Simple PT rating question

[pdf64]

...power transformer ratings can be all over the place, they can play games with temperature rise, regulation, duty cycle,
Hammond is probably the least conservative of the bunch, Triad is usually pretty good, pull a transformer out of a piece of HP tube equipment if you want something that runs very cool.

Bear in mind that PTs intended for use with a tube rectifier would hopefully have some degree of equivalent series resistance designed into their HT windings, so as to avoid the need to add external resistance. The resultant copper loss will cause them to run hotter than PTs designed without that constraint.
I suspect that designers of tube rectified amps whose marketing boasts of their cool running transformers may have neglected to take note of the plate series resistance limits noted on the tube info. External resistors are rarely added to compensate, whereas the HP would perhaps have been more likely to take such things into account.
When the rectifiers subsequently fail, bad QC of current production tubes then gets the blame, when in reality it may well due to design error.

[nsubulysses]

Well, you got it, I am in fact talking about Hammond 278CX. Again I appreciate all the answers I'm learning a lot.

Here's the real scenario. I am using this PT in a prototype build with 4 6CA7 power tubes. It has either 4 or 5 12AX7s depending on if the amp has an FX loop. So I'm running the 6A heater winding at 7.2 or 7.5A.

This shows that Escherton was correct, the PT was chosen for high B+ and that was the first design choice. This 535mA and 6A heater winding business is just a secondary thought you know what I mean ??

The PT gets kind of hot if you crank the amp full volume-ish for two hours straight (gain and master volume both on about 7) but nothing that really worries me. My V4 and Sound City 120 PTs get hotter in less time cranked and they have been going for 40 years.

So this PT is cool but when I make the amp again I'm going to have Heyboer make a version of this PT with 8A heater winding. Since i'm ordering custom anyway it got me wondering how good this 535mA HV rating is for my application and should that be adjusted as well?

So yeah then I wanted to just understand the whole topic more in general so I could know what to desire if I were to do a 2 or 6 power tube version as well. Or even in scenarios of ordering PTs for my customers for repairs. Seems like choosing a transformer listed by ratings which may be the same anyway but may NOT called "Fender Twin Reverb PT replacement) is sometimes cheaper. So I appreciate the answers.

[J M Fahey]

I think they get 55W by using regulated power supplies for the plate and screen voltages.

This.
I *often* fiund datasheet capabilities impossible to meet on Real World stuff (same in the SS World) until I realize that in their NASA type labs, white coated Engineers test everything fed from their expensive super regulated Hewlett Packard power supplies, so "450V" stays 450V no matter what ... while in our World, a 450V **at idle/no load** copper and iron>diodes>caps supply will easily drop 10% (at least) to 20%
Being that power is proportional to *square* of voltage, "small" vlotage drops become large *power* drops, relative to theoretical maximum.

Me? , I m happy to get 40W RMS all day long, with any brand of tube, from comfortable 420/430V +V which anybody can handle, and easy to drive 5500 ohms plate to plate, so I aim to that goal.
I *can* get 50W but I have to substantially rise voltage, use serious iron and capacitance, be choosy about tubes, the works.
Not really justified to get extra 10W, difference is not audible, I better invest in better speakers.

60W RMS? sure, no problem ... if you use scary high voltage, bias tubes Class B so they start colder, and **smash** them with cathode current drive (Music Man).
Problem is now they do not overdrive that nice, have more of a "clinical" sound, since tubes are tightly controlled, so better used for loud clean sound.
But then you might as well use SS

My average customer is happy with a real steady 40W RMS 2 x 6L6 combo and a single good 12" which can be easily carried by hand or in any car trunk and sings onstage when played hard.

[bob p]

^^^ Words to the wise -- 40W sounds as good as 50W and it's a more practical build. To get that 25% more power a 50W amp is going to be more expensive to build, will be harder on tubes, etc.

To put this into perspective -- there's not that much of an appreciable power difference between a Bandmaster / Super Reverb and a Bassman. The 40W amps make up for their lower power with their 3x10 and 4x10 speaker compliment. As it turns out I'm building an blonde bassman type amp right now and I'm scaling it down to the 40W platform with a set of efficient speakers because I don't really need 50W.

of course, if you gig in a place that has FOH support you might as well pack a 20W combo and mic it. We're at the point today where the amp is a tone generator, and you don't really need the amp for amplification purposes unless you're playing in BFE where there's no PA and you're trying to keep up with an animal on drums.

And by the way -- guitar amps with fully regulated supplies don't really sound all that good -- unless you want something to thrash.

[nsubulysses]

I understand the 20W, 40W thing but it does not apply to my friend/customer group. My friends/customers usually want to drop the bomb.

My friend/customer actually just posted this pic of my proto paired with his 200W orange for a nice dual fullstack rig to devestate everyone. Crowd will love it, sound guy may possibly have opposite response, unless they are into extra heavy metal. He wants me to build a 6 power tube version so he can dump the Orange

[catalin gramada]

Man, if you love to get power don.t look for fancy transformers. Take a 300V (you can supply screens and preamp stages from it) and a 200V one (making a stacked supply for the plates). 2 amps arround bridge rectification for 6 KT88 for instance, I get 115w clean sine for a pair at 620v with 365 on screens when full throttle (meant 8,5w dissipation on screens)into a 5k OT
This are the specs for one of transformers used for a KT88 pair. Voltages are for nominal currrent (the plates drop from 650V at idle to 620v under full load) Pretty good regulation for a cheap one.(It cost me 50 bucks equivalent )



I have 460V into transformer secondary (285+175) for rated current 600mA, meant 620V arround after rectification (460x1.35=621V) and I got 620v. That means transformer deliver the right amount of current demanded by circuit at full spec. Also the voltage difference between idling and full power is 30V meant under 5% which is a very decent regulation.
Oversizing can get a better regulation but cannot predict the voltages for demanded current in any way. This is not a problem for the plates but is a very important one for the screens when small diferences can double or halve the power disipated by screens...then limiters come into the role afecting the whole dynamics of power stage

[cjenrick]

"adding up the VAs for the 278CX (see below) I get 800 x 0.535 + 6.3 x 6 = 466VA which is close enough (2.5%) to its 454VA.."

remember that with a full wave circuit, you are only utilising 1/2 the sec wind at a time as the diodes only conduct 50% of the time,

now the wire they use in the sec wind will handle 535 ma, right? probably about #24-#25 AWG, so 535 ma at 800 volts might mean 428 VA, but since we can only use 1/2 the sec wind at a time to fill the caps, and since we are still limited by that #25 wire, our VA for the full wave circuit is now 535 x 400 = 214. in other words, we should not draw twice the current from through sec wind thinking that both halves are conducting current at the same time.

now we might run into something interesting with the pri wind. since Hammond assumes that their xfmr will be hooked up in a full wave circuit, and that it will only be using half the published VA, why not make the primary wire smaller? that way we can save money and have more room for insulation and maybe even bigger sec wire for less heat? they probably go for the smaller pri wire but i bet they do not increase sec wire size as copper is 15 bucks a pound and they sell their stuff for dirt cheap already.

so this means that if you were to wire a xfmr that is meant to be used in a full wave circuit and ended up using it in a bridge circuit, then you might want to watch to see how hot it gets if dragging max current out of it as that smaller pri wire might not like it.

you can search Transformer Utilization Factor to get all the ugly math for bridge vs full wave vs voltage doubler vs half wave, etc.

that transformer on the power pull can run your lights and your dryer, as soon as you turn the dryer off, it's utilization factor will go down.

here is a good article on this>

http://www.ietlabs.com/pdf/GR_Appnot...r%20Design.pdf

[bob p]

I understand the 20W, 40W thing but it does not apply to my friend/customer group. My friends/customers usually want to drop the bomb.

My friend/customer actually just posted this pic of my proto paired with his 200W orange for a nice dual fullstack rig to devestate everyone...

It sounds like "cool_ghoul" is compensating for a small dick.

[Enzo]

Little Richard?

[catalin gramada]

well, what think I made with 278cx to run in pentode/tetrode conection... I use 450v (400+50v) secondary with a bridge rectifier to supply the plates. and another small 300v PT to supply the grids and preamp stages. can use 278 heater supply for power tubes and the other tranny heater supply for preamps. a 50v winding is plenty for the bias. Of course it is a waste but think is only decent way to use it for other thinks but UL

[Dave H]

"adding up the VAs for the 278CX (see below) I get 800 x 0.535 + 6.3 x 6 = 466VA which is close enough (2.5%) to its 454VA.."

remember that with a full wave circuit, you are only utilising 1/2 the sec wind at a time as the diodes only conduct 50% of the time,

you can search Transformer Utilization Factor to get all the ugly math for bridge vs full wave vs voltage doubler vs half wave, etc.

I was just pointing out that Hammond says the 278CX is a 454VA transformer and that's how the secondary ratings add up.

I always avoid ugly math as much as possible
I didn't use the VA ratings to calculate the DC current, I used the 'Hammond Design Guide for Rectifier Use' I posted in #10. For Full Wave Capacitor Input is says I DC = 1.0 x Sec I AC so according to Hammond the 278CX with full a wave capacitor input rectifier should be able to provide 535mA DC.

For Full Wave Bridge Capacitor Input the Guide says I DC = 0.62 x Sec I AC. My 18W amp measures exactly like that but it has a 220R resistor in series with the PT secondary. I suspect the sec I AC (rms) would be higher with the 220R shorted.

EDIT
I should have said the I AC : I DC ratio would be higher with the 220R shorted i.e. higher than 1/0.62
The 220R is a sag resistor (poor man's tube rectifier)

[catalin gramada]

My 18W amp measures exactly like that but it has a 220R resistor in series with the PT secondary. I suspect the sec I AC would be higher with the 220R shorted.

I suspect those resistor is there to protect the thin wire into secondary to be stresed or burn due to startup inrush currents (you have caps there for a first half a cycle works as a short and also the rectifier have a current rating) , or maybe a sag resistor. A series resistor is anything but not a current limiter by itself. It can determine the current flowing in circuit in respect to voltage dropped on it so is a variable which can be considered as a extra winding resistance as well as we can talk about winding loses due to winding resistance wich is a voltage matter I suppose . On the other hand at startup the caps is a shunt for a fraction so a small resistance will determine a max amount of current just not to blow diodes or a fragile secondary winding. It is a good practice for bias circuits usually rated for 100 mA or so

[bob p]

So yeah then I wanted to just understand the whole topic more in general so I could know what to desire if I were to do a 2 or 6 power tube version as well. Or even in scenarios of ordering PTs for my customers for repairs. Seems like choosing a transformer listed by ratings which may be the same anyway but may NOT called "Fender Twin Reverb PT replacement) is sometimes cheaper. So I appreciate the answers.

It's interesting that the Twin Reverb PT appears to be cheaper than the 278CX, at least at Angela: $84 vs $113.

Angela Instruments Online Catalog - Search

Hammond Power Transformer 290FX

I don't know if that's the case with other suppliers, but it's interesting that the 290 series iron that's named as a Fender replacement costs less than the generic 270 series. Like you I thought the prices might be higher.

[nsubulysses]

Based on the things I've learned on this thread I am now wondering why this Marshall 50W OT that has 690V CT @ 150mA is only 150mA if it's for two power tubes. Seems like this PT would not do good when you play the amp primarily cranked. But isn't that what people do with a Marshall JMP, JCM800 etc?

http://www.classictone.net/NOS16575.pdf

It has about 490V B+ so it does 55W at clipping. I use a 50W OT with it. compared to other same amps with different transformer sets (4 power tube versions), it seems like it has a poorer range on the volume knob. Meaning that once I get to about 4 it's full volume. Increasing the volume more does almost nothing. But also it doesnt' sound worse as you exceed max volume and keep turning it up like you would expect. You know how when you go past probably 6 or 7 in volume on an amp, it usually doesn't sound any louder, and usually just sounds a little worse? It makes me guess that in that case you are exceeding the OT power handling, or PS is maxed, or NFB loop is now ineffective because power amp is distorting and it just sounds worse and not louder.

How possible does it sound that this amp may be current limited by PT? This is just something I thought of as I was perusing PT spec sheets and thinking about my amp. So maybe I'm way off but.....????

edit: Or would the PT just supply the current demanded if the power supply can supply it, but get hotter ??

[Merlinb]

remember that with a full wave circuit, you are only utilising 1/2 the sec wind at a time as the diodes only conduct 50% of the time,

now the wire they use in the sec wind will handle 535 ma, right? so 535 ma at 800 volts might mean 428 VA, but since we can only use 1/2 the sec wind at a time to fill the caps our VA for the full wave circuit is now 535 x 400 = 214. in other words, we should not draw twice the current from through sec wind thinking that both halves are conducting current at the same time.

so this means that if you were to wire a xfmr that is meant to be used in a full wave circuit and ended up using it in a bridge circuit, then you might want to watch to see how hot it gets if dragging max current out of it as that smaller pri wire might not like it.

For a two-phase cap input rectifier, the RMS current in each half of the secondary is not half but 1/(root2) times the total average ripple current. If you use that figure you'll find the numbers still add up when using a bridge rectifier.