transformer query Parallel SE use. - UK Vintage Radio Repair and Restoration Discussion Forum

bikerhifinut · 13th Nov 2014, 12:33 pm

Hi Gents,
After a bit of thought I have decided to get a pair of proper P-P output transformers for the PCL86 project. I can get a pair of Hammond ones from Bluebell audio for less than £40 a pair.

So my mind got to wandering and on checking the "Hygrade" specs, It suggests that by connecting tags 1-4 and 4-5 in parallel I can get a primary impedance of 2.2k.
I had an idea that I could use a pair of PCL/ECL86 in parallel Single Ended mode using a circuit I have to hand which would give 2W per Valve in triode mode and around 4W per valve pentode connected. I imagine i would be better off using the pentode mode as this would give a better match on the anode resistance.
The transformer is rated at 60mA, and the circuit i would use would pull in the order of 25 to 27 mA per valve from the supply.

Can anyone think of a reason why I shouldn't give this a go?

It seems a good use of the *CL86 valves and I could get a slightly better power output over using a single EL84.

These transformers are real "Old School" jobs and I feel I should use them in an application that gets the best out of them.

Andy.

TrevorG3VLF · 13th Nov 2014, 1:42 pm

You can only parallel windings if the turns ratio is exactly 1:1. This may not be the case and power transformers will often specify if paralleling is permissible.

Paralleling two identical windings will not alter the impedance. It will drop the resistance to half so allow higher currents.

bikerhifinut · 13th Nov 2014, 3:35 pm

I get you Trevor.
I'm going by the RS Hygrade information sheet. This indicates that I can get a primary impedance of 2.2k using windings 1-4 nd 4-5, this suggests to me connecting in parallel as the impedance from 1 to 5 is 8k, and 1-4-5 is 8k CT. So each section must be 4k and to get the 2.2k I would need to parallel the windings? I am not sure about this as connecting a CT mains transformer in this manner would end up with a short circuit across the secondary to the CT as I understand it as the windings are going in the same direction. I think that's how it works. I know from experience when I was messing about a good few years ago that connecting the 2 ends of a CT secondary together can end in tears.
The data sheet is very ambiguous on this and merely states connections "1-4 and 1-5 for optimum load 2.2k" I don't get it why it should be 2.2k for one section if its half of an 8k winding?
I have an inkling it should be either 1-4 OR 4-5 and not connecting 1 and 5 together.

Notwithstanding that, am I right in thinking a 2.2k load for 2 valves in parallel that would normally like to see 4k on the anode is about right? Seems like too good a chance to miss.

Andy.

Radio Wrangler · 13th Nov 2014, 3:36 pm

You're not trying to match the anode resistance of the valves. You are trying to create a load-line on thier characteristic curves which keeps the valves within their voltage and current ratings while extracting your chosen amount of power.

The impedance presented to the world by the anode is the slope of the lines on the anode characteristic plot, and it changes somewhat as the signal changes the grid voltage.

The impedance presented by the transformer to the anodes is called the 'load line' and it cuts diagonally across the family of anode curves, with the opposite slope (falling to the right in the usual presentation)

David

bikerhifinut · 13th Nov 2014, 4:05 pm

So basically it's not that simple.
I knew it was probably too good to be true. Seemed a good idea to run a pair of parallel PCL86's drawing 25mA each in SE on a primary rated at 60mA. I assumed also that i could run the triode sections in parallel, with the usual alterations of cathode and anode resistors.
Running into problems anyway as there seems to be not much room in the chassis due to the larger size of the transformers but I should be able to work my way round it. I had hoped to use a choke on the HT line.
Onwards and upwards.

Andy.

13th Nov 2014, 4:09 pm

Quote:

Seemed a good idea to run a pair of parallel PCL86's drawing 25mA each in SE on a primary rated at 60mA

A) the transformer is rated for 60mA balanced, not SE mode, it will just saturate.

B) why halve the output power?

julie_m · 13th Nov 2014, 4:15 pm

If you wanted to connect the primary segments in parallel, you would have to do a bit more work.

Pin 4 is obviously the centre tap -- the finish of the winding that starts on pin 1, and the start of the winding that finishes on pin 5. If you joined pins 1 and 5, you would be connecting the two primary halves "back-to-back"; and the two magnetic fields would cancel one another out, rather than adding together.

To get the phasing correct, you would need to separate the two wires connected to pin 4, and re-connect just the one which has continuity to pin 1 to some unused pin on the transformer (let's number it 6). Then connect 1 and 4 together, and connect 5 and 6 together.

And bear in mind, the First Law of Thermodynamics applies at all times: there are no free lunches to be had. Like the time a mate of mine tried bridging the outputs of a stereo amplifier, and then found he had to wire two speakers in series anyway just to present enough impedance to keep it happy .....

bikerhifinut · 13th Nov 2014, 4:22 pm

Quote:

Originally Posted by merlinmaxwell

A) the transformer is rated for 60mA balanced, not SE mode, it will just saturate.
B) why halve the output power?

A) It's an SE transformer, I don't understand that comment.

B) I thought that by paralleling the valves I could get more or less double the power output, all other things being equal and assuming twice the total current drawn from the PSU. Obviously I haven't thought this one through correctly.

Regards,

Andy.

bikerhifinut · 13th Nov 2014, 4:31 pm

Quote:

Originally Posted by ajs_derby

If you wanted to connect the primary segments in parallel, you would have to do a bit more work.

Thanks, that's a good explanation of things and bears out my concerns.
I won't mess about with the windings, I attached a copy of the data sheet and I think it does mean either 1-4 or 4-5 but NOT both.

So all things being equal I think? I can do a paralleled valve version as long as I don't pull more current through the transformer than it's designed for.

I realise one gets a bigger bang for ones buck with a p-p design but as I have these SE transformers I just wanted to maximise their capabilities.

Ah well it looks like I shall be just building a better version of the current SE EL84 amp I have in use. At least I know that works and gives me a nice 3 watts.

A.

13th Nov 2014, 4:47 pm

Quote:

It's an SE transformer, I don't understand that comment.

Because I saw this, must have had a brain glitch.

Quote:

After a bit of thought I have decided to get a pair of proper P-P output transformers

PJL · 13th Nov 2014, 6:33 pm

Why not go for push pull on 1-4-5? The 8K a-a impedance is about right for the ECL86.

kalee20 · 13th Nov 2014, 6:34 pm

Quote:

Originally Posted by bikerhifinut

I'm going by the RS Hygrade information sheet. This indicates that I can get a primary impedance of 2.2k using windings 1-4 and 4-5, this suggests to me connecting in parallel as the impedance from 1 to 5 is 8k, and 1-4-5 is 8k CT. So each section must be 4k and to get the 2.2k I would need to parallel the windings?

The data sheet isn't particularly clear, to be honest, about tags 1 - 4 and 4 - 5.

However, if tags 1 - 4 - 5 is 8kΩ CT, then 1 - 5 is the 8kΩ with tag 4 being half the turns. Now in a transformer, half the turns results in a quarter of the impedance (and NOT half!), so 1 - 4 would be 2kΩ and so would 4 - 5.

I suspect Radiospares have allowed for a bit of winding resistance which is why they specify 2.2kΩ load not 2kΩ (and it is not really much help anyway if this is the case).

It would be good practice, if using the 2kΩ connection, to be able to connect the other winding in parallel, you are dead right there, as the current would be shared between the windings. Unfortunately, if the finish of one section and the start of the next section are both internally soldered to tag 4, it's just not going to be possible.

bikerhifinut · 13th Nov 2014, 9:41 pm

Thanks Gents.

Now I get the reason why the impedance is a quarter, not half, I can see what it all means.

I think this transformer may have been originally sold as a "universal" type for use as a replacement in commercial radio and domestic Audio equipment where a small drop off in performance may not have been noticed if the gapped transformer was used in push-pull.

I shall use these as originally intended in a SE EL84 amp. I think they'll be at their best here.

Thanks again for all the positive comments.

The PCL86's won't go to waste, they'll end up in a P-P amp but with the correct transformers for the job.

Very best regards.

Andy.

13th Nov 2014, 12:33 pm	#1
bikerhifinut Octode Join Date: Nov 2013 Location: Penrith, Cumbria, UK Posts: 1,993	transformer query Parallel SE use. Hi Gents, After a bit of thought I have decided to get a pair of proper P-P output transformers for the PCL86 project. I can get a pair of Hammond ones from Bluebell audio for less than £40 a pair. So my mind got to wandering and on checking the "Hygrade" specs, It suggests that by connecting tags 1-4 and 4-5 in parallel I can get a primary impedance of 2.2k. I had an idea that I could use a pair of PCL/ECL86 in parallel Single Ended mode using a circuit I have to hand which would give 2W per Valve in triode mode and around 4W per valve pentode connected. I imagine i would be better off using the pentode mode as this would give a better match on the anode resistance. The transformer is rated at 60mA, and the circuit i would use would pull in the order of 25 to 27 mA per valve from the supply. Can anyone think of a reason why I shouldn't give this a go? It seems a good use of the *CL86 valves and I could get a slightly better power output over using a single EL84. These transformers are real "Old School" jobs and I feel I should use them in an application that gets the best out of them. Andy.

13th Nov 2014, 1:42 pm	#2
TrevorG3VLF Rest in Peace Join Date: Apr 2011 Location: Matlock, Derbyshire, UK. Posts: 1,378	Re: transformer query Parallel SE use. You can only parallel windings if the turns ratio is exactly 1:1. This may not be the case and power transformers will often specify if paralleling is permissible. Paralleling two identical windings will not alter the impedance. It will drop the resistance to half so allow higher currents.

13th Nov 2014, 3:35 pm	#3
bikerhifinut Octode Join Date: Nov 2013 Location: Penrith, Cumbria, UK Posts: 1,993	Re: transformer query Parallel SE use. I get you Trevor. I'm going by the RS Hygrade information sheet. This indicates that I can get a primary impedance of 2.2k using windings 1-4 nd 4-5, this suggests to me connecting in parallel as the impedance from 1 to 5 is 8k, and 1-4-5 is 8k CT. So each section must be 4k and to get the 2.2k I would need to parallel the windings? I am not sure about this as connecting a CT mains transformer in this manner would end up with a short circuit across the secondary to the CT as I understand it as the windings are going in the same direction. I think that's how it works. I know from experience when I was messing about a good few years ago that connecting the 2 ends of a CT secondary together can end in tears. The data sheet is very ambiguous on this and merely states connections "1-4 and 1-5 for optimum load 2.2k" I don't get it why it should be 2.2k for one section if its half of an 8k winding? I have an inkling it should be either 1-4 OR 4-5 and not connecting 1 and 5 together. Notwithstanding that, am I right in thinking a 2.2k load for 2 valves in parallel that would normally like to see 4k on the anode is about right? Seems like too good a chance to miss. Andy. Attached Thumbnails

13th Nov 2014, 3:36 pm	#4
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,902	Re: transformer query Parallel SE use. You're not trying to match the anode resistance of the valves. You are trying to create a load-line on thier characteristic curves which keeps the valves within their voltage and current ratings while extracting your chosen amount of power. The impedance presented to the world by the anode is the slope of the lines on the anode characteristic plot, and it changes somewhat as the signal changes the grid voltage. The impedance presented by the transformer to the anodes is called the 'load line' and it cuts diagonally across the family of anode curves, with the opposite slope (falling to the right in the usual presentation) David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

13th Nov 2014, 4:05 pm	#5
bikerhifinut Octode Join Date: Nov 2013 Location: Penrith, Cumbria, UK Posts: 1,993	Re: transformer query Parallel SE use. So basically it's not that simple. I knew it was probably too good to be true. Seemed a good idea to run a pair of parallel PCL86's drawing 25mA each in SE on a primary rated at 60mA. I assumed also that i could run the triode sections in parallel, with the usual alterations of cathode and anode resistors. Running into problems anyway as there seems to be not much room in the chassis due to the larger size of the transformers but I should be able to work my way round it. I had hoped to use a choke on the HT line. Onwards and upwards. Andy.

13th Nov 2014, 4:15 pm	#7
julie_m Dekatron Join Date: May 2008 Location: Derby, UK. Posts: 7,735	Re: transformer query Parallel SE use. If you wanted to connect the primary segments in parallel, you would have to do a bit more work. Pin 4 is obviously the centre tap -- the finish of the winding that starts on pin 1, and the start of the winding that finishes on pin 5. If you joined pins 1 and 5, you would be connecting the two primary halves "back-to-back"; and the two magnetic fields would cancel one another out, rather than adding together. To get the phasing correct, you would need to separate the two wires connected to pin 4, and re-connect just the one which has continuity to pin 1 to some unused pin on the transformer (let's number it 6). Then connect 1 and 4 together, and connect 5 and 6 together. And bear in mind, the First Law of Thermodynamics applies at all times: there are no free lunches to be had. Like the time a mate of mine tried bridging the outputs of a stereo amplifier, and then found he had to wire two speakers in series anyway just to present enough impedance to keep it happy ..... __________________ If I have seen further than others, it is because I was standing on a pile of failed experiments.

13th Nov 2014, 6:33 pm	#11
PJL Dekatron Join Date: Sep 2005 Location: Seaford, East Sussex, UK. Posts: 5,997	Re: transformer query Parallel SE use. Why not go for push pull on 1-4-5? The 8K a-a impedance is about right for the ECL86.

13th Nov 2014, 9:41 pm	#13
bikerhifinut Octode Join Date: Nov 2013 Location: Penrith, Cumbria, UK Posts: 1,993	Re: transformer query Parallel SE use. Thanks Gents. Now I get the reason why the impedance is a quarter, not half, I can see what it all means. I think this transformer may have been originally sold as a "universal" type for use as a replacement in commercial radio and domestic Audio equipment where a small drop off in performance may not have been noticed if the gapped transformer was used in push-pull. I shall use these as originally intended in a SE EL84 amp. I think they'll be at their best here. Thanks again for all the positive comments. The PCL86's won't go to waste, they'll end up in a P-P amp but with the correct transformers for the job. Very best regards. Andy.