Dang. May be good candidates for use with a full range speaker crossed to a sub at around 75hz so the LF isn't an issue?
I'm very interested in what you find with the Toroidys when the supply is not a question mark, and what your impression is of what they sound like.
Very good work , happy to read about your tests!
So what could be the right imput impédance of the OPT for this tube ?
Regards from France
Gilles
So what could be the right imput impédance of the OPT for this tube ?
Regards from France
Gilles
Nice, hope you get a chance to run the toroidy transformers with a better supply soon!
P.S. why do you always seem to be pushing tubes far beyond their limits 🙂
P.S. why do you always seem to be pushing tubes far beyond their limits 🙂
Hi George,
I’m using custom made opt’s from Toroidy. I can run a bode plotter test to see how output power differs. My test setup uses EL34’s in triode or UL-mode. In UL I get 36 Watts @ 0,3 % THD. I noticed THD to correlate with my generator’s THD.
Regards, Gerrit
I’m using custom made opt’s from Toroidy. I can run a bode plotter test to see how output power differs. My test setup uses EL34’s in triode or UL-mode. In UL I get 36 Watts @ 0,3 % THD. I noticed THD to correlate with my generator’s THD.
Regards, Gerrit
I assume that is a push pull amp. The OPT's that I have are made for an SE amp where there is unbalanced DC running through the OPT.
Even in a huge EI OPT there will be some saturation effects limiting the low frequency response. These are toroids, and they weigh 2.5 Kg each. I would not expect them to outperform some large 5 Kg EI transformers like the Hammond 1628SEA's.
The test amp that I used last night is something new, and I have seen it pull performance out of some less than stellar transformers already due primarily to its low plate resistance /impedance.
I may take a step back and hook up one of my TSE-II amps with the 300B's wired in parallel to see how these work in a well known design.
I don't know.....maybe it's the same reason the grandkids play video games....to see what happens if I try it.
I assume that you are asking what the ideal load impedance for the tube (36LW6) or other tube would be. Unfortunately there is no "correct" answer.
There are however many usable answers depending on what you want to accomplish, and what your constraints are.
The typical audio tubes like EL34, KT88, 6L6... have the plate on pin 3, and the heater on pin 2. This limits the maximum plate voltage that can be applied to the tube before it arcs over at the socket or inside the base. The plate voltage can hit 2X the B+ voltage in a normal amp which is not driven into clipping. I have seen 4X the B+ in a cranked guitar amp running into a speaker near it's resonant frequency.
I an using TV sweep tubes, that have no such limit. Kilovolt levels on the plate are normal in a TV set. The OPT however WILL arc over, or even catch fire if fed too much voltage. I was experimenting with the same tubes in the same amp with two different 5K ohm OPT's. I could get over 40 watts out of this amp with good efficiency, and distortion, but I was using almost 650 volts of B+. This is too much voltage for a standard off the shelf OPT, especially if it gets driven into clipping.
Last night I was using these 1.5K OPT's and getting over 40 watts, but I had to push the idle current somewhere beyond 300 mA (the limit of the meter on the power supply) to get there. The transformers are rated for 300mA nominal, 500mA max. B+ was 450 volts. I think this will be a decent choice for two tubes in parallel.
Given what I have seen, and some limited testing with a transformer that's too small, I would use a 2.5 to 3K ohm load and aim for 25 to 30 watts out. These are ALL preliminary SE amp testing, and may bump into the tube's plate dissipation rating. In a real amp, I would probably use 2 to 4 smaller sweep tubes in parallel, but I have over 100 25DN6's that I got for 50 cents each looking for a home.
Just for information, a decent made (Van der Veen, at 8ohm, instead of the 5ohm he cleverly uses to be able to state higher power, 8ohm = 4% Rdc loss) 2.5kg pp toroid will hit saturation (1.8T) at 11W (14Hz), so it should be rated as 22W (sold as 80W).
So I would not expect to much from a 2.5kg SE transformer.
So I would not expect to much from a 2.5kg SE transformer.
Last edited:
Hi George,
Thank you.
I am thinking about that:
C3g Triode strapped/interstage TangoNC20/36LW6/3.5k OPT Tamura F7002.
Anode voltage C3g 180v
I anode 10ma
Anode voltage 36LW6 450v
I anode 100 ma (OPT data)
Perhaps ok for around 15 watts ?
I have these transformers since 10 years, and just found some 36LW6 so perhaps an opportunity...
Gilles
Thank you.
I am thinking about that:
C3g Triode strapped/interstage TangoNC20/36LW6/3.5k OPT Tamura F7002.
Anode voltage C3g 180v
I anode 10ma
Anode voltage 36LW6 450v
I anode 100 ma (OPT data)
Perhaps ok for around 15 watts ?
I have these transformers since 10 years, and just found some 36LW6 so perhaps an opportunity...
Gilles
Last edited:
Are you sure that the transformers are rated at 500ma max? All that I've seen online says max 300ma. On the Toroidy site, the polish "Prąd maksymalny", meaning max current, got translated to nominal on the English site.
0.3A would suffice for SE 60W into 1.5k but 60W at low frequencies? No way if you want bearable distortion and copper losses.
No matter what iron, toroid or whatever, 2.5kg just wont cut it
No matter what iron, toroid or whatever, 2.5kg just wont cut it
Last edited:
A very good 10W transformer (but lousy at 15W), why would you not match its performance with
a 300B?
Last edited:
At lunchtime today Sherri decided to go play cards with the neighbors, which left me alone to play with my test amp. This means, abandon all orderly testing, connect an OPT to, and stick a tube in that second channel, verify that it works.
I connected the board up to my big OLD 60Hz switching HP power supply that goes to 650 volts at 1.7 Amps. Yes, it has current limiting, but the 1000uF of output capacitance, and the very slow control loop will guarantee that the limiter kicks in after there is a smoking hole in the circuit board where your parts once lived when things go wrong.
I set the supply voltage to 450 volts and set the idle current to 180 mA per channel. Total supply current was about 400 mA.
A quick test revealed that one channel would hit about 20 watts at 50 Hz before saturation was seen, but the other did a bit less. I saw THD in the 2% range at 30 watts on both channels, so shut it down and connected up speakers and a CD player.
I powered up, popped in a CD and hit play. I noticed that the current meter started dancing a lot at half volume and was hitting 1 amp on some VERY LOUD music peaks.....This is a class A amp, and should draw a nearly constant current unless it leaves class A. After about 2 minutes of head scratching, I deduced that I had connected my 4 ohm "indestructible" test speakers up to the 8 ohm taps. I moved them to the 4 ohm taps and put one scope probe across each speaker.
I cranked up some music and rattled the speakers for about an hour and decided to get my camera for some pictures. The first picture shows the setup with the lights on. The second picture shows why the two channels acted a bit different from each other.
Either I messed up in setting the bias, or the tube that I put in the second channel that had been sleeping for 50 years or so took some time to get up to speed. Either, or both are possible. I decided not to take time out to hook stuff back up to readjust the board, I just dialed back the power supply until the redness went away, which was around 400 volts.
I spent several more hours listening to this amp, and I really like how it sounds. I did turn the power supply up...sometimes to 500+ volts, when I played loud dynamic music, just because it SOUNDED better. The scope revealed that some transients, usually drums or bass guitar got clipped occasionally, or just missed clipping, but used up all the amp's dynamic range so that concurrent vocals or guitar got squashed. One channel was worse than the other in this regard, and the hotter tube correlated with more headroom.
The next picture shows a large bass guitar note at around 100Hz making it through the hot channel (top) at nearly 45 volts P-P across my 4 ohm speaker. The bottom trace carries extra baggage, a screaming lead guitar on top of the bass which is at a lower overall level (38 volts p-p), but its top gets abruptly chopped off.
The 4th picture shows a similar event. Twice a 40+ volt near sinewave from a bass guitar makes it through the top channel nearly unscathed, but gets clipped or squashed in the bottom trace.
There are two things going on here. Yes, the channel connected to the top trace is running hotter, presumably because of more current. However I have several pictures where the bottom trace gets through unscathed, while the top trace gets clipped or squashed, although it does appear to have a bit more headroom.
Did anyone figure out that 44 volts peak to peak into a 4 ohm speaker is over 60 watts? Is my amp really cranking out 60 watts? NO, it's not. What this means is that my speakers are NOT 4 ohms at 100Hz. 100Hz is near resonance for the pair of 6 inch drivers in the cabinet, which is tuned for 70 Hz, so there are two impedance peaks in the overall speaker impedance, and that's only considering a single swept tone at a constant level.
What happens to that speakers DYNAMIC impedance when it is in motion at say 100Hz and a drum hit or other transient tries to reverse the cone's motion. It drops. Conversely, if that transient is in phase with the existing come motion, the speaker's dynamic impedance will increase.
Dynamic speaker impedance is a subject that's rarely discussed. I tried to explore it in depth many years ago, and the further down that rabbit hole I went, the more strangeness I found. I decided to climb back out, and not go there again...But I was convinced that a large woofer with a heavy cone and a big voice coil can generate enough counter EMF to make it's impedance appear negative during that brief instant of cone reversal. I also learned NOT to stick a Shure SM-57 INSIDE the bass drum, but that's a different story.
SO....This means that If I build it, it will need TWO output tubes to avoid melting one. These OPTs can obviously SOUND good, and they can sound better when the tubes are running well beyond their dissipation capability.
I fully understand this. My test results would have been different if I used a speaker with a 15 inch woofer that doesn't exhibit resonance around 90 Hz.
If I decide to build this amp as is, it would likely be used to drive my Yamaha NS-10M Studio speakers, which have zero response and a high impedance below 70 Hz, so I can concentrate my efforts above that range.
So far I like these transformers for their USD $88 price. The
The Edcor CXSE25-1.25K for USD $94 which weigh in at 4.4 kg would probably do better in the bass range. I have used their 5K OPT in a 300B amp with the same test speakers, and the highs, realism, and definition didn't seem as good as the Toroidy.
The Hammond 1640SEA would do far better in the bass department at 5 kg, but it costs about $150, and may have a HF roll off problem. Again I tested a 5K Hammond 1628SEA with 300B's using the same speakers, and preferred the Edcors.
I connected the board up to my big OLD 60Hz switching HP power supply that goes to 650 volts at 1.7 Amps. Yes, it has current limiting, but the 1000uF of output capacitance, and the very slow control loop will guarantee that the limiter kicks in after there is a smoking hole in the circuit board where your parts once lived when things go wrong.
I set the supply voltage to 450 volts and set the idle current to 180 mA per channel. Total supply current was about 400 mA.
A quick test revealed that one channel would hit about 20 watts at 50 Hz before saturation was seen, but the other did a bit less. I saw THD in the 2% range at 30 watts on both channels, so shut it down and connected up speakers and a CD player.
I powered up, popped in a CD and hit play. I noticed that the current meter started dancing a lot at half volume and was hitting 1 amp on some VERY LOUD music peaks.....This is a class A amp, and should draw a nearly constant current unless it leaves class A. After about 2 minutes of head scratching, I deduced that I had connected my 4 ohm "indestructible" test speakers up to the 8 ohm taps. I moved them to the 4 ohm taps and put one scope probe across each speaker.
I cranked up some music and rattled the speakers for about an hour and decided to get my camera for some pictures. The first picture shows the setup with the lights on. The second picture shows why the two channels acted a bit different from each other.
Either I messed up in setting the bias, or the tube that I put in the second channel that had been sleeping for 50 years or so took some time to get up to speed. Either, or both are possible. I decided not to take time out to hook stuff back up to readjust the board, I just dialed back the power supply until the redness went away, which was around 400 volts.
I spent several more hours listening to this amp, and I really like how it sounds. I did turn the power supply up...sometimes to 500+ volts, when I played loud dynamic music, just because it SOUNDED better. The scope revealed that some transients, usually drums or bass guitar got clipped occasionally, or just missed clipping, but used up all the amp's dynamic range so that concurrent vocals or guitar got squashed. One channel was worse than the other in this regard, and the hotter tube correlated with more headroom.
The next picture shows a large bass guitar note at around 100Hz making it through the hot channel (top) at nearly 45 volts P-P across my 4 ohm speaker. The bottom trace carries extra baggage, a screaming lead guitar on top of the bass which is at a lower overall level (38 volts p-p), but its top gets abruptly chopped off.
The 4th picture shows a similar event. Twice a 40+ volt near sinewave from a bass guitar makes it through the top channel nearly unscathed, but gets clipped or squashed in the bottom trace.
There are two things going on here. Yes, the channel connected to the top trace is running hotter, presumably because of more current. However I have several pictures where the bottom trace gets through unscathed, while the top trace gets clipped or squashed, although it does appear to have a bit more headroom.
Did anyone figure out that 44 volts peak to peak into a 4 ohm speaker is over 60 watts? Is my amp really cranking out 60 watts? NO, it's not. What this means is that my speakers are NOT 4 ohms at 100Hz. 100Hz is near resonance for the pair of 6 inch drivers in the cabinet, which is tuned for 70 Hz, so there are two impedance peaks in the overall speaker impedance, and that's only considering a single swept tone at a constant level.
What happens to that speakers DYNAMIC impedance when it is in motion at say 100Hz and a drum hit or other transient tries to reverse the cone's motion. It drops. Conversely, if that transient is in phase with the existing come motion, the speaker's dynamic impedance will increase.
Dynamic speaker impedance is a subject that's rarely discussed. I tried to explore it in depth many years ago, and the further down that rabbit hole I went, the more strangeness I found. I decided to climb back out, and not go there again...But I was convinced that a large woofer with a heavy cone and a big voice coil can generate enough counter EMF to make it's impedance appear negative during that brief instant of cone reversal. I also learned NOT to stick a Shure SM-57 INSIDE the bass drum, but that's a different story.
SO....This means that If I build it, it will need TWO output tubes to avoid melting one. These OPTs can obviously SOUND good, and they can sound better when the tubes are running well beyond their dissipation capability.
I fully understand this. My test results would have been different if I used a speaker with a 15 inch woofer that doesn't exhibit resonance around 90 Hz.
If I decide to build this amp as is, it would likely be used to drive my Yamaha NS-10M Studio speakers, which have zero response and a high impedance below 70 Hz, so I can concentrate my efforts above that range.
So far I like these transformers for their USD $88 price. The
The Edcor CXSE25-1.25K for USD $94 which weigh in at 4.4 kg would probably do better in the bass range. I have used their 5K OPT in a 300B amp with the same test speakers, and the highs, realism, and definition didn't seem as good as the Toroidy.
The Hammond 1640SEA would do far better in the bass department at 5 kg, but it costs about $150, and may have a HF roll off problem. Again I tested a 5K Hammond 1628SEA with 300B's using the same speakers, and preferred the Edcors.
Local FB only around the output tube. It's a rather complex circuit that results in triode like curves with a low Rp.
There is no connection from the OPT secondary back to the amp at all except for a safety ground.
There is no connection from the OPT secondary back to the amp at all except for a safety ground.
You are right.
I am looking for more power than a 300b
But it seems that Toroidy could make a good match at a good price😉
They could built for 3k.
Gilles
Last edited:
Gilles,
Toroidy has an OPT with 3K secondaries:
TTG-KT88SE TOROIDY - Transformer: speaker | 40VA; O115x65mm; 0.01/56kHz; 250mA | TME - Electronic components
Toroidy has an OPT with 3K secondaries:
TTG-KT88SE TOROIDY - Transformer: speaker | 40VA; O115x65mm; 0.01/56kHz; 250mA | TME - Electronic components
@Tubelab_com
your verdict seems to be that the toroidy transformers are quite alright at the price? or am I misreading that
your verdict seems to be that the toroidy transformers are quite alright at the price? or am I misreading that
Here in the US the Toroidy and the Edcor would be about the same price once shipping is factored in. If I had to choose between the two based on what I have heard recently, I would pick the Toroidy in MY situation. As stated my Yamahas don't need anything below 70Hz, and their impedance in that range is over 20 ohms, so the amp is not heavily loaded.
The test speakers were designed for portability, budget, and ruggedness, I also wanted to use them with a portable music keyboard in an outdoors situation. Sound quality was not the prime criteria, however they have been a pleasant surprise in all respects. The sound quality is excellent, they have been used outdoors for a rock band PA fed by a 125 WPC tube amp pushed to the edge of clipping for hours without issue. They have also been used for loud guitar playing with the tweeters turned off. They also seem to sound quite nice with the Toroidy OPT's pushed right up to the point where the tubes start running out of headroom.
If you have big speakers with significant response below 70 Hz and a relatively low impedance below 70 Hz, then the Edcor may be a better choice.
Your choice of music, preferred listening level, and speaker efficiency must also be considered. IE if your speakers are highly efficient, so that your average listening level is well below 1 watt, then the Toroidy's could be the first choice.
Note that I have never actually tested the 1250 ohm Hammonds or Edcors. I have tested the 5K versions of both in the SAME amp with the SAME speakers, but I was running Sylvania 36LW6's instead of the GE's that I abused yesterday, and the B+ was 650 volts to accommodate the 5K load. Max power output was in the same range, 30 to 40 watts depending on B+ voltage.
I may get a pair of 1250 ohm Edcors since this amp design could be the next Tubelab PC board and I will need to test the board under all possible operating conditions. For now the budget does not support it, and Edcor's delivery times are long, so they will not be tested soon.
The big Hammonds, especially the 1628SEA's that I have exhibit a resonant notch in the high frequency range just above the audio band. If they are driven with a wimpy tube like a trioded 6L6GC, the notch can be deep and wide enough to cause a loss of high frequency response. It can be a dB or two down at 15 KHz and 3 or more dB down at 20 KHz. Adding feedback, or using a low impedance driving source can reduce the "Q" factor of the resonance, thus taming the notch. The 1628SEA measures around 3/4 dB down at 20 KHz in this amp.
Those were my first choice, but there was only one in stock at the time I ordered.
The test speakers were designed for portability, budget, and ruggedness, I also wanted to use them with a portable music keyboard in an outdoors situation. Sound quality was not the prime criteria, however they have been a pleasant surprise in all respects. The sound quality is excellent, they have been used outdoors for a rock band PA fed by a 125 WPC tube amp pushed to the edge of clipping for hours without issue. They have also been used for loud guitar playing with the tweeters turned off. They also seem to sound quite nice with the Toroidy OPT's pushed right up to the point where the tubes start running out of headroom.
If you have big speakers with significant response below 70 Hz and a relatively low impedance below 70 Hz, then the Edcor may be a better choice.
Your choice of music, preferred listening level, and speaker efficiency must also be considered. IE if your speakers are highly efficient, so that your average listening level is well below 1 watt, then the Toroidy's could be the first choice.
Note that I have never actually tested the 1250 ohm Hammonds or Edcors. I have tested the 5K versions of both in the SAME amp with the SAME speakers, but I was running Sylvania 36LW6's instead of the GE's that I abused yesterday, and the B+ was 650 volts to accommodate the 5K load. Max power output was in the same range, 30 to 40 watts depending on B+ voltage.
I may get a pair of 1250 ohm Edcors since this amp design could be the next Tubelab PC board and I will need to test the board under all possible operating conditions. For now the budget does not support it, and Edcor's delivery times are long, so they will not be tested soon.
The big Hammonds, especially the 1628SEA's that I have exhibit a resonant notch in the high frequency range just above the audio band. If they are driven with a wimpy tube like a trioded 6L6GC, the notch can be deep and wide enough to cause a loss of high frequency response. It can be a dB or two down at 15 KHz and 3 or more dB down at 20 KHz. Adding feedback, or using a low impedance driving source can reduce the "Q" factor of the resonance, thus taming the notch. The 1628SEA measures around 3/4 dB down at 20 KHz in this amp.
Those were my first choice, but there was only one in stock at the time I ordered.