Hi guys, I just finished building a tube amplifier, nothing fancy, a 3 tube design. A typical ECC83 voltage amplifier, and cathodyne followed by a push pull EL84 pair. Now, in the EL84 datasheet they give a Pmax value of 17W. I can hardly get anything above 5W, and everything measures pefectly (HV-300V, 36mA per tube, Za-a 8k). Ive been ruling out stuff and the only thing that keeps bugging me is the output transformer. I salvaged it from an old radio, where the output stage was SE, but coincidentally the output transformer had a center tap, maybe for ultra linear operation or something. Can the output transformer be responsible for the power drop ive been measuring, and if so is there a way to fix it (closing the air gap, or parallelising it with another transformer or something)
What was the output tube in the old radio? Rayma is right that even if the center tap is perfectly centered, the impedance match may not be right.
If it's actually a UL tap (usually not truly a center tap), that may be compromising the output power. Because it's gapped, you wouldn't necessarily hear the effects of the imbalanced current.
If it's actually a UL tap (usually not truly a center tap), that may be compromising the output power. Because it's gapped, you wouldn't necessarily hear the effects of the imbalanced current.
> I can hardly get anything above 5W
That's just what you expect from a *single* EL84. Clue??
> the output transformer. ...from an old radio, where the output stage was SE, but coincidentally the output transformer had a center tap, maybe for ultra linear operation
While UL has been known over 60 years, and SE UL is obvious, I do not know any practical application before the tube-fad heated-up in the late 20th century.
The most likely "extra tap" on an SE radio is a trick to reduce hum/buzz. A 3% tap on the OT Pri induces an opposite hum/buzz in the feed to G2 and small tubes. Filter the audio and the large stray buzz-leakage, you can get lower buzz in the speaker without raising the budget for capacitors. See attached.
If true push-pull is 50:50, and you have loaded with a 97:3 winding, sure, you only got one tube working effectively.
Hammond 125E (not 125ESE) is a good OT, though the price may be high.
OTs for Champ and Marshall 18W guitar amps are serious parts and often sell at competitive prices.
A 115V+115V:8V 10VA Power Transformer would "work", with correct leverage. Max output will be severely distorted below ~~200Hz, and response may fall badly above 1KHz, but some testing around 400Hz would show if the amp works except the OT, or if you have other problems.
That's just what you expect from a *single* EL84. Clue??
> the output transformer. ...from an old radio, where the output stage was SE, but coincidentally the output transformer had a center tap, maybe for ultra linear operation
While UL has been known over 60 years, and SE UL is obvious, I do not know any practical application before the tube-fad heated-up in the late 20th century.
The most likely "extra tap" on an SE radio is a trick to reduce hum/buzz. A 3% tap on the OT Pri induces an opposite hum/buzz in the feed to G2 and small tubes. Filter the audio and the large stray buzz-leakage, you can get lower buzz in the speaker without raising the budget for capacitors. See attached.
If true push-pull is 50:50, and you have loaded with a 97:3 winding, sure, you only got one tube working effectively.
Hammond 125E (not 125ESE) is a good OT, though the price may be high.
OTs for Champ and Marshall 18W guitar amps are serious parts and often sell at competitive prices.
A 115V+115V:8V 10VA Power Transformer would "work", with correct leverage. Max output will be severely distorted below ~~200Hz, and response may fall badly above 1KHz, but some testing around 400Hz would show if the amp works except the OT, or if you have other problems.
Attachments
No, I havent measured the impedances, I calculated it, with the reflected impedance formula. The transformer is marked and it has 2x1650 turns for the primary and 90 turns for the secondary. That gives me exactly 8k Za-a with a 6 ohm load. Its not a choke, I think, because I do have another SE transformer with a choke and its separate from the primary. I believe the transformer was made with a center tap, either for UL, but it was easier to make it at 50%, rather than 43%, or probably if you want the tube to work in a different mode, to make it easier for you or something. Anyway, the transformers are certainly for SE, I can see the air gap. I noticed the laminations can be rearranged, so I can close the air gap. What will I accomplish with this little exercise?
P.S. I`ve read that they are usually smaller in size than SE transformers, but I don`t have a reference on how big should a 15W push pull transformer be.
P.P.S. I dont think an inbalance causes the problem, because the transformers should be perfectly center tapped according to the markings, and the power tubes are matched and driven equaly. I cant remember from what piece of equipment i got the tranformers from, but i believe they were attached to a PCL86 SE output tube
P.S. I`ve read that they are usually smaller in size than SE transformers, but I don`t have a reference on how big should a 15W push pull transformer be.
P.P.S. I dont think an inbalance causes the problem, because the transformers should be perfectly center tapped according to the markings, and the power tubes are matched and driven equaly. I cant remember from what piece of equipment i got the tranformers from, but i believe they were attached to a PCL86 SE output tube
Last edited:
An externally hosted image should be here but it was not working when we last tested it.
Here's a pic of the circuit, and the limitation comes from the transformer, in this I'm sure. Probably the secondary being designed for a lower power has higher active resistance and the output kneels at heavier output levels.I was thinking of building my own transformer. I have a pretty decent C core I've been meaning to use for something. Anyway, I was thinking something along the line of this. How's it look?
An externally hosted image should be here but it was not working when we last tested it.
Have you terminated the secondary in a suitable resistor (like 8 ohms) and measured the inductance of the primary? If the transformer can be disassembled and the core removed without damaging anything, remove the core and remeasure the inductance again. The change of primary inductance with and without the core should give you an idea of the permeability of the iron. The primary impedance may be on the order of 20 henries or more with the core, so make sure the inductance meter works accurately at that range.
I don't think it's a saturation problem. I've never measured the output at anything less than 440Hz. I pretty much doubt the core will saturate at this frequency. I believe the secondary is made for a much lower power and the wire is not thick enough to carry sufficient current to a 17W load. It's simply kneeling under a heavy load. So I decided to make my own transformer. Judging by the transformers I have I decided to make the primary from two 1500 turn coils, tapped at 20% for maximum power distribution UL. Just need a yay or nay for the interleaving of the sections.
I think I got it. Basically I have 4 primary windings and 3 secondaries. I have to interleave them primary secondary primary secondary e.g. and keep all turns between layers the same. If I have 3000 turns for a primary I have to interleave it every 750 turns, but since my UL taps will be at 20% I'd have to tap twice for the UL and once more for the HT at the middle layer. I think I got it, thanks
For Hi-Fi I'd use a Hammond output tranny, and for guitar amp use I'd use a Fender tranny, which is about half the price of a similar Hammond. Edcore is said to be pretty good and reasonably priced too, but I have no experience with them personally. The output tranny has to be right for many reasons. Pick one that can handle the power, and has the correct primary impedance (when the secondary is properly loaded). If it will be mounted near other transformers, mount them so their cores are at 90 degree angles to each other, to minimize electromagnetic coupling.
A higher than normal primary impedance will give less distortion, but also less power out before clipping. A lower primary impedance will give more power output, but will have more real-time compression distortion, which some people might actually like (I do - especially in a guitar amp circuit). But you don't want to be very far from what is the proper primary impedance for the tubes you are using.
A higher than normal primary impedance will give less distortion, but also less power out before clipping. A lower primary impedance will give more power output, but will have more real-time compression distortion, which some people might actually like (I do - especially in a guitar amp circuit). But you don't want to be very far from what is the proper primary impedance for the tubes you are using.
I also have one unrelated question about EL84. A lot of the schematics online show the tube as having a 150R cathode resistor in a push pull circuit. With 300V power supply that gives around 16W plate dissipation, and the datasheet of EL84 shows a PaMax of 12W. What`s the deal with that. Have I missed something or can you use tubes at over 150% their maximum dissipation without problems?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.