Announcing a new article on Tronola.com: The amazing Supermods™ technique can double the output power of almost any push-pull, fixed-bias, hi-fi, tube amplifier! (Assuming it has room for more tubes.) "Hotrodding the Eico ST-70" produces 71-watts per channel from the original output and power transformers of this classic amp. It astonishes experts with excellent 20Hz performance.
This technique was pioneered by one of the diyAudio forum members and was mentioned in an earlier thread here. The article offers in-depth coverage of the technology, including analyses of how and why it works, along with verifications of the results.
See it at: http://www.tronola.com (The top story.)
Steve L.
Tronola.com
This technique was pioneered by one of the diyAudio forum members and was mentioned in an earlier thread here. The article offers in-depth coverage of the technology, including analyses of how and why it works, along with verifications of the results.
See it at: http://www.tronola.com (The top story.)
Steve L.
Tronola.com
You reminded me my company's lame spam filter.
One of our clients sent us a file to consult an error. The company's spam filter (at the server level) blocked off the email (due to the attachment), and no one knew that because the email didn't appear in the "personal" spam folder at all. A few day later, the client's VP escalated to our VP to complain that we didn't respond to customer's urgent matter.
Sometimes false alarm can be quite damaging
To: aardvarkash10
Sales pitch? Tronola is a non-commercial site and neither the article nor the site has anything for sale. A number of people worked hard to bring this surprising development to light, simply because they think it is interesting and important.
You and your company's spam filter were made for each other :-D
(Sorry, I couldn't resist. Gotta grow up
Steve Lafferty
Sales pitch? Tronola is a non-commercial site and neither the article nor the site has anything for sale. A number of people worked hard to bring this surprising development to light, simply because they think it is interesting and important.
You and your company's spam filter were made for each other :-D
(Sorry, I couldn't resist. Gotta grow up
Steve Lafferty
Nothing much wrong with what I saw there.
Assuming the power transformer can deliver the juice then it will work as advertised.
He's merely adding another pair of output tubes, so it is push-pull parallel.
Also changing the circuit a bit and the feedback.
That is questionable in terms of the subjective sound, but it might be ok or better than stock.
It might be a good idea sonically speaking to strap the output toobes for triode, you lose a bit of power but gain a bit in sonics... at least usually.
_-_-bear
Assuming the power transformer can deliver the juice then it will work as advertised.
He's merely adding another pair of output tubes, so it is push-pull parallel.
Also changing the circuit a bit and the feedback.
That is questionable in terms of the subjective sound, but it might be ok or better than stock.
It might be a good idea sonically speaking to strap the output toobes for triode, you lose a bit of power but gain a bit in sonics... at least usually.
_-_-bear
SO, qualifying here, "almost any" means any amp with a significantly overspecified PTX (by around 80%), and OPTs (by 100%), or the ability to beleive that running at or beyond the outer edges of specification will do little long-term harm.
And ok, if we want to get picky, in AB1 or 2, the OPT is running at maximum power only at peaks, and at those peaks the PTX is not having to deliver into half the output circuit, so my worst case is a little extreme.
But still. Some caveats need to be advised, surely?
I didn't realize that this was such a new and novel concept since I have been doing this and discussing it on this forum for years. I have built a SPP amp that uses 4 X EL84's pushed a wee bit beyond the specs to make 30 WPC using a 6600 ohm OPT. I simply wired a second board in parallel with the first to make a 60 WPC amp using the same OPT with the 8 ohm load tied to the 16 ohm tap. In either case the feedback is tied to the 8 ohm tap.
Doing this does increase the idle current drawn from the power transformer unless the bias in each output tube is reduced. This will increase the distortion at low volume levels which can be compensated for by boosting the feedback level. I chose to leave the feedback at a low level and run everything from a 400 VA Antek toroid with each output tube biased at about 10 watts idle. It just sounds better that way.
Oddly enough the power output at 30 Hz nearly doubled for a given distortion level. I have seen the same effect by using two BIG FAT output tubes per channel at 3300 ohms instead of 4 smaller ones. The available power output at low frequencies goes up on these particular 6600 ohm ultra cheap OPT's that were designed for guitar amps. The effect IS OPT speciffic and it will not occur on all OPTS.
No. Cutting the distance in half would be the "ultra power mod" that quadruples the power.
If you only want double the power move closer to roughly 3/4 the distance. (square root of 2 over 2 is close enough to 3/4)
Remember the inverse square law.
To astouffer:
To TheGimp:
To aardvarkash10:
Nevertheless, there ARE caveats at the beginning and end of the article. Guess you missed those too.
To Tubelab.com:
Steve L.
--- You are missing the point: What is new here is using the combination of four techniques to yield a result that experts said was impossible. If any of the four elements is removed, the result is far less.
To TheGimp:
--- If you had read the article, you would find out that it delivers an amazing amount of power at 20Hz. The article also explores that in the lab and explains why that is possible.
To aardvarkash10:
--- No. As pointed-out in the article, the peak-to-average power ratio in music is typically 14dB. So if you set the level for 124W clipping on peaks, the average output power is less than 5W. That is why the power transformer need not be over-specified. You also miss the point that the output transformers are, if anything, under-specified, as the original design could only deliver 10W at 20Hz (32W at 1kHz). The amazing thing is that this technique delivers 53W at 20Hz from those same output transformers.
Nevertheless, there ARE caveats at the beginning and end of the article. Guess you missed those too.
To Tubelab.com:
--- If you have previously disclosed the combination of all four elements mentioned in the article, used to get 2X the amount of power, simply and cheaply, out of an existing amplifier, that would be prior art and I would like to see a link to that. Otherwise, no, you have not been doing this for years.
Steve L.
After reading the article and digesting it I feel that it was a good modification.
That being said, it is not a tactic that would be desirable for every amp built. You may not get so lucky with every output transformer you use and the same goes for the power supply circuit.
The words "prior art" are normally used in a patent context. Let me assure you there is nothing even close to patentable here, because it fails both the "novel" and "non-obvious" requirements.
To dgta:
As far as obviousness, most inventions seem obvious, after they have been published and the courts are wary of that. The classic test of obviousness is whether the prior art expresses some "teaching, suggestion or motivation" to combine the previously known elements into the invention.
At least two of the elements of the Supermods™ technique were actively discouraged by conventional wisdom, rather than suggested by it. The notion that ordinary output transformers could perform much better at low frequency and put out much larger amounts of power, if operated at half impedance and from push-pull-parallel tubes is surprising to experts and challenges conventional wisdom.
Another element which is deprecated by conventional wisdom is the use of extraordinary amounts of negative feedback (30dB). Because the Supermods technique requires operation into the nonlinear saturation region of the output transformer, the high negative feedback is essential for high fidelity performance. For example, if the open loop distortion is 30% at 20Hz and we had a conventional 20dB of NFB, one would expect 3% distortion. That would not meet the 1% distortion criteria, at which power is commonly measured. With 30dB of NFB, it would beat 1% and be acceptable.
So we have two of the elements of the invention which certainly were not suggested by conventional wisdom. Hence, the invention is not obvious. Your comments fail on both counts.
--- Dead wrong. I have had quite a bit of experience in my engineering career with patents and patent law. I headed up the effort to produce a patent portfolio at one company. Of the many patents that I have dealt with (including my own), this invention is in the top third in terms of patentability. There are four main elements to this invention. If you think there is prior art which embodies all four elements, I challenge you to provide a link to it.
As far as obviousness, most inventions seem obvious, after they have been published and the courts are wary of that. The classic test of obviousness is whether the prior art expresses some "teaching, suggestion or motivation" to combine the previously known elements into the invention.
At least two of the elements of the Supermods™ technique were actively discouraged by conventional wisdom, rather than suggested by it. The notion that ordinary output transformers could perform much better at low frequency and put out much larger amounts of power, if operated at half impedance and from push-pull-parallel tubes is surprising to experts and challenges conventional wisdom.
Another element which is deprecated by conventional wisdom is the use of extraordinary amounts of negative feedback (30dB). Because the Supermods technique requires operation into the nonlinear saturation region of the output transformer, the high negative feedback is essential for high fidelity performance. For example, if the open loop distortion is 30% at 20Hz and we had a conventional 20dB of NFB, one would expect 3% distortion. That would not meet the 1% distortion criteria, at which power is commonly measured. With 30dB of NFB, it would beat 1% and be acceptable.
So we have two of the elements of the invention which certainly were not suggested by conventional wisdom. Hence, the invention is not obvious. Your comments fail on both counts.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.