Toroid Output Transformer for Bass Amp

[CJ]

not too many toroid output transformers out there for power amps,

why? probably a lot of reasons, cost, need for a toroid winder and somebody to run it, large size, fab work needed for can and mounting,

but the main reason is the lack of a sufficient air gap to prevent any DC imbalance current from saturating the core,

however there are some good things about these cores that make an effort of this kind well worth it, low excitation current? that is certainly a plus but not something to worry about in a power OPT,

what about very low leakage inductance? these cores allow the wire to surround them very well, with an E-I lam, there is quite a bit of core (the legs) which has no wire around it, you can fake a toroid by using a U-I lam so that the two coils surround the alloy more completely, but  there is no wire wrapped around the top and bottom of the core,

so a toroid core should make a great bass amp OPT as you get better coupling and thus more Henries for an incredible low frequency response,

somebody came into the shop the other day with an Eggnater Tourmaster guitar amp that had a blown pwr trans due to improper winding, this is a known factory defect for these amps, so we get a replacement at cost, boy, this transformer has about 30 leads coming out of it for the triple pwr range and international pri voltage options, plus low voltage winds for the solid state and logic, plus tube heaters, sheesh, what a pain to do the leads,  but an OPT will only need 5,

so now we have a junk xfmr that came off the amp that we can use for our experiment,

do we want to unwind all those turns off by hand, hell no, that would take forever,

hack saw? getting too old,

so, we call up Uncle Dewalt and put on some ear goggles,

looks like a  pumpkin, or Moe's hairdo,

grind til you see a few sparks, that would be the core,

 

[gyraf]

Looking forward to the hack!!

My local transformer guy tells me that for toroid power output transformers you will really REALLY want the cheapest Chinese toroid core material - the type that is low-grade because it's wound from left-over strips from "real" core manufacturing. This leaves kinda semi-gapped properties, making the core less hysteric about slight bias mismatches and offsets...

Jakob E.

 

[CJ]

thanks Jakob! i believe that is what we have, so that is good to know,

so we have the wire off, this core measures about 5" OD, 3" ID, and 2.5" high, probably about 8 lbs, plenty good for a pair of 6550's  KT120's? sure, but those new Tungsol's blow up if you look at them wrong,

we hand wrap about 76 turns of #26 to get an AL number,

what is AL? it is a core constant that we can use to tell us how many turns we need,

our 76 turns give us about 1.8 Henries at 20 Hz and 7 volts AC,  about 31 ma current,

Step 1 - Calculate the Henries from 76 Turns:

crank up generator to max at 20 Hz, measure volts and current across 76T wind,

7 V/.031 ma = 225.8  take this XL inductance  reactance  number and use the formula XL=2pifH to get Henries,    225.8=6.28 x 20 x H    so  225.8 = 125.6 H,  H=1.8

Step 2 - Calculate the AL number

Henries = N^2 x AL (N=number of turns)

so for our test wind of 76 turns, we have 1.8H = 76^2 x AL, 1.8=5776AL,  AL= 1.8/5776= 0.00031

some folks put H in nano Henries to get a bigger AL number, but we can pick our own units for DIY,

now we can estimate turns for our desired inductance at 20 Hz,

but what is our desired inductance?  well, we want to beat out every OPT out there, even the Ampeg bass amp stuff,    that ay we can have an amp that will blow away anybody on the circuit, even the funky SWR Class D 2000 watt shoe box amps you see nowadays,  ;D

going through our OPT scrapbook, we see that the highly regarded Ampeg V4 amp, which was the inspiration for the SVT (just add 2 more tubes)  has a pri ind of about374 Henries at 20 Hz,

we got a Traynor YGM-3 OPT with about 274 H at 20HZ,

that Eggbeater Tourmaster has about 126 H, Marshall Super Lead clocking in at about 206 at 20,

what if we shoot for something crazy like 1000 Henries?  ;D

sure, why not, how may turns? remember inductance goes up as a function of turns Squared,
so you get an expotential graph, like the mass of an object as it approaches the speed of light, (WTF?)

Step 3 -  Calculate Pri Turns

1000 Henries,  AL=.0003  so 1000=.0003 x N^2 

N^2 = 1000/.0003 so N^2=3,333,333 so take the root of that and we have

N pri=1825,    not bad at all,  we have a push pull amp,
so coming of the center tap will be 912 turns, going each way,  this will fit easily onto this core,

what will be the sec turns?  just pick a common turns ratio out of the scrap book, most OPT's for big tubes hover around the 20:1 mark, so

Sec Turns = 1825/20  so

Sec Turns = 91 ,  seems like a common number for an 8 ohm winding,

so what will the tubes see? 20:1 = 400:1 Z ratio,

8 times 400 = pri  Z = 3200 Ohms,    too low for 6550's?

remember we have 1000 H,  so  our low freq XL= 6.28 x 20 x 1000 = 125,600 Ohms,

so how much current gets wasted into the filter caps?

what is our plate voltage about 450?  somebody said we can swing twice that with an inductive kick,

so full blast, amp on 12,

Wasted current is  = 900V/125,600Ohms  = 0.00716 so we waste 7 ma,  at 900 that is about 40 milliwatts , if we juice the tubes for 150 ma, we can get .15 x 900m = 135 watts, peak power,

what size pri wire?  anything on the shelf that will take about 100 ma, we got an 8 inch turn length,
1200 feet of wire, call it 1000 for out lookup table, #26 gives 40 ohms, at 255 ma,
#30 gives 103 ohms at 100 ma, so that or anywhere in between,

sec wire will be 8" x 91 turns = 60 feet,  #18 will come in at about 6 ohms per 1000 feet,

so sec DCR = 60/1000 x 8 = .48 ohms, kind of high but we will be crankin out some volts,

900 volts peak/20 = 45 volts sec voltage,    how many watts?  45/8 ohm speak = 5.65 amps,

voltage drop is 5.65 x  .48 ohms wire plus cables, = 2.7 volts,

so figure 40 volts and 8 ohms when the tubes are new,  40/8 = 5 amps, I^2R= 25 x 8 = 200 (peak),

maybe a pair of Electrovoice SRO 15's, or an SWR Godzilla cab with six 10's, 

now we have to build a quick and dirty toroid winder, just a turn table with a hole in it, and a pulley wheel we can cut a slit in to load wire onto, .....hmmm

 

[CJ]

here is a link for an amp that uses the Plitron toroid designed by Vanderveen i believe,

http://www.normankoren.com/Audio/TENA.html

article by Mr Vanderveen.

http://www.next-tube.com/articles/Veen/VeenEN.pdf

 

[gyraf]

Seems like TENA uses a DC-servo to ensure balanced bias: Intuitively a good idea here, I think..?

 

[CJ]

yes, gotta keep the tubes balanced, maybe a twin bias pot with a little nulling meter,

using this for bass helps as the high flux due to low f will hopefully provide some degaussing action,

also have to be careful to keep the symmetry of the pri wind,

 

[CJ]

here is a KT88 amp which uses the toroid OPT>

https://web.archive.org/web/20130805141923/http://www.jumpjet.info/Pioneering-Wireless/eMagazines/VTV/VTV19.pdf

and we just ordered some #16 magnet wire for the speaker winding to get the DCR down to around 0.2 ohms instead of 0.48,

damping factor improved? and we can use some big wire for the speaker cable,

here is a pdf spec sheet on the Plitron 4142>


http://r.search.yahoo.com/_ylt=AwrTHRkTLihWpRwA57RXNyoA;_ylu=X3oDMTByb2lvbXVuBGNvbG8DZ3ExBHBvcwMxBHZ0aWQDBHNlYwNzcg--/RV=2/RE=1445502611/RO=10/RU=http%3a%2f%2ftec-sol.com%2fproducts%2felec%2fplitron%2fspec%2fPAT-4142.pdf/RK=0/RS=Wohu2yoZRteXZJgq0VbKtyiNovo-

 

[gyraf]

yes, gotta keep the tubes balanced, maybe a twin bias pot with a little nulling meter

Am I detecting a motorized multiturn pot here?  ;D

Jakob E.

 

[CJ]

you mean for intentionally inducing saturation with DC offset via a foot controller?  ;D

almost done with the toroid winder, using a bicycle rim for the wire shuttle,  big sucker, 26 inch,

rubber rim liner will work for wire feeder,

used empty 6 inch wire spools cut in half and lined with foam for the core core clamps,

thinking of winding opposing pies for the primary, wind half way around the core for each side of the P-P,

W.E. does the 111C like this,

 

[gyraf]

Quasi-handwinding toroid?

WE WANT PICTURES!!!!

 

[Gus]

Are you going to run the output class B? 

 

[CJ]

probably class A = cathode resistor, 

 

[CJ]

you shall have pictures!  ;D

 

[CJ]

ok here is the core holder being thrown together,

found the wood on the side of the neighbor's house, that's fair game up here,

and those spools were excess baggage that we did not want to claim,

use shipping tape because it sticks to foam real good,

 

[CJ]

so the core holder spools simply slip over some threaded rod, you could go hi-tek with sealed bearings and such, we are trying to do this on the cheap, trying to stay out of the hardware store,

cut the rim at the seam, it bends easily around the core, one shuttle wheel mounted,

Jovil want $8,000 for a winder, a little bit out of our range for just one coil,

using a stick loaded with wire works, but can be a pain when you need 800 turns each pri,

 

[CJ]

2nd shuttle wheel bolted in, 2 x 4 conects both wheels so they stay on the same plane,

wood dowel is used for axles, file down the radius to fit spools and then use wood screw to hold it to the back plate,

one more wheel needed on the bottom of the rim, and we need to trim out some wood with the jig saw, have to make a cutout for the wire to go thru in order to wrap around core,

we can get six feet per rev on the bike rim, and enough capacity to string up a small suspension bridge,

probably do  1/2 primary at a time, 800 turns, mean length per core turn is about 8 inches, so

800T x 8" = 6400"

inches to feet:  6400"/12= 533 feet of wire per primary, 

our wheel holds 6 feet ea. so 533/6= 88.8 wraps of wire on the rim,

what if we run out at turn #783?  so add a few revs, maybe 95 turns of #29 which will easily fit on that rim,

when you load up a commercial toroid shuttle too full, you have some real problems, like a birds nest happening right in the middle of the wind,

we could add a stepper motor and belts to spin the cores but that would defeat the purpose of doing this cheap,

that foam keeps the spool hub from mashing the turns as they get put on, a commercial machine has the core holder wheels on a spring system so the wheels expand outward as the wire is put on, that could easily be implemented if you are going to do these on a regular basis,
that core rotates when you turn on of the wheels, holding a core by hand  while passing a stick thru it is kind of a pain as it wants to roll all over the place,

a motor with a rubber wheel on the rim could help wind it up, gotta check the neighbors garage,  ;D

 

[gyraf]

Outright fantastic.

Thanks.

..and please more pics of the process if possible...

Jakob E.

 

[Gus]

Found this
http://www.aguilaramp.com/pdf/support_manuals_db359.pdf

 

[CJ]

thanks Gus! good to know that this will work,

here is another one from Eden who makes some good bass amps>

 

[CJ]

well, we finished the winder tonight, but it was a dismal failure,  ???
tooi big a wheel to make the wire too hard to manage,
it works, but takes too long to do one turn, the only good thing about it is the wire storage, easier to wind up that rim than a threading stick,

we do have some shots of the effort,