Your first statement is the more accurate when you take account of the way we load transformers.
Your second looks more like an "idealised" situation trying to prove a point and make me look stupid,
A centre tapped transformer can only load both sides equally when there is no current in the centre tap. The current flowing out of one side flows back into the other side and then swaps around on the next halfwave.
Almost every loading situaion will put some current through the centre tap on each halfwave and when that centre tap current flows the winding loads are unbalanced. Series connection transformer do not work properly in that situation.
A centre tapped feeding a rectifier and smoothing capacitor/s will always have alternate sides charging the the capacitors. That requires centre tap current during each half wave charging pulse. That is exactly the unbalance that leads to dropping output voltage of the more loaded side. But the capacitor is trying to charge, it tries to pull more current in an attempt to charge up. The primary impedances work against this.
Series connected transformers do not work properly. The previous discussion will have aired some of these points. I think I was a minor player/poster in that old Thread. But it may be worth looking back to see the consensus at that time.
Last edited:
Andrew, would you be able to search out that thread? It may be a bit difficult to find by others. Ta.
Thanks for the explanation on what happens when the secondaries are feeding a bridge rectifier rather than simple resistors. That was one aspect I was still not 100% clear on.
The mains voltage is 240V rms nominal.
The transformer (as far as I am aware) is only designed to operate on that voltage. As I mentioned previously it has no provision to operate on 120V by parallel connection of the primaries.
I made an error in my earlier LTSpice simulation - I used the mains rms value when I should have used the peak (339V). It doesn't affect the results significantly as far as the imbalance issue is concerned - it's just a scaling factor. The inductance values I used were pretty much a guess as I have no accurate means of making the measurements. I simply tweaked the primary and secondary inductance ratios until the secondary voltages from the simulation matched what I measured in practice.
BTW I think the earlier suggestion about the magnetic flux cancelling in the centre arm is correct. I found some information on a similar looking 3 phase shell transformer with three separate primary and secondary coils that confirms this. The three phase transformer uses primaries connected in parallel.
The mains voltage is 240V rms nominal.
The transformer (as far as I am aware) is only designed to operate on that voltage. As I mentioned previously it has no provision to operate on 120V by parallel connection of the primaries.
I made an error in my earlier LTSpice simulation - I used the mains rms value when I should have used the peak (339V). It doesn't affect the results significantly as far as the imbalance issue is concerned - it's just a scaling factor. The inductance values I used were pretty much a guess as I have no accurate means of making the measurements. I simply tweaked the primary and secondary inductance ratios until the secondary voltages from the simulation matched what I measured in practice.
BTW I think the earlier suggestion about the magnetic flux cancelling in the centre arm is correct. I found some information on a similar looking 3 phase shell transformer with three separate primary and secondary coils that confirms this. The three phase transformer uses primaries connected in parallel.
I was just wondering what would happen with this type of transformer if there were a fault condition in which the current in one rail became very much higher than the other. For example something like 1 Amp difference so that it doesn't blow the secondary fuse.
Presumably the voltage across the associated primary would drop while the other would increase considerably. If this increase were sufficiently large it would saturate that section of the core and the transformer would get hot.
Would the transformer be protected by the mains fuse in this situation (There is a 1 Amp mains fuse in my amp).
Presumably the voltage across the associated primary would drop while the other would increase considerably. If this increase were sufficiently large it would saturate that section of the core and the transformer would get hot.
Would the transformer be protected by the mains fuse in this situation (There is a 1 Amp mains fuse in my amp).
Are you in the UK?
I said in an earlier post
I was thinking you had 110/120Vac transformers that needed to be series connected to run on a 220/240Vac supply.
If you have 240Vac transformers you can safely connect each one to your 240Vac Mains supply. That allows you to connect the two transformer primaries in parallel. Will the secondary voltage be right when you do this?
The two transformers each have a separate magnetic circuit.
These two circuits are separated by a small air gap.
Any mutual coupling will be tiny.
Yes - its UK mains.
No - It's a single 'shell' type transformer with two separate sections as shown in my original photo. Each section has a primary coil overwound with a secondary. The primaries in each section are connected in series for 240V operation. Parallel connection of the primaries is not an option.
There in not enough room in the existing case to replace it with two separate 240V transformers with primaries in parallel or a conventional centre tapped torroidal.
It may not be the best arrangement but it's been working fine for over 30 years so I see no reason to change it. I just wanted to understand why I was seeing the supply rail imbalance and maybe understand the fault that caused the transformer and capacitor to get hot.
I want to be sure that the transformer will not present a fire hazard if I get a fault where there is a gross current mismatch in the two secondaries. I know some transformers have a built in thermal protection fuse. I have no idea if this one has.
I had another look at your pic. I saw a "join" in the iron and misread that as two separate transformers.
I see now that it is just the link between the E & I of the halves of the dual core transformer.
They are coupled. My mistake.
It has been designed to operate with the primaries in series. Quite different from two separate transformers.
Fuse the primary feed with a close rated T fuse.
What is the VA?
Close rated fuse = VA/Vac.
eg for a 100VA running on 220/240Vac the fuse will be 100/230 = 0.434Aac. Use a T500mA fuse.
That prevents a house fire.
Many transformers will not start on a close rated fuse.
A normal fuse for motors and transformers is 3*VA/Vac = 3*100/230 = 1.3Aac, use a T1.25A or T1.6A without a soft start.
In a fault situation, this higher rated fuse can take ten times longer to rupture. But would even that lead to a fire?
I see now that it is just the link between the E & I of the halves of the dual core transformer.
They are coupled. My mistake.
It has been designed to operate with the primaries in series. Quite different from two separate transformers.
Fuse the primary feed with a close rated T fuse.
What is the VA?
Close rated fuse = VA/Vac.
eg for a 100VA running on 220/240Vac the fuse will be 100/230 = 0.434Aac. Use a T500mA fuse.
That prevents a house fire.
Many transformers will not start on a close rated fuse.
A normal fuse for motors and transformers is 3*VA/Vac = 3*100/230 = 1.3Aac, use a T1.25A or T1.6A without a soft start.
In a fault situation, this higher rated fuse can take ten times longer to rupture. But would even that lead to a fire?
Last edited:
That center arm makes differential impedance high, it would be better not to be there. I suggest you to eliminate its effect by making a shorted turn around it! A strip of copper may fit over the windings, and can be connected as 2 secondaries in parallel.
It is taken account: "in this connection" includes the load.
My second is the completion of your incomplete statement. If somebody "idealised" it, it was you. I specialised it to make true. Intentions are not the subject of a DIY forum. I absolutely don't care how you look like.
This topic is about the relation of the primaries and secondaries. Telling something about primaries alone basically tells nothing.
This is the rare, special case when your beloved ampere*turns plays an important role. But unfortunately saturation also...
in effect its 2 transformers sharing an outside leg. changes the aspect ratio & saves a little space, mounting hardware, and iron. the phase must correct so the flux doesn't double up. otherwise nothing special, but a revealing look into the minds eye of a 'boffin' amp. post the schematic for more weirdness.
It really shares each other's complete magnetic circuit - the common 'outer arms' are easy to appreciate, but the longer magentic path length through the other's winding core and around is also present - although like 2 parallel resistors, the resistor with the lowest value dominates the final resistance achieved. The 'outer arm' is only a minor % of the magnetic path length for each transformer.
The voltage sharing between the series connected primaries is very dependant on loading of each transformer - and although that can be rather worrying, should be fine when the loadings are sufficiently balanced, as would be expected for a typical amp set up for symmetrical clipping.
The voltage sharing between the series connected primaries is very dependant on loading of each transformer - and although that can be rather worrying, should be fine when the loadings are sufficiently balanced, as would be expected for a typical amp set up for symmetrical clipping.
- Status
- Not open for further replies.