|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
68.6.131.58
In Reply to: RE: Schematic? posted by Triode_Kingdom on June 29, 2021 at 19:14:15
The '3 tube' solution is apparently what both the ST70 and ST120 have.
I don't remember the order of 'gain' for the interchangeable tubes of the '12 series'. Can you put them in order from LOW to HIGH gain?
I don't know enough to know WHY pin #7 has a 1 meg resistor to it and is @110v while pin #2 has no such resistor and is at 115v.....
Shouldn't these be symmetrical? Do you adjust the AC balance pot ONCE per tube change to get the SAME 370 volts OUT? The 20k resistors are listed as 2 watts, which is a LOT, IMO, Is the DC balnance pot for 'bias'? That makes a little sense because I see downstream of that .2v whcih might be the bias voltage.....whatever the current works out to?
Bob Lantino mentions 5 tubes OK for this circuit. 12AU7 / 12BH7 / 5963 / 5814 / 6189 AND has a CCS in the circuit in the form of an LM334 from TI....... I do NOT know what tube the amp ships with OR if any others are 'optional' or which to choose should I be given such a choice.....
Too much is never enough
Follow Ups:
Just to be clear, the fact that the driver has three tubes isn't sufficient to determine the configuration. It could be a dual triode driving a cathodyne, a single triode or pentode driving a long tailed pair, or something else. And within each circuit configuration, there are numerous possible variations in design that can be significant.
Considering the VTA amps come in kit form and would be of interest not only to audiophiles, but also to electronics enthusiasts, omission of the schematics is poor practice in my opinion. I'd rather restore an older amp that I can assess prior to the purchase than gamble on an unknown circuit built on a circuit board that can't be easily modified.
About the various 12XXX tubes, you can look up multiple data sheets for almost any tube at the link below. Look for the tube's "amplification factor" and multiply it by 70% to derive the approximate in-circuit gain. Some data sheets will also have resistance-coupled gain charts showing in-circuit gain under various operating conditions. Note that tubes used in a long tailed pair will behave differently. The in-circuit gain of a triode used in a long tailed pair is approximately 30% of the amplification factor.
"I don't know enough to know WHY pin #7 has a 1 meg resistor to it and is @110v while pin #2 has no such resistor and is at 115v..... Shouldn't these be symmetrical?"
They do in fact have the same voltage. The difference shown on the schematic is created by the loading effect of the voltmeter that was used. Pin 7 reads less because of the small voltage drop across the 1M resistor when the voltmeter probe was connected.
"Do you adjust the AC balance pot ONCE per tube change to get the SAME 370 volts OUT?"
There are a few ways to adjust AC balance. The most precise is to make the adjustment while monitoring distortion with an analyzer. Best balance creates minimum harmonic distortion. Another method is to inject a sine wave at the amplifier input, connect a sensitive AC voltmeter across the two cathodes in that output channel, and adjust for minimum voltage. This adjustment should probably be checked every few months to accommodate tube and component drift, but it's not extremely critical. Just being able to make the adjustment when new tubes are installed is a step forward.
The dissipation of the 20K resistors is a function of the current passing through them. In this case, that can be reasonably extracted by noting that 425V is applied to the 5K potentiometer and 370V is measured at each anode. If we assume the adjustment is centered, that means there's 55V across 22.5K ohms (2.5K + 20K), which creates 2.44 mA. At that current, each 20K is dissipating approximately 120 mW. 2W is indeed lager than necessary from this perspective, but the larger rating might have been chosen to provide better stability over time or temperature. There are more factors involved here than just dissipation, and I would be very hesitant to question HK's usually excellent engineering decisions.
The CCS is probably indicative of a long tailed pair being used, but not necessarily. Again, a schematic would answer many questions. Incidentally, I'm not fond of using an IC for this. It's a shortcut that saves time and money, but which can also compromise ultimate signal quality in my opinion.
--------------------------
Buy Chinese. Bury freedom.
First? I honestly don't understand a fraction of what you wrote. I do understand elementary stuff like voltage drop across resistors.
Yes, if you read the documentation of the tubes4hifi website, they DO talk about 'long tailed pair'....
Several thousand in Kit Form have been sold. I don't know if that's the ST70 or the ST120 or combined.
I went to tube depot. The number of the REVILED 12AU7 is about 4x or 5x the the 12BH7 which tells me something.....not sure what.....Several others, like the 5965 or 5814 or 6189 are also listed in small numbers. Just gotta look harder than I'm willing to do.
I thought modern voltmeters were VERY high input impedance and did away with most of those loading effects? My Simpson Senior Voltohmist was a VOM, but that was 30 years ago...or more. I have a WESTON Selective Analyzer up on the shelf as an 'object'......It's a model# 5xx / type 3, if that means anything to anyone.....
To safeguard their intellectual property, you do NOT get the schemaitc without the REST of the kit. Or drawings of point-2-point or the complete parts list, for that matter. But you CAN get a 'sample' upon request. Which I have done. Also? A YouTube video of the first part of the assembly including the 'stuffing' of the driver board including the Current Regulators....a SS device.....
One day you'll have to let me explain semiconductor processing to you. Easilly as arcane as tube design / execution to the uninitiated. Just dressing UP to go into the fabrication area is a treat. Head-2-toe coverall with mask, safety glasses and neoprene or latex gloves........along with shoe coverings. NO PENCILS allowed. A class 10 room is the cleanest place anyone is ever likely to be. Dust NOT sterile......
The places Ive seen where they assemble spacecraft are NOWHERE near that good........
Too much is never enough
Yes, the 12AU7 has been used in a large amount of gear over the years. As a cathode follower or cathodyne inverter, it provides good service. However, some amplifiers have used it as a gain stage, and they often get away with this - sonically speaking - because A) distortion at the output of the amp equals the square root of the sum of the squares of THD generated within each stage, and B) negative feedback typically reduces the harmonics by a factor of 10. Both of these serve to reduce the damage on paper, but this doesn't mean you won't hear the shortcomings of the tube.
The schematic I posted shows a 5V drop across a 1M resistor. That implies it was measured with a meter having a 22M input impedance. The amplifier is from the '60s, so they probably used a VTVM for the measurement.
Sorry, I had to laugh about the intellectual property concept. We're talking about tubes designed many decades ago and circuits invented before most of us were born. Any manufacturer who thinks a long tailed pair on a Dynaco driver board contains intellectual property is seriously paranoid. The value in that product lies elsewhere.
About the CCS, it's necessary to understand that the RMS characteristics of the LTP are being modified in accordance with the frequency response of the IC, as well as by instantaneous response to the changing values of the tubes. I'm not a fan of mixing SS devices and tubes in the signal path, but if I were, this job would be much better managed by a MOSFET or two. It's also worth pointing out that the purpose of the CCS is to balance the LTP, but that eliminates the AC balance control, so the AC characteristics of the output tubes can no longer be manually balanced. Instead, we have to rely on the application of feedback and its attendant limitations.
I'm no expert when it comes to SS processing, but I did provide technical writing services to a company that manufacturers the fabs themselves. Always had to wear a bunny suit to avoid contaminating the equipment prior to it being shipped out. Spacecraft? Some of the products I helped design and test over the last fifteen years at my last position are now on Mars. I guess that's why I got a chuckle from the intellectual property statement. : )
Good luck with your search, there are many routes to success with tubes, old and new.
--------------------------
Buy Chinese. Bury freedom.
Another search this evening located two schematics for the VTA ST-70 driver board. One uses the LM334 current source in the LTP. Tube options include 12AU7 and 12BH7. The other schematic appears to be older. It also uses the LTP configuration, but there's a balance pot instead of a CCS. Tubes are all 12AT7. This is more traditional, it eliminates the risk of the LM334 injecting its own personality, and this one would be my choice if I was buying the kit.
--------------------------
Buy Chinese. Bury freedom.
Not that you have anything beyond intellectual interest in the VTA circuit, but in a purely self-serving way, I sure wish you had one. I have no deep objective knowledge of what improvements my (pre-CCS) 12AU7 based VTA might possibly need to improve its objective and / or subjective performance. But I can give you a list of the subjective things I've tried in my VTA, which include:
Increasing the value of the capacitance in the "on board" B+ caps over stock.
Using Panasonic, rather than Xixon electrolytics on the board.
Using a variety of well regarded resistor brands and types instead of Xicon metal films, including Dale RN, PRP, and Riken Ohm carbon films, all the same values as the stock schematic calls for.
Several different brands of NOS 12AU7s
12BH7s
Upgraded coupling caps
Output tubes tried included Gold Lion KT-77, Russian 6P3S-E, and original (not reissue) Mullard EL-34
Despite the above, I still find the amp "musically uninvolving" with the only significant change being the use of the 6P3S-E, which were the best of the tubes I tried.
I email another VTA ST-70 owner who has the later, CCS version. He's taken things further. In addition to the some of the above, he's increased and decreased the amount of feedback, removed the additional "on board" electrolytics, and attempted to change the circuit to use the more linear 12AX7, to name a few that I can recall.
Although we both have very different speakers and musical tastes, we're in agreement in that we both have other amps, both tube and solid state, which we enjoy listening to more than the VTA ST-70.
You can't please all the people all the time, so I thought I'd post my results to add some balance to this thread.
Wow, you've really put in a lot of time and work. I've also owned amplifiers that were "musically uninvolving," but the vast majority of those were solid state. :)
Back in the day, I was privileged to hear many different tube amplifiers - Scott, Fisher, Dynaco, Heath, McIntosh, etc. I never cared for the sound of the ST-70, but that could have been the result of many things that have been addressed in more modern times, such as the power supply, coupling caps, etc. Nevertheless, when I do hear complaints about this amplifier, I'm reminded of my experiences years ago.
Are you familiar with anyone's work to run the ST-70 in pentode mode, rather than UL? This requires slightly more modification than the usual triode conversion (which I believe is entirely ill-advised), but it might remedy some of the sonic grievances. I have always thought the reduction in output stage gain created by UL operation might be more of a detriment in this particular amplifier than a benefit. In fact, it appears to be one of the reasons aftermarket driver circuitry has gradually morphed into something that bears almost no resemblance to the original product.
Converting the ST-70 to pentode might be counterintuitive, but it's been my impression through the years that pentode operation is more dynamic and engaging. Conversion of the ST-70 output stage would be relatively straightforward, owing to the characteristics of the EL34. The feedback network could remain as-is for initial tests or revised to provide the more typical 20 dB of NFB. In any event, it's not too difficult to undo if the results aren't beneficial.
--------------------------
Buy Chinese. Bury freedom.
...It really wasn't a lot of time or work. I bought a blank board from Tubes4hifi and stuffed it with the previously mentioned caps and resistors. I didn't mean to imply that I had tried, and then removed all the Xicon components.
To a pure objectivist, I certainly wasn't doing any worse by using (mostly) Mil-Spec Dale RN's over the Chinese Xicons VTA was using, even if they "sounded" the same. The exception to the Dales are Riken as grid stops, Kiwame on the plates, and a few PRP in the B+
The same with the Panasonic electrolytics I built it with. They have low ESR, are rated at 105 ºC, and had higher capacitance than the Xicon value VTA used. Possibly a lateral move, perhaps, but certainly no worse.
The point of all that is, I feel that I've given the stock circuit a fair chance, and then tweaked via tube rolling, if not the passive components.
I'm not familiar with a pentode input stage, but it sounds interesting.
Oddly enough (or perhaps not) my most involving amp is SS, but it's a (Nelson Pass) First Watt F1J, so it's hardly a typical SS circuit. If the late Art Dudley's Stereophle review is accurate, I'd need to use a good 2A3 SET circuit to better the F1. Maybe it's time to explore that a bit further.
"I'm not familiar with a pentode input stage, but it sounds interesting. "
I'll assume you meant "pentode output stage," being as that was what I was describing.
Essentially, you would disconnect the screens from the screen taps on the output transformers and connect the screens at each channel together. The screens at each channel would then be connected to B+ through a common screen grid resistor with a value as indicated in the data sheet. Decoupling at the screen pins is not required in this configuration.
The above modification places the amplifier output stages into pentode mode. The gain of the output stage will increase considerably, as will available output power. Below is a chart showing this operating condition with voltages similar to those in the ST-70.
Several things to note...
The value of the common screen grid resistor is 1K. It should be rated 4W or more. Maximum output in the chart is 55W. This is based on the point at which the tubes produce 5% distortion. 20 dB of NFB will reduce that to 0.5%, and it will be less if output power is constrained to the original 35W rating. At listening levels of a few watts, distortion won't be audible at all, except maybe to a golden-eared few. The additional global feedback can be accomplished by adjusting the series resistor in the feedback loop, and this will help to normalize the gain of the amplifier relative to its previous UL operation. All else will stay the same. No changes are needed to the bias circuit or driver, and unlike triode conversions, the same OPT impedance as is used for UL is appropriate for this mode.
You've tried everything else. This is all that's left, and it's a modification that's easy to implement and easily reversed if you don't like the sound.
--------------------------
Buy Chinese. Bury freedom.
My mistake. I read what you wrote, but was focused on something else.
I'll save what you wrote, and the next time the VTA is on the bench (which may be a while), I'll try connecting the output in pentode. You're right, I've got nothing to lose by trying it.
As always, thanks for your time and expertise, TK.
Well, I know what I said, but using EL34s and those output transformers in straight pentode mode is much more than just a last ditch effort. I want to be clear about this. It's exactly the way I'd go if not for the vintage iron I'm currently using up, and I'd do it with or without an actual ST-70 amplifier. Some of the replica Dynaco transformers on the market look very good, and I wouldn't hesitate for a second to build a straight pentode amp around them. If you ever get a chance to do this, I think you'll be pleasantly surprised.
--------------------------
Buy Chinese. Bury freedom.
Thank you for the suggestion. It's the first time I've had any enthusiasm for doing any work on the amp in quite a number of years.
My friend would email me, telling me all the things he'd been trying, to coax music out of his amp, to no avail, and I mentally kept pushing it further and further down my list of interesting projects. "Pleasantly surprised" would be nice.
Thanks again, TK!
We can walk through this while you're doing the work if you want. Offline is OK too. Just say the word.
--------------------------
Buy Chinese. Bury freedom.
As I mentioned to you elsewhere, I have a PP 6BQ5 build which interests me more, but after that's done, I'll put the ST-70 on my list.
Thank you for the offer of help, that's greatly appreciated.
Oh no, not the 2a3. Just when I thought I was out, you pull me back in.Now I have to put back my Bottlehead bookmark.
Edits: 07/01/21
As the Eagles sang, "You can check out any time you want, but you can never leave".
The most sobering aspect of a 2A3 SET for me is its low power, and yet their siren song lures me toward the rocks.
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: