PUSH PULL AUDIO AMPLIFIERS

[W2PFY]

Need a little input here from the experts.

If I were to build a mono amplifier for class A or AB1 would I be able to use two single ended transformers? I know there would not be a common core but would it work? The secondary's in my thought would be tied together

 

[k4kyv]

What will you be using it for?  Modulator driver?  To drive a speaker?

[W2PFY]

That would be for driving a speaker Don.

[KX5JT]

SO... Single ended mono... or Push Pull?  It's either one or the other but not both right?

[W2PFY]

The idea is for push pull using single ended transformers for mono. Would it work?

Both transformers would be used but only for one channel as if it were a transformer with a common core.

[WD5JKO]

Terry,

   Maybe you can post a sketch of what you are thinking about doing.

Remember though that the output transformer in a conventional P-P circuit is part of how the 2nd harmonic distortion is nulled out (when balanced). I'm not sure your idea will accomplish the same.

If you made the AMP class A, or A2, using both transformers in parallel would work fine as a moderate power single ended mono amplifier so long as both transformers are identical.


Regards, and good luck with your idea.

Jim
WD5JKO

[N2DTS]

I dont see why it would not work, but I dont think it would work well.
There will be differences between transformers, and combining the 8 ohm side does not sound like a good idea.
Any differences will show up, all you need is a little phaze difference and since the transformers operate over a wide frequency range, the output waveform might look a mess.
Hard to see two different transformers combining into a clean waveform from 20 to 20KHz...

Brett

[WBear2GCR]

Terry,

Should work ok, assuming identical output trannys. But the question is why? PP output iron is easy enough to find. Cheap enough too... Decent SE output iron is more expensive, so if what you have is decent then you could just sell it and buy good PP iron with no difficulty.

How many watts?
Which toobes??

                     _-_-bear

[KM1H]

I wonder if an artificial center tap could be used? That works on TX tube filaments. Finding the correct balancing resistor values may be a problem??  Dunno, never considered the problem before.

Carl
KM1H

[N4LTA]

My Guess is that it will work pretty well but that the power handling will be about 160% of one of the transformers (not double )- the same as when this is done with a power supply transformer - when two untapped transformers are combined to make a common ct transformer connection. See the ARRL Handbook for an example.  

Since the SE transformers are designed to work with DC flowing through them -  it should work OK.

A push pull transformer has no net DC current flowing in it to saturate the core. If you use two SE transformers - that is not the case - each transformer has the whole DC  current flowing in it for each tube and it will tend to saturate the core - but if you use SE transformers - they are made to work with DC current flowing.

The transformer should be the same type. If they are not equal in wire size and turns ratio - then there is a possibility that one will "hog" most of the power and overheat.

I really don't see any other problems  (like phase problems or generation of spurious signal and distortion)

Try it and let us know how it works.


Pat
N4LTA

[W2PFY]

Well guys so far a lot of good comments. This is purely a hypothetical exercise to solicit comments about the feasibility of such a circuit and weather it can be done. I have never hear of anyone doing it before. The closest circuit I have seen is when someone uses the primary only of a transformer in push pull hysing connected modulator. I believe I have also seen a modulator circuit using large inductors in push pull again with the audio tapped directly from one of the modulator tube plate caps connected directly to the PA final.

[K5UJ]

Terry how much audio power are you looking for to drive the speaker?

I have had good luck with these little 10 to 20 watt old  vac. tube amps like the Heathkit EA 2 and EA 3.  Manuals have all the info needed to fix them up.  Enough power to drive a decent bookshelf speaker.  I use them to boost the receive detector audio and they make a new s.s. rig sound more like an old tube boatanchor.  12AX7 preamp works great on the prod. detector tap.  Only prob. is those pair 6BQ5 p.p. get pretty hot.  But these amps can be cheap.  I find them at hamfest fleamarkets but here's a couple on eBay:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=230411982970

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=250546682443

maybe this better than JSing two amps?  unless you need more wats 

Rob

[W2PFY]

hi Rob, I don't need an amplifier right now. I was just looking for ideas for a project down the road using Hammond single ended transformers. I have seen the quality of these transformers. One is at Bill's KC2IFR's QTH. I think its good for over 100 watts but I'm not sure.


I hope that since this is not an immediate project for me that ideas, answers will still come in.

That comment presented by Pat about transformers in parallel would only produce 160 percent of the total current current available from each transformer was news to me. I wonder if that would hold true of two of the same transformers in series as far as kva is concerned?

[N4LTA]

It was news to me until I was studying the ARRL handbook on tapped vs bridge rectifiers a few days ago.

From the ARRL HB 1965 (and most other years)

" Two separate transformers, with their primaries connected in parallel and and secondaries in connected is series (with the proper polarity) may be used in this circuit" (the full wave center tap circuit)" However, if this substitute is made, the primary volt ampere rating must be reduced to about 40% less than twice the rating of one transformer.

I often use the Hammond SE series transformers. The largest is very conservatively rated at 25 watts and 100 MA (the 125GSE). I just got one from UPS today to use in modulating a single 807 with a solid state amp. I have used these in se audio amps and they are very conservatively rated.

The 16xxSE series are rated up to 75 watts but they are expensive. I am a Hammond dealer and can get these to experiment with at cost.

Two of these would make a good push pull transformer - but why do it?

Two 125ESe s cost about $51 each for a total of $102 at a 30 watt rating (assuming you don't need to de-rate to 24 Watts.

A single 1645 rated at 30 watts PP cost about $66 and has better frequency response.


Pat
N4LTA

[W1GFH]

I have had good luck with these little 10 to 20 watt old  vac. tube amps like the Heathkit EA 2 and EA 3.  Manuals have all the info needed to fix them up.  Enough power to drive a decent bookshelf speaker.  I use them to boost the receive detector audio and they make a new s.s. rig sound more like an old tube boatanchor.  12AX7 preamp works great on the prod. detector tap.  Only prob. is those pair 6BQ5 p.p. get pretty hot.  But these amps can be cheap.  I find them at hamfest fleamarkets but here's a couple on eBay:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=230411982970

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=250546682443

Those look like great amps for folks who want to pick off audio from their AF-challenged receiver's volume control like many R390A owners do to have the best of both worlds. $26 doesn't sound bad. But I wonder what the eBay auction prices eventually get to.

[W7TFO]

Gates Radio did this sort of thing in their BC-1J model AM 1kW transmitter.  It had two sets of 6L6's in push-pull, driving through seperate output transformers,  one into each grid of the 833A modulators.  Balance was achieved by looking at a scope on the demodulated output waveform and making adjustments to the input audio splitter and the respective bias on the 6L6 stages.

Not very popular, it was used on only the one model.

Dennis

[Gito]

Hi

Since  now we find stereo solid state amplifiers( 2 independent audio amplifier) used for mono amplifier, using a bridge  configuration  and in theory it can give 4 times the power output of a single amplifier,if the mono audio can deliver the current needed,and the rating of the power transistor is enough

Can we used it in mono tubes Amplifier?I believed it can.
The theory is, to used a phase splitter in front of those mono amplifier ,that makes the audio injected to it has a 180 phase difference ,each audio output with 180 % phase input injected in this 2 mono amplifier gives an output with 180% phase difference.
Since the two output winding( Loud speaker connection) from the two out put has a common ground.
and if we used it as an audio amplifier the speaker is connected from the "live" (ungrounded secondary winding) first Transformer output to the "live" of the second Transformer output.
Than we got more  output than two output amplifier in parallel.(in theory?)
Maybe I'm wrong.


Gito

a picture of what i'm trying to explain

[K5UJ]

Those look like great amps for folks who want to pick off audio from their AF-challenged receiver's volume control like many R390A owners do to have the best of both worlds. $26 doesn't sound bad. But I wonder what the eBay auction prices eventually get to.

It depends on the usual eBay random auction factors but unfortunately, if one or two glass audio nuts get involved they can go way too high.   I've seen one get up there around $200 which is ridiculous but these people think that's a steal because their audio salons are selling tube AF amps for thousands of dollars.  $50 or $60 for one that's complete and clean isn't bad.   They're even higher now at ham flea markets because of the audiophool effect.   Last summer this guy at a hamfest (he wasn't a ham) had a pair of the Heathkits for sale.  Wanted $125 each.  Wouldn't bargain and wouldn't split them up.  I said "Buddy, you're at a hamfest, not some antique show.  Good luck."  At the same fest a friend beat me to a beautiful little Grommes tube amp for around $60, to pick off the detector audio and boost it.   If you have a hamfest coming up I'd wait for that unless something is on eBay and not getting bid way high.

[W2XR]

Gates Radio did this sort of thing in their BC-1J model AM 1kW transmitter.  It had two sets of 6L6's in push-pull, driving through seperate output transformers,  one into each grid of the 833A modulators.  Balance was achieved by looking at a scope on the demodulated output waveform and making adjustments to the input audio splitter and the respective bias on the 6L6 stages.

Not very popular, it was used on only the one model.

Dennis

Dennis, actually this audio driver amplifier configuration was used in Gates' 250 and 500 watt broadcast rigs from that period as well, along with a 1000 watt 2-30 Mhz HF transmitter Gates sold in very limited quantities for shortwave broadcast applications. The entire series of 250 to 1000 watt plate modulated AM broadcast rigs were referred to as the "Gates Hi-Watters".

I pulled apart a Gates BC-1J a number of years ago, and it had that funky audio driver deck. In my opinion, the design was a real POS, and it subsequently went to a fellow in Texas who was restoring a BC-1J. The Gates rigs of that period (circa 1956) marked the beginning of the decline in Gates Radio transmitter quality. What a departure from the superb build quality of the BC-1F!

73,

Bruce

[W7TFO]

I agree fully.

DG

[WBear2GCR]

I was going to suggest the bridged configuration, rather than the parallel one, but Gito got there first.

Still seems like a waste of good SE iron...

                                _-_-bear

[WD5JKO]

It was news to me until I was studying the ARRL handbook on tapped vs bridge rectifiers a few days ago.

From the ARRL HB 1965 (and most other years)

" Two separate transformers, with their primaries connected in parallel and and secondaries in connected is series (with the proper polarity) may be used in this circuit" (the full wave center tap circuit)" However, if this substitute is made, the primary volt ampere rating must be reduced to about 40% less than twice the rating of one transformer.
Pat
N4LTA

   I recall trying this many years ago as part of a quick R&D effort I was part of at an old job. We needed a transformer to provide 60V CT at 200 amps into a two diode FW rectifier with choke input filter. At the time we only had two 30V 200 amp transformers lying around, each about the size of  CVM5 transformer. So we tried the two as Pat described. The arrangement worked GREAT! But then we noticed a subtle side effect of doing this. The combined primary current of the two transformers was exactly TWICE what we were expecting. At this kind of power level seeing 2X normal amperes was a show stopper. Oh, and nothing got hot.

   I'm still not sure what went on, but some theories tossed about include:

* Each transformer sees a pulsing half wave load, and possible core saturation was occurring.
* The KVA at the primary had a large imaginery component. Still Amps were Amps, and exceeded our Power Distubition Amp budget.
* Proper operation required a single core with the C.T. secondary winding.

Not sure how relevant my example is to the audio transformer configuration proposed in this thread.

That transformer we ultimately used was 60V C.T. with 150,000 V insulation primary to secondary. Yes I am in one crazy industry, ION IMPLANTATION. I vividly recall developing that transformer. At one point, after many tries taking apart, adjusting turns ratio, or winding configuration, we applied primary power to it but forgot to put the 'C' shapped iron core back on. We were using two 120v circuits (unfused) from a building with 208v 3 phase feed. This gave us 208v. With no core on the widings, the 60 HZ scream was deafening! The power to the entire building went dark. Oops!

Jim
WD5JKO

[Gito]

Hi Jim


What You are doing ,is combining two half wave rectifier power supply and in Motorola silicon rectifier handbook stated that in Half wave diode configuration the  average current needed is twice  the current of full wave center tap configuration / twice the full wave bridge configuration,

So a half wave rectifier power supply needs twice the needed current than a full bridge power supply.

Actually what You do, is combining two independent half wave rectified power supply,so  the current must be high (twice a full wave/full bridge power supply)

Maybe if we used a full  bridge configuration for each power supply and than combining them( 30v--0V in series 30v--0V)  ,than we get the right current like a normal one power supply with  full wave configuration.

Or putting the primary in parallel ,secondary in series but left the "artifical"
Center Tap floating ( not grounded) and used a full bridge configuration.
AM I right?

Gito

[WD5JKO]

Hi Jim
What You are doing ,is combining two half wave rectifier power supply and in Motorola silicon rectifier handbook stated that in Half wave diode configuration the  average current needed is twice  the current of full wave center tap configuration / twice the full wave bridge configuration,
So a half wave rectifier power supply needs twice the needed current than a full bridge power supply.
Actually what You do, is combining two independent half wave rectified power supply,so  the current must be high (twice a full wave/full bridge power supply)
Maybe if we used a full  bridge configuration for each power supply and than combining them( 30v--0V in series 30v--0V)  ,than we get the right current like a normal one power supply with  full wave configuration.
Or putting the primary in parallel ,secondary in series but left the "artifical"
Center Tap floating ( not grounded) and used a full bridge configuration.
AM I right?
Gito

  Gito, Yes I think you are right, makes sense to me. That old Motorola diode book is a classic that I also own and proudly put on my technical bookshelf. On the project we discussed, going to a full wave bridge, or two full wave bridges would heve been a challenge at the 200 ampere current level. Those diodes were pretty large, and expensive at that time (1981). We had to water cool them, and play games with a R-C snubber so that transients did not avalanche them. Big diodes like that back then were very slow to turn off, about 1ms reverse recovery time; so with 60 hz, 1ms out of 8.333 ms would place a dead short across the transformer. Combine that with a high leakage inductance transformer, the situation was challenging. The power supply load was a huge electromagnet, and the current was controlled with a water cooled linear regulator composed of 32 2N6259 transistors in parallel. The power FET's of that era were not up to the task althouh the IRF-100 was available and came close to what we could use if a bunch were combined.

   Terry, If you try your idea with those SE audio transformers, by all means please post any test results you get.

Regards,
Jim
WD5JKO

[W2PFY]

Gito, I'm going to try your schematic. I don't know when and I don't have the xformers but will try a couple filaments transformers as output transformers to test the feasibility of the circuit.

Thanks again Guys.

Terry