing pair
How can I calculate the output impedance using a c4s?
Is it equal to the plate resistance or the transistor change the impedance?
Diego Nardi, an italian desiger (maybe someone knows him, he write on soundpractice sometimes), use often an hybrid srpp with the upper tube substitued by an mje340 (see lik for an example).
http://www.webalice.it/jlc891/page10.html
To polarize the bjt he use a special led made by HP. This led works good with 1mA and with this current it provide a voltage reference of 1.6V ca.
http://www.fairchildsemi.com/ds/HL/HLMP-K150.pdf
I think that he use this led because he don't want too much current that run from the reference chain to the plate.
Look here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=43986&perpage=10&pagenumber=6
Ec8010 said:
"The Nardi circuit works in a very similar way, but is very slightly compromised by passing the current from its reference chain into the load."
The difference between c4s and Nardi circuit is mainly the output impedance, I'm right? With this hybrid srpp is the impedance something like the bjt collector one?
How does this solution looks?
Any comments? Nardi say that his solution is better, particularly for the musical results.
Thankyou in advance
Mark
PS. I'm rebuilding the circuit to test the 6sn7
Is it equal to the plate resistance or the transistor change the impedance?
Diego Nardi, an italian desiger (maybe someone knows him, he write on soundpractice sometimes), use often an hybrid srpp with the upper tube substitued by an mje340 (see lik for an example).
http://www.webalice.it/jlc891/page10.html
To polarize the bjt he use a special led made by HP. This led works good with 1mA and with this current it provide a voltage reference of 1.6V ca.
http://www.fairchildsemi.com/ds/HL/HLMP-K150.pdf
I think that he use this led because he don't want too much current that run from the reference chain to the plate.
Look here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=43986&perpage=10&pagenumber=6
Ec8010 said:
"The Nardi circuit works in a very similar way, but is very slightly compromised by passing the current from its reference chain into the load."
The difference between c4s and Nardi circuit is mainly the output impedance, I'm right? With this hybrid srpp is the impedance something like the bjt collector one?
How does this solution looks?
Any comments? Nardi say that his solution is better, particularly for the musical results.
Thankyou in advance
Mark
PS. I'm rebuilding the circuit to test the 6sn7
mark_titano said:
The output impedance of a BJT cascode current source is equal to Re x hfe(upper) x hfe (lower) + 1/hoe lower, where Re is the current set resistor. 1/hoe is analagous to the plate resistance of a valve (~50K) for a MJE350.
If you're asking about the output impedance of the stage as a whole, yes. The output impedance is equal to the plate resistance in parallel with the output impedance of the current source. Since the output impedance of the current source is approximately infinite, it simplifies to the plate resistance of the valve.
EDIT:
Drat, I was being too hasty. The output impedance is equal to the plate resistance only if the cathode is bypassed. In this case it is not, so the output impedance is equal to rp + rk x (mu +1)
Mark, I had another look at your schematic (post no. 54). It seems you're using a high-voltage type (MJE350) in the upper transistor position. Someone might have said this before (it's been a while since I've read this thread) You may get better performance with a small-signal type, say BC558, since these have a higher hfe, making the output impedance of the CCS higher (making it more ideal).
Nonetheless, I hesistate in posting this suggestion since once you load the stage with the input impedance of your amp it all comes undone....
Nonetheless, I hesistate in posting this suggestion since once you load the stage with the input impedance of your amp it all comes undone....
audiousername said:
Thankyou very much. Yes I'm using an mje 350 for the upper transistor. I've bought some 2n2907 to substitute it, I'm rebuilding the circuit also to change these bjt. I've understood how to calculate the impedance of the cascode, what I had not understood was how this can modify the circuit impedance.
So thankyou
Mark
SY said:
Thankyou SY, I understand this issue
Have you some favorite bjt? I look at your 6sn7 schematics, I saw that you used an 2n3702...
Just a question: You said me that ""recovery from overload is one of the most important and neglected aspects of tube amp design" and that this recovery is affected by the ccs. So, is a faster bjt better?
Please comment post 66
Mark
"Faster" is better from a circuit bandwidth standpoint- compared to RC recovery times, even the slowest transistor is blazingly fast.
My transistors were picked with the critereon I stated- what was sitting in my coffee can full of transistors. That WEP unit has to be twenty years old!
Post 66: Exactly what audiousername answered- the CCS is essentially infinite Z, so the output Z becomes the tube's effective plate resistance, rp + (mu + 1)Rk.
My transistors were picked with the critereon I stated- what was sitting in my coffee can full of transistors. That WEP unit has to be twenty years old!
Post 66: Exactly what audiousername answered- the CCS is essentially infinite Z, so the output Z becomes the tube's effective plate resistance, rp + (mu + 1)Rk.
SY said:
I'm going to trying different tubes and I'll not change the power supply so to use an optimize supply will not possible. I understand what you're saying and I agree. In a line stage we don't need voltage swing of 80/100V so we could optimize the circuit to have across the ccs only a voltage drop that is the summarize of the voltage swing plus 10V ca to make the ccs working. Right? With a swing of 30/40 V the voltage across the bjt wuold be 40/50V ca...we only would have to match the power dissipation of the bjt to the needed.
I'll go with the mje350, I want to try different tubes and different working points too, an higher power supply voltage will suit me
Mark
Hybrid SRPP, please help
Dear mark titano and forum all,
I'm looking for a schematic of Hybrid SRPP, very easy like the CCS at http://www.machmat.com/sales/kits/ccs.htm, using mje 350 or mje 340, like Diego Nardi does.
Hybrid SRPP, not mu-follower.
I need the advantages of CCS with low output impedence (let say 100-200 Ohms, with a 6h30 or ecc99 or 6922 paralled)
Thank you very much, G.M.
Dear mark titano and forum all,
I'm looking for a schematic of Hybrid SRPP, very easy like the CCS at http://www.machmat.com/sales/kits/ccs.htm, using mje 350 or mje 340, like Diego Nardi does.
Hybrid SRPP, not mu-follower.
I need the advantages of CCS with low output impedence (let say 100-200 Ohms, with a 6h30 or ecc99 or 6922 paralled)
Thank you very much, G.M.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.