Suggested Dynaco PAS mods

the thermionic watercooler

Seduction phono stage with lots of red LEDs

Postby EWBrown » Thu Feb 12, 2009 9:10 am

Here is a peek under the hood of a Bottlehead "Seduction" 6DJ8 based phono stage, with LED cathode "resistors" and C4S plate loads. Photo is fromWardsweb's site:

Image

LEDs are all HLMP6000s. The C4S replaces the four original 18K, 1W plate resistors . B+ is 150 to 160VDC, unregulated.

/ed B
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Re: Seduction phono stage with lots of red LEDs

Postby jukingeo » Thu Feb 12, 2009 10:17 am

EWBrown wrote:Here is a peek under the hood of a Bottlehead "Seduction" 6DJ8 based phono stage, with LED cathode "resistors" and C4S plate loads. Photo is fromWardsweb's site:

LEDs are all HLMP6000s. The C4S replaces the four original 18K, 1W plate resistors . B+ is 150 to 160VDC, unregulated.

/ed B


I see lots and lots of silicon in that pup. Too much for my tastes. Looks 'pretty' though with all the lit LEDs.

I am hoping Terry gets back soon, I am curious to hear about some tests he has run on that PAS mod.

I wonder if this would work in a power amp voltage amp as well. For instance. This is the amp I am building:

http://www.angela.com/catalog/how-to/SE.EL34.html

If you scroll down to the schematic you will notice that the cathode of the voltage amp has a 1k resistor bypassed by a 100uf cap. The voltage on the cathode is expected to be 1.7volts. I would assume that this amp is a high candidate for a LED (or battery) correct? If so, then I can save the use of a very large cap that would normally sit here. (I normally wouldn't use an electrolytic here...so a poly cap can get pretty big).

Geo
jukingeo
 
Posts: 198
Joined: Wed Jan 21, 2009 11:45 am

Postby EWBrown » Thu Feb 12, 2009 11:05 am

The CCCCS board on top runs at approximately 4 mA per 6DJ8 triode. The silicon is essentially outside of the signal path, but it can still "color" the sound, some folks like it better, some prefer plain old fashioned resistors. My ears aren't good enough to tell the difference... Yellow_Light_Colorz_PDT_05 Yellow_Light_Colorz_PDT_02

The C4S has too many "moving parts" for my liking, I'd probably go with an IXYS IXCP10M45S and a current set resistor, use two components, instead of six, per tube plate. There are some neat 10M45S "tricks", but I'll go into those elsewhere.

The reason for the C4S (or any CCS) plate load is that with its very high effective impedance, the tube stage can allegedly deliver more gain, and wider bandwidth.

That EL34 SE would be a perfect one to try the red LED on the 6SL7 cathodes. The voltage is in the right range for common red LEDs.

It runs the two sections of each 6SL7 in parallel, which doubles the gm and halves the rp, and the mu stays the same. This allegedly increases the dynamic range and lowers noise. Looks like they run at 1.7 mA (1.7V thru a 1K resistor). Or 0.85 mA per section.

I'd start t with the red LED (cathode to ground, anode to the tube's cathodes).

It is easy enough to take the LED out and install the resistor and cap, and then compare which sounds better.

Or you could build one channel with the LED and the other with the resistor and cap and get a real-time side by side comparison.

/ed B
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Postby jukingeo » Thu Feb 12, 2009 11:30 am

EWBrown wrote:The CCCCS board on top runs at approximately 4 mA per 6DJ8 triode. The silicon is essentially outside of the signal path, but it can still "color" the sound, some folks like it better, some prefer plain old fashioned resistors. My ears aren't good enough to tell the difference... Yellow_Light_Colorz_PDT_05 Yellow_Light_Colorz_PDT_02


I have done some tests a while back in regards to ss v.s. tube power supplies, and the diodes in the power supply can create noise and other 'artifacts' in the sound, but this is more true of the power supply in the presence of higher gain stages. This usually is remedied by using more filter sections or even a choke filter. So that can negate that benefit of the tube rectifier. In a guitar amplifier, a class AB amp can 'pull down' a tube rectifier and as such, when you nail the amp, the voltage drops further enhancing that 'playing the amp' feel. This is why many guitar amp purists WANT a tube rectifier. However, this only works with class AB amps. Class A amps draw constant current and the effect isn't as noticeable. BUT still a tube rectifier makes less noise.

I think that one of the biggest assets of a tube rectifier is that you can create a 'slow heat' circuit on the filament. This way you only need one power switch when you turn on the amp. In other words you can set the amp so that the B+ comes up after the rest of the amp has warmed up. You also don't have the nasty inrush as you would with silicon rectifiers.

But despite all of this, I still don't mind having silicon rectifiers in the power supply. But when you put them in the signal path, then I pull on the reins a bit.

The C4S has too many "moving parts" for my liking, I'd probably go with an IXYS IXCP10M45S and a current set resistor, use two components, instead of six, per tube plate. There are some neat 10M45S "tricks", but I'll go into those elsewhere.

The reason for the C4S (or any CCS) plate load is that with its very high effective impedance, the tube stage can allegedly deliver more gain, and wider bandwidth.


Agreed, this design falls out of the "KISS" concept and would normally be a design I would avoid. As for high impedance plate loading...well, that is what an SRPP gain stage is for...and it is all tube :).

That EL34 SE would be a perfect one to try the red LED on the 6SL7 cathodes. The voltage is in the right range for common red LEDs.


Good, then I can build my breadboard with one channel using the resistor and cap and the other using a LED. Then I can compare.

It runs the two sections of each 6SL7 in parallel, which doubles the gm and halves the rp, and the mu stays the same. This allegedly increases the dynamic range and lowers noise. Looks like they run at 1.7 mA (1.7V thru a 1K resistor). Or 0.85 mA per section.


Yeah, they mentioned the benefits of paralleling the sections in the document with the amp. I actually prefer this to 'sharing' stereo amp sections in one bottle. I always believed that there could be some crosstalk or 'bleed over' when building two sections of a stereo stage in one bottle. However, I have not tested that theory, so I am not sure what the truth is in that matter.

I'd start t with the red LED (cathode to ground, anode to the tube's cathodes).

It is easy enough to take the LED out and install the resistor and cap, and then compare which sounds better.

Or you could build one channel with the LED and the other with the resistor and cap and get a real-time side by side comparison.

/ed B


Yup, that is what I mentioned above and that seems like a good way to go. I have a dual trace scope too, so feeding the same signal into both amp sections, I can lay one trace on top of the other and do simultaneous comparisons.

Last night I was orienting some parts on the 'board' that I am going to put everything on. However, since I was using shelving material, I realized that my board was not actual pine, but a wood 'image' overlay glued onto press board (El Cheapo wood). Given the weight of the transformers, I am sure the board would snap in half if I tried to carry it with that weight. So I need to find a nice solid piece of pine to lay everything out on.

Geo
jukingeo
 
Posts: 198
Joined: Wed Jan 21, 2009 11:45 am

Postby TerrySmith » Thu Feb 12, 2009 5:00 pm

I have about 5 hours on it now with no ill effects. The only thing is the gain is slightly lower. Voltage check is 212v, it needs 250v according to the RCA book.

The linestage is basically like Soren's (tubebuilder.com) 6SN7 circuit but with 7AU7 tubes instead. Required some jumpering to make that circuit work. Here's a link to the schematic:

http://www.tubebuilder.com/images/schem ... n7_sig.gif

I removed the cathode bypass caps, and added feedback to reduce gain.
T. Smith
User avatar
TerrySmith
KT88
 
Posts: 973
Joined: Sat Jan 31, 2004 12:51 pm
Location: Maryville TN

Postby TerrySmith » Thu Feb 12, 2009 5:56 pm

How can you get circuit boards made to your circuits? The boards in the PAS pictured in this thread are pretty butchered up, and it would be nice to have some custom boards made.
T. Smith
User avatar
TerrySmith
KT88
 
Posts: 973
Joined: Sat Jan 31, 2004 12:51 pm
Location: Maryville TN

Postby Quad » Thu Feb 12, 2009 11:40 pm

How can you get circuit boards made to your circuits?


Lay them out in a PCB CAD tool like eagle and etch them yourself using toner transfer.

http://www.5bears.com/pcb.htm
Quad
KT88
 
Posts: 251
Joined: Mon Jan 14, 2008 6:54 am
Location: India

Postby EWBrown » Fri Feb 13, 2009 6:13 am

FWIW, LEDs would not be appropriate inan SRPP circuit.

Possibly for the "lower" tube's cathode, but definitely not for the upper resistor, as it would defeat the whole intent of an SRPP design.

If your plate voltage reads low, it is because the LEDs are biasing the cathodes at a somewhat lower voltage than the original resistor & cap, which will cause increased plate current. Probably not much of a real issue, unless you are trying to extract every last dB of gain out of it.

If you measure the DC voltage across the LEDs, and the plate resistors you can figure out the actual plate current being consumed.

The HLMP6000s would generate too low of a cathode voltage (1.56 to 1.57V) which would increase the plate current in this circuit.

If your DVM has a diode check function, then you can select the LEDs which have the closest-to-correct vF drop.

/ed B
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Postby TerrySmith » Sat Feb 14, 2009 2:01 am

The LED's are only in the phono stage. Yellow_Light_Colorz_PDT_06
T. Smith
User avatar
TerrySmith
KT88
 
Posts: 973
Joined: Sat Jan 31, 2004 12:51 pm
Location: Maryville TN

Postby jukingeo » Sat Feb 14, 2009 3:07 pm

TerrySmith wrote:I have about 5 hours on it now with no ill effects. The only thing is the gain is slightly lower. Voltage check is 212v, it needs 250v according to the RCA book.


But that is fault of the power supply, not the LED's right?

The linestage is basically like Soren's (tubebuilder.com) 6SN7 circuit but with 7AU7 tubes instead. Required some jumpering to make that circuit work. Here's a link to the schematic:


I am going to assume the 7AU7 is a 7volt version of the 12AU7, correct?

I removed the cathode bypass caps, and added feedback to reduce gain.



Nice SRPP stage. I LOVE simplicity!
That is interesting...so the circuit had too much gain?

EWBrown wrote:FWIW, LEDs would not be appropriate inan SRPP circuit.

Possibly for the "lower" tube's cathode, but definitely not for the upper resistor, as it would defeat the whole intent of an SRPP design.


It is funny you mentioned that of the lower tube in the SRPP design. I have thought of an LED there as well. SRPP's are picky about the supply voltage and I was curious if an LED in that lower spot might stabilize things a bit.

However, for the most part I understand that we are talking about LED in typical voltage amp circuits. As Terry said, he just has the LED's in the phono stage.

If your plate voltage reads low, it is because the LEDs are biasing the cathodes at a somewhat lower voltage than the original resistor & cap, which will cause increased plate current. Probably not much of a real issue, unless you are trying to extract every last dB of gain out of it.


I did ask Terry if this was due to adding the LED or was it because of his power supply.

If you measure the DC voltage across the LEDs, and the plate resistors you can figure out the actual plate current being consumed.

The HLMP6000s would generate too low of a cathode voltage (1.56 to 1.57V) which would increase the plate current in this circuit.


So then you would recommend a LED that has a higher trigger voltage?

If your DVM has a diode check function, then you can select the LEDs which have the closest-to-correct vF drop.

/ed B


Yes, my DVM does have a diode check function...but the funny thing is that I never used it as such. The 'beeper' in the meter is connected to the diode check function and it doubles as a continuity checker function and that is how I mostly use the diode check function. But I have noticed that there is a reading on the display when I do check diodes and transistors. I never paid it any mind because as I said, I just use the function to check for shorts.

Geo
jukingeo
 
Posts: 198
Joined: Wed Jan 21, 2009 11:45 am

Postby TerrySmith » Sat Feb 14, 2009 5:20 pm

Yes, I need to adjust the power supply voltage slightly.

Yes, the 7AU7 (cheap!) is a 7 volt version of a 12AU7. There is also a 6AU7 and a 9AU7, and cheap for NOS too!

Look up "Sorens 6SN7 linestage" in this forum, this linestage follows that one, except the few mods and the *AU7 instead of a 6SN7.
T. Smith
User avatar
TerrySmith
KT88
 
Posts: 973
Joined: Sat Jan 31, 2004 12:51 pm
Location: Maryville TN

Postby jukingeo » Mon Feb 16, 2009 9:27 am

TerrySmith wrote:Yes, I need to adjust the power supply voltage slightly.

Yes, the 7AU7 (cheap!) is a 7 volt version of a 12AU7. There is also a 6AU7 and a 9AU7, and cheap for NOS too!

Look up "Sorens 6SN7 linestage" in this forum, this linestage follows that one, except the few mods and the *AU7 instead of a 6SN7.


Hello Terry,

Wow! Lots of reading (info) in that "Sorens 6SN7 linestage". I just finished getting through it just now.

Overall I think your approach to the PAS situation seems like a nice simple approach that would yield excellent results for the RIAA and line stages. Far simpler than the original circuit. So the only thing I would like to entertain is the possibility for tone controls. More then likely I have to go the 'passive' route as most tone controls rely on feedback. Feedback reduces gain even further (not to mention the 'sterility' problem). So I would like to avoid that.

I do remember reading something in the Mullard book and they had a passive tone control circuit that could be used (or not used) with most of their amp designs. They did this so that you could either design a dedicated power amp OR a fully integrated amp using pretty much the same basic circuit design.

I will say that I am probably not going to use a 12AU7 for the PP line stage, but rather a 5687. I have a bunch of NOS black plates and these tubes can drive into a VERY low impedance (on the order of 100 ohms), this is so low that I could double the line stage's use as a direct driver for a headphone.

Here is a similar schematic as I cannot find the original. The original was called "Pat's Ultimate SRPP Preamp".

http://www.bonavolta.ch/hobby/en/audio/5687line.htm

In Pat's design he uses 1.5k resistors in the cathode positions. There is a group of caps totalling 50uf for the output, and there isn't a 300 ohm resistor present on the input. Everything else is the same. The link to "Pat's Site" is bad. However, I do have the original schematic on my system and could work from that.

In one regard with a proper set of headphones, I could use the modified PAS as a double duty preamp AND headphone amp.

So I think going with the 5687 for the SRPP line stage is a better bet than going with the 12AU7. I have quite a few 5687's, so availability isn't a problem.

EDIT: Oh! Terry, I meant to ask you. How do you compare the quality of the audio output of the SRPP linestage to the prior one you had installed in the PAS? Thus far I really never compared other low impedance line stages to the SRPP. Mostly the pre-amps I messed with were for tube amps only and had high impedance coupling only. So I really don't know what a dedicated VA/Cathode follower sounds like. I did have an old Bell Labs amplifier that had a pre-out mainly for hooking up a tape deck. This DID have a cathode follower and as such I was able to use a standard SS deck. The Bell amp was nice sounding, but I didn't have it long to do any meaningful tests. It was only 11 watts per channel utilizing 2 6V6 output tubes on each channel. The amp I had was a revised version and I didn't have the correct schematics for it it properly re-build it. I ended up swapping that amp out for a Fisher 500B (which I still have to restore). But little did I know how complex the Fisher is!

Geo
jukingeo
 
Posts: 198
Joined: Wed Jan 21, 2009 11:45 am

Postby jukingeo » Wed Feb 18, 2009 7:24 am

TerrySmith wrote:I have about 5 hours on it now with no ill effects. The only thing is the gain is slightly lower. Voltage check is 212v, it needs 250v according to the RCA book.

The linestage is basically like Soren's (tubebuilder.com) 6SN7 circuit but with 7AU7 tubes instead. Required some jumpering to make that circuit work. Here's a link to the schematic:

http://www.tubebuilder.com/images/schem ... n7_sig.gif

I removed the cathode bypass caps, and added feedback to reduce gain.


Do you happen to have a schematic of the phono stage with the LED's??
jukingeo
 
Posts: 198
Joined: Wed Jan 21, 2009 11:45 am

Postby TerrySmith » Wed Feb 18, 2009 3:23 pm

It's the same circuit as seen in the RCA book, the LED's are in place of the 2.7k resistors amd 25uf caps connected to the cathodes.
T. Smith
User avatar
TerrySmith
KT88
 
Posts: 973
Joined: Sat Jan 31, 2004 12:51 pm
Location: Maryville TN

Postby jukingeo » Sat Feb 21, 2009 11:15 am

TerrySmith wrote:It's the same circuit as seen in the RCA book, the LED's are in place of the 2.7k resistors amd 25uf caps connected to the cathodes.


Ok, thanx. I DLed the version with the batteries that was posted earlier, I can then just sub the batteries for the batteries and I would be good to go.

So overall this is what I am going to do. I am pretty much going to build the phono stage exactly as you have. I DID find a couple of Mullard 12AX7's and I am going to set them asside for this stage.

For the line stage I am going to once again go with the SRPP design I used with my first pre-amp...however this time I am going to go with a 5687 tube. I have about 2 dozen of these tubes...one dozen is all black plate tubes and the other dozen is all JAN NOS tubes. I didn't even know I had this many, BUT when you pack up a house to move, you never know what you find! Thus, having an ample supply of 5687's makes this a good choice for the line stage.

Now comes the question once again about the tone control. Since I am going to build the stages around the original PAS boards (new ones of course), I will only have a tube per stage per side. Thus I need a VERY low signal loss method for tone control.

The PAS doesn't use ganged tone controls, thus you have separate left, right, bass & treble controls. So with four mounting positions this does allow many possibilites.

What I thought of doing is perhaps using all ganged controls (for left/right together), I thought of creating some kind of tunable filter. This way I can center in on a frequency for the bass and also the treble. I got this idea when I saw a 'tunable' loudness contour on a Yamaha receiver the other day. Thus I wondered if the same thing could apply to the bass/treble controls of my pre-amp. I could also use the "filter" switch to bypass the tone control if I want.

However, I am not sure what I would be getting myself into in terms of complexity. And given that I need a low loss tone control does add to the challenge.

I am hoping in the end I will have a nice full functional pre-amp that will have NO negative feedback. This should really sound good with the SE amps I intend to build.

Thank You,

Geo
jukingeo
 
Posts: 198
Joined: Wed Jan 21, 2009 11:45 am

PreviousNext

Return to diy hifi

Who is online

Users browsing this forum: No registered users and 2 guests