ToolShedAmps Modified 4S Tube Pre-Amp Build

[ToolShedAmps]

Looks like up to 22db gain. I ran some quick simulations using the same parameters. Average source input of 1Vrms, 1kHz sine wave. The distortion spectrum looks better, mainly 2H, but still rather high at 1.8%.

 

Ok then, this is Version #1, straightforward while still retaining the pot outboard of the gain-stage. Now lets rub it a hair. Move the pot to the input, hang a 1 Meg resistor off the output to ground, throw out the 1K bias resistor and replace it with a TL783 (keep the bypass cap as it provides a good AC path to ground) use a 1/2 watt 240 Ohm resistor to set the quiescent current draw @ 5.2mA........Now what do you have?

 

Matt.

Edited November 11, 2015 by ToolShedAmps

[xxJPMxx]

So current source the cathode at 5.2mA, fully AC bypassed. Volume moved to front end. Same gain 22db. 1kHz sine wave input. Output 1Vrms  is < .2% THD.

 

[mike stehr]

I call them effect boxes and people get offended sometimes. If you are intentionally adding sound (2H is an octave above fundamentals) to your recordings it's not far from the truth. I have no problems with people doing this, some hi fi guys do though. I try to design stuff with low distortion myself, but there will always be some......second harmonic is the most harmonious.

 

This linestage project would be great for lessoning the harshness of some CDs, mp3s, etc. The linestage, (with it's second harmonics) would be nice for use in front of a SS amplifier, that may have been harsh/bright sounding. Where the tube linestage helps smooth out the overall sound of the system...something like that...

 

But when using a tube amplifier, like a single ended class A that has it's own 2nd harmonic distortion, and then adding a tube linestage...would that seem like too much of a good thing?

I used a stepped attenuator nowdays, because I don't need the extra gain when using a CD player or my Raspberry Pi2/HiFiBerry DAC+.

 

Sometime ago, I lashed up a pair a Aikido linestages, right and left. The circuit has a sort of self cancelling of noise from the way it's designed. It worked/sounded rather well actually, but I didn't need the extra gain and they eventually got shelved. Had to really watch out for microphonic 6SN7s, some of them really howled in circuit with no signal.

Edited November 11, 2015 by mike stehr

[xxJPMxx]

FWIW the CCS in the cathode doesn't really help or hurt things. If you want to linearize it you would have to CCS the plate. I actually like LED's at the cathode, they are cheaper than capacitors.

[ToolShedAmps]

So current source the cathode at 5.2mA, fully AC bypassed. Volume moved to front end. Same gain 22db. 1kHz sine wave input. Output 1Vrms  is < .2% THD.

 

Now this is the pre-amp we are going to build. I think there are a fair number of DIY'ers who can handle this build without too many problems. If they have any, hopefully everyone here is willing to give them a hand. I will modify the existing Power Supply design to increase voltage to an acceptable level for this circuit. Additionally, I will provide a 2nd Power Supply design of my own (for those DIY'ers who would like to extract every iota of performance from this circuit) which offers approx. 0.4mV of ripple at stated current.

[ToolShedAmps]

FWIW the CCS in the cathode doesn't really help or hurt things. If you want to linearize it you would have to CCS the plate. I actually like LED's at the cathode, they are cheaper than capacitors.

I like the "sound" of a purely resistive load. And, I'm pretty sure everyone else will too. By "forcing" the quiescent operating current out of the tube, we can then "tweak" the supply voltage till we achieve the most "pleasing" operating point. Additionally, because I'm not a "rich" guy by any stretch of the imagination, by using a current "sink" to bias the tube we have the added benefit of NOT HAVING TO BUY "MATCHED" TUBES. That's right, as long as the tubes are of the same construction and known to be "good", we can just stick them in there and enjoy. This will be a "Killer" preamp. One that I'm sure everyone will really take satisfaction in owning/building. 

 

Matt.

[ToolShedAmps]

So, over the next couple of days, we will put together a "Bill Of Materials" for those of you who would like to build this Pre-Amplifier. The "BOM" will be divided into three different parts: 1) Circuit, 2) Less Expensive Power Supply (Tweaked supply from the original project) and 3) More Expensive Power Supply (my own design, typical of my amplifiers). Additionally, some suggestions will be provided for the other necessary hardware and chassis parts.

 

Cheers!

 

Matt.

[xxJPMxx]

That's cool, this isn't a phono pre, we don't want it too linear/sterile and a CCS plate load might do just that. I think the distortion level and spectrum look good now.

 

As far as setting bias goes here is how I look at it. Transconductance and plate resistance vary from tube to tube, mu is the most consistent parameter. If you set current through the device you will still have different voltages at the plate and cathode due to the slight differences in plate resistances from tube to tube. The plate load resistor and internal plate resistance form a potential divider, the current set by the CCS will develop different voltages due to the different plate resistance. So to summarize current will remain constant (it does anyway because it's class A) but voltage will change due to plate resistance differences. I think I did some experiments with this by setting up a test jig. I set current at cathode via CCS, and set a fixed plate resistor. I then placed different tubes in and measured different plate voltages at quiescent operation. This was on a regulated bench power supply. If anyone wants to try this too that would help to make sure I am not crazy. I will try and set up a test jig later this evening.

 

 

Matt, I think Mike and William bring up a good point in that someone with a very sterile SS setup might actually want 1-2% THD added. Others who run SET like amps w/ 5% THD might want the preamp to be cleaner. Maybe as a prerequisite they could inform you of their setup and maybe they can build or buy it either way, volume control in the rear or in the front. Both circuits are fine but have different characteristics.

[tube fanatic]

 

 

Alright, here we go again, I guess I just like a certain amount of "SAG" in my PS.

 

What are you measuring for a current change in the PS during signal?  Class A operation generally doesn't sag since the average current draw doesn't change.

 

 

 

 

 

And thanks for the schematic!!  Any valve choice is better than the 12AU7 IMHO     

 And, you are absolutely correct. ALL Class-A, Single-Ended designs draw "constant current" regardless of load, that's why forum members should be informed to NEVER turn on a Single-Ended Class A amplifier without speakers attached as you may damage your output transformers.

 

Matt.

 

Matt, I need to take exception to your statement as I don't want users of single ended amps to have any undue concerns. The constant current drawn by class A single ended amps is a dc condition only.  Under zero signal conditions (i.e. no music playing) it is of no consequence whatsoever whether the speakers are connected or disconnected.  Both SETs and SEPs can be open circuited with no damage to either the tubes or opts.  Under signal conditions (such as someone accidentally forgetting to reconnect their speakers), a triode and its opt is very unlikely to be hurt, and pentode amps using voltage feedback "should" be able to handle it as well.  Obviously, one should always check things over before firing up an amp.  But if, as we all occasionally do, we forget to connect the speakers, there's really no need to panic.

Maynard

[ToolShedAmps]

That's cool, this isn't a phono pre, we don't want it too linear/sterile and a CCS plate load might do just that. I think the distortion level and spectrum look good now.

 

As far as setting bias goes here is how I look at it. Transconductance and plate resistance vary from tube to tube, mu is the most consistent parameter. If you set current through the device you will still have different voltages at the plate and cathode due to the slight differences in plate resistances from tube to tube. The plate load resistor and internal plate resistance form a potential divider, the current set by the CCS will develop different voltages due to the different plate resistance. So to summarize current will remain constant (it does anyway because it's class A) but voltage will change due to plate resistance differences. I think I did some experiments with this by setting up a test jig. I set current at cathode via CCS, and set a fixed plate resistor. I then placed different tubes in and measured different plate voltages at quiescent operation. This was on a regulated bench power supply. If anyone wants to try this too that would help to make sure I am not crazy. I will try and set up a test jig later this evening.

 

 

Matt, I think Mike and William bring up a good point in that someone with a very sterile SS setup might actually want 1-2% THD added. Others who run SET like amps w/ 5% THD might want the preamp to be cleaner. Maybe as a prerequisite they could inform you of their setup and maybe they can build or buy it either way, volume control in the rear or in the front. Both circuits are fine but have different characteristics.

I COMPLETELY agree with everything you just said. That is why I presented the design the way I did, Version #1 is the design that would pair best with SS amplifiers (Higher distortion, mostly 2nd order), and the "Revision" which, to be honest, is actually MY Pre-amp design (actually revised for Version #1 to be more compatible with SS Amps). So, no more "funny business". DIY'ers can build the Pre as either V1 for use with SS Amps or V2 for use with S.E. Tube amps (During the build, we will allow for connection points that the DIY'ers can use to CHANGE their build from V1 to V2 in the event they buy or build a Tube Amp).

 

As for the comment about the voltage change depending on a "specific" tubes' plate resistance, I agree, however, we as designers (just like voters) have to make informed decisions and select the topology (or candidate) that most closely relates to our Utopian ideals. What I have stated regarding non-matching tubes has been proven by my own empirical data over the course of many years now. I started using the lowly LM317 as a current sink several years ago, and in fact, they still exist in several of my old builds, none the worse for wear.

 

So, we will get on to the Power Supply discussion shortly, perhaps later this afternoon.

 

Matt.

[ToolShedAmps]

 

 

 

Alright, here we go again, I guess I just like a certain amount of "SAG" in my PS.

 

What are you measuring for a current change in the PS during signal?  Class A operation generally doesn't sag since the average current draw doesn't change.

 

 

 

 

 

And thanks for the schematic!!  Any valve choice is better than the 12AU7 IMHO     

 And, you are absolutely correct. ALL Class-A, Single-Ended designs draw "constant current" regardless of load, that's why forum members should be informed to NEVER turn on a Single-Ended Class A amplifier without speakers attached as you may damage your output transformers.

 

Matt.

 

Matt, I need to take exception to your statement as I don't want users of single ended amps to have any undue concerns. The constant current drawn by class A single ended amps is a dc condition only.  Under zero signal conditions (i.e. no music playing) it is of no consequence whatsoever whether the speakers are connected or disconnected.  Both SETs and SEPs can be open circuited with no damage to either the tubes or opts.  Under signal conditions (such as someone accidentally forgetting to reconnect their speakers), a triode and its opt is very unlikely to be hurt, and pentode amps using voltage feedback "should" be able to handle it as well.  Obviously, one should always check things over before firing up an amp.  But if, as we all occasionally do, we forget to connect the speakers, there's really no need to panic.

Maynard

 

Hi Maynard,

 

We can agree to disagree. I've seen a couple of OT's go "open" in my time, that's why many old timers still put an 8 Ohm/12 Watter across the outputs "internally". 

 

Matt.

[xxJPMxx]

I never sweat low powered tube amps with an open load even under signal conditions. I don't practice it; i.e. i always have a load hooked up but in theory most windings insulation are rated for over 1kV, anything with a plate voltage of under 500v can't swing that high to cause shorting.

[ToolShedAmps]

I never sweat low powered tube amps with an open load even under signal conditions. I don't practice it; i.e. i always have a load hooked up but in theory most windings insulation are rated for over 1kV, anything with a plate voltage of under 500v can't swing that high to cause shorting.

Good practice is good practice, you never know when someone in the DIY community is going to salvage a paper and shellac tranny out of some old console and "forget" to hook up the speakers.

[xxJPMxx]

Just an idea since you are doing two octal sockets. I messed around with a 6V6 for a linestage in the past. You could also offer a brute force linestage.  Less gain and plenty of current to drive heavy capacitive loads, lower rp than the 6J5 too. With 1Vrms input you get 8.5Vrms out @ .2% THD. Not too shabby. Also if your cheap like me you could even use some cheap tv sweep tubes.

 

Edit: R1 and R2 is my model of a volume pot so use appropriate grid leak and volume pot.

 

 

 

Edited November 12, 2015 by xxJPMxx

[ToolShedAmps]

Just an idea since you are doing two octal sockets. I messed around with a 6V6 for a linestage in the past. You could also offer a brute force linestage.  Less gain and plenty of current to drive heavy capacitive loads, lower rp than the 6J5 too. With 1Vrms input you get 8.5Vrms out @ .2% THD. Not too shabby. Also if your cheap like me you could even use some cheap tv sweep tubes.

 

Edit: R1 and R2 is my model of a volume pot so use appropriate grid leak and volume pot.

 

 

 

That looks pretty interesting, I may stick one together at some point and give her a listen, but, for now we will go with the linestage I have designed. The 6J5 is a very "musical" tube that has yet to hit the "endangered species" list, in fact, in addition to using it in a linestage, I use it as a cathode follower in my 45 amp.

[ToolShedAmps]

Good Morning All,

 

For those of you who are interested in building the "low-distortion" version of the pre-amp (which is the version I would suggest for pairing with any of my amplifiers, SET's in general, OR many of Nelson Pass's Solid State S.E. Designs) see the schematic attached below.

 

 

I will draft the two different Power Supplies at some point today and attach them as well.

 

Cheers!

 

Matt.

Edited November 12, 2015 by ToolShedAmps

[ToolShedAmps]

Howdy All,

 

My initial plan had been to offer two different versions of Power Supply to drive this circuit. The less expensive supply offering economy and a more expensive supply featuring ultimate performance. Well......to avoid confusion, I have decided to design a Power Supply from scratch to drive this Pre-Amplifier. This supply offers FANTASTIC performance at LESS THAN 1.0mV ripple at stated current and voltage, AND, I have "upped the ante" in terms of "Cool-Factor" (especially for those "getting their feet wet") in that the rectifier will be either an 80 or for those of you who think that a "DeForest Audion CX-280 Large Globe from the 20's" isn't "cool enough", the PS will also accept the 83 Mercury Vapor Rectifier.

 

Yep, super cool, super performing, AND....if built with 'Lytics, it won't break the bank. However, because the capacitor values are so low, if you would like to go "Whole-HOG" and dump the electrolytics in favor of film or oil, have at it!

 

Take a peek.

 

]

 

 

We'll open it up for comments and questions (if any). 

 

Matt.

[xxJPMxx]

Good afternoon Matt,  First thing I would like to comment on is that there is no safety bleed resistor. Not necessary but a good idea to add one.

 

That power transformer's 6.3 winding is rated for 2.5A and you will be only pulling 600mA. I calculate that the voltage will be 6.6V across the heaters, IMHO that's too much and will shorten the life of the tubes. I personally like to keep it under 6.3V and above 6V. To get 6.2V across the heaters I suggest using a .6 ohm resistor in series to drop .4V.

 

The HV secondary winding resistance of that transformer is 175 ohms from red to center tap, you could up the capacitance to a max 100uF without stressing the rectifier but the ripple is already low enough it might not matter.

 

-JP

Edited November 12, 2015 by xxJPMxx

[tube fanatic]

The voltage output of the Hammond power xfmrs tends to run fairly high, especially if wired for 115V.  I recommend using the 125V primary of the 269JX in any situation where the line voltage is above 117 or so.  Even doing that, the 6J5 filaments are likely to be presented with over 7V necessitating a higher value dropping resistor than JP suggested above.  Same for the rectifier filament xfmr which has a primary designed for 115-117V- voltage will have to be dropped for that tube as well.  The resistors should be installed in a way which allows the generated heat to escape.  I'm not a fan of heat build-up under the chassis.  And, as JP pointed out, a bleeder should absolutely be installed for safety.  The 6J5s are not going to draw enough current to allow for even a moderate bleed as their cathodes cool off.

 

Maynard

[ToolShedAmps]

Good afternoon Matt,  First thing I would like to comment on is that there is no safety bleed resistor. Not necessary but a good idea to add one.

 

That power transformer's 6.3 winding is rated for 2.5A and you will be only pulling 600mA. I calculate that the voltage will be 6.6V across the heaters, IMHO that's too much and will shorten the life of the tubes. I personally like to keep it under 6.3V and above 6V. To get 6.2V across the heaters I suggest using a .6 ohm resistor in series to drop .4V.

 

The HV secondary winding resistance of that transformer is 175 ohms from red to center tap, you could up the capacitance to a max 100uF without stressing the rectifier but the ripple is already low enough it might not matter.

 

-JP

Howdy JPM,

 

Now this is the kind of "discussion" I think most forum members can learn something from. 

 

I have not added a "bleed" resistor in the Power Supply for a couple of reasons,  as I normally like to include a "Standby" switch, I have found that dumping the voltage in the caps really isn't necessary when you run get yourself another beer (or get rid of the one you just had ), additionally the capacitance is so low that they will naturally "bleed-down" after a listening session in less than an hour. Furthermore, if you have ever measured the temp. of a PT (under load) with and without bleed-resistors you will find the PS runs hotter "with" them.

 

Great catch on the filament voltage! Wow, what a waste of perfectly good current. , As this PS was designed this morning (and has yet to be built for practical application and Empirical study), along with the decision to include a 6.3v power indicator lamp or not to include one. As well as still trying to decide (and YOU would be an awesome help for this), using a large portion of this winding (perhaps through a voltage-doubler) to provide the power to a motorized volume control. (It seems as though EVERYONE wants one these days!) Also, I have to apologize, I normally use the Hammond 300 Series PT's which have dual-primaries and have taps for 120v/60Hz instead of the 115v/60Hz primary on this PT. Anyway, NORMALLY what I do with a new design is this: Build it, then measure voltage "under load", do the math and then add the series resistor (even if the current draw puts the heater voltage at exactly 6.3v I will add a resistor anyway as I think the tube "sounds" better running them slightly "starved" at approx. 6.1-6.2v).

 

The reason that I am not running MAX capacitance is for the reason I stated when I posted the schematic. I have designed an excellent power supply with LOW capacitance so that DIY'ers (And MYSELF) can use either Film and/or ASC Oil Filled caps instead without having to build a separate chassis for the power supply. Also, I really didn't want "rise-time" to be impacted negatively by the increase in unnecessary capacitance, this Power Supply is LIGHTNING FAST, coming up in 0.5ms So, If someone (DIY'ers or MYSELF) wanted a Line-Stage they could live with indefinitely (provided there are enough inputs, two pair of outputs, AND a motorized volume control), this is a pretty good candidate. 

 

Anyway, I am waiting on the TKD motorized pot and Glasshouse IR Controller kit to arrive from the UK that I will be using in this build, so if you have any ideas on how to power it with the unused current in the 6.3v winding it would be much appreciated. If not, I'll just have to order up the 12v DC "wall-wart" and add the connector to the back panel.

 

Thanks,

 

Matt.