Question about Back to Back Boost & Buck Transformer connection

[kagliostro]

To use a Back to Back connection with transformers and take heater voltage between the two transformers is an often used trick



But sometime we require a higher or lower voltage than that we can achieve with that connection, so I've think to a Buck or Boost Back to Back connection

Said this the firs step is to ascertain if this connection can work as expected

Book:


Boost:


If those connection are feasible the further step will be to know if also with such kind of connection is possible to connect between transformers as to take the Heater voltage without problems being the B+ de facto connected to the heater wiring



What do you think about all this thing ?

Thanks

K

[eleventeen]

All I can say is, if you try this, make sure your connections are properly fused!

[kagliostro]

if you try this, make sure your connections are properly fused

I'm not so sure about that way of connection with Back to Back architecture


http://sound.westhost.com/articles/buck-xfmr.htm#s30


http://sound.westhost.com/articles/buck-xfmr.htm#s40


K

[sluckey]

I'm not so sure about that way of connection with Back to Back architecture

I highly recommend using a variac for this experiment.

[kagliostro]

I highly recommend using a variac for this experiment.

That can be done

Ciao Steve

Franco

[vibrolax]

As long as neither of the filament connections are connected to the secondary of the second transformer, the filament  will be independent of the B+.

Other connections are possible.  You can use 1 secondary of the first transformer for the filaments, and the other secondary connected in parallel to both primaries of the second transformer for an AC voltage approximately equal to the mains.  If you connect the primaries of the second transformer in series, you will get about 1/2 the AC mains voltage.

If you want substantially more DC high voltage out, you can use a FW voltage doubler rather than a FW or bridge rectifier.

You can also use a different second transformer if you want a different step-up ratio.

My own favorite hack using a step-down transformer for a tube circuit supply is to use a dual primary, dual secondary transformer.  One primary is connected to the mains, the other "primary" becomes the high voltage "secondary".  The two low voltage windings supply the filaments.

I show how to do this here (also using the FW doubler):
http://www.frontiernet.net/~jff/af2b/Dual_Channel_Blackface-style_Preamp_for_Bass_Guitar.html

The VA rating of the transformer is essentially cut in half, because only one primary is connected to the mains.  This 12 VA transformer works fine for a couple of 12AX7's.

I don't like the "back-to-back" thing, because 2 transformers take up more space in the chassis.  I suppose it's ok if you want to power your circuit "wall wart" style with the first transformer outside the chassis, and only low voltage AC entering the chassis.

If you have 220 VAC mains, a full wave bridge will probably give you a high enough B+.  You didn't say how high you wanted the high voltage to be.

You can dump excess B+ voltage (and filter the output well) with an extra RC stage for your filter.  Once again, if you're just powering 12AX7's, then it's usually no big deal to dissipate the excess with resistance.

Jon

[kagliostro]

Ciao Vibrolax 

As long as neither of the filament connections are connected to the secondary of the second transformer, the filament  will be independent of the B+.

But they are connected

You can use 1 secondary of the first transformer for the filaments, and the other secondary connected in parallel to both primaries of the second transformer for an AC voltage approximately equal to the mains

This way there will be a loss of 50% in current (although can be that is sufficient for B+)

You can also use a different second transformer if you want a different step-up ratio.

This can be done if you want a lower output voltage (220v/12v-15v/220v) but not if you want a higher output voltage (220v/15v-12v/220v) this way there is core saturation and heat development, we discussed this thing here in the forum some time ago

If you want substantially more DC high voltage out, you can use a FW voltage doubler rather than a FW or bridge rectifier

Here we have 220v mains so is easy to find transformer with 220v windings, if you use a doubler with 220v AC you obtain ~600v DC, too high for a lot of applications

My own favorite hack using a step-down transformer for a tube circuit supply is to use a dual primary, dual secondary transformer.

Don't feasible 220v+220v primary transformers are like Unobtanium here

..... 2 transformers take up more space in the chassis.

Yes, you are right, but can be a way to reuse something is under hand (and cheap)

If you have 220 VAC mains, a full wave bridge will probably give you a high enough B+.  You didn't say how high you wanted the high voltage to be.

with 220v AC I can have ~300v DC, if I want to clone a preamp (or a small amp) often 300v B+ are less than required

You can dump excess B+ voltage (and filter the output well) with an extra RC stage for your filter.  Once again, if you're just powering 12AX7's, then it's usually no big deal to dissipate the excess with resistance.

Yes, so it is, but not feasible if we are talking of a small amp like Champ

---

This morning a new idea come to me, this one





I think this will work (I hope)

there is always to find out a way to obtain the voltage for the tube heaters

Franco

[tubeswell]

This morning a new idea come to me, this one





I think this will work (I hope)

there is always to find out a way to obtain the voltage for the tube heaters

Franco

Do the rectifiers on the 30 VAC supply not need to be filtered (and smoothed?) before providing a reference for the rectifier on the 200VAC supply?

[kagliostro]

I think they must .......... only I forgot to draw the necessary electrolytic capacitor, SORRY


Thanks for pointing that Tubeswell

my error apart, do you think this thing will function ? are you able to give a council about how to obtain the voltage for the heaters ?





Thanks

Franco

[tubeswell]

Not sure that you would shave enough VA in those transformers for running heaters. Aren't they intended for transistor-only circuits?

[kagliostro]

I didn't remember exactly which is the disposable current on the 15v windings, however the transformer are toroidal for halogen lamp supply, if I'm not wrong there are 3A or 5A disposable on each 15v winding


I was thinking to use a circuit like this for the heater, but I'm not sure this way I can get out 21v DC (15v x 1.4)



Some time ago I asked something that was similar, but my brain refuses to collaborate and i don't remember what was told at that time 


K
 

[vibrolax]

Your calculation for the full wave bridge filament circuit does not include the forward voltage drop from a pair of diodes (2 diodes out of 4 are conducting during each half of the AC cycle).  The drop depends on what diodes kind of diodes you use.  For a 1N5404 the forward voltage drop is 1.2V.  So a pair drops 2.4V.  http://www.vishay.com/docs/88516/1n5400.pdf

There is also the ripple voltage that depends on the filter capacitance and current.  Sorry I can't provide a reference for this calculation at this moment.  I suspect you'll get around 18 VDC, unless you use diodes with a lower forward drop.

On the B+ side, the diode voltage drops are usually negligible, but as you can see they are >10% at 21V.

I would think about arranging my tube filaments 12.6V and 6.3V in series+parallel in order to avoid dropping filament voltage with resistors.  A 6V6 filament draws 600 mA. 4 x 12AX7 draw 600 mA at 12.6 V.  All filament groups connected in a series circuit need to see the same current.  If you can come up with a sensible series-parallel arrangement, you won't need to drop any filament voltage with a resistor.

A couple follow-up comments about your response to my earlier post:  About dual primary, dual secondary TX's:
I don't think one really needs dual 220 + 220 primaries.  I'd feed 220 VAC into a 120 VAC primary. (and get 2x the voltage on the secondaries).  The primary winding insulation is probably safe when overvolted by at least 2X.  You could check the detailed CE conformance directives listed on the transformer data sheet.  The primary to core and primary to secondary insulation is usually good for 1200V or more.

Anyway, I've only powered preamps this way.  I think when you need more than 20 mA B+, then properly designed transformers are just more space and cost efficient.  All the extra diodes and caps add cost and take more space.  But I understand wanting to see how far one can go with parts that one has laying around.  I enjoyed making a tube preamp power supply with a ~3 euro surplus step down power transformer.

Cheers,
Jon

[darryl]

A 6V6 filament draws 600 mA.

It's actually 450mA.

I'd feed 220 VAC into a 120 VAC primary. (and get 2x the voltage on the secondaries).  The primary winding insulation is probably safe when overvolted by at least 2X.  You could check the detailed CE conformance directives listed on the transformer data sheet.  The primary to core and primary to secondary insulation is usually good for 1200V or more.

This is very dangerous advice. Running a 120VAC primary at 240VAC will almost certainly cause core saturation and severe overheating.

[vibrolax]

You're right, it is dangerous to significantly overvolt a transformer due to core saturation, so don't do it.  I wasn't thinking about V/turn for the core area.

Right also about the 6V6 filament.  One could place 6V6 filaments in series with  3 x 12AX7's @ 12.6V in parallel to run off an ~18 or ~19 VDC filament circuit.

Jon

[kagliostro]

Yes, I forgot the drop in the diodes, I'll keep it in mind

A guy told me that the voltage won't be near 20v but will be near 40v because the diodes in the green circle will never conduct



 

I'm confused   

---


For tube preamp (also in U1 rack) I can get this (and I've 3 at the moment)

http://www.musikding.de/Toroidal-transformer-230V-270V-14V

But for a small amp that is not usable (only 80mA B+)

However, to reuse the PT I've may be important, but much more important is to learn how to do things


K

[Merlin]

Do the rectifiers on the 30 VAC supply not need to be filtered (and smoothed?) before providing a reference for the rectifier on the 200VAC supply?

No capacitor is needed. It works just like a boost winding, except we have mooved the boost to the pulsating DC part of the circuit rather than the AC part.

I'm confused   

Here:

[darryl]

You could forgo the back-to-back arrangement, and use one transformer into a voltage octupler - or more correctly an eight times voltage multiplier:  http://en.wikipedia.org/wiki/File:Stacked_Villard_cascade.svg  The second transformer could then be used just to power the heaters, removing the concern of having the heater supply tied to the B+ supply. Although the octupler requires eight filter capacitors, they only have to be rated at 100 volts.

The amplifier described here uses this arrangement, and has done so successfully for over five years of regular use.

http://www.guitargear.net.au/discussion/index.php?topic=17009.0

[kagliostro]

Many Thanks Merlin and Darryl

@ Merlin

As far as i can understand there is a full wave rectifier connected to a "virtual" (15v-0-15v) winding with CT

(I say virtual 15v-0-15v because the transformer has 2 x 15v separated windings not a single 30v (15v+15v) CT winding)

with this configuration, assuming each 15v winding has 3A of disposable current, which will be the disposable current for the heaters 



---

@ Darryl

I know you have a large experience with doublers but this octupler is like a rabbit pulled from a top hat 

is correct to assume that the B+ current disposable is ~1/8 of the current disposable on the PT ?

---

Before I read your posts, this morning I found that I've no more the 2 15v transformers, but I've 3 (one is in excess) 220v/12v+12v 3A transformers

12v + 12v is a bit less than 15v + 15v, so a new Idea come to me, this one (I pulled the unnecessary capacitor as stated by Merlin)



At this point I know I can have it done as from Merlin schematic, but I would like to know if this last schematic is correct or wrong

Many Thanks

Franco

[Merlin]

with this configuration, assuming each 15v winding has 3A of disposable current, which will be the disposable current for the heaters 

It depends. How much DC current does the high-voltage supply need to provide?

[sluckey]

Franco, is this just a brain exercise or do you have a real world application in mind?

[kagliostro]

Grazie Merlin

Assuming the B+ requirement in current is ...... (??) ...... say 100mA

EDIT: but we can discard the B+ part of the circuit, the current will be 3A or 6A 

Franco

[kagliostro]

@ Steve

the simplest schematic draw by Merlin will be used

the other schematic is more a brain exercise as to know what can and what cannot be done 

Franco

[6G6]

Wouldn't it be a lot simlper to just use a small filiment transformer?
It has been my experience, in general, that the more parts you have, more chance of one of them failing.
Sometimes, simplicity is a virtue in itself. 

[Merlin]

This would also be simpler. With 90VA transformers you could easily allow 100mA for the high voltage, and 2A for the heaters.

[kagliostro]

Thanks

Franco

[darryl]

I know you have a large experience with doublers but this octupler is like a rabbit pulled from a top hat 

is correct to assume that the B+ current disposable is ~1/8 of the current disposable on the PT

That is correct. Power supplies don't come with free lunches included.

The transformer used in the amplifier linked to above is rated: 30VAC@2A

The approximate B+ voltage in an octupler is 30 x 8 x 1.4 = 336VDC

The approximate available current is ( 2000 ÷ 8 ) ÷ 1.4 = 178mA DC

[kagliostro]

Thanks


Franco

[darryl]

Regarding a octopler, I believe Darryl is being optimistic, as he hasn't accounted for the voltage drops across the diodes, and current losses through those imperfect capacitors.

I'm practicing to become a politician - I did include the disclaimer "approximate". 

The real-world result is included here in the thread I linked to: http://www.guitargear.net.au/discussion/index.php?topic=17009.msg161314#msg161314

Both back-to-back transformers, and voltage multipliers are inefficient, but they are ways of utilising components which are either easily obtained, or already on-hand.

[kagliostro]

I'm practicing to become a politician - I did include the disclaimer "approximate". 

 

  

Franco