ok I'm going back and forth on this I know the advantage of parafeed is to keep the output transformer from saturation.
The question is how do you decide when saturation is going to be a real problem.
The output transformer I'm going to use has a max dc current of 806ma per side and my current draw is approx. 600ma per side of the O.T. winding.
Is this close enough where I will get a gain in spectrial purity of the output or not.
Nick
The question is how do you decide when saturation is going to be a real problem.
The output transformer I'm going to use has a max dc current of 806ma per side and my current draw is approx. 600ma per side of the O.T. winding.
Is this close enough where I will get a gain in spectrial purity of the output or not.
Nick
I can't answer your actual question, but the advantage of parafeed is not avoiding saturation as such, but that you can eliminate the air gap in the transformer. You then get to figure out whether you like an air gap or a capacitor more, but using an airgaped transformer in parafeed mode is the worst of both worlds.
I don't know what spectrial purity is.
I don't know what spectrial purity is.
What tube or tubes are you using that need 600mA of current?
Remember that you will be dealing with resonances in the bass region whose characteristics are determined by your capacitor's interactions with the inductance of the transformer. This allows you to: 1) tune the bass to suit your tastes, and 2) it allows you to pull your hair out from spending too much money on different capacitor sizes trying to tune the bass to your tastes. YMMV
I would try it without parafeed, if your transformer can really take 800mA...they must be some big mommas.
Remember that you will be dealing with resonances in the bass region whose characteristics are determined by your capacitor's interactions with the inductance of the transformer. This allows you to: 1) tune the bass to suit your tastes, and 2) it allows you to pull your hair out from spending too much money on different capacitor sizes trying to tune the bass to your tastes. YMMV
I would try it without parafeed, if your transformer can really take 800mA...they must be some big mommas.
Well, FWIW I'd built a series of EL84 based SE amps a year or so ago, as an exercise in vintage parts recycling and progressive upgrading. I.E. with the right parts in the scrap pile, how good a sound can you get? The answer is pretty damned fine, even if you don't know wtf you're doing - just build a proven circuit.
Starting with very basic parts quality and shameless copy of the Decware SE84CS amp, I progressed through parts quality and circuit revisions.
The final result was a pair of mono-blocks incorporating a combination of Alex Kitic's RH84 feedback topology and Bottlehead inspired parafeed connection. Far and away the best sounding of the series of 4 versions - even with a fairly "pedestrian" Clarity serving as the parafeed cap .
I also own a pair of Blottlehead ParamourII - which sound spectacularly good out of the box with a definitely budget OPT - (SPECTRO 70V line matching transformer) carefully selected for the fortuitously appropriate primary and secondary ratios - and you know this iron isn't optimized for full bandwidth audio. Of course you can get big time spendy with custom Magnequest OPT and plate load chokes, and Mundorf Silver/Oil caps - I'd dare say no compromises there by most folk's standards.
Starting with very basic parts quality and shameless copy of the Decware SE84CS amp, I progressed through parts quality and circuit revisions.
The final result was a pair of mono-blocks incorporating a combination of Alex Kitic's RH84 feedback topology and Bottlehead inspired parafeed connection. Far and away the best sounding of the series of 4 versions - even with a fairly "pedestrian" Clarity serving as the parafeed cap .
I also own a pair of Blottlehead ParamourII - which sound spectacularly good out of the box with a definitely budget OPT - (SPECTRO 70V line matching transformer) carefully selected for the fortuitously appropriate primary and secondary ratios - and you know this iron isn't optimized for full bandwidth audio. Of course you can get big time spendy with custom Magnequest OPT and plate load chokes, and Mundorf Silver/Oil caps - I'd dare say no compromises there by most folk's standards.
I've never heard of an OPT with nearly an amp of permissable DC current. If there's one transformer that at first glance wouldn't seem to justify parafeed, that would be it. Conversely I recently bought a pair of toroidal 5k SE OPTs that can't withstand any DC current. The bass rolls off fast below 100 Hz with as little as 10 ma. Those will be used parafeed.
Do you have any way to test yours for LF roll-off with the considered DC bias? It wouldn't take much and you could probably kludge something together on the bench with an audio generator and bench power supply.
Do you have any way to test yours for LF roll-off with the considered DC bias? It wouldn't take much and you could probably kludge something together on the bench with an audio generator and bench power supply.
George is right, parafeed is of benefit only in an SE application. In PP applications the dc flux in the core will be cancelled leaving no net dc flux on the core provided the tubes are operating at exactly the same idle current. It is not unusual to specify a PP transformer to operate with up to 10% idle current imbalance.
PP transformers do not usually have air gaps. (I never seen one)
I think you ought to perhaps be focusing on a simpler, less ambitious project, unless I read you wrong this is your first project and you are bravely attempting something I would not be inclined to try with 20+ yrs of tube design experience. Still maybe that is why I am not doing that. I guess I have no imagination.. LOL
How about a quartet of 6550 for 120Wrms or a sextet for about 200W. IMO this is more likely to end up well. Have you ever heard a 400W tube stereo amp? Louder than you can imagine with just about any currently available speaker. .
And all off the shelf parts..
edit
PP transformers do not usually have air gaps. (I never seen one)
I think you ought to perhaps be focusing on a simpler, less ambitious project, unless I read you wrong this is your first project and you are bravely attempting something I would not be inclined to try with 20+ yrs of tube design experience. Still maybe that is why I am not doing that. I guess I have no imagination.. LOL
How about a quartet of 6550 for 120Wrms or a sextet for about 200W. IMO this is more likely to end up well. Have you ever heard a 400W tube stereo amp? Louder than you can imagine with just about any currently available speaker. .
And all off the shelf parts..
edit
Hey-Hey!!!,
I have done some listening to grid chokes. Moderate inductance devices. The most eye-opening experience I had with them was that they sound best with a gap. the rest of the circuit must match the lowered L value, or the coil must be modded to deliver higher L but I am not going to do anything else. The L is more constant with a gap. It is not so variable to frequency either. The combination of air and steel is a lot more linear than steel with a minimum gap.
I have plans to build PP with a gap as soon as the design work is complete.
cheers,
Douglas
I have done some listening to grid chokes. Moderate inductance devices. The most eye-opening experience I had with them was that they sound best with a gap. the rest of the circuit must match the lowered L value, or the coil must be modded to deliver higher L but I am not going to do anything else. The L is more constant with a gap. It is not so variable to frequency either. The combination of air and steel is a lot more linear than steel with a minimum gap.
I have plans to build PP with a gap as soon as the design work is complete.
cheers,
Douglas
Not a bad thought, but how wedded are you to using these particular tubes? And at these power levels? I know that they are cool looking, and high power is also appealing, but it seems like a herculean design task to get something together that will be reliable, sound good, and not take the entire contents of the Fort Knox vault to build..
How much power do you really need? What kind of speakers are you using and how much space do you have to fill? Perhaps even more importantly do you have major $$$$ to devote to opt development as it is unlikely that the first protos will be acceptable?
A friend of mine designed and developed a c core based opt for an 845 SE amplifier he was designing, it took him three attempts to get a transformer he could live with.. Design target was 10K, L >100H, 100mA, 3KV insulation system, and a first resonance above 50kHz - what he ended up with was good, but he didn't quite make his bandwidth requirement. I don't remember his leakage inductance target, but I do remember he did not attain it, nor the stray capacitance either. This fellow was doggedly determined and the end result was probably about as good as you can get. He told me afterwards that he would think long and hard before he ever attempted to wind another large opt..
How much power do you really need? What kind of speakers are you using and how much space do you have to fill? Perhaps even more importantly do you have major $$$$ to devote to opt development as it is unlikely that the first protos will be acceptable?
A friend of mine designed and developed a c core based opt for an 845 SE amplifier he was designing, it took him three attempts to get a transformer he could live with.. Design target was 10K, L >100H, 100mA, 3KV insulation system, and a first resonance above 50kHz - what he ended up with was good, but he didn't quite make his bandwidth requirement. I don't remember his leakage inductance target, but I do remember he did not attain it, nor the stray capacitance either. This fellow was doggedly determined and the end result was probably about as good as you can get. He told me afterwards that he would think long and hard before he ever attempted to wind another large opt..
Well Im working on a schematic right now but the problem is I have yet to get curve captor to work.
Once that is done It should go alot faster. Becasue you that 813 is a common everyday tube
But I'm making a basic schematic the old fashion way with a pencial and when it's done I will post it.
Nick
Once that is done It should go alot faster. Becasue you that 813 is a common everyday tube
But I'm making a basic schematic the old fashion way with a pencial and when it's done I will post it.
Nick
Hi Nick,
Download LTspice if you haven't done so already - you can use it to make your schematics.
Curve Captor incidentally will only work with triodes and diodes, there is no option for pentodes or tetrodes. Not a problem of course if you triode connect the 813 in which case CC makes very good models.
Paintcc has a sub program known as paint_kip that is for pentodes, but it doesn't seem capable of making very accurate captures. I have had limited success.
Download LTspice if you haven't done so already - you can use it to make your schematics.
Curve Captor incidentally will only work with triodes and diodes, there is no option for pentodes or tetrodes. Not a problem of course if you triode connect the 813 in which case CC makes very good models.
Paintcc has a sub program known as paint_kip that is for pentodes, but it doesn't seem capable of making very accurate captures. I have had limited success.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.