Add an input transformer and voila, meet Sue Parker's Zeus. She tried many xfrmrs, Sowters, wind your own EI, Toroids. Check out her thread - Zero Feedback Impedance Amplifiers - still active since 2004. 
Then you miss her use of power toroids as output inductors in Zeus Amp - Toroid Transformer Version.
Hi Daanve,
Since this is a center tapped transformer, the currents cancel and no air-gap is needed. Can we simply use a closed toroidal core such as these wound silicon stell ones found here? How do we specify the correct size toroidal core?
https://www.semic.cz/!KATEGORIE/6K/Catalogue Silicon steel cores.pdf
Since this is a center tapped transformer, the currents cancel and no air-gap is needed. Can we simply use a closed toroidal core such as these wound silicon stell ones found here? How do we specify the correct size toroidal core?
https://www.semic.cz/!KATEGORIE/6K/Catalogue Silicon steel cores.pdf
Bridgeport Magnetics OA-500 GOSS (grain oriented silicon steel) toroidal core good for 600-1200VA and 5.56in OD x 2.56in ID and weighs 6.4lbs ea is available as a blank core for $50. I think I will get a couple to try winding my own. 14ga magnet wire from Remington Industries is on its way. Doubling up the strands as a bifilar and then winding it will ensure identical numbers of winds for a center tap with matching DCR and inductances.
These Bridgeport’s are nice in that the vary they sheet steel to give a round cross section. This prevents sharp corners and pinch points so that you have fewer issues with insulation getting cut at the sharp corners. Also, easier to wind and more compact.
I think this transformer is going to be the key to “Finding SuSyLu”!
These Bridgeport’s are nice in that the vary they sheet steel to give a round cross section. This prevents sharp corners and pinch points so that you have fewer issues with insulation getting cut at the sharp corners. Also, easier to wind and more compact.
I think this transformer is going to be the key to “Finding SuSyLu”!
Last edited:
Generally speaking, bifilar winding is not such a great idea for balanced operation due to parasitics being maximized, but in this case the choke only needs a few hundred turns so there won't be much of total parasitics. However, I'm not that knowledgeable on iron design and tricks, you want the iron hardcore like 50ae, 45 or wavebourn if you want better tips.
In the old days I wound a similar output transformer serving in a MOSFET push-pull amp that still works. At that time there existed tables for different core sizes stating turns/per volt specified for 50Hz (Europe). So I calculated my max output voltage, multiplied with turns/volt and at the end multiplied by a factor of 2 gaining safety against core saturation down to 25Hz. Done
No. Obviously you are talking of tube xformer with their high impedance of several kOhms. In a solid state output impedances are decades lower, inter-winding capacitance is no issue at all, but stray inductace should be kept low. Bifilar windings are the best you can do here.
Yes, agree, bifilar is by far the best, and also ensures that your two windings are identical. The parasitic capacitance will not be more than 200-300pF, no more, and your turns should be around 200-300.
To find CT, you connect one start winding to the end of the other.
The problem with winding toroids is that building a decent winding you have a lot of difficulty routing the wire. A commercial shuttle is needed for speed, and not economical for low numbers. EI is FAR easier as you can wind onto a former, then simply slip the former with winding into the core, then close it with interleaves lams. Much easier to build.
HD
To find CT, you connect one start winding to the end of the other.
The problem with winding toroids is that building a decent winding you have a lot of difficulty routing the wire. A commercial shuttle is needed for speed, and not economical for low numbers. EI is FAR easier as you can wind onto a former, then simply slip the former with winding into the core, then close it with interleaves lams. Much easier to build.
HD
And using a c-core is even easier, as you don't have to fill in all those EI lams!
Also much more practical to tune the airgap.
My friend is using the SU range for these applications, that is with two bobbins on the core.
When wound with care these two separate coils have exactly the same DCR.
Bobbins are straight, so they can be wound with maximum control, unlike toroidals / 0-cores where winding "cross".
HD, take a look at your late fellow countryman's site Patrick Turner; he used c-cores whenever possible for a reason.
I should have mentioned this page sooner; very well documented examples of inductively coupled hybrid amplifiers:
solidstateamps5+tube-input
Also much more practical to tune the airgap.
My friend is using the SU range for these applications, that is with two bobbins on the core.
When wound with care these two separate coils have exactly the same DCR.
Bobbins are straight, so they can be wound with maximum control, unlike toroidals / 0-cores where winding "cross".
HD, take a look at your late fellow countryman's site Patrick Turner; he used c-cores whenever possible for a reason.
I should have mentioned this page sooner; very well documented examples of inductively coupled hybrid amplifiers:
solidstateamps5+tube-input
Last edited:
That's the theory, yes
.Practice however is different: you must cope with some DC current unbalance as it will be there sooner or later (or always).
Non-gapped toroidals O-cores easily shoot into saturation.
EI chokes have more or less a "natural" airgap between E and I parts, and with a c-core you have full control over the gap.
So, also for push-pull applications, output chokes/transformers need some air-gapping, be it for tube amps or solid state.
Yes, the gap is in the middle of the two c-core halves and thus "screened" by the coils.
Gap material can be paper, plastic, pertinax as long as it does not conduct.
Mounting material and band clamps are available (sorry German):
https://www.arzner.de/assets/uploads/Gruppe7.pdf
Gap material can be paper, plastic, pertinax as long as it does not conduct.
Mounting material and band clamps are available (sorry German):
https://www.arzner.de/assets/uploads/Gruppe7.pdf