Antek's new trannys?

[WA4SWJ]

An example:



Once the core is sawn various techniques are used to set the air gap and then wind the coil with a toroid winder.

There are other ways to do it as well.

Regards,

[Blair]

Here is a small concept drawing for a triode 100w amp. I can get the B+ up a little if necessary by dropping the 15 ohm resistor all together. For some reason Duncan's PSD II was bogging the B+ down to 465-80v when the CC was set at 450ma. I'm not sure why. The Antek tranny has 800ma available so I figure it will not sag, but for general purposes, I had to change the value of the CC to 250ma.

Also, some suggestions as far as the choke is concerned. The 193Q 10H 500ma is a large choke, and it yeilds a nice smooth B+. The 193N 3H 500ma is much smaller, and gives a similar ripple. Is there any justification for spending the extra $60 on the bigger chokes?



Does all of this look OK as far as layout and grounding schemes are concerned. I'd like to do this once if possible Yellow_Light_Colorz_PDT_02

Thank you,

Blair

[Geek]

I dunno just how well a sawn toriod would work (or even stay together)

Wouldn't it just act like any other C-core then, just woefully more expensive?

[Blair]

I took this power stage and added another set of tubes. Will this work? I want to use a simple driver stage, and I'll probably use something from an early dynaco. Shannon's boards are nice, but I was wanting to try point to point on the entire amplifier.


Any thoughts or suggestion?

P.S. Does anyone have the data sheet for this Sovtek 6550WE?

http://www.tubedepot.com/so-6550we.html

Will it handle the 500v B+?

Blair

[EWBrown]

The circuit basically looks god, with the follwing caveat(s):

Having a single fixed bias adjustment for a set of three tubes pretty much mandates that the tubes be very well matched, else one or two of the tubes will "hog" more current than the remainder. The best approach for the fixed bias is to make three of your exixting bias circuit, so that each tube can be set properly.

I'd change the 30 ohm cathode resistors to 10 ohms, and skip the string of "horizontal" cathode resistors, and take the current readings at each cathode, rather than using the combined readings as the "horizontal" resistor network would do. The 10 ohms simply makes reading the cathode current easier, as there is no extra math involved, With 10 ohms, 50 mA = 0.5V across 10 ohms, etc. With 30 ohms it would be 1.5V for 50 mA, etc.

Which leads to:

I'd either go with individual bias adjustments for each tube, or use cathode resistor biasing, which has the benefit of being more "forgiving" but at the same time, it lowers the plate to cathode voltage which may or not be a disadvantage.

AS far as a driver circuit, anything which can drive a PP pair of 6550S should also drive the sextet of them.

/ed B

[Blair]

OK Ed,

Based on your suggestions, this is what I did. Does this look better?


Thank you for looking at this. I was wondering about the two bias pots for six tubes.

Blair[/img]

[EWBrown]

I am sure you already know this, the inputs to the G1 grids of the third through sixth tubes should connect back to the 0.22 uF input coupling caps (like the first pair).

Other than that, it looks 100% good to go!

The actual bias current readings can be taken from the "top" of the 10 ohm cathode resistors, I'd probably use some inexpensive through-panel "pin jacks" for that purpose, one for each tube. And one connnected to the common ground return.

I presume (without actually trying to calculate it out) that the resistance values in the bias networks will give the correct biasing range for the chosen tubes.

To answer an earlier question, the 10H 500 mA choke would give 3.33X the ripple reduction of the 3H 500 mA choke, so it is a matter of choice.

The effective ripple reactance would be 2*PI*F*L or 6.28X120X3 (2262 ohms Z) or 6.28X120X10 (7540 ohms Z) ohms at the 120Hz ripple, while it also maintains the choke's DC resistance which would be much lower.
Be sure to use some hefty rectifiers, the usual uF4007s won't hold up too well in a huge PSU like this one, the bare minimum would be uF5408s, and some 6A or better 1600V HEXFREDs, though not exactly inexpensive, would perform very well. Don't forget a 270K, 2 or 3 watt bleeder resistor to help discharge those electrolytics, so there aren't any HV stored charge "surprises" :o

Any VA / LTPI driver circuit should work with these tubes, something based on Shannon's ST70 or Poseidon driver, the VTA driver, or anything similar to those. A 7247 (like ST35) won't have enough drive voltage capability, and the classic ST70 or MKIII with 7199 or 6AN8 needs the oentode VA in order to have enough drive voltage, and most pentodes aren't the best choice for decent sound quality.

You could also refer to the Eiclone manual, and use that input and driver circuit pretty much as-is.

A 12AX7 / 12AU7 (12BH7 or 6CG7) or 12AT7 / 12AT7 driver would be the best choice.

The Sovtek (and Elektroharmonix) 6550s should be pretty close to most of the newer manufacture 6550s, and the vintage "old school" 6550s would probably be a bit more rubust, but the Sovteks, EHs or Svetlana 6550Cs should all be good for this application.

It looks like what you ill have here is essentially 3/4 of a VTL (Vacuum Tube Logic) MB450 monoblock, set to triode mode. It should do 100WPC easily, and perhaps a tad more, up to 125-150Watts before it gets driven too hard.

/ed B

[EWBrown]

I ran a quick calc, (using SE CAD) and for a B+ voltage of 500VDC, and a cathode current of 50 mA, each 6550 would require a 1K, 10W ressistor (properly bypassed) which indicates a peak grid drive voltage of +/- 50V PK tp PK. That runs the 6550s with around 450V (or somewhat less, depending on the PT's primary resistance) from plate to cathode voltage, which is well within its normal operating range.

This is only an approximation, give a +/- 5% "fudge factor" so somewhere between 47.5V and 52.5V cathode-to-grid differential.

I'll try it with fixed bias later...

/ed B

[Shannon Parks]

Hey Shannon,

Just a few questions.

1) Are you working on the Ikezilla in fullboard form...i.e.. will there need to be all of the lead wires like in the forum post?

I'd like to kill two birds with one stone and basically revise the Ike but add some tie-ins for a second push-pull pair. As far as the extra bias circuits I haven't brain-stormed tha tpart yet. Yellow_Light_Colorz_PDT_06

2) Which Antek trannys are you looking at? Are your going to be pentode amps with a triode switch? How audible is the difference in distortion between pentode and UL? I woudl prefer not using 6 KT88s per amp! :o

Probably something similar to the Hammond PTs from the Ike project. I won't want to have a B+ more than 525V. Actually pulling the test-Ikes from the closet at this time and moving them to the new diytube lab this weekend!

[Blair]

I'm not sure why it is not letting the new URL change the schematic just like I did earlier Yellow_Light_Colorz_PDT_21 , but if you could look at it and see if I highlighted the test points in red correctly please.

Also, I pulled the bias circuit off another amp schematic. What is the proper range? -60v or so?

As far as the chokes are concerned, I may as well put the big ones on if I'm going to build this. May as well pay the doctor as well. These shoudl weight about 60lbs each :o

Do you have a link to the Eiclone? I cannot find the manual....

Also, you said these may get a little over 100w. Will that damage the OPT is the opt is only rated at 100w. Not that I will use 100w much for anything.

Thanks Ed for all your help!

Blair

[Blair]

Thanks Shannon!

Unfortunately, I'm leaning towards a Poseidon driven 100w PPP triode design now. Because god knows I need to buy one more quad of tubes!!

Blair

[kheper]

Antec 100W trannies on ebay: $170 BIN

Not a bad price considering the rusty rot that goes for plenty more.

If they only had screen taps...

http://cgi.ebay.com/100W-Push-Pull-Toro ... 24810201QQ

[Blair]

Those are the ones I'm going to use. Since I'm wiring the amp with six tubes per channel in "triode" they are fine. I hope he begins producing screen tapped trannys too though. If I find anything out, I'll let everyone know. I'll call him when I place my order.

Blair

[EWBrown]

the main concern with the OPTs is to not exceed the rated plate currents through the windings - I didn't see a published current rating on the e-bay or ANTEK specifications. If the OPTs are engineered like his power trannies, they should have plenty of reserve capability.

Exceeding 100 watts isn't a real problem, most OPTs can go well over the published power ratings, , just the lowest and highest octaves will be effected first. The difference betwen 100 and 150 watts isn't all that audible, anyway, just a matter of a couple of dB. SPL. Hardly the difference between earaches and broken eardrums :o Yellow_Light_Colorz_PDT_02 Yellow_Light_Colorz_PDT_08 Yellow_Light_Colorz_PDT_04

Best to run it a little conservatively (why I used 50 mA as a target figure) That would keep the current at 150 mA per siide. .

The VTL 6550C based amps use higher B+ voltages, but then they run the cathode current slower, around 30 or 35 mA for 6550Cs. They can be run up to 75 mA, if the OPTs can handle 225mA per side.

looks like the PSU would hardly "break a sweat" with that 400 VA power transformer.

/ed B

[Blair]

Thanks Ed,

Does the new schematic in the previous post look better with the connections to the bypass cap and the bias test points in red?

Blair