The parafeed topology couples the OPT using a capacitor. Doesn't this then mean that the primary impedance of the OPT is pretty much a moot point? As long as the capacitor is sized properly for LF cutoff does the output tube really care any more what load it is working into?
For example lets say I had a 45 tube and I had three OPT's, one has a 10K primary another has a 7K primary and the last one has a 5K primary. If it was not parafeed then the 5K would be the best choice for the tube to work into and the 10K and 7K the poor choice. With parafeed is that still true?
For example lets say I had a 45 tube and I had three OPT's, one has a 10K primary another has a 7K primary and the last one has a 5K primary. If it was not parafeed then the 5K would be the best choice for the tube to work into and the 10K and 7K the poor choice. With parafeed is that still true?
My dog pulls a log. How big a log? Does it matter if I couple her by a rope, a chain, or by tying her tail (direct coupled)? There's still an optimum log size and weight. The hitching makes nearly no difference.
If the cap is sized properly, it can be considered almost a dead short for all relevant audio frequencies ... otherwise they wouldn't pass unaltered.
In other words: the cap is open circuit for DC but nearly like a piece of wire for AC.
A relevant thread, a bit on the technical side: https://www.diyaudio.com/community/threads/parafeed-l-c-calculations-wrt-z.293826/ see post 4
The most useful spreadsheet when it comes to calculating parafeed caps values: https://www.diyaudio.com/community/threads/01a-question.282272/page-27#post-6786912 Voltsecond used to have a website detailling this, which is sadly closed.
The most useful spreadsheet when it comes to calculating parafeed caps values: https://www.diyaudio.com/community/threads/01a-question.282272/page-27#post-6786912 Voltsecond used to have a website detailling this, which is sadly closed.
Windcrest77,
Now you have been given 2 very good explanations.
Just use the same primary impedance for the Parafeed output transformer, as you would use for a non-parafeed output transformer (OPT).
Parafeed OPT uses interleaved laminations, such as interleaved Es and Is. No air gap required, since there is no DC current in the primary.
Non-Parafeed OPT uses alternate stacked laminations, such as E then I then E then I . . . They require an air gap, since there is DC current in the primary. All the Es on one side of the air gap, and all the Is on the other side of the air gap.
A 'C core' single ended OPT is similar, there are air gaps between the two 1/2s of the laminations, to allow for the DC current.
Now you have been given 2 very good explanations.
Just use the same primary impedance for the Parafeed output transformer, as you would use for a non-parafeed output transformer (OPT).
Parafeed OPT uses interleaved laminations, such as interleaved Es and Is. No air gap required, since there is no DC current in the primary.
Non-Parafeed OPT uses alternate stacked laminations, such as E then I then E then I . . . They require an air gap, since there is DC current in the primary. All the Es on one side of the air gap, and all the Is on the other side of the air gap.
A 'C core' single ended OPT is similar, there are air gaps between the two 1/2s of the laminations, to allow for the DC current.
Ok thanks, makes sense now that I see its all AC. Unburdening the OPT from DC is the main difference so it might expand my transformer choices based on that but not the impedance.
Windcrest77, I hope you are also trying to get a good appreciation of the LF response quirks that may arise in a parafeed amp, especially if you want to use feedback around the output stage or even from the speaker side. The parafeed choke introduces phase shift that can bite depending on the LF response you are aiming for.
The parafeed circuit low frequency phase and low frequency response depend on several things:
Plate choke
Parafeed coupling cap
Inductance of the parafeed output transformer primary (normally high in a parafeed OPT).
And . . .
The Reflected load of the loudspeaker's various low frequency impedances versus frequency (these reflected load impedances are in parallel with the normally high inductance of the parafeed OPT primary).
Example, an 8 Ohm load on a 5k to 8 Ohm OPT reflects 5k to the primary; and a 4 Ohm load on the same 5k to 8 Ohm OPT reflects 2.5k on the primary.
Many 8 Ohm rated loudspeakers can have minimum impedances of 6 Ohms or even 4 Ohms at some bass frequencies.
Plate choke
Parafeed coupling cap
Inductance of the parafeed output transformer primary (normally high in a parafeed OPT).
And . . .
The Reflected load of the loudspeaker's various low frequency impedances versus frequency (these reflected load impedances are in parallel with the normally high inductance of the parafeed OPT primary).
Example, an 8 Ohm load on a 5k to 8 Ohm OPT reflects 5k to the primary; and a 4 Ohm load on the same 5k to 8 Ohm OPT reflects 2.5k on the primary.
Many 8 Ohm rated loudspeakers can have minimum impedances of 6 Ohms or even 4 Ohms at some bass frequencies.
I see, so using a gapped OPT is kind of a waste if you're doing parafeed, may as well save that OPT for a different conventional SE project? Not having to create the air gap you can get a transformer with more exotic materials, more inductance, etc. right? So who makes "parafeed OPT's" these days? Or do you just order a transformer from Jack Elliano with no gap? I see why parafeed is popular with headphone amps, you can use line level output transformers, for that matter even line level input transformers as long as the secondary matches the phones and you're reflecting the right tube load. Lundahl should have something too.
Sowter made parafeed transformers.
https://www.sowter.co.uk/cgi-bin/ss000001.pl?page=search&SS=parafeed&PR=-1&TB=O&ACTION=Go!
https://www.sowter.co.uk/cgi-bin/ss000001.pl?page=search&SS=parafeed&PR=-1&TB=O&ACTION=Go!