Never Get Old,
Please take 5 minutes out of your day, and measure the DCR of your Parallel Vifa "8 Ohm" Woofer/Midrange.
Measure at the speaker input terminals, not at the driver terminals; you do not need to disassemble the cabinet.
And, that will include the driver DCR plus the woofer crossover inductor DCR in series with it (which is what you really want to know).
Then, Please take another 5 minutes out of your day, and post the results of the DCR measurement.
Then, Please take another 5 minutes, and tell me the model number of your Vifa driver.
Thank You!
Please take 5 minutes out of your day, and measure the DCR of your Parallel Vifa "8 Ohm" Woofer/Midrange.
Measure at the speaker input terminals, not at the driver terminals; you do not need to disassemble the cabinet.
And, that will include the driver DCR plus the woofer crossover inductor DCR in series with it (which is what you really want to know).
Then, Please take another 5 minutes out of your day, and post the results of the DCR measurement.
Then, Please take another 5 minutes, and tell me the model number of your Vifa driver.
Thank You!
Never Get Old,
After re-reading your Post # 153, I have some further questions.
Is your speaker like this? . . .
Vifa 8 Ohm parallel woofers with a crossover coil in series?
Vifa 8 Ohm parallel midranges with a crossover coil and capacitor bandpass in series?
Tweeter, single, with a capacitor in series, and a coil across the tweeter?
I asked you to list the model number of the Parallel Vifa woofers.
Now, Please list the model number of the Parallel Vifa Midranges.
What is the speaker cabinet, ported, closed, etc.?
Thanks!
After re-reading your Post # 153, I have some further questions.
Is your speaker like this? . . .
Vifa 8 Ohm parallel woofers with a crossover coil in series?
Vifa 8 Ohm parallel midranges with a crossover coil and capacitor bandpass in series?
Tweeter, single, with a capacitor in series, and a coil across the tweeter?
I asked you to list the model number of the Parallel Vifa woofers.
Now, Please list the model number of the Parallel Vifa Midranges.
What is the speaker cabinet, ported, closed, etc.?
Thanks!
@quadtech you sir deserve an award! 🏆
For anyone reading this thread in the future, here are relevant excerpts:
To get better adjustment range:
Change R1-R4 from 470 ohms to 510 or even 560 ohms, that will lower the EL84s' cathode current adjustment range. Changing the 10K trimpots to 20K would not significantly widen the adjustment range. Typically, the best adjustment range would be evenly centered around 400 ohms to ground.
and
35 ma. of cathode current is an approximation in Dyna ST-35 amplifiers, but is subject to variations. Plate and screen current combined, is equal to the cathode current in the 6BQ5 output tube. An example from the original Dyna manual-- Plate voltage: +370VDC Cathode voltage: +13.5VDC. Subtract the cathode voltage from the plate voltage to arrive at the working plate voltage. 370-13.5= 356.5. Multiply that by the cathode current (.0355A) to get the combined plate and screen grid dissipation of 12.66 Watts. Subtract an estimated 1.5 watts of screen dissipation to arrive at 11.16 Watts plate dissipation. These parameters were set to give good performance with good tube life.
Independently increasing either the cathode current or B+ voltage will raise the dissipation, and also result in a slight increase in the operation of the output stage under Class A conditions. This might (but not necessarily) result in a slight improvement in sound. In practice it will also slightly decrease tube life, but that’s subject to a lot of other variables, particularly the quality of the output tubes. Some 7189 tubes (mostly no longer available) were rated as high as 15.4 Watts of combined plate and screen dissipation.
As for my situation, I discovered a "sweet spot" (sort of) for the 6P14P-ER tubes at about 37mA. At 35 (amplifier spec) to 36 they are harsh and no good, and from 38 to 39 they aren't very good either in the bass. However, they really "sing" and come to life at 40mA!!! Given that the tube has maximum ratings of 14W plate plus 2W screen, 40 mA would translate to roughly 99% using the calculator here: https://robrobinette.com/Tube_Bias_Calculator.htm Gee, that would explain a lot. The original Dynaco design ran the output tubes very, very hard too. If I manage to live with the little sweet spot at 37mA, I think they would be just fine as far as tube life. It's a 10,000 hr rated tube. 37 mA might rough out to about 90%. It is possible that after these tubes break in they will sound fine at 37mA. The 6P14P-ER takes a long time to break in according to owners, and I only have about 100 hours on this set. I have a feeling that the NOS General Electric 12DW7's will sound more to my liking, so I may actually be fine. Time will tell. I'm going to crank the tubes up to 40mA where they sound the best for now. In the dark, there is no red glow on the plates (yet). Directions are above to resolve the issue I have of maxing out the potentiometers at 40mA, if I need to do that.
I also did (unintentionally) verify that my meter is reading mV correctly by verifying my test results against the tube bias calculator above.
So, I'm feeling a little better now. I hope it lasts! A little persistence and testing and research go a long way. Thanks again @quadtech
And now, to test and break in the new Chinese DAC/preamp that just arrived from ... China! I love the other ones I have, but I wanted a new toy, and this one may be a little higher quality.
Last edited:
Yes, I'll do it next time I want to drag one of my massive speakers aside. Each one has two 8 ohm bass drivers in parallel and two 8 ohm midrange drivers in parallel, with first order crossovers. Like I said, my other two tube amps and a half dozen solid state amps have never had a problem with them, even with the tube amps using the 8 ohm outputs which although "wrong" in theory still sounds better.
Two Vifa P21WO-20-08 (I just bought the last four ScanSpeak replacements to be shipped to the US I am told) in a large ported enclosure
Two Vifa P13WH-00-08 (which need to be replaced within a year or two, discontinued, likely will use ScanSpeak 15W8434G00-CSD) in a sealed sub-enclosure
One Morel 8 ohm soft dome tweeter of some sort with a second order crossover to that
I chose this configuration some 30 years ago after a lot of sound quality testing by a dozen people including musicians and a half dozen different experimental designs. We even went so far as to set up drums in the room with them for comparison. 😀 A man who played the pipe organ helped me a lot with tuning them. They have good sensitivity and tremendous bass and sub-bass in their large ported enclosures, and I haven't ever felt any need to change them other than replacing the drivers every 10 years or so because the rubber surrounds harden over time. I won't go off topic and delve into speaker design choices here. A lot of people with different speakers wouldn't ever hear or feel what I am hearing and feeling in the bass and sub-bass though. A lot of speakers simply can't do it. That was one of my primary design criteria. For years I called them the Eviction Model 1 because neighbors likely agree.
See what I just posted regarding bias at either 37mA or 40mA and sound quality. I may be fine now that I have more knowledge and info.
Last edited:
"For years I called them the Eviction Model 1 because neighbors likely agree."
LMAO- I needed a good laugh today...
It seems apparent to me- relevant to this thread- that you possess above average (never get old) hearing😉
Also-
Your writings indicate that you've a bevy of musicians as an auditory eval panel... As such- don't be suprised if your requested 'oscilloscope session' (if occurs) will be disagreeable w/ your auditory panel. In this case, a picture might not be worth much- not the usual 1000 words.
Jim
LMAO- I needed a good laugh today...
It seems apparent to me- relevant to this thread- that you possess above average (never get old) hearing😉
Also-
Your writings indicate that you've a bevy of musicians as an auditory eval panel... As such- don't be suprised if your requested 'oscilloscope session' (if occurs) will be disagreeable w/ your auditory panel. In this case, a picture might not be worth much- not the usual 1000 words.
Jim
I knew someone was going to ask that. I have written this up somewhere before, but I forget where I posted it.
I spent a summer about 30 years ago housesitting for 3 months, taking care of two cocker spaniels while the owners were away (no comment on the dogs' intelligence) and I took that time isolated to do math and drawings for different designs about 8 hours per day most days. I was getting paid, ROFL. I spent the rest of my time reading papers by Paul Klipsch and various pro audio engineering references when people still used something called a public library. Very informative. Remaining time was spent in live jazz clubs listening (no comment on anything else done there). No computer. All math done by hand and with a simple calculator. I still have all of the math written down in a folder.
I then tried a number of different designs including open baffle because I had access to a neighbor's wood shop and I could knock out as many enclosures as I wanted any time I wanted. I learned about distance-related acoustic coupling between drivers, and I learned the critical nature of the baffle step, and then I finally admitted to myself that 99% of all loudspeakers need to account for a room with a floor and a rear wall somewhere. Go figure. 😀 Once we accept that fact and plan for it, we can design for the rest of the intended room. Every room I ever will have will have wall-to-wall carpet and/or heavy area rugs covering everything. Almost every room I ever have will be have side walls at different distances. So we have to account for that. Then there is the ceiling. Yes, we need to account for reflections and interactions there too. I always have a standard ceiling, not vaulted, so add that in. As for the rear wall, I don't live in a warehouse and can't ever have speakers 1/3 of the way out into the living room. So we need to account for interactions with the rear wall and if we are smart and planning well then use it to our advantage. Notice the method to my madness - you can't plop any speaker down in any room and expect to get the same result as someone with a different room.
I also developed a speaker technology that I have thought about patenting for decades. It hasn't ever been done before. I implemented it for testing in a small speaker but don't use it currently, sadly. Maybe one day. I just don't see a market for it due to the extreme cost and engineering and construction difficulty of implementing it. It's all written down and will be made public by a relative should anything happen to me.
I have always wanted to design and build a floor-to-ceiling line array speaker for fun, but it's huge and time consuming and expensive and I no longer have access to a wood shop.
Enough of that.
The picture of the Eviction Model 1 Loudspeaker
It may not look fancy as it's "just" a large 3-way floorstanding speaker. Roughly 4 feet tall. It's the engineering and voicing during development that makes them work. Boston Acoustics did a similar physical configuration and so did JBL, but they both overlooked one critical factor that I didn't overlook. They work for me. They can reproduce a pipe organ or a snare drum accurately, and with a 300 watt per channel amp, they will do a snare drum at doggone near correct volume in a small room. Very efficient. Use the "next room test" on any loudspeaker. Put a drum set in the living room. Go into another room. The drums will still sound live. Go to the other end of the house and close three doors. They will still sound live. Go out to the street. They will still sound live. What about the speakers? Yes. I have done it all. I have even taken them outside for testing to eliminate side walls and the ceiling. I had to notify the neighbors first of course.
Your mileage may vary.
PS: Why would anyone in the real world want a speaker where the ideal listening spot is exactly between the speakers with them pointed straight at that spot? Only ONE person can sit there, and they have to be glued to the chair. What I want is a speaker that sounds good EVERYWHERE EXCEPT that tiny spot. 😱 I probably just offended a lot of people, LOL. We all have different approaches to life, and audio. Do what you like and what sounds good to you.
Last edited:
There's no reason you can't change those if you want. There are good alternatives.
And mica capacitors are well known to have high dielectric absorption.
If changing to the little see-through polystyrenes like Mallory, MIAL etc. be sure to make yourself a clamp-on heatsink for soldering the capacitors' leads. A bit of Romex mashed kinda flat and soldered into a clip lead end works, but still be careful - they melt like there's no tomorrow.
One explanation for your dissatisfaction with your new amplifier is that it's being improperly loaded. The extreme sensitivity to tiny changes in idling current isn't normal and indicates something fundamentally wrong. Pretty clear that it's the known wrong choice of OPT secondary taps. Don't matter if you've gotten away with it on other amplifiers; that's other amplifiers. With correct loading the new amplifier sounds fine (right?) except for some lack of low bass.
So, what could cause this change in bass response? I'd posit that the speakers, with their four pole ported response, are quite sensitive to source resistance (all speakers are, and ported/passive/drone are especially). An 8 Ohm OPT secondary tap gives a higher source resistance than the 4 Ohm, changing the speaker's frequency response. Your speakers want a higher source resistance to give you the frequency response you're used to. You've had decades to learn how those speakers sound, so can hear the difference of even a few dB at very low frequencies, where the greatest changes in frequency response with source resistance happen.
In the bad old days, amplifiers often had adjustable output resistance to accommodate varying speakers' requirements, but modern (post Kennedy era) amplifiers all strive for as close to zero output Z as possible, and speakers are designed for that. For homebrew speakers, especially ported or other four pole designs, might be good to include something like that old tech again. In your case a series resistor, adjustable from zero up to maybe 2 Ohms or 3 Ohms, should be about right. Maybe a 5 Ohm fixed in parallel with an 8 or 10 Ohm adjustable. This would allow you to operate the amplifier at its design load and still provide the speakers their desired source resistance. Obviously this could be outside of the amplifier chassis.
All good fortune,
Chris
One explanation for your dissatisfaction with your new amplifier is that it's being improperly loaded. The extreme sensitivity to tiny changes in idling current isn't normal and indicates something fundamentally wrong. Pretty clear that it's the known wrong choice of OPT secondary taps. Don't matter if you've gotten away with it on other amplifiers; that's other amplifiers. With correct loading the new amplifier sounds fine (right?) except for some lack of low bass.
So, what could cause this change in bass response? I'd posit that the speakers, with their four pole ported response, are quite sensitive to source resistance (all speakers are, and ported/passive/drone are especially). An 8 Ohm OPT secondary tap gives a higher source resistance than the 4 Ohm, changing the speaker's frequency response. Your speakers want a higher source resistance to give you the frequency response you're used to. You've had decades to learn how those speakers sound, so can hear the difference of even a few dB at very low frequencies, where the greatest changes in frequency response with source resistance happen.
In the bad old days, amplifiers often had adjustable output resistance to accommodate varying speakers' requirements, but modern (post Kennedy era) amplifiers all strive for as close to zero output Z as possible, and speakers are designed for that. For homebrew speakers, especially ported or other four pole designs, might be good to include something like that old tech again. In your case a series resistor, adjustable from zero up to maybe 2 Ohms or 3 Ohms, should be about right. Maybe a 5 Ohm fixed in parallel with an 8 or 10 Ohm adjustable. This would allow you to operate the amplifier at its design load and still provide the speakers their desired source resistance. Obviously this could be outside of the amplifier chassis.
All good fortune,
Chris
Never Get Old,
Wow!
That is a very nice loudspeaker.
The idea that the speaker is not dependent on the listener having to sit in the sweet spot is probably one of the most important factor of your speaker.
I am impressed.
Only once or twice in my life have I been able to walk from the listening room to another room, and had it sound "Live".
But I was curious about the parallel woofers and the parallel midranges, and how that might affect the "different sound" of the ST-35 versus other amplifiers.
The Vifa woofer DCR is 5.7 Ohms, 2 in parallel is 2.85 Ohms DCR.
The woofer impedance graph shows 6 Ohms from 200Hz to 600 Hz (3 Ohms for 2 in parallel; pretty close to the 2.85 Ohm DCR).
I expect that at the port tuning frequency, the parallel woofer impedance may be somewhere between 3 to 4 Ohms.
The Vifa midrange DCR is also 5.7 Ohms, 2 in parallel is 2.85 Ohms DCR.
The midrange impedance graph shows 8 Ohms from 400Hz to 900Hz (4 Ohms for 2 in parallel).
I am sure there are a lot more factors that affect the impedance versus frequency, such as the crossover, series resistor for equalizing sensitivity, the cabinet, stuffing, etc.
You know all that about your speaker.
It does appear that some amplifiers might sound different, depending on your speaker's impedance versus frequency.
I hope that the ST-35 can be adjusted to sound the way you want it to sound with your speakers.
Good Luck, and happy listening!
Wow!
That is a very nice loudspeaker.
The idea that the speaker is not dependent on the listener having to sit in the sweet spot is probably one of the most important factor of your speaker.
I am impressed.
Only once or twice in my life have I been able to walk from the listening room to another room, and had it sound "Live".
But I was curious about the parallel woofers and the parallel midranges, and how that might affect the "different sound" of the ST-35 versus other amplifiers.
The Vifa woofer DCR is 5.7 Ohms, 2 in parallel is 2.85 Ohms DCR.
The woofer impedance graph shows 6 Ohms from 200Hz to 600 Hz (3 Ohms for 2 in parallel; pretty close to the 2.85 Ohm DCR).
I expect that at the port tuning frequency, the parallel woofer impedance may be somewhere between 3 to 4 Ohms.
The Vifa midrange DCR is also 5.7 Ohms, 2 in parallel is 2.85 Ohms DCR.
The midrange impedance graph shows 8 Ohms from 400Hz to 900Hz (4 Ohms for 2 in parallel).
I am sure there are a lot more factors that affect the impedance versus frequency, such as the crossover, series resistor for equalizing sensitivity, the cabinet, stuffing, etc.
You know all that about your speaker.
It does appear that some amplifiers might sound different, depending on your speaker's impedance versus frequency.
I hope that the ST-35 can be adjusted to sound the way you want it to sound with your speakers.
Good Luck, and happy listening!
On the 4 ohm output, the ST-35 is not good overall, just like the ST-70 ii and the Chinese EL84 tube amp kit are not good on the 4 ohm output. It will give a midrange without harshness on the ST-35, but that's at the expense of everything else.
Yes, I always figured that to be just about correct across major portions of their curve, not just the bass. I was concerned about it when I designed them, but it never turned out to be an issue so I stuck with the design.
Tuning by ear. It's all we had back then outside of speaker manufacturers' and stereo magazines' testing labs. Who listens in an anechoic chamber anyway? I still think too many people overlook "voicing" a speaker by ear and focus on test results too much. It is a whole lot of work, but it has to be done. My friend who played the pipe organ back then had me adjust the mids and tweeter by 1/2 dB here and there and also adjust the damping material a number of times. Theoretically that small a change (1/2 dB) is measurable but "inaudible" right? 😉 Damping material and its placement also make an enormous difference. A lot of times, less is better. People used to mindlessly stuff speakers almost full of fiberglass back then.
It's interesting to note that I have tried the ST-70 ii with a lot of different speakers over the years, toting it around to various friends' houses and also to various stereo shops for demos. It was a novelty back then in the heyday of ultra-high-end solid state gear from the likes of Krell and Levinson. I had a Threshold amplifier back then too. I paid $2,000 for the Threshold - used. Yikes. I knew a lot of people who had never heard a tube amplfier, like me at the time. So I hauled it around a lot. Every time it sounded best on the 8 ohm output, regardless of what speaker I atttached to it. In theory I should be using the 4 ohm outputs on all three of my tube amps, but in reality it just doesn't work. I always have suspected that it has something to do with the NFB being attached to the 8 ohm output.
I tried the amplifier again at 37mA, and there is a sweet spot there for these tubes. I also tried it again at 40mA, and the amplifier is impressive there. I'm going to call the bias setting a non-issue now that I realize that 40mA is roughly 100% of rated dissipation for the tubes. I can run it at 37mA and live with that if the tubes die early, but I doubt they will. There also is a fix posted above to expand the adjustment range if I want to swap a few resistors.
The General Electric 12DW7 drivers will be here in a few days. We will see how they impact the sound. I threw my JJ 12AU7 and EL34 tubes in the trash, so I question their 12DW7.
The new DAC from China was fun to listen to last night. It has a fiery and sparkling top end, but it's interesting. It may settle down over time, and I have a bunch of different op amps to experiment with. Fun fun. 😀
Last edited:
I don't have to tell you what you have to do, but that's a shame.
I'm not a tube star but I have a lot of various tube amps and above all a lot of tubes and what I can say is that a tube that sounds bad in an amp can sound very good in another and it is not the tube that is in question, it is the assembly and the adjustments that play.
Finally, I use a lot of JJ tubes and sometimes I even torture them and I have never had a problem with them, except their rectifier tubes.
With all that cash & sweat you have invested in various amplifiers & speaker systems over several years you could
easily have bought a first rate scope, SG & other test equipment. And learned how to use it all to advantage.
On this amp there could be bursts of ultrasonic oscillation that could be detected with a scope.
And fixed with low value, carbon composition resisters in the plate & screen leads close into the tube sockets.
But difficult to do on a PCB.
Running up close to the dissipation limit is never a great idea, if the line goes high the result could be red plating or worse.
The heat is a problem for the PCB unless it is MIL Grade. 🙂
They are. And the tube sockets are not soldered to the PCB fortunately. There is a gap around the sockets.
And use it exactly once. Never needed one. I have no intention of ever building another amplifier, and I now am not convinced there is anything wrong with this one.
To those of you proposing that I modify the amplifier, it would look like this, in red, correct? The 100 ohm would be hanging loose in the air, which is not ideal. The 1k ohm is very easy and would have the most benefit, correct? I could do the 1k ohm resistors in just a few hours, if I order the resistors. They would simply replace an existing jumper wire from the PCB to the tube socket. Other than $10 (because of shipping cost) and a few hours of time, is there any downside to adding those 1k ohm resistors? Both the ST-70 ii and the cheapo Chinese EL84 amp kit do have them in place.
Last edited:
Never Get Old,
1. Agreed, use the 1k resistors at pin 2; use the 100 Ohm resistors at pin 9.
2. An easy modification that should fix the 8 Ohm and 4 Ohm tap problem.
Change R10 from 20k to 14k (or 15k, whichever is easier to purchase)
Change C7 from 39pF to 50pF.
Then remove R10 and C7 (now 14k and 50pF from the 8 Ohm Tap, and re-connect them to the 4 Ohm Tap.
Now, wire your speakers to the 4 Ohm Tap.
The negative feedback is now coming from the Tap that your speaker uses.
I bet that fixes the problem.
1. Agreed, use the 1k resistors at pin 2; use the 100 Ohm resistors at pin 9.
2. An easy modification that should fix the 8 Ohm and 4 Ohm tap problem.
Change R10 from 20k to 14k (or 15k, whichever is easier to purchase)
Change C7 from 39pF to 50pF.
Then remove R10 and C7 (now 14k and 50pF from the 8 Ohm Tap, and re-connect them to the 4 Ohm Tap.
Now, wire your speakers to the 4 Ohm Tap.
The negative feedback is now coming from the Tap that your speaker uses.
I bet that fixes the problem.
6A3sUMMER's suggestion moves the feedback tap down 3dB and then decreases feedback network impedance 3dB to keep closed loop gain at its original value. This is almost not even a modification, merely an adjustment to optimize for desired OPT tap. If you were to keep the existing values and simply move the feedback source to the 4 Ohm tap you'd have 3dB higher closed loop gain and 3dB higher source resistance to the speaker (from both taps), and 3dB higher distortion. Super easy to try, but maybe or maybe not what you want. Amplifiers have been made with feedback taken from multiple taps and mixed together, or from dedicated feedback windings (although this is for distortion reduction and a step away from your goal).
Grid (G1) stops are also not a modification and are only left off output valve grids for cost savings. David Hafler ("The cheapest part is the best part.") knew how to keep his stuff affordable, but he had those exact Eisenhower era parts and layout right in front of him to work with, so could confidently bend Best Practice in ways that DIYers shouldn't (and needn't - parts are cheap). It was Japanese fashion in the 1980s to omit grid stops, probably because of how schematics would look, then it became fashionable in the New York Noise set and both were popularized by Sound Practices magazine in the 1990s. Foolishness, but we're only young once, if at all.
Screen (G2) stops are less critical but if used many would recommend that they be very low wattage or even "fusable" resistors, the hope being that in some catastrophic overcurrent failure mode they'd bake open before the screens get too cooked. I've heard of a situation where this actually saved the valve, and therefore a matched quad of valves. Easy enough anyway.
All good fortune,
Chris
Grid (G1) stops are also not a modification and are only left off output valve grids for cost savings. David Hafler ("The cheapest part is the best part.") knew how to keep his stuff affordable, but he had those exact Eisenhower era parts and layout right in front of him to work with, so could confidently bend Best Practice in ways that DIYers shouldn't (and needn't - parts are cheap). It was Japanese fashion in the 1980s to omit grid stops, probably because of how schematics would look, then it became fashionable in the New York Noise set and both were popularized by Sound Practices magazine in the 1990s. Foolishness, but we're only young once, if at all.
Screen (G2) stops are less critical but if used many would recommend that they be very low wattage or even "fusable" resistors, the hope being that in some catastrophic overcurrent failure mode they'd bake open before the screens get too cooked. I've heard of a situation where this actually saved the valve, and therefore a matched quad of valves. Easy enough anyway.
All good fortune,
Chris
Last edited:
The ST35 boards are sometimes prone to oscillation , different tubes might be all that is needed. Even inslalling
EFB kit can provoke this unless screen and grid stoppers are installed.
Yes, building amp lacking oscilloscope is asking for trouble.