Full power at 20 Hz is unnecessary for music if your listening habit is “undistorted”. Flat response to 20 Hz is, but the power capability of the amplifier and power handling of the speaker need not be. There is so little 20 or even 30 Hz actually present that the power and excursion demand is small. Most of the power is contained in higher frequencies. Up rating is probably a good idea. That lets you increase levels before saturation, but low end small signal is still ultimately limited by primary inductance. Can’t have EVERYTHING.
Cranking to rave levels with solid state - full power at 20 Hz is a side benefit of amps that don’t have other gross problems when being used for that sort of duty. Tight controlled bass even when driven solidly into clipping. You’re not doing THAT with a push pull EL84. Even then I doubt full power 20 Hz capability in the OPT is necessary. When I run my 200 watt monoblocks with 30 Hz Hammonds OPTs that hard I can find nothing lacking. The OPTs are up rated, and the power supply follows the same rules needed for hard solid state duty.
Cranking to rave levels with solid state - full power at 20 Hz is a side benefit of amps that don’t have other gross problems when being used for that sort of duty. Tight controlled bass even when driven solidly into clipping. You’re not doing THAT with a push pull EL84. Even then I doubt full power 20 Hz capability in the OPT is necessary. When I run my 200 watt monoblocks with 30 Hz Hammonds OPTs that hard I can find nothing lacking. The OPTs are up rated, and the power supply follows the same rules needed for hard solid state duty.
I do need to do a more in depth follow up, these transformers exceeded my expectations as did this pair of amps!
What are the dimensions of the space you have available to reproduce this frequency? That does set the LF limit required. And usable.
The LF is required at low power to aid in the design of an amplifier that is stable with NFB. Triodes do much better than pentodes in that respect.
Nothing suggested here seems to satisfy you. Perhaps you are simply looking for attention.
And a lot of wheel spinning is the result sofar. IMO🙄
The OP wants 20Hz. We may not agree with the reasons, but he wants 20. We may agree or disagree, express our opinion, but it is his thread.
https://www.diyaudio.com/community/threads/carnhill-made-sowters.367324/
Despite the buy out I believe they still keep most popular lines. Keep an eye out though, as custom stuff gets quite deare. Carnhill took over St Ives windings, old English house names holding reliability and tradition.
BB
If he wants it he’s just got to fork out $800. If it’s just a matter of not fully understanding the requirements, he may be able to get away with spending less.
Toroidy?
Frequency bandwidth (-3dB) : 8 Hz - 58 kHz
https://sklep.toroidy.pl/en_US/p/TT...sformer-8kOhm-2xEL84-Push-pull-or-similar/561
Frequency bandwidth (-3dB) : 8 Hz - 58 kHz
https://sklep.toroidy.pl/en_US/p/TT...sformer-8kOhm-2xEL84-Push-pull-or-similar/561
There is a rule in hitech sales (HP, R&S & Others), don't give the client too many alternatives,
They will give up & go elsewhere. In this case it appears the OP does not know what he wants, 🙂
I know exactly what I want. Sleep and work sometimes interfere with hobbies. 😉
I have heard from two manufacturers today with their options. That makes three manufacturers so far. Updates later or tomorrow as I get clarification on a few things, namely size and orientation. All have "affordable" options. In other words, not $800+ per pair.
Thank you! I'm glad someone said it for me.
Well, Hammond specifies it (30-30Khz) at full rated power.
That's called "full power bandwidth".
Very different from "bandwidth", which is sometimes also called frequency response.
Very different from "bandwidth", which is sometimes also called frequency response.
Last edited:
Comparing apples and oranges isn't ideal, and I know better than that fortunately. Thermionic has a 10 watt ultra-linear pair rated, "Frequency Response: 8hz-54Khz -3bB at 1W output power."
Thermionic Lab ebay listing
That's not the transformer I need, but I give Thermionic Lab credit for posting detailed specifications, and they say that they test every transformer they make, which is a service worth paying a little extra for in my case. I know it was mentioned that testing would add to the cost, and that's fine with me.
I think I have a good proposed solution from one of the manufacturers, physically bigger than the Thermionic 10 watters above, and without ultra-linear that I don't need. However, the orientation of the dimensions they sent me needs to be clarified because they are significantly larger, in part due to the specifications and in part due to the end bells, which I do require. As for toroids, no as I said they won't fit and I need the "classic" iron appearance with end bells.
There are still two links provided earlier that I haven't examined yet. More homework.
I'm surprised nobody has commented on the note that the "same" transformer with and without ultra-linear outputs requires a slightly different design for optimum performance. Two manufacturers have told me that now. Something about losses or something like that. I'm no expert obviously. I just listen to what they say. I always assumed, incorrectly perhaps, that they were identical internally. Most manufacturers seem to follow the approach of simply not using the ultra-linear outputs if they are present and not needed.
Now I need a drink! 😵
Testing frequency response at high power is a bit obsolete, due to the fact the result will largely depend on the nature of your output stage.
Reactive transformer components, such as main inductance, shunt capacitance and leakage inductance will form differrence in loadline shape and result into variaton in frequency response and distortion, but one has to mention the conditions of the stage which was used to deliver the obtained results.
On top of that, you will get different frequency response behaviors with different speakers loads. To somewhat achieve more HF response uniformity, manufacturers tend to prefer capacitance roll-off dominant transformers compared to leakage inductance dominant.
Reactive transformer components, such as main inductance, shunt capacitance and leakage inductance will form differrence in loadline shape and result into variaton in frequency response and distortion, but one has to mention the conditions of the stage which was used to deliver the obtained results.
On top of that, you will get different frequency response behaviors with different speakers loads. To somewhat achieve more HF response uniformity, manufacturers tend to prefer capacitance roll-off dominant transformers compared to leakage inductance dominant.
This is very valuable information. Thermionic Lab seems to charge a bit more than some other manufacturers, but there may be a reason for that such as their testing. They are very responsive and helpful as well.
Primary Windings looked promising until I tried the dimensions of their 25 watt model. It is a few millimeters bigger than I have available. I might be able to squeeze them in with 1mm between them and the power transformer, but I likely won't try it. There is nothing between 10W and 25W on their site.
Edcor said to use their off-the-shelf model or pay a $300 setup fee, plus the cost of the transformers, which isn't going to happen.
Lundahl LL1682.
It is sold as 5K for 5R speaker but the ratio is precisely 32:1 so it is 8K when used with 8R load. Power rating is 40W @30Hz into 5R which means 450V rms signal at the primary. So those 450V become 375V and 300V rms @25Hz and 20Hz, respectively. If the speaker is 8R the power rating is 17.5W @25Hz and 11W+ @20Hz . More than good enough for EL84 PP. On the other side it easily gets to 60-70 KHz.
It has 33% UL tap as well.
It doesn't use end bells being C-core but you can always make a cover which screws in like an OPT with end-bells.
Not really cheap but affordable with excellent performance in any aspect.
It is sold as 5K for 5R speaker but the ratio is precisely 32:1 so it is 8K when used with 8R load. Power rating is 40W @30Hz into 5R which means 450V rms signal at the primary. So those 450V become 375V and 300V rms @25Hz and 20Hz, respectively. If the speaker is 8R the power rating is 17.5W @25Hz and 11W+ @20Hz . More than good enough for EL84 PP. On the other side it easily gets to 60-70 KHz.
It has 33% UL tap as well.
It doesn't use end bells being C-core but you can always make a cover which screws in like an OPT with end-bells.
Not really cheap but affordable with excellent performance in any aspect.
Last edited: