Output transformer essential? - UK Vintage Radio Repair and Restoration Discussion Forum

YT2095UK · 3rd Sep 2009, 8:01 pm

I was wondering during one of my inumerable experiments if a special output transformer was really needed for an audio amp, I didn`t have one and so I used a simple mains transformer taken out of an old CB psu, it was a 2 amp 12 volt power supply, so I esstimate it gave about 17(ish)VAC on the secondary.
the Primary winding was roughly 45 ohms if memory serves.
cut a long story short I used the mains side (primary) directly to the plate of an EL34, and the secondary to drive a 4ohm 10W speaker taken from a pair of old Labtec computer speakers (again, that`s all I had to hand).
the Cathode of the EL34 was fed with a 220R resistor and had a 8Uk 450v cap across that, and was driven with half an ECC83 (I haven`t decided what if anything I`m going to do with the other half yet).

the sound out if it lovely! I can`t hear any distortion at all, and nothing gets hot (other than that which should) when it`s left running for a few hours playing radio fed from the Ext Spkr socket on a Cossor 520.

was I just Lucky or am I missing something vital in my "Newness" to this technology?
do you really Need a special transformer?

Thanks in advance

G8HQP Dave · 3rd Sep 2009, 9:05 pm

It all depends. A mains transformer is designed to give efficient transfer of 50Hz power. An audio output transformer is designed to give a balance of efficient, low distortion, wide bandwidth audio power and, in most cases, coping with a significant DC current through the primary. These two roles lead to different design compromises.

A mains transformer has to have very good insulation for safety reasons, as the mains can put almost unlimited current through a short circuit. A valve output transformer is in a slightly safer environment, but it has to cope with a wide variation in frequency - maybe 20Hz to 50kHz for hi-fi. The upper limit is well above the normal human audible range to give some scope for stable negative feedback. In a radio set it might only have to cover 100Hz to 5kHz, as often no feedback is used.

The DC in the primary of an output transformer can cause reduction in inductance (which affects the low frequency end) or core saturation which stops it working altogether. This problem is avoided by adding a small air gap to the core, but this then means many more turns to get the required inductance.

So if you are not too bothered about frequency response or distortion then a mains transformer can be used as a low power output transformer. If your cathode decoupling capacitor is only 8uF then you won't get much bass anyway - you need more like 25uF for a radio or 500uF for hi-fi.

YT2095UK · 3rd Sep 2009, 9:22 pm

funny that you should mention Bass, I did indeed notice that without the cap the bass simply Wasn`t.
and a direct short of the cathode to 0V, gave the most bass, but the Really perculiar part was that adding a large cap across the HT supply boosted the whole thing entirely as well as general overall clarity.
I assumed (perhaps wrongly it would seem) that HT smoothing caps were only to buffer the ripple and reduce hum, there was no hum so I didn`t have the second one there.
I didn`t follow any known circuit diagram, and the whole thing was just knocked up out of bits-at-hand and designed with pure Logic wrought from trial & error.
it has presented me with a whole plethora of questions however, but I think I`ll put off asking them here for a while until I get a better feel for the place (there really is plenty of questions), as I don`t think it would be fair to everyone else.

I must admit though, the sound from this amp is astonishingly clear to my ears anyway, and my wife agrees, I have bought a Proper output transformer a couple of days ago and look forwards to trying it out in place of these Mains one, just to see if there really IS a difference, I`m hoping so!

oh, 8uF is the largest small cap I have at that sort of voltage, all the others start at 220uF and upwards, or are 0.1uF downwards at several KV rating.
I hope to rectify this in future as this is a totaly New area of technology for me and I`m ill equiped for parts suitable.

G8HQP Dave · 3rd Sep 2009, 10:25 pm

Quote:

adding a large cap across the HT supply boosted the whole thing entirely as well as general overall clarity

As well as reducing hum the HT smoothing capacitor also provides an AC ground for the HT, or a return path for AC currents (same idea, different description). This is particularly important for a single-ended output; a push-pull stage balances out this current anyway.

Quote:

8uF is the largest small cap I have at that sort of voltage

For an EL34 cathode decoupler you only need a voltage rating of 50V as the cathode voltage will be somewhere around 30-40V.

YT2095UK · 3rd Sep 2009, 10:49 pm

Quote:

Originally Posted by G8HQP Dave

For an EL34 cathode decoupler you only need a voltage rating of 50V as the cathode voltage will be somewhere around 30-40V.

Now That`s interesting to note, as I have several caps of the 63V range, I think I`ll try a few of these out also.
I just figured that since the Anode load was only ~45 ohms rather than in the 100`s or 1000`s ohms, that keeping the cathode Higher would compensate somewhat.

one of the questions I did have RE; cathode decoupling was Does the other end of the cap have to go to deck? or can it go to the +HT (obviously reversed inpolarity as well).
need it always be in paralel with the resistor, or can it be in series with, to the Other rail +HT instead.
otherwise I saw it as setting up a sort of Timebase LC set-up and that would also give it a center frequency (something I wanted to avoid).

jjl · 4th Sep 2009, 11:17 am

The purpose of the cathode bypass cap is to provide a steady DC bias voltage at the cathode of the output valve. Without it, the signal current flowing through the cathode resistor causes the voltage at the cathode and hence the grid bias to vary in sympathy with the signal, this variation acts as negative feedback and reduces gain.

You're correct in stating that the capacitor and cathode bias resistor act as a high pass filter, so for best bass response you need to use a large value capacitor.

The capacitor should be connected from cathode to 0V and must be in parallel with the resistor otherwise no DC current will flow in the cathode. Connecting the other end to HT would provide no benefit, would require a much higher working voltage part and and would possibly cause the output valve to oscillate.

If you want to learn a good deal on valve amplifier output stages, Valve Amplifiers by Morgan Jones is an excellent read.

John

YT2095UK · 4th Sep 2009, 11:31 am

Thanks both! I have noticed very subtle changes when I alter certain values of resistor and caps in my amp, and it`s great to learn the reason behind these changes on a more technical level from those that Know.

my Logic for using a simple mains transformer was after having seen the small audio transformers that you get on these Science fair 100 kazillion in one electronics sets from the 70`s 80`s this mains transformer looks just like a scaled up version of one of these and I knew that the 180 volts at a few ma in my amp was never going to blow the mains side or be dangerous/fire hazard as it`s designed for 250VAC and the secondary was capable of driving a 10W speaker without damage, That and the fact I didn`t have a propper speaker output transformer handy

slightly off topic, the other half of the ECC83, would it be worth doubling up and wiring it in paralel with the other half?
or should I just leave it with the heater off and have a switch as a backup in case the other halfs fillament failed?

thanks again each!

G8HQP Dave · 4th Sep 2009, 11:48 am

Quote:

slightly off topic, the other half of the ECC83, would it be worth doubling up and wiring it in paralel with the other half?
or should I just leave it with the heater off and have a switch as a backup in case the other halfs fillament failed?

Yes, it is OT so let's hope the mods are feeling flexible! No point in parallel wiring. Just leave it disconnected until you find a use for it.

I would like to second the suggestion that you read Morgan Jones' book on Valve Amplifiers. There is plenty of information on the web too, although some of it is ignorant nonsense! One useful site is http://www.freewebs.com/valvewizard/index.html; although aimed at guitar amps most of what it says applies to any valve amp.

YT2095UK · 4th Sep 2009, 11:53 am

I know where next months "Science Budget" is going, Thanks for that

Herald1360 · 4th Sep 2009, 1:38 pm

[QUOTE=YT2095UK;269518]

Quote:

Originally Posted by G8HQP Dave

I just figured that since the Anode load was only ~45 ohms rather than in the 100`s or 1000`s ohms, that keeping the cathode Higher would compensate somewhat.

The 45 ohms is only the DC resistance of the transformer primary winding. Providing the core is not saturated by the DC anode current, the actual load resistance seen by the valve anode will be the speaker impedance times the transformer winding ratio squared. In this case (240:17 transformer and 4 ohm speaker) the load seen will be about 800 ohms which is rather less than the optimum (around 3500 ohms) but will still work OKish with a reduced output power capability. FWIW a transformer with a 9V secondary would be about right, ratio wise.....

Chris

YT2095UK · 4th Sep 2009, 1:47 pm

You are a Darling! it just so happens that I have 9V transformer here too

it did seem a bit Over-engineered and a waste to be using that 2A transformer just to drive a speaker when it could be used to power Filaments

I`m going to write down what you posted and see if I can work out a YT2095 friendly formula from it for future ref, Cheers!

Station X · 4th Sep 2009, 1:53 pm

Look up the optimum anode load for the valve. Divide it by the speaker resistance to give the impedance ratio. Take the square root of that to give the turns ratio of the output transformer.

YT2095UK · 4th Sep 2009, 2:04 pm

I only have access to the Max ratings for the EL34 as: Va 800v Maa 91.
I`m running at 180vDC (that drops to 120vDC when heated), and the supply is 380vDC through a 2k7 resistor.
ideally I need a curve graph of optimum voltage over current to plot my specs, any resources other than the National Valve Museum, would be welcome at this point

G8HQP Dave · 4th Sep 2009, 2:26 pm

Google "EL34 valve data"

YT2095UK · 4th Sep 2009, 2:46 pm

fair enough, and I have done this last week in fact, however there is much there that I don`t understand, for instance I know that mine Single ended, but what Class is it?
in fact what does class even mean?
for instance: http://tdsl.duncanamps.com/show.php?des=el34 in the application data, there`s no mention of 180V so I guess I`m already outside of parameters, BUT that`s not always a bad thing is it? as long as it`s not dangerous, it could just mean that I`m Under driving it and outside of their Specific test parameters for that trial.
I think I`ll just finish off the few experiments I have left to do with it, namely that new OP transformer when it arrives, and the cathode decoupling cap (raising it`s value), and then just tidy up the temp wiring and be done with, it`s already far exceeded my expectations anyway.
it also works amazingly well with my guitar and a quick crystal radio that I knocked up this morning.
I may post a picture of it when I can figure how to get my cam down to 200k, I think a trip to photoshop may be in order *sigh*

again, thanks each for your help and info

ppppenguin · 4th Sep 2009, 2:55 pm

Reasonable explanation of amplifier classes:
http://en.wikipedia.org/wiki/Electronic_amplifier

All single ended audio amps must be class A. Push-pull amps can be class A, AB or B. C is for RF power amps only while anything other class involves switchmode technology. All true for valve or transistor.

If you're using XP then the windows powertoys image resizer is the simplest thing to use. http://www.microsoft.com/windowsxp/D...powertoys.mspx
Otherwise Irfanview is very simple. Photoshop is a sledgehammer to crack a nut for this.

YT2095UK · 13th Oct 2009, 12:13 pm

I`v been continuing with this experiment, and have some interesting news to share, I managed to get hold of a Stentorian speaker transformer, it has a brass arm whereby you can select which winding sounds best for your speaker, and the Primary is center tapped also.
so I`v been experimenting with this and different configurations and by far the Best ever (after studying Ultra Linear designs) was to have the HT on one end of the primary, the EL34s Anode on the Other end, and the center tap going to Grid 2.
so I guess it`s a sort of single ended ultra linear type, and the sound is fantastic and Very hard to get any sort of distortion out of it (beyond abusing it!).
I did finally parallel the ECC83s triodes, but I think I`ll undo them and have a go at something called Cascode next and see what that sounds like.

but the Final stage I`m very happy with and recommend this idea to anyone that needs to make a simple amp out of scrap bits and pieces that sounds really nice!

murphyv310 · 13th Oct 2009, 7:31 pm

Hi.
There's also another way to leave out the speaker tranny, A higher impedance speaker could be hung on the cathode of a cathode follower, two triodes paralleled together will give a lower cathode impedance and greater power handling. A couple of low impedance triodes may even drive a 15 ohm speaker fairly well.
cheers
Trevor

Herald1360 · 14th Oct 2009, 12:58 pm

May be a problem with the cathode current offsetting the speaker voice coil unless power out from the amp is << speaker rating. RC coupling would be inefficient and LC no cheaper than using a transformer would have been

Chris

YT2095UK · 14th Oct 2009, 4:40 pm

other than converting the PSU to a Valve rectifier (it`s the 1n4007`s currently) and tidying up the wiring a little, my Amp is complete now, but I`m also interested in putting into practice some of the things I have been reading about and also experimenting with ideas that I haven`t read or heard of, and so RC/LC coupling and pre-amp configurations will all be tried out on this.
Some will make the amp Better than it is and some worse than it is currently, but I`ll always know how to put it all back together again in a working configuration should anything unfortunate happen.

the Cost isn`t anything I`m concerned with as I have all the parts I need anyway, and I can`t think of a better way to use them than to Learn from them, that in itself justifies any costs in my opinion

the Cathode coupling seems Very intriguing, a gain of only Unity at best but Higher current, sounds almost ideal for direct driving a speaker, i`v read that LEDs can be used to bias a cathode (I still don`t quite beleive that yet though), so maybe Via a 1000uF electrolytic to the speaker may work also?

3rd Sep 2009, 8:01 pm	#1
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Output transformer essential? I was wondering during one of my inumerable experiments if a special output transformer was really needed for an audio amp, I didn`t have one and so I used a simple mains transformer taken out of an old CB psu, it was a 2 amp 12 volt power supply, so I esstimate it gave about 17(ish)VAC on the secondary. the Primary winding was roughly 45 ohms if memory serves. cut a long story short I used the mains side (primary) directly to the plate of an EL34, and the secondary to drive a 4ohm 10W speaker taken from a pair of old Labtec computer speakers (again, that`s all I had to hand). the Cathode of the EL34 was fed with a 220R resistor and had a 8Uk 450v cap across that, and was driven with half an ECC83 (I haven`t decided what if anything I`m going to do with the other half yet). the sound out if it lovely! I can`t hear any distortion at all, and nothing gets hot (other than that which should) when it`s left running for a few hours playing radio fed from the Ext Spkr socket on a Cossor 520. was I just Lucky or am I missing something vital in my "Newness" to this technology? do you really Need a special transformer? Thanks in advance

3rd Sep 2009, 9:05 pm	#2
G8HQP Dave Rest in Peace Join Date: Sep 2008 Location: Solihull, West Midlands, UK. Posts: 4,872	Re: Output transformer essential? It all depends. A mains transformer is designed to give efficient transfer of 50Hz power. An audio output transformer is designed to give a balance of efficient, low distortion, wide bandwidth audio power and, in most cases, coping with a significant DC current through the primary. These two roles lead to different design compromises. A mains transformer has to have very good insulation for safety reasons, as the mains can put almost unlimited current through a short circuit. A valve output transformer is in a slightly safer environment, but it has to cope with a wide variation in frequency - maybe 20Hz to 50kHz for hi-fi. The upper limit is well above the normal human audible range to give some scope for stable negative feedback. In a radio set it might only have to cover 100Hz to 5kHz, as often no feedback is used. The DC in the primary of an output transformer can cause reduction in inductance (which affects the low frequency end) or core saturation which stops it working altogether. This problem is avoided by adding a small air gap to the core, but this then means many more turns to get the required inductance. So if you are not too bothered about frequency response or distortion then a mains transformer can be used as a low power output transformer. If your cathode decoupling capacitor is only 8uF then you won't get much bass anyway - you need more like 25uF for a radio or 500uF for hi-fi.

3rd Sep 2009, 9:22 pm	#3
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Output transformer essential? funny that you should mention Bass, I did indeed notice that without the cap the bass simply Wasn`t. and a direct short of the cathode to 0V, gave the most bass, but the Really perculiar part was that adding a large cap across the HT supply boosted the whole thing entirely as well as general overall clarity. I assumed (perhaps wrongly it would seem) that HT smoothing caps were only to buffer the ripple and reduce hum, there was no hum so I didn`t have the second one there. I didn`t follow any known circuit diagram, and the whole thing was just knocked up out of bits-at-hand and designed with pure Logic wrought from trial & error. it has presented me with a whole plethora of questions however, but I think I`ll put off asking them here for a while until I get a better feel for the place (there really is plenty of questions), as I don`t think it would be fair to everyone else. I must admit though, the sound from this amp is astonishingly clear to my ears anyway, and my wife agrees, I have bought a Proper output transformer a couple of days ago and look forwards to trying it out in place of these Mains one, just to see if there really IS a difference, I`m hoping so! oh, 8uF is the largest small cap I have at that sort of voltage, all the others start at 220uF and upwards, or are 0.1uF downwards at several KV rating. I hope to rectify this in future as this is a totaly New area of technology for me and I`m ill equiped for parts suitable.

4th Sep 2009, 11:17 am	#6
jjl Octode Join Date: Jan 2003 Location: Ware, Herts. UK. Posts: 1,082	Re: Output transformer essential? The purpose of the cathode bypass cap is to provide a steady DC bias voltage at the cathode of the output valve. Without it, the signal current flowing through the cathode resistor causes the voltage at the cathode and hence the grid bias to vary in sympathy with the signal, this variation acts as negative feedback and reduces gain. You're correct in stating that the capacitor and cathode bias resistor act as a high pass filter, so for best bass response you need to use a large value capacitor. The capacitor should be connected from cathode to 0V and must be in parallel with the resistor otherwise no DC current will flow in the cathode. Connecting the other end to HT would provide no benefit, would require a much higher working voltage part and and would possibly cause the output valve to oscillate. If you want to learn a good deal on valve amplifier output stages, Valve Amplifiers by Morgan Jones is an excellent read. John

4th Sep 2009, 11:31 am	#7
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Output transformer essential? Thanks both! I have noticed very subtle changes when I alter certain values of resistor and caps in my amp, and it`s great to learn the reason behind these changes on a more technical level from those that Know. my Logic for using a simple mains transformer was after having seen the small audio transformers that you get on these Science fair 100 kazillion in one electronics sets from the 70`s 80`s this mains transformer looks just like a scaled up version of one of these and I knew that the 180 volts at a few ma in my amp was never going to blow the mains side or be dangerous/fire hazard as it`s designed for 250VAC and the secondary was capable of driving a 10W speaker without damage, That and the fact I didn`t have a propper speaker output transformer handy slightly off topic, the other half of the ECC83, would it be worth doubling up and wiring it in paralel with the other half? or should I just leave it with the heater off and have a switch as a backup in case the other halfs fillament failed? thanks again each!

4th Sep 2009, 11:53 am	#9
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Output transformer essential? I know where next months "Science Budget" is going, Thanks for that

4th Sep 2009, 1:47 pm	#11
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Output transformer essential? You are a Darling! it just so happens that I have 9V transformer here too it did seem a bit Over-engineered and a waste to be using that 2A transformer just to drive a speaker when it could be used to power Filaments I`m going to write down what you posted and see if I can work out a YT2095 friendly formula from it for future ref, Cheers!

4th Sep 2009, 1:53 pm	#12
Station X Moderator Join Date: Jan 2003 Location: Ipswich, Suffolk, IP4, UK. Posts: 21,288	Re: Output transformer essential? Look up the optimum anode load for the valve. Divide it by the speaker resistance to give the impedance ratio. Take the square root of that to give the turns ratio of the output transformer. __________________ Graham. Forum Moderator Reach for your meter before you reach for your soldering iron.

4th Sep 2009, 2:04 pm	#13
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Output transformer essential? I only have access to the Max ratings for the EL34 as: Va 800v Maa 91. I`m running at 180vDC (that drops to 120vDC when heated), and the supply is 380vDC through a 2k7 resistor. ideally I need a curve graph of optimum voltage over current to plot my specs, any resources other than the National Valve Museum, would be welcome at this point

4th Sep 2009, 2:26 pm	#14
G8HQP Dave Rest in Peace Join Date: Sep 2008 Location: Solihull, West Midlands, UK. Posts: 4,872	Re: Output transformer essential? Google "EL34 valve data"

4th Sep 2009, 2:46 pm	#15
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Output transformer essential? fair enough, and I have done this last week in fact, however there is much there that I don`t understand, for instance I know that mine Single ended, but what Class is it? in fact what does class even mean? for instance: http://tdsl.duncanamps.com/show.php?des=el34 in the application data, there`s no mention of 180V so I guess I`m already outside of parameters, BUT that`s not always a bad thing is it? as long as it`s not dangerous, it could just mean that I`m Under driving it and outside of their Specific test parameters for that trial. I think I`ll just finish off the few experiments I have left to do with it, namely that new OP transformer when it arrives, and the cathode decoupling cap (raising it`s value), and then just tidy up the temp wiring and be done with, it`s already far exceeded my expectations anyway. it also works amazingly well with my guitar and a quick crystal radio that I knocked up this morning. I may post a picture of it when I can figure how to get my cam down to 200k, I think a trip to photoshop may be in order sigh again, thanks each for your help and info

4th Sep 2009, 2:55 pm	#16
ppppenguin Retired Dormant Member Join Date: Dec 2003 Location: North London, UK. Posts: 6,168	Re: Output transformer essential? Reasonable explanation of amplifier classes: http://en.wikipedia.org/wiki/Electronic_amplifier All single ended audio amps must be class A. Push-pull amps can be class A, AB or B. C is for RF power amps only while anything other class involves switchmode technology. All true for valve or transistor. If you're using XP then the windows powertoys image resizer is the simplest thing to use. http://www.microsoft.com/windowsxp/D...powertoys.mspx Otherwise Irfanview is very simple. Photoshop is a sledgehammer to crack a nut for this.

13th Oct 2009, 12:13 pm	#17
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Output transformer essential? I`v been continuing with this experiment, and have some interesting news to share, I managed to get hold of a Stentorian speaker transformer, it has a brass arm whereby you can select which winding sounds best for your speaker, and the Primary is center tapped also. so I`v been experimenting with this and different configurations and by far the Best ever (after studying Ultra Linear designs) was to have the HT on one end of the primary, the EL34s Anode on the Other end, and the center tap going to Grid 2. so I guess it`s a sort of single ended ultra linear type, and the sound is fantastic and Very hard to get any sort of distortion out of it (beyond abusing it!). I did finally parallel the ECC83s triodes, but I think I`ll undo them and have a go at something called Cascode next and see what that sounds like. but the Final stage I`m very happy with and recommend this idea to anyone that needs to make a simple amp out of scrap bits and pieces that sounds really nice!

13th Oct 2009, 7:31 pm	#18
murphyv310 Dekatron Join Date: Sep 2006 Location: Kilmarnock, Ayrshire, UK. Posts: 5,422	Re: Output transformer essential? Hi. There's also another way to leave out the speaker tranny, A higher impedance speaker could be hung on the cathode of a cathode follower, two triodes paralleled together will give a lower cathode impedance and greater power handling. A couple of low impedance triodes may even drive a 15 ohm speaker fairly well. cheers Trevor __________________ Cheers, Trevor. MM0KJJ. RSGB, GQRP, WACRAL, K&LARC. Member

14th Oct 2009, 12:58 pm	#19
Herald1360 Dekatron Join Date: Feb 2007 Location: Leominster, Herefordshire, UK. Posts: 16,535	Re: Output transformer essential? May be a problem with the cathode current offsetting the speaker voice coil unless power out from the amp is << speaker rating. RC coupling would be inefficient and LC no cheaper than using a transformer would have been Chris

14th Oct 2009, 4:40 pm	#20
YT2095UK Retired Dormant Member Join Date: Sep 2009 Posts: 487	Re: Output transformer essential? other than converting the PSU to a Valve rectifier (it`s the 1n4007`s currently) and tidying up the wiring a little, my Amp is complete now, but I`m also interested in putting into practice some of the things I have been reading about and also experimenting with ideas that I haven`t read or heard of, and so RC/LC coupling and pre-amp configurations will all be tried out on this. Some will make the amp Better than it is and some worse than it is currently, but I`ll always know how to put it all back together again in a working configuration should anything unfortunate happen. the Cost isn`t anything I`m concerned with as I have all the parts I need anyway, and I can`t think of a better way to use them than to Learn from them, that in itself justifies any costs in my opinion the Cathode coupling seems Very intriguing, a gain of only Unity at best but Higher current, sounds almost ideal for direct driving a speaker, i`v read that LEDs can be used to bias a cathode (I still don`t quite beleive that yet though), so maybe Via a 1000uF electrolytic to the speaker may work also?