Some agreement on reasonable affordable and well performing design builds?

[devilsindetails]

Here's my conundrum...been reading on this site for years attempting to determine if there are agreed upon topologies for an excellent sounding tube-based phono stages, pre-amps and S/E amps.

While I certainly appreciate and marvel at the infinite array of topologies and designs I feel many are either fairly esoteric, expensive to build or subject to heated debate regarding their pros or cons.

For those of us who have the interest, reasonable expertise and curiosity, but not the budget, to engage in full-out experimentation or to indulge in "ultimate" builds we're looking for designs that would get us up and running quickly with simple, yet quality, circuits and builds.

The Tubes for newbies thread has been wonderful as a bracer.

So here are my basic design parameters:

Amplifier: SET/5 watts and under
Phono-stage MM and MC capable

Budget: Next to nothing: iron, OTs salvaged from various EL84 and 6GW8 based chassis (chassi?-not sure what the plural of chassis is)

Tubes: Rectification, gain. driver, output tubes limited to those on hand or easily sourced for minimum rubles. 6CA7s/12AX7s (7025), 6SN7s, 6GW8s EL84s etc.

Caps/resistors/sockets etc: NEW of course

Chassis: own construction

Skill level: intermediate


Surely there are some agreed upon phono-stage. pre-amp and amplifier designs that achieve excellent performance relative to their overall cost.

It's always confusing, for example, to find so many poo-pah-ing 12AX7s in phono stages when RCA manuals have these in their circuit designs or dismissive comments about simple-low powered 6 based topologies.

I have compared many of the simple 6 based SET amps with more elaborate, esoteric and powerful builds, and while there are clearly sonic improvements not enough in my mind given the difference in overall cost or ease of design and construction.

Thanks in advance for any direction you might be able to provide

 

[SY]

It's always confusing, for example, to find so many poo-pah-ing 12AX7s in phono stages when RCA manuals have these in their circuit designs

Remember, the RCA diagrams were mostly for cheap, mass market stuff. The phono stages they show are the finest low-bid cheap stages they could make in 1955, i.e., the performance is pretty lousy. You can make a good phono stage using 12AX7, but it won't be a simple one and it won't look like the ones in the RCA manual.

Doing this on a shoestring budget? I wouldn't use tubes. They work superbly with good design and execution, but good design and execution are not cheap.

 

[DF96]

I assume your salvaged OPTs come from decent amplifiers and not radio sets? Otherwise you may be disappointed. A good OPT is expensive, which is one reason why SS can be cheaper.

We all have our prejudices. Some don't like 12AX7. Others don't like SE.

 

[planet10]

Amplifier:

The RH84 is designed to give good results with low budget iron. The console you got the IRELs from should have sufficient iron. EL84 outputs, calls for a 12AT7 on the front. You could likely use 12AX7, but there is a synergy between 12AT7 and EL84 (SY's Red Light District uses this combo, and it is a very good option for Eli's El Cheapo). It is pretty simple to build, and a salvage build should be doable for <$50 (much of that for a 10H 100mA choke). When we went all out -- all Solen power supply caps, mono construction, bought hammond chassis, and nice 5-way posts we were still <$150 cash outlay (free iron from 2 identical Grundig consoles). 3.5-5 watts depending on B+ you end up with.

Preamp... just use a 100k pot on the RH84 (and a switch if needed). AIFF & Line Level out of my iPod with the stereo RH84 made a quite decent budget hifi when coupled with reasonable efficient vintage drivers or a set of Fonken, A126, or Frugel-Horn.

Phono: The RCA phono that SY poo-poos can be very good, particularily if you build Eli's version. Granted the one Chris built that i am well familiar with had a power supply that was beefier than the ones in the RH84 monoblocs. Key is to match the EQ parts (true for any RIAA). If you build it in the same chassis or keep real close, you can skip the output bufer and still get good results.

Careful salvaging and parts aquirement should allow you to put together something quite enjoyable for a seriously frugal-phile(tm) budget.

dave

 

[TerryO]

The OP asks:

Some agreement on reasonable affordable and well performing design builds?

In the 35+ years that I've been involved with Audio, I've never seen much, if any, agreement on design. I might add this is among the designers themselves, leaving out many talented hobbyists that also have their own ideas.

You need to narrow the field down considerably before any sort of reasonable responses can be offered.

Here's just a few questions:
What speakers are you using and how efficient are they?
What type of Music do you listen to?
How big is your listening room?
What is the normal sound pressure level that you use?
Can you Bi-Amp your speakers?
Are you going to use subwoofers?
What is a realistic budget for you?

Best Regards,
TerryO

 

[Eli Duttman]

MC is OUT, when the budget is low. Don't argue!

Some MM carts., along with some SUTs, don't work well into the high C_Miller of the 'X7. IIRC, that's SY's primary objection to the type. A solution is to pick a MM cart., that's somewhat bright, say an Audio Technica.

The 2 primary weaknesses of the original RCA setup (schematic attached) are wretched drive capability and mediocre bass extension. I've uploaded a tweaked version of the RCA setup that deals with those issues. Contact Jeff Yourison for his experience in building the tweaked version. A good performing, low cost, PSU can be built around an inexpensive Allied power trafo and SS regulator ICs. Buy a matched pair of 220 nF. Soviet surplus K40 PIO caps., for the interstage couplers and don't look back.

Triode wired SE EL84s are good for approx. 2 WPC. While I prefer 6V6 family tubes in inexpensive SE amps, there's nothing wrong with the EL84. Wire your sockets up for the Russian 6Π15П (6p15p), AKA SV83, and increase your options. Sockets set up for the 6Π15П work fine with EL84s, but the other way around fails. The 6Π15П-EB (6p15p-ev) makes a FINE triode.

Check this thread out. You should be able to get some hints. An important point is the use of the 12AV7 (a 5965 is good too). That yields sufficient gain, along with providing the high gm and low R_P needed to drive a triode mode "final". The 'X7 can't fight its way out of a wet paper bag.

 

[Miles Prower]

It's always confusing, for example, to find so many poo-pah-ing 12AX7s in phono stages when RCA manuals have these in their circuit designs...

The A Number One problem with the 12AX7 is its EEEEEENORMOUS plate resistance. It's in the same league as a small signal pentode. However, it's not a pent, and like every other triode, it needs to work into as light a plate load as possible. That means some serious DC rail voltage (see attached).

Less than satisfactory performance from this type always goes back to passive plate loads that are too small, and barely adequate.

If you don't have the voltage reserve, then your only remaining option is to bite the bullet, and employ that Spawn of the Devil, and include a solid state, active plate load. Otherwise, you'll be operating into a too-vertical loadline.

 

[SY]

The A Number One problem with the 12AX7

You got that right. Number two is high input capacitance- too high for many MMs (like mine). Noise and distortion performance of that stage is... ummm... wellllllll.......

I've built a dozen different versions of that circuit over the years, some better than others, none really satisfactory.

 

[Miles Prower]

You got that right. Number two is high input capacitance- too high for many MMs (like mine). Noise and distortion performance of that stage is... ummm... wellllllll.......

High Cmiller is going to be problematic, but you're always gonna have that with any triode stage. The above design was meant for an LTP driven from the usual suspects: CD players, DVD players, tuners, and so forth, and Ci isn't a problem here.

As for the MM problem, I've given it some thought, and about the only solution is more Spawn of the Devil: good quality, low noise, low Ci op-amp, or JFET source follower with active tail load ahead of the triode.

 

[devilsindetails]

Remember, the RCA diagrams were mostly for cheap, mass market stuff. The phono stages they show are the finest low-bid cheap stages they could make in 1955, i.e., the performance is pretty lousy. You can make a good phono stage using 12AX7, but it won't be a simple one and it won't look like the ones in the RCA manual.

Doing this on a shoestring budget? I wouldn't use tubes. They work superbly with good design and execution, but good design and execution are not cheap.

Well I'm certainly not an expert but it seems to me that I'm not proposing anything that's so utterly unachievable as to wave me off tubes all together.

What I was attempting to get at, albeit obviously so clumsily, and without the detail many would need to render, what they would consider to be informed opinions, was this idea that everything had to be uber-$$$, uber-complex and uber high-end to attain very credible performance.

Having been a die hard reasonably high-end S/S aficionado for decades I was struck by the fact that my multi-thousand dollar system wasn't multi-thousands of times better sonically than a very simple little S/E 6GW8 integrated amp driving some sensitive but more importantly tube-friendly speakers.

Did it have as much bass extension...no....or top-end? Probably not. But for a few hundred dollars including DIY speaker build certainly within striking distance.


The 12AX7 and its variants are uber=ubiquitous in so many designs past and current, (commercial and DIY) that this tube can't be all bad as a starting point for an economical and well conceived topology under the right direction. If not the 12AX7 then what about other economical and available tubes?

After hearing and reading about the 26 and 27 based designs and the associated bugaboos inherent with these tubes and the expense associated with their procurement there has to be something more budget and sourcing friendly.

I understand the cost efficiencies utilized by mass manufacturers in speakers, TT`s, amps etc but when tweaked or reconfigured as many vintage pieces have been over the past and freed from their original constraints yield amazing results...Lenco Idler drives, Philips Drivers and the EL84 come immediately to mind...

Am I completely out in left field to think that credible tube-based performance on a real tight budget can`t be realized?

I`m not suggesting using old caps, resistors, wire or the like...just keeping investment down to the bare and essential minimum.

 

[devilsindetails]

I assume your salvaged OPTs come from decent amplifiers and not radio sets? Otherwise you may be disappointed. A good OPT is expensive, which is one reason why SS can be cheaper.

We all have our prejudices. Some don't like 12AX7. Others don't like SE.

Thanks DF...I consider them to be credible candidates as they were pulls from Telefunken, RCA, Phillips and other stereo console chassis using EL84 and 6GW8 topologies.

 

[SY]

What I was attempting to get at, albeit obviously so clumsily, and without the detail many would need to render, what they would consider to be informed opinions, was this idea that everything had to be uber-$$$, uber-complex and uber high-end to attain very credible performance.

...
Am I completely out in left field to think that credible tube-based performance on a real tight budget can`t be realized?

Depends on your definition of "tight."

You can use solid state to get a VERY good level of performance at relatively low cost. You can use tubes to get a superb level of performance and quite a bit more cost and complexity. For example, you can get around the input capacitance issue of triodes by cascoding- but then you need an ultra-high performance power supply because of the next-to-zero power supply rejection of cascodes. Your BOM has just doubled to tripled.

Or you can use the same circuits used in cheap 1950s Magnavox consoles. JMO, but I've built a LOT of phono stages over the years, so "voice of experience."

 

[devilsindetails]

Amplifier:

The RH84 is designed to give good results with low budget iron. The console you got the IRELs from should have sufficient iron. EL84 outputs, calls for a 12AT7 on the front. You could likely use 12AX7, but there is a synergy between 12AT7 and EL84 (SY's Red Light District uses this combo, and it is a very good option for Eli's El Cheapo). It is pretty simple to build, and a salvage build should be doable for <$50 (much of that for a 10H 100mA choke). When we went all out -- all Solen power supply caps, mono construction, bought hammond chassis, and nice 5-way posts we were still <$150 cash outlay (free iron from 2 identical Grundig consoles). 3.5-5 watts depending on B+ you end up with.

Preamp... just use a 100k pot on the RH84 (and a switch if needed). AIFF & Line Level out of my iPod with the stereo RH84 made a quite decent budget hifi when coupled with reasonable efficient vintage drivers or a set of Fonken, A126, or Frugel-Horn.

Phono: The RCA phono that SY poo-poos can be very good, particularily if you build Eli's version. Granted the one Chris built that i am well familiar with had a power supply that was beefier than the ones in the RH84 monoblocs. Key is to match the EQ parts (true for any RIAA). If you build it in the same chassis or keep real close, you can skip the output bufer and still get good results.

Careful salvaging and parts aquirement should allow you to put together something quite enjoyable for a seriously frugal-phile(tm) budget.

dave

Thanks Dave...i think you get where I am coming from...there are a lot of enthusiasts such as myself who are weary of apologizing for the lack of financial wherewithal to engage in the pursuit of the penultimate performing design.

I am not foolish enough to believe I could approximate the performance of an Akido design or many other breathtaking topologies with an assortment of twisted barb wire, a couple of light bulbs and a battery charger.

I`ve watched a friend absolutely twist himself into knots over securing an extra couple of khz in freq response or bottom end and maxxing out his credit limit to boot for results that are far from overwhelming.

Much like some of the recent complaints regarding certain speakers I am not one to think a single 12`` Marsland PA driver with a nearly non-existent motor, zero cone excursion, and other limitations is going to equal a 12`Tannoy Monitor Gold Dual Concentric driver.

But could it be massaged to punch above its weight and limitations with some clever and well-thought out enclosure design and some auxilary driver and crossover support...I think yes. And I feel the same could be done with what I am looking for regarding available, plentiful, economical and reasonably well-designed tubes.

 

[devilsindetails]

The OP asks:

In the 35+ years that I've been involved with Audio, I've never seen much, if any, agreement on design. I might add this is among the designers themselves, leaving out many talented hobbyists that also have their own ideas.

You need to narrow the field down considerably before any sort of reasonable responses can be offered.

Here's just a few questions:
What speakers are you using and how efficient are they?
What type of Music do you listen to?
How big is your listening room?
What is the normal sound pressure level that you use?
Can you Bi-Amp your speakers?
Are you going to use subwoofers?
What is a realistic budget for you?

Best Regards,
TerryO

Hi Terry, yes my vagueness is probably my own undoing. I believe my speakers are quite sensitive (Isophons, Telefunkens, Foster F104s, a couple of Coral variants, several of the 1950s and 1960s Philips and some early 1960s Wharfedales.)

My understanding is that sensitivity is but one yardstick and that the stability of the impedance load is actually more important to it`s ultimate utility as a reproducer for tube applications. But, these drivers are all arguably well suited for tube applications.

My musical tastes run the gamut classical, jazz, country-western and more.

Listening room dimensions would typically be around 20 ft by 14 ft.

No bi-amplification and no-subs.

My mantra would be ``simplicity and deliberate understatement with finesse``

 

[devilsindetails]

MC is OUT, when the budget is low. Don't argue!

Some MM carts., along with some SUTs, don't work well into the high C_Miller of the 'X7. IIRC, that's SY's primary objection to the type. A solution is to pick a MM cart., that's somewhat bright, say an Audio Technica.

The 2 primary weaknesses of the original RCA setup (schematic attached) are wretched drive capability and mediocre bass extension. I've uploaded a tweaked version of the RCA setup that deals with those issues. Contact Jeff Yourison for his experience in building the tweaked version. A good performing, low cost, PSU can be built around an inexpensive Allied power trafo and SS regulator ICs. Buy a matched pair of 220 nF. Soviet surplus K40 PIO caps., for the interstage couplers and don't look back.

Triode wired SE EL84s are good for approx. 2 WPC. While I prefer 6V6 family tubes in inexpensive SE amps, there's nothing wrong with the EL84. Wire your sockets up for the Russian 6Π15П (6p15p), AKA SV83, and increase your options. Sockets set up for the 6Π15П work fine with EL84s, but the other way around fails. The 6Π15П-EB (6p15p-ev) makes a FINE triode.

Check this thread out. You should be able to get some hints. An important point is the use of the 12AV7 (a 5965 is good too). That yields sufficient gain, along with providing the high gm and low R_P needed to drive a triode mode "final". The 'X7 can't fight its way out of a wet paper bag.

Thanks ELI, like many of the others here, who take the time to steer relative novices like myself clear of wrong-headed assumptions, I appreciate your willingness to weigh in.

This thing about the 12AX7 keeps coming up and I can`t help wondering why it`s used so often if it`s so dreadful. Your explanations are extremely helpful and DIRECT.

What about the 6sn7s, of any use to me in this regard.?

Any thoughts about the use of the 6GW8 with 6CA7 rectification?

Thanks..Leo

 

[devilsindetails]

The A Number One problem with the 12AX7 is its EEEEEENORMOUS plate resistance. It's in the same league as a small signal pentode. However, it's not a pent, and like every other triode, it needs to work into as light a plate load as possible. That means some serious DC rail voltage (see attached).

Less than satisfactory performance from this type always goes back to passive plate loads that are too small, and barely adequate.

If you don't have the voltage reserve, then your only remaining option is to bite the bullet, and employ that Spawn of the Devil, and include a solid state, active plate load. Otherwise, you'll be operating into a too-vertical loadline.

Thanks Miles. So what`s a good candidate in its stead keeping in mind easy availability, economy and synergy and simplicity?

 

[kevinkr]

You might want to take a look at the 6DJ8 and its variants. Employed properly it is capable of very good performance, it's cheap, low noise, has high transconductance and reasonable mu. Rp is very low - low enough in fact that it will drive loads that many other tubes will not.

Current production 6DJ8 from JJ sound pretty good, and PQ Amperex (expensive) are more or less stellar IMLE.

In phono stages its moderate mu can be overcome with a cascode input stage which works much better than I expected based on very recent experience.

 

[devilsindetails]

You might want to take a look at the 6DJ8 and its variants. Employed properly it is capable of very good performance, it's cheap, low noise, has high transconductance and reasonable mu. Rp is very low - low enough in fact that it will drive loads that many other tubes will not.

Current production 6DJ8 from JJ sound pretty good, and PQ Amperex (expensive) are more or less stellar IMLE.

In phono stages its moderate mu can be overcome with a cascode input stage which works much better than I expected based on very recent experience.

Thanks Kevin as it happens I have more than a few NOS on hand...any concensus on this tube...anyone? Eli?

 

[Eli Duttman]

Kevin,

JJ is having their, unfortunately, usual QC problems and recent examples of the JJ E88CC have been unsatisfactory. It's a shame, as JJ can make good tubes. Your guess is good as mine, as to why they are shipping junk.

For phono section service, the safe 6922 variant is Russian 6Н23Π-ЕВ (6n23p-ev), including the tubes labeled ElectroHarmonix (EH) 6922.

----------------------------------------------------------------------------------------------------------------------------------------------

All,

Adequate gain can be touch and go using "simple" common cathode 6922s in phono preamps. OTOH, the 6GK5 also has high gm and low R_P, but it's a high μ type, unlike the medium μ 6922. So, adequate gain ceases to be a problem. The down side to the 6GK5 is a single triode per 7 pin mini bottle.

FWIW, I'll vigorously defend the MOSFET buffered, grid leak biased, (possibly CCS @ 1 mA. loaded) 'X7 section 2nd gain block. Those folks better than me at crunching the numbers can (PLEASE) work up an EQ network appropriate to the 6GK5. IMO, a 6GK5 1st gain block and a buffered 'X7 2nd gain block will be an excellent setup. Gain will be quite sufficient and C_Miller is not a problem. Use any MM cart. or LOMC SUT that tickles your fancy.

 

[jono1]

What about cap-coupled parafeed, as a low-cost alternative?