Output Transformer Notes, December 2008.

In this page below I have some useful notes.....
Speakers, amplifiers and the output transformer,
Description of output transformers for sale,
Output Transformer Design Parameters,
Sale conditions for transformers.

On a separate page I have provided analysis of the compared frequency performance of my own designs and those produced by Mr Flanagan at
output transformer frequency behaviour.

If you wish to buy a pair of output  transformers from those listed or from amoung many not yet listed but in stock, you need to decide if you want PP or SE, and how much total power you will be happy with. I also have considerable stocks of power transformers and chokes to suit.

I can arrange to supply a compatible set of iron wound parts including a power transformer, an output transformer and choke for B+ filtering to suit a monoblock or dual channel chassis.  Let me know what you intend doing with your project and I can assist to ensure what I have will suit you.


Speakers, amplifiers and the output transformer.


Tube amps you want to build probably will produce much less power than the latest solid state offering from Krell or tube behemoth from ARC. So when less power is considered, it becomes important to match the amplifier to the speakers, rather than rely on a huge amp capable of driving any type of speaker between 2 and 50 ohms to deafening levels even at 10% of its maximum power rating.

Output transformers operate just like a car gearbox to match the engine RPM and torque to get lower rear wheel RPM but with higher torque. Output transformers allow a typical 6550 output tube to work ideally with an anode signal voltage of 220Vrms and signal current of 50mArms which if transformed becomes a speaker signal of 8.1Vrms at 1.35Arms. The tube is happy making 11 Watts into a load of 4,400 ohms to provide 11 watts into 6 ohms at the transformer secondary. Good music depends on a good choice of turn ratios in the transformer just like smooth motoring depends on the right gear ratio in the gear box. 

You need to have a mission plan, know what you want, why you want it, and realize you have to spend some real money and a lot of time to make a tube amp.
You should know the technical specs for your speakers you have now or of those you wish to buy. 

Most people will have speakers with perhaps "nominal impedance 8 ohms, sensitivity of 88dB SPL at 1metre with one watt of input power."
I build amps to cope with the needs of most people. Under normal listening level conditions, most people find that two stereo speakers which are each being powered by 1.0 watt each of *average power* will be too loud, and their wife will complain. SPL average would be 91dB, with peaks perhaps rising to 101dB, when 10 Watts is needed in each channel. Most people would find 10 watts is enough power per channel, but its best to pay for more than what you need because the fidelity of music produced with tube amps becomes far nicer to listen to with an amp with 3 times the power needed for the loudest peaks of 101dB. This would give the amplifier system an ability to produce peaks up to 106dB with speakers rated for 88dB, 1W/1M.
30W is a starting number if you can afford it.

Loudspeakers are notorious for having varying impedances at different frequencies bands between say 5 ohms and 60 ohms for what may be
a nominal "8 ohm speaker". For something nominally 4 ohms, its common to find an impedance dip to 2 ohms.
The frequency band where such low impedances lurk are often right between the transition between bass and midrange where most audio energy is located.
The speaker maker's figures for quoted sensitivity of so much SPL per watt could be only a rough indication of the speaker's performance and at some frequencies the power required from an amp may much greater than the average quoted.
Dynamic Loudspeakers are very inefficient, and the efficiency varies.
You don't want the variations in your loudspeaker impedance to cause high amplifier distortion at some bands of frequencies.
Therefore it is useless to shop for a tube amp which was made to produce its absolute maximum power with an 8 ohm load if you buy speakers which dip to 2 ohms at some part of the audio band.
Tube amps need to cope with modern speakers. They have rapidly falling power maximums either side of the impedance value where the maximum power is produced. And at low speaker impedances, distortion is usually far worse than at high impedances.
So I always try to make my PP amps with maximum AB power produced when the load is 3 ohms if the range of loads to be tolerated at normal listening levels is above 2 ohms, ie, any speaker with nominal impedance above 4 ohms. So if you have an UltraLinear output stage with 2 x 6550 and Ea at =500V, you get about 60Watts absolute max at 3 ohms but only 34 Watts at 8 ohms but the % of class A1 will cover all your average listening needs.
See my web pages on load matching and the 5050 stereo amp and see that a transformer with an impedance ratio of 8k:8 ohms makes perfect sense with Ea = +500V and a pair of 6550 in UL mode.
Some makers have 4 ohm and 8 ohm outlets which makes it easier to match a tube amp to a speaker. Such amplifiers usually produce maximum power when 4 ohms is at the 4 ohm outlet or 8 ohms is at the 8 ohm outlet. Doing things my way there isn't any need for two lots of output terminals. Since only low power is actually needed for 90% of average listening, there will be enough power for 8 ohms speakers even though the power maximum occurs at 3 ohms.
SE amps operate always in pure class A and they are not able to produce more than their maximum class A output power as the load value reduces.
Therefore their maximum output power should be arranged to occur when the load is about 5 ohms, not 3 ohms, to suit most modern speakers. Even though an SE amp will have falling maximum output power below 5 ohms it is still all class A power and the SE amp has good tolerance of non ideal loads.
A 30 watt SE amp will usually perform just as well or better than a PP amp capable of 30 watts of AB power and with only 8 watts of class A power before it begins to work in class AB at above the 8 watt class A threshold. Most SE amps cop a lot of criticism about their poor technical performance. This is because so many SE amps are vastly under powered for what owners may expect of them with modern insensitive speakers. If you have normal modern speakers of 88dB/W/M sensitivity then don't expect one lone EL34 or 300B or with 5 to 8 watts to ever sound very good when turned up loud especially if there is "zero negative feedback".
If you have say 4 x EL34 or 3 x 300B for over 20watts and with moderate NFB, then you will begin to hear the glories of music via tube gear.
If you have horn loaded speakers of perhaps 100dB/W/M then the single EL34 or 300B is more than enough and a lone 2A3 or 45 can be just fine.
 
Nearly all tube amps operate in Push Pull mode where at least the first 1/4 of the total power maximum is what is called "pure class A" and is the most naturally clean undistorted power possible from vacuum tubes. So any typical PP amp with a capability of 32 watts AB maximum will make the first 8 watts or more in clean class A power and you want this amount of power to cover most of your listening levels with the remaining less clean "class AB power" used for the transient peaks of drum beats etc which are the noisy bits in music and which are little changed by the instantaneous rise in distortion.

Where an amplifier is classified as "Single Ended", ie, SE, it means that ALL of its power is type class A power. This also means that the distortion harmonics at all power levels may be higher than the PP amp but usually that the first 1/4 of the SE power has harmonics less displeasing to the ear than the lower artifacts of the PP amp.  This is not always the case, and I have made many SE amps which measure lower THD/IMD than a PP amp rated for the same power, but then I understand how to do this better than many makers.

The SE amplifier power is always pure class A, and if you want 32 watts maximum from an SE amp, you'll need to have at least 75 watts of anode power from the power supply at all times while the amp is turned on. So you'd need 3 x 6550 with each idling at 25 watts of anode dissipation. If you had a PP amp from which you wanted a maximum of 32 watts of pure class A then you also would need the same 75Watts of anode input power. With only 2 x 6550, you'd have Pda = 37.5W and this is far too high and uncomfortable and unreliable and likely to cause much reduced tube life. So you would have to use 4 x EL34 or 6550 perhaps with each having Pda = 19 Watts.
You could well have a PP class AB amp rated for a maximum of 20Watts of class A power but with the capability to make up to 60W AB into 3 ohms. The input power required at idle will be say 45W, so you may happily use just one pair of 6550 each idling with Pda = 22.5W. With Hi-Fi and music, the use of levels above the class A threshold for any load will be brief, and the anode power supply draw may only briefly rise to a maximum of say 80Watts only very occasionally for high volume levels. So when you build an amp using 2 x 6550, allow for the 80W of possible anode power and the excess capability means a cooler transformer.
The costs of building an SE amp rated for 32 watts is about the same as those for a PP AB amp of 60 watts. Hence you pay more for less watts with a class A SE amp compared to a class AB amp. If either PP or SE amp are rated to make the same class A power there should be little difference in cost. Using class AB operation lessens construction and operation costs without ruining music so it is why class AB has been used by so many for so long.

Most of my customers prefer a well done SE amp of maybe only 1/2 the power ability of a class AB PP amp because the sound they hear seems superior, so the cost isn't a problem.

If you have been using a 15W amp for the last few years without enduring the high distortions at high levels, then you'd find many SE amps with say at
least 2 x 6550 or 3 x EL34 to be quite sufficient.

If you think that nothing is as good as the first watt from a 200W class A amp, then you won't be swayed by those happy with 20 watts. Let me tell you, a 200W class A power amp is really heavy, hot, and expensive. Anode input power must be at least 500 watts which means 20 x 6550. Cathode heating power is 226W.
I don't have the transformers for sale for such a greenhouse un-friendly amplifier.

Tube amps cost a lot to make, and the cost per watt is appallingly high for 4 Watts from a lone 2A3 or EL34. But as the watts rise, the cost per watt rapidly falls, and a 300W class AB amp is not much more expensive than a 100W amp depending on the anode input power. 
                                                         __________________________________________________________


Description of output transformers for sale.


Most transformers have have grain oriented silicon steel double C-cores, made by AEM in Sth Australia before about 1999
when AEM ceased manufacturing C-cores.
The maximum permeability of the C-cores, µ-max, with no air gap is above 4,500 and entirely adequate.
Music sounds excellent with these cores. These GOSS cores give excellent technical performance with low iron related distortion artifacts which are much lower than harmonic artifacts produced in the vacuum tubes even at high levels and down to 50Hz where transformers produce most distortion at LF.
They are most suitable for tube amp diy projects or perhaps repair/upgrade replacements. Makers such as McIntosh used C-cored transformers.

The low loss low distortion C-cores in output transformers sustain negligible rise in temperature and offer lower distortion than ordinary E&I laminations and are high quality items for audiophile grade amplifier performance. They are of negligible benefit in guitar amps where harmonic distortion in the amplifier is much desired.
The transformers all have carefully layer wound wire with at least 0.15 mylar insulation between every layer of wire. Leakage inductance is extremely low due the extensive amount of primary and secondary interleaving, resulting in very wide potential bandwidth if the driving source resistance is low enough.
The bobbins have a 3mm base wall thickness with ends of wire layers all kept back 3mm from the edge of the insulation to maximize creepage distance.
Insulation between anode windings and earthy core and secondary windings is excellent.

Some listed transformers may be used for Push Pull or Single Ended projects.
The necessary clamp yokes to allow bolting to a chassis will be provided with each transformer as demand requires.
Some extra terminations are required for provision for local CFB tertiary windings and UltraLinear screen taps.
For PP operation, transformers will be supplied with C-cores clamped tight as they appear in images.
For SE operation, cores will be supplied loosely clamped cores allowing final assembly and air gap adjustment by the amp constructor
unless prior arrangements are made with me to set a gap to suit the proposed use, and provide clamps which also act as brackets for mounting.

Where I have stated the possible SE operation for what was originally designed to be a PP transformer, the maximum primary winding dc idle current 
has not been allowed to to exceed 2Amps dc per square millimetre. So if the wire size is only 0.4mm dia, you may only safely have Idc = 256mAdc.
If the winding concerned had R = 100 ohms, you get 5.6Watts of heat liberated in that winding and any more would raise the winding temperature too high.
Where one did have Idc = 256mA, the signal current maximum could be 0.7 x 256mA = 179mA, and if Ea was +330V then the anode signal voltage could be 297 peak volts, or 207Vrms, and the load is thus 207V / 179mA = 1,156 ohms and the signal winding losses in the primary wire resistance of 100 ohms will be 8%, and getting rather wasteful of available 40 watts of SE class A power. And this does not include losses in the secondary windings. However, in nearly all these transformers for sale the secondary winding losses are very low compared to the primary losses because there are so many paralleled secondary windings with so much interleaving.
Where I have listed the SE operation conditions for anode supply voltage and currents, I have aimed to have maximum power produced when
the load to be driven as 5 ohms. This means the SE amp can drive any speaker impedance above 3 ohms.

Most tube amps give their lowest distortion and best damping factor when the speaker load is a high number of ohms. So when a 16 ohm speaker is connected to an amp with a 5 ohm rated output terminal, the music will be better than if the load was 5 ohms, as long as the amp does not operate near clipping levels.

The secondary windings of most of the OPT have only two useful ways of connecting the secondaries to vary the load matching. Where there are say 8 secondary windings using one layer of wire across the bobbin, one can only have all of them in parallel or two lots of 4 windings in parallel in series. With all secondaries in parallel, you will get a match for an anode load of perhaps 5k at the primary where a 3 ohm load is connected at the at secondary. If half all the secondaries are in series with the other half, you will get 5k loading for the anode at the primary only if the secondary load is 12 ohms.
It is fundamental to understanding output transformers that you know that the resistance load ratio of primary to secondary is the turn ratio squared.
Because these OPT have at least two loading possibilities, and because it is always possible to alter the Ea and Iadc conditions, variable but useful load matches can be found. In general the winding losses in these transformers are usually low because most have a large winding window area to central iron core area compared to wasteless E&I laminations, and the designer reckoned on the saturation frequency to be not lower than 25Hz at maximum mid band signal voltage. This means that the primary turns are 66% of a same sized transformer with Fsat = 17Hz at the same signal voltage. The primary wire dia is large for PP operation when primary and secondary winding loss % will be low. SE use always results in higher winding losses. The interleaving between primaries and secondaries is what I would call excessive, and in most OPT the windings have alternative single layers of primary and secondary from the bottom of the bobbin to the top. Thus these C-cored output transformers have extraordinarily low leakage inductance far below what is regarded as low by the experts of 1953 who wrote the Radiotron Designer's Handbook and who spelled out once and for all what mattered about output transformers. However, as a result the shunt capacitance is higher than what is regarded as low by the same experts, but for most applications where the source driving resistance is at least one pair of high transconductance beam or pentode output tubes or triodes, the capacitance does not present a phase shift and stability problem, and does not begin to adversely load the amp until you reach 12kHz and the resonances between the leakage inductances and shunt capacitances within a given transformer are at frequencies well above 50kHz and well away from the audio band. The output impedance of the amplifier with these transformers is more capacitive because it is capacitance which causes the frequency attenuation at above 20kHz rather than the presence of leakage inductance. In most normal output transformers there is less interleaving than has been used by Elson Silva and his master craftsman winding expert, Bruce Flanagan. I have prepared a page for deeper design comparison at output transformer frequency behaviour.
                                                            ________________________________________________

Output Transformer Design Parameters.

The figures in the table for OP1 are what I have worked out to be optimal for OP1 only. One could always work out that more or less power is available depending on loading and power supply voltage and current conditions.

I have given what I think is best.

Keen prospective buyers will do the necessary loadline analysis for the use of the transformers.

It would be easy for those capable of designing OPTs to copy my table and fill in all the details if they have a good micrometer or pair of calipers to measure wire diameters, and they can measure the P to S turn ratio by measuring the voltage ratio and can count the layers and estimate the turns.

The transformers came to me for sale without much information and it took me about 4 hours to work out all the figures 1-54 for OP1.

At other pages at my website the methods for designing output transformers are outlined. The winding losses and bandwidth performance can be calculated from the listed information supplied with each transformer.

If you do any design  for any transformer, you should be able to copy the tables below and fill in the right hand column with relevant figures for the PP or SE design you propose. When you have all 54 listed items filled in, you should be able to calculate the shunt capacitance, the leakage inductance, and the winding resistance losses to confirm that your design will not go up in smoke after turn on, and that you will get low distortion, enough power, and wide enough bandwidth.
Good sounding music follows good numbers.

The 2 tables have the same 54 items listed and some are Not Applicable depending on whether you are working something out for either PP or SE.
Feel free to copy and paste a table to a folder where you can delete all the right hand numbers, then print a copy of the blank template you will have so that your calculations on paper in your exercise book may be written down, then copied for a PC folder storage. I doubt you will come up with better suggestions than i have included for the 16 varieties of OPT listed. 

EXAMPLE, Table 1.
1
Output transformer number,
OP1
2
Push Pull or Single Ended ? PP
3
Weight in Kg 7.4
4
Overall size, length x width x height, mm 176 x 132 x 138
5
Turn ratio, TR, Pri to Sec 23.7 : 1
6
Impedance ratio, Pri to Sec, ohms
562 : 1
7
Primary load x secondary nominal centre load value, ohms.
2,845 x 5.0
8
Primary turns, NP, and wire size, Cu dia, mm 1,496 x 0.45
9
Primary layers x turns per layer 11 at 136t
10
Primary section number, all in series,
11
11
Secondary turns, NS
63
12 Secondary layers x turns per layer x wire size, Cu dia, mm 12 x 63 x 1.0
13
Number of parallel Secondary sections x Secondary turns per section 12 x 63
14
Grain oriented steel core with max µ > 4,500, description double C-cores
15
Afe, Tongue x Stack, or total strip build up x strip width, T x S, mm x mm
52 x 52
16
Window size, L x H, mm x mm
76 x 28
17
Iron path Magnetic Length, ML, mm 296
18
Insulation, Nomex or other, P to S, mm
0.22
19
Average turn length, TL, mm 300
20
Primary winding resistance, RwP, ohms 50
21
Secondary winding resistance, PwS, ohms
0.036
22
Total winding resistance RwP + RwS at primary, ohms
70
23
Total winding losses, %
2.5
24
Maximum allowable current density in any winding at idle, Amps dc per sq.mm 2
25
Maximum continuous allowable dc or ac current, primary wire, Amps
0.48
26
Maximum design value for dc current, primary wire, for SE operation, Amps NA
27
Load values for tests, primary anode load x secondary load, ohms.
2,845 x 5.0
28
Nominal Anode Source resistance ( between highest and lowest Ra possible ) for tests, ohms
2,845
29
Maximum primary inductance, PP operation, Lpmax, no air gap, µ > 4,500, Henrys, H 128
30
Minimum Primary inductance, PP operation, Lpmin, approximate, no air gap, µ = 1,000 at low Vo, Henrys, H 28.4
31
Effective permeability, µe, with air gap for SE operation, Bdc = 0.75Tesla NA
32
Air gap calculated for reducing maximum iron µ to above wanted µe, mm, ( confirmed by experiment in amp )
NA
33
Non magnetic material thickness used to gap cores on both breaks in magnetic loops, mm
NA
34 Primary inductance, Lp, for SE operation, for µe, Henrys, H NA
35 Leakage Inductance, at primary input, LL, mH 0.24
36 Shunt capacitance, at each anode primary input/s, pF
3,000
37 Maximum audio power at anodes for THD < 2%, at 1kHz, at rated RL loaded, Watts
135.0
38 Maximum anode signal voltage across primary at THD < 2%, Vrms
620
39 Maximum class AB1 power, PP operation, at rated RL, Watts 135
40 Maximum class A1 power at rated RL, Watts
30
41 Frequency of core saturation at max audio power signal across primary at Bac + Bdc = 1.5 Tesla, Hz 22.3
42 Low frequency cut off at 0.1 x max Vo, source resistance = 1/2 anode RL, Hz
8
43 High frequency cut off with Secondary loaded. Primary source resistance = primary RL, kHz 75
44 Tubes usable, type numbers
6 x KT88
45 Tube configuration, Ultralinear, UL, Cathode Feedback, CFB, Pentode, P, Beam Tetrode BT, Triode, T. UL, CFB
46 Maximum Idle dc anode and screen dissipation power setting for each tube, Pda, Watts
25
47 Maximum total tube idle dissipation power Pda, Watts 150
48 Anode to cathode dc supply voltage ( excluding possible cathode biasing voltage ), Ea, Vdc 500
49 Total Idle condition anode and screen supply dc current for all output tubes, Idc, mAdc
300
50 Anode plus screen Idc per tube, mAdc
50
51
Actual anode plus screen dissipation in each output tube ,Watts
25
52
Maximum Idc needed for maximum output power for stated load, approximate, mAdc
500
53
PP tolerated loads with the tubes above, Primary load : Secondary load, ohms 1,686 : 3.0
2,248 : 4.0
2,800 : 5.0
3,372 : 6.0
4,496 : 8.0
54
SE tolerated loads with the tubes above, Primary load : Secondary load, ohms NA


EXAMPLE, Table 2.
1
Output transformer number,
OP1
2
Push Pull or Single Ended ?
SE
3
Weight in Kg 7.4
4
Overall size, length x width x height, mm 176 x 132 x 138
5
Turn ratio, TR, Pri to Sec 11.85 : 1
6
Impedance ratio, Pri to Sec, ohms
140.5 : 1
7
Primary load x secondary nominal centre load value, ohms.
711 x 5.0
8
Primary turns, NP, and wire size, Cu dia, mm 1,496 x 0.45
9
Primary layers x turns per layer 11 at 136t
10
Primary section number, all in series,
11
11
Secondary turns, NS 126
12 Secondary layers x turns per layer x wire size, Cu dia, mm 12 x 63 x 1.0
13
Number of parallel Secondary sections x Secondary turns per section 6 x 126
14
Grain oriented steel core with max µ > 4,500, description double C-cores
15
Afe, Tongue x Stack, or total strip build up x strip width, T x S, mm x mm
52 x 52
16
Window size, L x H, mm x mm
76 x 28
17
Iron path Magnetic Length, ML, mm 296
18
Insulation, Nomex or other, P to S, mm
0.22
19
Average turn length, TL, mm 300
20
Primary winding resistance, RwP, ohms 50
21
Secondary winding resistance for NS, RwS, ohms
0.144
22
Total winding resistance RwP + RwS appearing at primary input, ohms
70
23
Total winding losses, %
9.0
24
Maximum allowable current density in any winding at idle, Amps dc per sq.mm 2
25
Maximum continuous allowable dc or ac current, primary wire, at 2Amps/sq.mm Cu,
0.324
26
Maximum design value for dc current, primary wire, for SE operation, Amps
0.324
27
Load values for tests, primary anode load x secondary load, ohms.
711 x 5.0
28
Nominal Anode Source resistance ( between highest and lowest Ra possible ) for tests, ohms
711
29
Maximum primary inductance, PP operation, Lpmax, no air gap, µ > 4,500, Henrys, H
NA
30
Minimum Primary inductance, PP operation, Lpmin, approximate, no air gap, µ = 1,000 at low Vo, Henrys, H NA
31
Effective permeability, µe, with air gap for SE operation, Bdc = 0.75Tesla
278
32
Air gap calculated for reducing maximum iron µ to above wanted µe, mm, ( confirm by experiment in amp! )
1.0
33
Non magnetic material thickness used to gap cores on both breaks in magnetic loops, mm
0.5
34 Primary inductance, Lp, for SE operation, for µe, Henrys, H 7.9
35 Leakage Inductance, at primary input, LL, mH 0.3
36 Shunt capacitance, at each primary input/s, pF
5,000
37 Maximum audio power at anodes for THD < 2%, at 1kHz, at anode rated RL loaded, Watts
36.0
38 Maximum anode signal voltage across primary at THD < 2%, 
160
39 Maximum class AB1 power, PP operation, at rated RL, Watts
NA
40 Maximum class A1 power at rated RL, Watts
36
41 Frequency of core saturation at max audio power signal across primary at Bac + Bdc = 1.5 Tesla, Hz 14.4
42 Low frequency cut off at 0.1 x max Vo, source resistance = 1/2 anode RL, Hz 7
43 High frequency cut off with Secondary loaded. Primary source resistance = primary RL, kHz 75
44 Tubes usable, type numbers
4 x KT88,6550
45 Tube configuration, Ultralinear, UL, Cathode Feedback, CFB, Pentode, P, Beam Tetrode BT, Triode, T. SEUL, SECFB
46 Maximum Idle dc anode and screen dissipation power setting for each tube, Pda, Watts
25
47 Maximum total tube idle dissipation power Pda, Watts
150
48 Anode to cathode dc supply voltage ( excluding possible cathode biasing voltage ), Ea, Vdc 295
49 Total Idle condition anode and screen supply dc current for all output tubes, Idc, mAdc
350
50 Anode plus screen Idc per tube, mAdc
86
51
Actual anode plus screen dissipation in each output tube ,Watts
27
52
Maximum anode Idc needed for maximum output power for stated load, approximate, mAdc
330
53
PP tolerated loads with the tubes above, Primary load : Secondary load, ohms
NA
54
SE tolerated loads with the tubes above, Primary load : Secondary load, ohms
423 : 3.0
564 : 4.0
711 : 5.0
846 : 6.0
1,128 : 8.0



Sale conditions for transformers.

Prices are in Australian dollars excluding freight and packaging.

Payments for transformers can be via direct deposit into my bank account or money order or cheque or cash and always after
email negotiation with me, Patrick Turner, info@turneraudio.com.au

Output transformers can only be sold only in pairs, because there are at least 1 pair of each type described, and if I sell just one I cannot easily sell the other remaining one because most people will want a pair for a pair of channels.

The transformers will be supplied with mounting brackets, extra terminals for added tappings and with all information. There are multiple connections of the primary wire layers  made outside the bobbin and which are accessible for use if cathode FB windings are wanted or UL screen connections.

There is no warranty for transformers which have been damaged from being subject to excessive current flows.
To prevent such damage occurring, users must connect suitable fuses between each output tube cathode and 0V where fixed bias is used. Slow blow fuses of 5 times the idle current are about right, so where you have 60mAdc at idle for a 6550, use a 250mA fuse.
For cathode biased amps with R&C cathode biasing networks, the only sure way to prevent transformer damage is to fit an active protection board  which turns off the mains supply if the cathode Vdc rises to 50% above the idle value for longer than 4 seconds. Protection schematics are available elsewhere at this website.
Because I have no control over how anyone might use the transformers being offered, I cannot offer a warranty included in the price unless you can easily prove there is a defect in the way the transformer was wound.
The winding method used is inherently reliable, and much better than anything wound in China or other places where little quality control exits and workers are
allowed to place turns of wire in an irregular "random winding pattern" where the wire is not neatly wound on layer by layer with insulation carefully placed between layers and with not one single crossed over turn.

Buyers should carefully design their amp schematics with regard for anode supply voltages and anode load values.
The design should allow for some flexibility for the anode supply voltage.
 
To easily calculate current ability of any wire size based on 2 Amps per square mm, Current allowed in Amps  = diameter squared  x 1.6 where the diameter is the bare copper dia and measured in millimetres. Hence a 0.4mm dia wire has a current ability = 0.4 x 0.4 x 1.6 =  0.256Amps, or 256mA. If you had 4 x EL34 each with 50mA of Ia connected, idle current would be 200mA total. If you then had bias failure of all four EL34, and the Ia in each tube went up to 300mAdc, the total would be 1.2Adc and if the primary winding resistance was 50 ohms, the heat dissipated in the primary would rise from 2 watts at idle to 72 watts with 1.2 amps, and this *will* damage the transformer if the situation lasts longer than a minute. 

For PP operation, C-cores will be supplied tightly band clamped as they appear in images and with fixed support brackets ready for bolting to a chassis.
For SE operation, C-cores may be supplied loosely clamped allowing final assembly and air gap adjustment by the amp constructor unless prior arrangements are made with me to set a gap to suit the proposed use.

Good luck!

Back to Output Transformers for sale.