Magnetic Components B+ Transformer Specs

[jojokeo]

Something caught my eye today: How is it that these transformer's B+ mA ratings are not under spec'd, will not run overly hot, and will not likely burn out under full output power during stage use when cranked up? These look more appropriate for 6V6s not EL34s?


I was looking at power transformers from Classic Tone intended to be used for Marshall 50w amps using two EL-34 tubes (in the Marshall amps category). Model numbers: 40-18023 and 40-18089 as two examples with listed B+ current ratings both being 150mA. (the JCM900 and 2000 PT spec sheets do not list their PT B+ mA ratings)


Without also considering the preamp & phase inverter tubes, etc and going by EL34 datasheets by Mullard, Telefunken, and others in Class AB-1 - they list in both cathode and fixed bias situations (under full signal) to ALL be well over what these power transformer's specifications provide!?

[pompeiisneaks]

Seems the datasheet I'm reviewing shows EL34's at max dissipation at 55mA per tube, so 110mA for two tubes, the preamp tubes don't pull much like 10 mA  each max, so seems well under the specced limit for the HT/B+ winding.

~Phil

[tubenit]

The tube data cheat sheet we put together (maybe 8+ yrs ago?) shows 120ma to 160ma for push pull.

I just looked at a data sheet and it was stating an EL34 with 75ma to 95 ma per tube.

with respect, Tubenit

[pompeiisneaks]

This is from a datasheet for a Mullard:

Am I looking at the wrong area?

~Phil

[DummyLoad]

transformers delivers specified voltage at specified current draw. that relates to some VA spec that is good for part to deliver those parameters 100% of the time with a long term core temp rise that won't self destruct part. transformers do not "blow up" or burn out" if we exceed the current ratings (1.5x or 2x ok, intermittently, but certainly not something like 4x-10x) as most are built very conservative to begin with. 

you don't run AB1-2 P-P amp at 100% power 100% of the time: take 1/2 to 2/3 the full power current draw for an average and see where the part spec lays in on the full power conditions as published in the data sheets. in SE or class A P-P, we'd want to spec the transformer current supply nearer to or greater than the full power current draw. for continuous service for AB1-2 P-P we'd spec the transformer with a current rating equal to or greater than the full power current demand that would allow the part to run 100% on at 100% power output indefinitely. 


--pete

[PRR]

A pair of EL34 may be run at 69mA idle and 260mA full-roar sine wave.

G-amp OVER-load approaches square-wave which will be more.

It takes many minutes to cook a large transformer. So over the course of several songs, how much time is spent near idle (low volume), how much slammed (recalling that guitar decays quickly after the pluck), and how much SLAMMED (non-stop boosted OVER-drive)?

If we assume 50% at less than max and 50% at MAX, we do come to a bit less than 150mA.

Some head-bangers play louder longer. I think these Marshall-design PTs do have a reputation for toasting every few years, tho I do not know how many were hard-worked and how many just were not over-built from new and died of assorted breakdowns.

[jojokeo]

Glad to hear the explanations and perspectives, my thoughts prior are I wouldn’t want to use anything rated less than 200mA?! And Mag Components think 75% is just fine? Technically for running the amp hard - as to be expected in demanding situations , ruggedness & reputation is important (not blowing amps up in smoke intentionally a la Ritchie) seems 250mA rated to be more appropriate? But I didn’t think of or count on the amp being used like PR explains. I guess I Imagine it to be capable of being either in a drag race to an Indy 500 on any given night?!

[jojokeo]

Reading (below) about speaker testing trials supports a similar thought process expressed (to reliably operate at high power levels over time without breakdown). Is that 150mA rating truly to be trusted with your name and/or band playing and on the line? Hardly seems worth it to skimp here since the PT is the very heart n soul of an amplifier.


What dictates power handling? Testing a speaker to find its power handling gives us an insight into how the speaker copes with a sustained input both mechanically (in the repeated back-and-forth motion of the cone, voice coil, and suspension) and thermally (with the heat generated when you send a powerful signal into the speaker) over an extended period of time.

At Celestion, we test speakers in a number of ways: We play guitar through them, we play recorded music through them, and basically we do everything possible to find the limits of the speaker’s performance and what it takes to break it. More scientifically, we also feed the speaker with an RMS test signal—random noise with a consistent level of sound energy that’s averaged over time.

By subjecting the outcomes of these tests to experienced interpretation, we can arrive at a meaningful power-handling value. “Averaged over time” is perhaps the most critical statement here. For example, if we know that a guitar speaker can cope both mechanically and thermally with a constant 50-watt input signal for 100 hours with no adverse effects, then we can safely assume that the same speaker will have little problem withstanding a cranked 50-watt tube-amp signal that’s apt to produce higher transients from time to time

[pompeiisneaks]

A pair of EL34 may be run at 69mA idle and 260mA full-roar sine wave.

So where did you get this 260mA value?  Just from practical knowledge? I couldn't find anything on the datasheet I found that indicated this was possible for the tube, but maybe I missed something?

~Phil

[jjasilli]

The Celestion quote lacks definitive meaning, because it fails to stated the rated wattage of the supposed speaker.  If it can handle 50W continuous for 100 hours, or 50W tubed cranked (figure 100W) for a while, then is Celestion rating that speaker for 50W; 100W??? W/o that info we're right back at jojo's original question!


Why is a cranked power amp considered double the nominal rated wattage?  Two reasons, both based on clipping -- the distortion of a sine wave to more like a square wave:  1.  It takes more power to produce a square wave; 2.  A square wave has no cooling-off period peak-to-peak.


When a PT manufacture says xxxmA, what does that mean?  From a reputable manufacturer, I assume it means under continuous conditions.  Otherwise specs would be largely meaningless, and incapable of comparison.  This comports with the general notion that it's OK to under-spec a reputable guitar amp PT by up to 50%.  As PRR points out, a musical program, particularly a guitar signal, pulses and is not continuous like a test signal.  However, under-spec'ing the PT may introduce sag and saturation which is subjective as a design goal; also, it is not risk free.

[jojokeo]

A pair of EL34 may be run at 69mA idle and 260mA full-roar sine wave.

So where did you get this 260mA value?  Just from practical knowledge? I couldn't find anything on the datasheet I found that indicated this was possible for the tube, but maybe I missed something?

Phil,
Prior you thought - incorrectly - that a tube's Pdiss aka power dissipation rating was how to figure this out. Rather, you need to consider the entire amp's current draw that is placed on the power tranny. In this discussion we are mainly factoring the power tubes' draw placed on it. You need to consider the type of operation (bias), operating voltage, number of tubes, etc to get an idea which to base your pre-build calculations derived from the datasheets and use the total summed mA's drawn by the plates and screens under full signal listed. That is what PRR did and what I was pointing out.

[pompeiisneaks]

A pair of EL34 may be run at 69mA idle and 260mA full-roar sine wave.

So where did you get this 260mA value?  Just from practical knowledge? I couldn't find anything on the datasheet I found that indicated this was possible for the tube, but maybe I missed something?

Phil,
Prior you thought - incorrectly - that a tube's Pdiss aka power dissipation rating was how to figure this out. Rather, you need to consider the entire amp's current draw that is placed on the power tranny. In this discussion we are mainly factoring the power tubes' draw placed on it. You need to consider the type of operation (bias), operating voltage, number of tubes, etc to get an idea which to base your pre-build calculations derived from the datasheets and use the total summed mA's drawn by the plates and screens under full signal listed. That is what PRR did and what I was pointing out.

That's the point of my question, you just said "You did it wrong" but now how to do it?  How do you derive the 260mA factor for 2 EL34 tubes if the datasheet says max diss rating is 2x63 at normal operating points?  What does that value mean, and how do you go from 126mA to 260mA?  is there some known 'factor' you multiply datasheet values of max diss to real world max diss?  I get the square wave 'slamming the system' type max diss rating, so does that mean datasheets give you a 'max diss' based on a more normal musical use case, and not sine or sq wave max diss?

Sorry, I'm still just trying to figure out how that's supposed to be calculated, I thought it was always based on max diss from the datasheet, but I'm obviously mistaken, and want to understand how.

~Phil

[jojokeo]

The Celestion quote lacks definitive meaning, because it fails to stated the rated wattage of the supposed speaker.  If it can handle 50W continuous for 100 hours, or 50W tubed cranked (figure 100W) for a while, then is Celestion rating that speaker for 50W; 100W??? W/o that info we're right back at jojo's original question!

I believe that they want you to understand how they go about deriving their power rating system for their speakers but also showing that it is an inexact science? As the last paragraph shows, "By subjecting the outcomes of these tests to experienced interpretation, we can arrive at a meaningful power-handling value."

However, under-spec'ing the PT may introduce sag and saturation which is subjective as a design goal; also, it is not risk free.

I don't think that this is wise to do. Rather than getting sag via voltage drop derived via tube rectification and resistance built into the B+ filter string - running a strained PT will only increase it's chances of running hot and experiencing thermal breakdown at some point sooner or later. It will also adversely affect headroom, punch, compression, and other characteristics prior to breakdown.

[jojokeo]

That's the point of my question, you just said "You did it wrong" but now how to do it? 

See the modified datasheet below:

[jojokeo]

I know that most manufacturers under rate/over design/provide a buffer of their transformers to provide a safety margin of maybe ~20% just like capacitors/resistors & other components -  but 40% is significant!!!

The explanation is in:
Mark Sacketti of Magnetic Components was kind enough to quickly reply with this info:

"Our 40-18023 and 40-18089 will safely handle the current requirements for a 50W EL34 amp. The 150mA rating was what an original customer rated it (but did not specify the operating condition). The true current rating is 250mA."

[pompeiisneaks]

DOH, sorry, I didn't even see the lower set... welp, I larned tings :P

Thanks for the information, I figured it would be something obvious, but not that obvious.

~Phil

[jjasilli]

Quote from: jjasilli on December 27, 2017, 10:24:50 amHowever, under-spec'ing the PT may introduce sag and saturation which is subjective as a design goal; also, it is not risk free.


I don't think that this is wise to do. Rather than getting sag via voltage drop derived via tube rectification and resistance built into the B+ filter string - running a strained PT will only increase it's chances of running hot and experiencing thermal breakdown at some point sooner or later. It will also adversely affect headroom, punch, compression, and other characteristics prior to breakdown.

Bargain tube guitar & hi-fi amps did this all the time, and PT's ran very hot.  Yet, Marshal probably has a worse failure rate!  Personally I would under-spec, but respect someone else's desire not to do so.  Running an amp near or over "the edge" is one way to go.  But maybe not with someone else's amp.

[jojokeo]

Agree JJ, but I wouldn’t want an amp breaking down on someone or for myself, then have to fix it & pay for another. I’d feel like a real dufus for knowing better but allowing for the high probability unless it was for experimental purposes and only used in the shop or home. Then yeah, what the hell right?! Sometimes it’s fun pushing things to their limit 😈😱😇

[shooter]

pushing things to their limit

and it keeps tech employed

I do mostly _SE, I try and spec my builds for.. just.. 1more power tube, then 20ish% over that if the price point is close

[PRR]

> where did you get this 260mA value?

See jojokeo. A different Mullard sheet gives 254mA (I added in my head and came out 4 thumbs long). It does look like every new engineer was asked to re-measure EL34 and got slightly different answers, oh well.

I think the "150mA" is copied from Marshall's original specs? If you gonna make clones, you must clone the spec numbers also, or buyers will be confused.

The "150mA" *could* be a long-term rating. Maybe 8 hour steady 150mA. ANY transformer will take short-term overloads.

I have a 6A battery charger with a 50A booster function. It will do a large part of 6A for days straight (I have a big old battery). The 50A rating is only for short-term boost for starting. It would fry in a few minutes; but it has a 30 second cut-out so it can help the battery start the engine quickly (and cut-out if the engine doesn't start quick).

The electric company pole-pig for my house and my neighbor is a 25KVA transformer but has 50KVA connected load. They know neither house is likely to pull 15KVA and then only a couple hours. If we both did our big Holiday cooking, laundry, and welding in the same 4 hours, a 2X "overload" would make the transformer hot but not dangerously so. 4 hours at 2X on a big iron maybe takes a day off its expected 20 year life. Since 99% of the time we each draw 1KVA, the tranny mostly runs cold and should last a lot longer than 20 years (it may be 30 already).

[jjasilli]

pushing things to their limit

and it keeps tech employed

I do mostly _SE, I try and spec my builds for.. just.. 1more power tube, then 20ish% over that if the price point is close

Yes, but this is a manufacturer's / tech's dilemma.  A race car driver may push his car over the limits, and knows something will eventually break, and he'll have to pay to fix it with money and down time.  The typical guitarist may like pushing his amp, but then resent the manufacturer or his tech when it breaks.

[shooter]

resent the manufacturer or his tech

my career has been fixing broke stuff, one of the things you learn early on, now how to de-fuse the customer
If that don't work, charge/bill more for their attitude

[jojokeo]

The electric company pole-pig for my house and my neighbor is a 25KVA transformer but has 50KVA connected load. They know neither house is likely to pull 15KVA and then only a couple hours. If we both did our big Holiday cooking, laundry, and welding in the same 4 hours, a 2X "overload" would make the transformer hot but not dangerously so. 4 hours at 2X on a big iron maybe takes a day off its expected 20 year life. Since 99% of the time we each draw 1KVA, the tranny mostly runs cold and should last a lot longer than 20 years (it may be 30 already).

Awesome (as usual)!!!