"Wound core needs to be heat annealed, so plastic ribbon is very likely to melt and burn."
More likely MgO paint coating on ribbon faces for a distributed gap. But as you say, it's not adjustable.
A cut toroid gap is likely to change dimension with time, since the ribbon is like one big coil spring. Usually held by a final spot weld (but not working with a cut). Cut C cores must have some kind of bonding between the layers.
More likely MgO paint coating on ribbon faces for a distributed gap. But as you say, it's not adjustable.
A cut toroid gap is likely to change dimension with time, since the ribbon is like one big coil spring. Usually held by a final spot weld (but not working with a cut). Cut C cores must have some kind of bonding between the layers.
LinuksGuru,
Thanks for the insight on Toroids!
By the way, are you a Linux user?
For my first SE amplifier, I took a push pull output transformer, removed all the lams, and re-stacked them as E, air gap, I. I know, the inductance was reduced significantly, but they worked pretty well in the triode wired pentode circuit I designed, and later in a 300B output stage.
Thanks for the insight on Toroids!
By the way, are you a Linux user?
For my first SE amplifier, I took a push pull output transformer, removed all the lams, and re-stacked them as E, air gap, I. I know, the inductance was reduced significantly, but they worked pretty well in the triode wired pentode circuit I designed, and later in a 300B output stage.
I was so curious about this case that I contacted Toroidy and asked them about the air gap in their SE OPT toroids and the result is: Yes, there is the air gap.
There is a commercial tube amplifier manufacturer in Poland that uses Toroidy OPTs and they also manufacture SE amplifier with 300B. (Mira Ceti 300B)
FezzAudio
There is a commercial tube amplifier manufacturer in Poland that uses Toroidy OPTs and they also manufacture SE amplifier with 300B. (Mira Ceti 300B)
FezzAudio
PLITRON
I do not know exactly what material they use for their SE OTP toroids, nobody knows it's their trade secret, but they use distributed air gap toroids.
Toroidy's are great!
I'm currently using this one...see attached in a Tubecad Ultra-empathic circuit.Thanks to Wavebourn,AndyEvans,Ale Moglia,Rod Elliot,ShinyEyes....for the 4P1L Soviet pocket rocket.
Goodness this valve has scary bass for such a small thing.
Here's an internal of Toroidy SE Txformer.
I'm currently using this one...see attached in a Tubecad Ultra-empathic circuit.Thanks to Wavebourn,AndyEvans,Ale Moglia,Rod Elliot,ShinyEyes....for the 4P1L Soviet pocket rocket.
Goodness this valve has scary bass for such a small thing.
Here's an internal of Toroidy SE Txformer.
Attachments
Is the problem solved? Regardless of the core type, the SE Output needs to work with DC. SE output that doesn't take DC ain't SE Output.
So, it's either you got duped or you did something wrong in your design.
Anyone has similar experience with Toriod Core SE output not taking DC in the primary?
I'm also considering this OPT from Toroidy for a 6bm8 SEP amp. The amp works great with James transformers but I'm saving those for another project. I was concered about the SE performance of the Toroidy's without using chokes. What was the upshot here, was there a mis-wiring or are these OPTs not usable in SE?
+1
I just picked up a pair of toroid output transformers made by SumR. They were advertised as single ended output transformers, and have two primary wires (3200 Ohm), and three secondary wires (0-4-8 Ohm). This is labeled on the pair, which are very nicely constructed. What else could they be with that wiring? I have not installed them in the project yet, and could put in a different transformer pair if these are going to be a problem. I doubt that they will be, but I would like to know what is going on before proceeding.
I just picked up a pair of toroid output transformers made by SumR. They were advertised as single ended output transformers, and have two primary wires (3200 Ohm), and three secondary wires (0-4-8 Ohm). This is labeled on the pair, which are very nicely constructed. What else could they be with that wiring? I have not installed them in the project yet, and could put in a different transformer pair if these are going to be a problem. I doubt that they will be, but I would like to know what is going on before proceeding.
Parafeed is their only application. Parafeed can work really well but it depends on whether you have the cash for some plate chokes, or enough headroom on the B+ to run CCS as plate loads. You need at least the peak to peak signal voltage across the CCS to make it work.
Shoog
Plate chokes are made by several companies.
Llundahl, and One Electron, are two I know of.
K and K audio in the US, and Antique Radio supply in Arizona, sell them (in that order).
I am sure there are others.
I once used a Hammond 20 Henry 100 mA power supply choke in that kind of application (as a current source).
The impedance is a little low at 20 Hz (2,500 Ohms), but rises all the way past 20 kHz.
The 2A3 or 300B have low plate resistance (800 and 700 Ohms respectively), but a pentode amp that does not use negative feedback would not work very good with only 20 Henry (5,000 Ohms at 40 Hz).
The impedance of the choke is effectively in parallel with the output transformer primary impedance.
But of course, all the above will work better than using the non air gapped toroid bare bones (with the plate DC current flowing through it).
You will need a good high voltage cap to couple from the 'plate & choke' to the 'toroid primary'. For safety sake, the cap voltage rating should be 2 times the B+ voltage.
A 4 uF 440VAC motor start capacitor can take a peak voltage up to 1.414 times the 440V,
which is 622V. That would work with a 311V B+, and will not arc over even if the tube plate swings down to 0V and up to 2 times B+.
A 660VAC cap x 1.414 = 933V, so it will be safe for a 466VDC B+.
Let us know how the modified amplifier sounds.
For more on this circuit topology, search the web for parafeed. My first memory of parafeed was back about 1997 at the Vacuum State of the Art Conference in Silverdale, Washington.
Doc Bottlehead's company has several models of parafeed amps.
Llundahl, and One Electron, are two I know of.
K and K audio in the US, and Antique Radio supply in Arizona, sell them (in that order).
I am sure there are others.
I once used a Hammond 20 Henry 100 mA power supply choke in that kind of application (as a current source).
The impedance is a little low at 20 Hz (2,500 Ohms), but rises all the way past 20 kHz.
The 2A3 or 300B have low plate resistance (800 and 700 Ohms respectively), but a pentode amp that does not use negative feedback would not work very good with only 20 Henry (5,000 Ohms at 40 Hz).
The impedance of the choke is effectively in parallel with the output transformer primary impedance.
But of course, all the above will work better than using the non air gapped toroid bare bones (with the plate DC current flowing through it).
You will need a good high voltage cap to couple from the 'plate & choke' to the 'toroid primary'. For safety sake, the cap voltage rating should be 2 times the B+ voltage.
A 4 uF 440VAC motor start capacitor can take a peak voltage up to 1.414 times the 440V,
which is 622V. That would work with a 311V B+, and will not arc over even if the tube plate swings down to 0V and up to 2 times B+.
A 660VAC cap x 1.414 = 933V, so it will be safe for a 466VDC B+.
Let us know how the modified amplifier sounds.
For more on this circuit topology, search the web for parafeed. My first memory of parafeed was back about 1997 at the Vacuum State of the Art Conference in Silverdale, Washington.
Doc Bottlehead's company has several models of parafeed amps.
Last edited:
That's just theoretical.
The choke will measure 20 Henry around 100 Hz, but at frequencies above some 1 kHz (depends on quality of the core) inductance will already be substantially lower, and there will not be much inductance left at 20 Khz; it is not a linear function following 2PixFxL.
An other, and likely better, option to be able to use the gapless output transformer, is a 1:1 interstage transformer instead of the classical parafeed LC combination.
pieter t,
1. I actually measured the Impedance (inductive reactance in parallel with distributed capacitance) of the Hammond 20 Henry 100 mA power supply choke.
I did the measurements to predict how well it would work in my amp design.
At higher frequencies the impedance did continue to rise to 20kHz.
At some point the distributed capacitance and core losses will limit the impedance.
I measured it because another audio expert said it might not work at higher frequencies.
I would not recommend using that choke to do a full 100 mA standing DC in that kind of an application. I used about 20 mA, and it worked fine.
It may be that a new Hammond 20 Henry 100 mA choke may not measure the same as the one I have. Vendors can change materials and processes.
Your mileage may vary, so measurements of any part that is used this way is paramount.
2. Purchasing an air gapped EI interstage transformer to allow the use of a non air gapped toroid seems a bit too expensive (generally more than a plate choke, but a parafeed coupling cap can be as inexpensive or expensive as is wanted, i.e. a surplus motor start cap, or 'special' performance Botique cap).
Instead of using an Air Gapped interstage transformer, just use a real Air Gapped EI output transformer, and sell the toroid transformer.
Now it is an SE design, not an SE Parafeed design.
1. I actually measured the Impedance (inductive reactance in parallel with distributed capacitance) of the Hammond 20 Henry 100 mA power supply choke.
I did the measurements to predict how well it would work in my amp design.
At higher frequencies the impedance did continue to rise to 20kHz.
At some point the distributed capacitance and core losses will limit the impedance.
I measured it because another audio expert said it might not work at higher frequencies.
I would not recommend using that choke to do a full 100 mA standing DC in that kind of an application. I used about 20 mA, and it worked fine.
It may be that a new Hammond 20 Henry 100 mA choke may not measure the same as the one I have. Vendors can change materials and processes.
Your mileage may vary, so measurements of any part that is used this way is paramount.
2. Purchasing an air gapped EI interstage transformer to allow the use of a non air gapped toroid seems a bit too expensive (generally more than a plate choke, but a parafeed coupling cap can be as inexpensive or expensive as is wanted, i.e. a surplus motor start cap, or 'special' performance Botique cap).
Instead of using an Air Gapped interstage transformer, just use a real Air Gapped EI output transformer, and sell the toroid transformer.
Now it is an SE design, not an SE Parafeed design.
Unfortunately proper plate choke's will cost as much as many SE OT, though there is a fair chance they will perform better than the gapped SE transformer.
The demand for plate choke is so low and the concept of parafeed is considered a step up from standard SE (by many) that the companies that make them can, and do, charge a massive premium.
Using a power supply choke is fraught with potential dangers as the parasitics will likely interact with the parallel toroidal parasitic's and the coupling cap to produce a very choppy response curve.
Also the parafeed cap can get expensive as most people grossly under-specify it and introduce low frequency resonances. Think 10uf rather than the typical 1-4uf which many people suggest.
All in all parafeed really only makes sense to my mind when used in flea powered amps where the costs can be managed. Here the CCS is the better option and a good 1-3watt design can be achieved with a 250-300V power supply. I fairly much guarantee that the measured result will be far superior to just about any standard gapped SE design.
Shoog
The demand for plate choke is so low and the concept of parafeed is considered a step up from standard SE (by many) that the companies that make them can, and do, charge a massive premium.
Using a power supply choke is fraught with potential dangers as the parasitics will likely interact with the parallel toroidal parasitic's and the coupling cap to produce a very choppy response curve.
Also the parafeed cap can get expensive as most people grossly under-specify it and introduce low frequency resonances. Think 10uf rather than the typical 1-4uf which many people suggest.
All in all parafeed really only makes sense to my mind when used in flea powered amps where the costs can be managed. Here the CCS is the better option and a good 1-3watt design can be achieved with a 250-300V power supply. I fairly much guarantee that the measured result will be far superior to just about any standard gapped SE design.
Shoog
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.