More DIY Amp Design & Fettling.

[Diabolical Artificer]

Built this monster this year after winding a 1KVA toroid ( it didn't need to be as big but has numerous unused rails for use in possible future amp builds) by hand to supply two monoblocks at 500/600v @ 1.5A. Also wound the choke myself using a no maths, no design approach that consisted of finding a big transformer with a good bobbin, taking off the windings then winding as many turns at the right gauge as would fit. Result a 30H 1.5A choke.

Monoblocks are parallel push pull 807's in Class A driven by a 6BL7 - a valve capable of 10w by itself with a DET 20 on the front end configured for a high open lop gain of 120. There's a few transistors in the circuit to annoy purists configured as constant current sinks and an active load for the DET20. THD is 0.3% @ 50w frequency response flat to 60khz mainly in part to quite a bit of NFB and excellent OPT's made by Joe Bogatek in Oz.

The PSU is delayed start this is done by a 6N30 delay valve as used in Tek 500 series scopes. The 300v for the 807's screen grids is regulated by a valve PSU. Interconnects are those military heavy duty jobbys made by Cannon etc, a bit overkill for the job but the price was right. Wooden boxes are oak & were made for me by the bloke I sold my EL34 120w amps to as part of the price, saved a lot of hard work.

There's still some work to be done, protection circuits as well as few little tweaks to make them unconditionally stable. See attached pics and video's on Utooob - https://www.youtube.com/playlist?list=PL...mCD7Jn5J92 .

Andy.

[PerdioPal]

That is particularly daft, I love it! Hand winding (did you?) big toroidal transformers is much easier than little ones, thicker wire, more room and fewer turns, win, win, win. One valve amp I did used an off the shelf 55-0-55 toroid, that was doubled for the HT using silicon diodes and a few turns added for the LT, I recall 6 turns a volt not too arduous. The HT turned out at about 330V, spot on.

[Diabolical Artificer]

Thanks. Yep, hand wound the toroid, as you say the bigger the core the less turns per volt, I think in this case 1.7T per volt. Hipot tested to 10kV between each primary and secondary. The only thing I didn't do was fit a screen as I had no copper tape, but no worries, power supply hum & noise is low.

[Kalee20]

Also wound the choke myself using a no maths, no design approach that consisted of finding a big transformer with a good bobbin, taking off the windings then winding as many turns at the right gauge as would fit. Result a 30H 1.5A choke.

That's a serious choke! But with no-maths, are you sure it's 30H 1.5A? It may well carry 1.5A, but without exhibiting 30H due to core saturation.

I'm looking at the design of a 2.2H choke at 2.5A, which is much less energy storage than yours, and it's going to be over 10kg.

[CambridgeWorks]

That should wake Molly up if you need to!
Where I used to work we used quite a few 28V 20A toroids. Primary was wound on the outer layer.
Rob

[PerdioPal]

I wonder why the primary was on the outside? Surly you could stock up on pre wound primaries and then add suitable secondaries.

[CambridgeWorks]

I don't know. They were decent quality, fabric tape and varnish over windings. Designed for Mil spec transceivers. However, same looking spec also used for ordinary commercial products.
I only found out when a radio in for repair a non-spec screw about 20mm too long that had been inserted and run straight into the winding. A black hole and blown fuse. Transformer was scrap.
Rob

[ppppenguin]

When I started to read Rob's story I thought it was going to be the rookie error of creating a shorted turn by means of a screw through the middle of a toroid and both halves of a chassis. Can also be done by using a metal bracket to hold down a toroid.

Easily done, and you only do it once. I was educated about it before making the mistake so I was lucky.

[CambridgeWorks]

Ours used to have a pair of long very shallow "u" brackets held on under the final layers of fabric tape. These brackets had fixing holes either end. The largest transformer I saw was about 9" diameter. Used in a 19" rack psu, 2 of which when in parallel powered a 265W am solid state vhf tx.

Rob

[Diabolical Artificer]

"That's a serious choke! But with no-maths, are you sure it's 30H 1.5A?" No, measured inductance is at no load but I did test it with DC trying to saturate the core, I got to 2.5A and gave up which surprised me, because as you say chokes for this sort of load are usually much bigger. I wound it to answer a question,"what would I get if I wound a load of turns on a bobbin" the answer 30H at 100hz.I didn't even bother much with the gap, I used the inners out of Rizla's, bout 2 thou at a guess as gaps. My main concern was can it supply two 50w ish amps at full whack + a bit when in a CLC filter, the answer was yes. The load on it will be nowhere near 1.5A, OP stage quiescent is 2 * 200mA rises to 300mA full whack, before clipping . So, yes no maths involved in the choke "design" but some fudging of specs.

I have several toroids out of Satelite PSU's that have the primary on the outside, which is a pain as normally it's just a case of removing a 100 turns or so of heavy gauge wire, popping on a fewsecondarys and bish bash bosh you have a good mains tfmr.With the primary on the outside you have to unwind to the core.That said that means you can overwind the tfmr at a lower Tesla rating to get better regulation and offten a better wound tfmr. Some of the toroids I've re-used are badly wound. this toroid does dim the lights and the PC's has a hissy fit despite a heavy slow start circuit to limit I into the HT big caps and a big thermistor to limit inrush into the toroid.

Andy.