[Patrick Turner]

Sorry I have not responded earlier, been busy, work, kids you know the
drill.
I got my c-cores from ebay, I know it was silly but they dont seem
bad. The guy I got them from dissapeared from ebay. He or they seem to
be back now as JLW tools. The cores arrived from China in five days to
the UK!!. Be carefull though if you want some they seem a bit
optimistic with the specs. I went for the next size up from what I
needed and Im glad I did. I know you dont need as much iron with c-
cores but the ones I have seem to be about the same size/weight as the
core of a 150VA torroid so im thinking they should be fine. They are
their "core 40" size. They came with bobbin tubes which I will use
with added side cheeks. The cast ally (if you can call it that) base
they supply is very poor. It could be cleaned up but I think it is too
rough to be worth the effort. The cores on the other hand seem OK.
I agree about the control tranny. I just have a few lying about as
Im an industrial electrician by trade. I might wind my own on torroids
after reading a few posts on winding them. Your use of variouse
trannys to get the correct voltage/power is a good way to get what you
need at a reasonable cost. I have been thinking of building a line
stage(eek) to drive some long cables. It will need plus and minus 100V
or so and 4 10VA 0-18-0-18V transformers work out perfect.
Mat

Good luck with your OPT winding. If it all comes out right, you'll deserve a medal because so few ppl are brave enough to discuss their OPT winding activities here.

C-cores can be deceptive because when you see a pair of O's on the bench the pair may look bigger than an equivalent core using E&I. But on close examination, the Afe of the double C-cores can be much smaller than the Afe on E&I lams. The rules about Afe and turns must be maintained on any sort of core you use. Usually, the ratio of window size to Afe for C-cores favors using more turns and thicker wire than for E&I, thus making it possible to reduce winding resistances. The idea is that most transformers are for mains and the makers want the weight and winding losses to be minimal while driving the iron to Bac = 1.4Tesla. This makes C-cores compete with toroids. But with OPTs, The greater winding traverse width slightly reduces leakage inductance. The higher window height allows thicker P-S insulation which reduces shunt C.
Using my OPT design methods should give you an OPT that's better and bigger than most amp company accountants will allow into their products.

The other good thing about OPT C-cores is that they allow air gapping.
Maybe you think this strange because air gaps belong on SE amps only.
Well, the permeability, µ, of a pair of C-cores can be up to about 12,000..
This would be for fairly good GOSS material used. Coming from China, maybe the µ is unknown, and you'd have to measure it after assembling the tranny and applying say 100Vac across primary at 50Hz and with 10r0 in series to measure the primary coil current, all with no loads connected.
The XLp reactance = Vac across coil / Current flow.
From that, the Lp = XLp / ( 6.28 x 50Hz )
Then you need to look at the inductance formula at my site and use that to find what µ must be to give you the Lp you have measured. Now µ may anywhere from about 3,000 to 12,000 depending on the C-core material used, which should be cold rolled GOSS, well glued up, and with cuts polished. I have measured 4,500 in one lot of C-cores made in Oz which I bought in 1996. Obviously, the grade of GOSS was lowish, but it did not matter one bit because those C-cores were used in a pair of SE32 amps with 13E1, and then the gap reduces µ to under 500.
However, for PP, having very high µ is of little benefit because any Idc imbalance can cause core saturation with the net Idc flow in one direction. The DC magnetization, Bdc, is proportional to turns x µ x Idc.
So, its best practice to keep the µ down to around 3,000, and if you have followed all my rules on Fsat 20Hz, then µ = 3,000 should give you plenty of primary inductance. To adjust the µ downwards you may need a very thin air gap, thinner than an average sheet of paper which is 0.05mm. I have used some shopping bag which is thinner, but just use a small bit for the gap, and leave enough space for varnish to run in around edges of C-core cut surfaces. At F below 20Hz, the core will act as an inductance rather than a saturated reactance and stray VLF signals have less chance of causing saturation, and recovery after overload is better and without the severe effects of possible RF bursts generated in OP tubes when faced with trying to drive a signal into a short circuit, for part of the wave cycle.

Patrick Turner.