Hi Gregg,
I asked Mike LaFevre about putting an 8 ohm speaker on the 16 ohm tap of his RH series OPT's on the Magnequest forum. You can follow the thread here:
http://www.audioasylum.com/forums/magnequest/messages/129.html
If you apply this to the XSE25-16-7.6K, you should get a 3.8K primary, but with the larger core size of the 25 watt series and the additional inductance due to the coil being set up for 7.6K. There will be some losses, but you might get better bottom end response. Whether this would sound better than the XSE15-8-5K, who knows? I would have thought that the 125ESE would be better than the 125CSE, but the consensus seems to be with the "C".
Kyle
I asked Mike LaFevre about putting an 8 ohm speaker on the 16 ohm tap of his RH series OPT's on the Magnequest forum. You can follow the thread here:
http://www.audioasylum.com/forums/magnequest/messages/129.html
If you apply this to the XSE25-16-7.6K, you should get a 3.8K primary, but with the larger core size of the 25 watt series and the additional inductance due to the coil being set up for 7.6K. There will be some losses, but you might get better bottom end response. Whether this would sound better than the XSE15-8-5K, who knows? I would have thought that the 125ESE would be better than the 125CSE, but the consensus seems to be with the "C".
Kyle
If you think about how the Hammond universal (125 series) works, there is one single primary and a secondary with 4 taps. The "ratio" depends upon which secondary tap is used. As explained by Mike LaFevre the primary is optimized for one impedance and all other impedances are a compromize.
That being said, I have done this many times and the results concour with what Mike said. Running a transformer with an 8 ohm load on the 16 ohm tap usually costs some high end frequency response, and some times improves the bass and power handling. The input impedance is 1/2 the usual value. Operating a transformer with an 8 ohm load on the 4 ohm tap usually costs quite a bit of low end frequency response, and power handling. The input impedance is double the usual value.
The results depend on the quality of the transformers. You can get away with more "ratioing" if you start with good transformers. I have a large quantity of cheap surplus 6.6K ohm 80 watt P-P transformers that were built for guitar amps. When running 2 KT-88's, with the 8 ohm load on the 8 ohm tap, I get 50 Hz to 18KHz response at 50 watts. If I run 4 KT-88's with the 8 ohm load on the 16 ohm tap, I get 45 Hz to 14 KHz response at 50 watts. The amp will make 100 wats but the transformer saturates at high power and low frequency. Running 2 6LW6's with the 8 ohm load on the 16 ohm tap, I get 50 Hz to 15 KHz response at 50 watts. The amp will make 140 watts. The Edcors have plenty of HF response, so it should be OK.
We can conjecture about what would happen, or I can simply put the 300B's back in the amp, hook up a 4 ohm load and test it. As soon as this CD is finished, I will do that.
That being said, I have done this many times and the results concour with what Mike said. Running a transformer with an 8 ohm load on the 16 ohm tap usually costs some high end frequency response, and some times improves the bass and power handling. The input impedance is 1/2 the usual value. Operating a transformer with an 8 ohm load on the 4 ohm tap usually costs quite a bit of low end frequency response, and power handling. The input impedance is double the usual value.
The results depend on the quality of the transformers. You can get away with more "ratioing" if you start with good transformers. I have a large quantity of cheap surplus 6.6K ohm 80 watt P-P transformers that were built for guitar amps. When running 2 KT-88's, with the 8 ohm load on the 8 ohm tap, I get 50 Hz to 18KHz response at 50 watts. If I run 4 KT-88's with the 8 ohm load on the 16 ohm tap, I get 45 Hz to 14 KHz response at 50 watts. The amp will make 100 wats but the transformer saturates at high power and low frequency. Running 2 6LW6's with the 8 ohm load on the 16 ohm tap, I get 50 Hz to 15 KHz response at 50 watts. The amp will make 140 watts. The Edcors have plenty of HF response, so it should be OK.
We can conjecture about what would happen, or I can simply put the 300B's back in the amp, hook up a 4 ohm load and test it. As soon as this CD is finished, I will do that.
OK, experiment done. Before testing the 300B's with a 4 ohm load, I ran a few measurements with the 45's. As I had stated earlier these transformers need the most current of all the transformers tested to be happy. With the 300B's (8 ohm) the 3rd harmonic totally dissapears at 80 mA. When I tried the 45's the third started dropping at 40 mA and I stopped at 50 mA. At this point the distortion at 1 watt was 1.4%, but you can't run 45's at 50 mA for long. At 30 mA the distortion at 1 watt was just over 4%. Clearly the 45's measure much worse than the 300B's, but I prefer the sound of the 45's on all but the loudest music. So much for conventional measurements. Now back to the topic.
I put back the 300B's and hooked a 4 ohm load to the 8 ohm Edcors. Then I ran the same tests at the same power (not output voltage) levels and bias (80 mA) as before. 300B's do not mind a 2500 ohm load. This load resulted in 8.4 watts at soft clip. 305 volts B+.
The 1 watt frequency response under these conditions is now 12.2 Hz to 43.3 KHz. Distortion is 1.1% mostly second harmonic. The third harmonic disapears at 105 mA.
Frequency response at 6 watts is 23.7 Hz to 40.8 KHz.
Gee we lost some highs, and gained a little bit of bass, kind of like what we said would happen. In this case there is plenty of high frequency to give up. So far my only complaint with these transformers is their desire for current. Maybe the gap is just a little too wide for small tubes like the 45.
I put back the 300B's and hooked a 4 ohm load to the 8 ohm Edcors. Then I ran the same tests at the same power (not output voltage) levels and bias (80 mA) as before. 300B's do not mind a 2500 ohm load. This load resulted in 8.4 watts at soft clip. 305 volts B+.
The 1 watt frequency response under these conditions is now 12.2 Hz to 43.3 KHz. Distortion is 1.1% mostly second harmonic. The third harmonic disapears at 105 mA.
Frequency response at 6 watts is 23.7 Hz to 40.8 KHz.
Gee we lost some highs, and gained a little bit of bass, kind of like what we said would happen. In this case there is plenty of high frequency to give up. So far my only complaint with these transformers is their desire for current. Maybe the gap is just a little too wide for small tubes like the 45.
With FFT software (winMLS) and a high end sound card these type of measurements literally take 5 minutes, just long enough to plug in the cables and run the program. I already have the Edcor connected into the amp. I plan on listening to it for a while while I decide what to build with them. I am leaning toward a typical low cost audio pentode (6L6, EL-34, EL84 etc) switchable for triode, UL or pentode). I won't get to it for a few weeks, but it will be posted here and on the web site.
I got back from a 2500 mile road trip Tuesday. I knew that 10 to 12 hour work days awaited me when I got back(I was right), so I left 1 day earlier to have 1 day of experiment time.
The TubelabSE amplifier that I have been using for science experiments was still wired up for 6AV5's in UL with cathode feedback (CFB). I listened to it for a while before deciding to make some changes. The 6AV5's really sound great, but they are too inconsistent in their power handling capabilities, and they are not available in large parts of the world. I think that the 6AV5's are more suited for screen driven P-P, since I am currently getting 80 watts out of a pair without abusing them.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=37403&highlight=
http://www.diyaudio.com/forums/showthread.php?s=&threadid=74650&highlight=
With mixed emotions I rewired the sockets to fit the popular audio tubes of today, 6L6, EL-34, KT-88, 5881, 6550, 6V6, and others. The amp is configured for triode, UL or pentode mode, with or without CFB. It is still connected to the Knight variable power supply which goes from zero to 400 volts. I then assembled a collection of tubes to roll through this amp, and proceeded to try them all for a few minutes each. After picking a few favorites I gave those an extended listening session. I have made no measurements yet.
These transformers really like CFB. I tried each tube with and without CFB in all 3 modes. In ALL cases the sound was far better with CFB. The bass was tighter (expected) and transients were more realistic. No loss of realism was detected in any case when adding CFB. The loss of gain (I have plenty) was different with each tube and greatest in pentode mode. I was not impressed with pentode mode with any of the tubes tried either.
Attachments
The tubes were selected to try mostly low cost tubes, a few favorites and a pair of NOS GE 6550A's (expensive by my standards).
Bottom left (and above it) are Svetlana winged C 6L6's, new. These sounded good. Triode mode good for most music, UL for loud rock or techno. I ran them at voltages from 250 to 400, adjusting the current to keep the dissipation under 25 watts. I liked 400 volts, 60 Ma in UL best overall.
Bottom row, right of the Svet is a pair of Sylvania 5932's, definitely used. A sleeper in guitar amps, except that they keep popping out of the sockets in a Bandmaster. These sounded a lot like the Svets, but I could push them to 30 watts without any signs of distress. Same deal, I liked 400 volts 70 mA UL.
Next, JJ EL-34's, new. Definitely different. In triode mode, these sounded the most like a DHT. The most detail in the music was revealed with this combination. As with many DHT's there was less bass and not as much punch. UL adds back some of the bass and punch, but takes away some of the DHT sound. In triode, I liked 325 volts at 80 mA, conveniently this is the same operating point that I like for a 300B with the Edcors.
Bottom row, far right is the pair of NOS GE 6550A's. This is the polar opposite of the EL-34. Best for loud rock, techno, and anything else that you would turn up all the way! I liked these best in UL with the power supply maxed at 400 volts and the current at 75 to 80 mA. Best bass, punch and dynamics. These sound the most like my 300B push pull amp.
Second row above the 5932's are Sovtek 6550WA's. These are used and 5 years old, but I also tried a NOS pair. Almost as good as the GE's but some loss of realism. The used pair suffered from bias creep. The current kept climbing.
Next up is a crusty old pair of GE 6L6GB's. I pulled these out of a Fender Bandmaster where they had lived a hard life for at least 10 years. These sounded remarkably like the other 6L6 types and worked happily at 25 watts despite a 19 watt rating. I played them for 2 hours at 400 volts and 60 mA, no complaints. These are proof that these old GE's were made well.
Far right is a pair of new Svetlana winged C EL-34's. I bought 4 from Antique Electronics Supply. One was bad out of the box. The tube works and sounds good, but there is a screen grid wire that is brightly glowing at any voltage above 250 volts. I don't feel comfortable using this tube because it will likely fail catastrophically. Other than that they sound a lot like the JJ EL-34's. I think I like the sound of the JJ's better, but they are close.
The fat guys on the left of the top row are Chinese KT-88's. These were made at least 10 years ago. I bought about 50 of these for an average price of $12 USD when a local tube dealer discontinued them. He said that he got too many returns and he warantees all of his tubes. I bought them all and found that they sound great and are reliable at 400 volts or less, and 30 watts or less. If you put these in a Marshall Major or an SVT, you will be reminded that the Chinese invented fireworks! I tried two pair and found that they do indeed sound great, as good as The GE 6550A's, maybe better bass, but they start to glow at 30 watts (400 volts 75 mA). If left alone at 30 watts they can go into a runaway condition. This WILL happen (4 out of 4) at 400 volts and 85 mA.
To the right is a pair of Chinese 6L6's see the next post to find out about these.
Bottom left (and above it) are Svetlana winged C 6L6's, new. These sounded good. Triode mode good for most music, UL for loud rock or techno. I ran them at voltages from 250 to 400, adjusting the current to keep the dissipation under 25 watts. I liked 400 volts, 60 Ma in UL best overall.
Bottom row, right of the Svet is a pair of Sylvania 5932's, definitely used. A sleeper in guitar amps, except that they keep popping out of the sockets in a Bandmaster. These sounded a lot like the Svets, but I could push them to 30 watts without any signs of distress. Same deal, I liked 400 volts 70 mA UL.
Next, JJ EL-34's, new. Definitely different. In triode mode, these sounded the most like a DHT. The most detail in the music was revealed with this combination. As with many DHT's there was less bass and not as much punch. UL adds back some of the bass and punch, but takes away some of the DHT sound. In triode, I liked 325 volts at 80 mA, conveniently this is the same operating point that I like for a 300B with the Edcors.
Bottom row, far right is the pair of NOS GE 6550A's. This is the polar opposite of the EL-34. Best for loud rock, techno, and anything else that you would turn up all the way! I liked these best in UL with the power supply maxed at 400 volts and the current at 75 to 80 mA. Best bass, punch and dynamics. These sound the most like my 300B push pull amp.
Second row above the 5932's are Sovtek 6550WA's. These are used and 5 years old, but I also tried a NOS pair. Almost as good as the GE's but some loss of realism. The used pair suffered from bias creep. The current kept climbing.
Next up is a crusty old pair of GE 6L6GB's. I pulled these out of a Fender Bandmaster where they had lived a hard life for at least 10 years. These sounded remarkably like the other 6L6 types and worked happily at 25 watts despite a 19 watt rating. I played them for 2 hours at 400 volts and 60 mA, no complaints. These are proof that these old GE's were made well.
Far right is a pair of new Svetlana winged C EL-34's. I bought 4 from Antique Electronics Supply. One was bad out of the box. The tube works and sounds good, but there is a screen grid wire that is brightly glowing at any voltage above 250 volts. I don't feel comfortable using this tube because it will likely fail catastrophically. Other than that they sound a lot like the JJ EL-34's. I think I like the sound of the JJ's better, but they are close.
The fat guys on the left of the top row are Chinese KT-88's. These were made at least 10 years ago. I bought about 50 of these for an average price of $12 USD when a local tube dealer discontinued them. He said that he got too many returns and he warantees all of his tubes. I bought them all and found that they sound great and are reliable at 400 volts or less, and 30 watts or less. If you put these in a Marshall Major or an SVT, you will be reminded that the Chinese invented fireworks! I tried two pair and found that they do indeed sound great, as good as The GE 6550A's, maybe better bass, but they start to glow at 30 watts (400 volts 75 mA). If left alone at 30 watts they can go into a runaway condition. This WILL happen (4 out of 4) at 400 volts and 85 mA.
To the right is a pair of Chinese 6L6's see the next post to find out about these.
Attachments
I was looking up the prices of the popular audio tubes being tested above on several web sites, when I found a bold statement. On Antique Electronics Supply's web site there is an add for a "6L6GC China Coke bottle" for $7.95, followed by the statement "We could not blow up this tube, no matter how hard we tried." Now this seems like a challenge to me. I dug through my collection and found a pair of tubes that look like theirs, except they have the Ruby Tubes logo screened onto them.
I put these into the amp and proceeded to turn up the juice. If you look carefully in the attached picture you can see the voltmeter on the power supply ix maxed at 400 volts. The digital meter on the right shows 110 mA. This is 44 watts per tube. There is an even glow on the plates, and the tube is stable with no hint of runaway. It ran at this current for an hour with no problems. I turned it down to 80mA and there is no glow. I am sure that I could blow up this tube if I really tried, but they are really tough.
I have been listening to these for 3 days and find them to be the best overall compromise. In triode they are almost as good as the EL-34's and in UL they are close to the 6550's.
Overall the Edcors have surprised me with their sound and price. I am working on a new amp design using these transformers. The idea is to make a low cost SE amplifier that sounds almost as good as a TubelabSE.
I put these into the amp and proceeded to turn up the juice. If you look carefully in the attached picture you can see the voltmeter on the power supply ix maxed at 400 volts. The digital meter on the right shows 110 mA. This is 44 watts per tube. There is an even glow on the plates, and the tube is stable with no hint of runaway. It ran at this current for an hour with no problems. I turned it down to 80mA and there is no glow. I am sure that I could blow up this tube if I really tried, but they are really tough.
I have been listening to these for 3 days and find them to be the best overall compromise. In triode they are almost as good as the EL-34's and in UL they are close to the 6550's.
Overall the Edcors have surprised me with their sound and price. I am working on a new amp design using these transformers. The idea is to make a low cost SE amplifier that sounds almost as good as a TubelabSE.
Attachments
If there's any blowing up to be done, you're the man!!
El-Cheapo OPTX
Go to rswww.com do a search for part number:- 210-6475
I can't link the page as it never seems to work with that particular site.
They are a universal OPTX UK£7-75+vat
I tested one a couple of years ago, they roll off at @13kHz.
They are pretty good replacement for single-ended o/p stages in AM radios!
And good fun to play with.
Most tube amp builders seem to favor a 2.5 to 3.6K ohm load with 300 or so volts B+ on the 6CA7 / EL-34. This is close to what the original Mullard data reccommended. A 2.5 K load can easilly be obtained with the $18 Edcor 5K transformers by using an 8 ohm load with a 16 ohm (secondary impedance) transformer. You could also use a 4 ohm load with an 8 ohm transformer. I have tested this with the Edcor transformers and the results were good.
I have been working on the Simple SE amp mentioned above and I prefer the sound of the EL-34 (EI and Svet winged C) with a 5 K load and 400 to 500 volts of B+. I have not measured the power output or distortion on any of these experiments yet due to a heavy work schedule. It will be posted in a couple of weeks.
Currently using 1/2 of a 12AT7 as a driver.
I have been working on the Simple SE amp mentioned above and I prefer the sound of the EL-34 (EI and Svet winged C) with a 5 K load and 400 to 500 volts of B+. I have not measured the power output or distortion on any of these experiments yet due to a heavy work schedule. It will be posted in a couple of weeks.
Currently using 1/2 of a 12AT7 as a driver.
cheap OPTs
I put a set of the $17 Edcors in a JE Labs 2a3 just to see what came out. Not bad, considering the price, excelent. After a few beers my listeners thought they sounded as good as the magnaquests it was designed for, O scope indicated they were close to correct. An american company building a good product for a fair price, Novel concept.
I put a set of the $17 Edcors in a JE Labs 2a3 just to see what came out. Not bad, considering the price, excelent. After a few beers my listeners thought they sounded as good as the magnaquests it was designed for, O scope indicated they were close to correct. An american company building a good product for a fair price, Novel concept.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.