Sven 6P1P PP Amp - diyAudio
Go Back   Home > Forums > Amplifiers > Tubes / Valves
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Thread Tools Search this Thread
Old 28th September 2009, 04:29 AM   #1
TheGimp is offline TheGimp  United States
diyAudio Member
TheGimp's Avatar
Join Date: Aug 2009
Location: Johnson City, TN
Default Sven 6P1P PP Amp

I've started the design of my first amp. It is meant to be a small PP stereo amp to drive a set of Klipsch Heresey Speakers (circa 1979).

I plan on using all surplus Russian tubes, 6N23P for the input for their low noise, 6N2P for drivers, and 6P1P for the output tubes.

I've modeled it with Lt-Spice, but am a bit frustrated with a couple of things.

(1) There is a pronounced peak around 7hz. This probably won't matter in the real world with a set of Hammond 1609 transformers. Unless, the system acts like the old BOSE 901 series II systems with the active filter between the line out and in. The amp could end up driving the heck out of the bottom end because of the inefficent speakers. With the feedback I have, this could result in the low frequency sucking up a lot of power.

(2) I had to tweak R6 to get a balance between the two differential drives to the grids of U11 and U12. In theory this should not be required.

The schematic does not take into account the UL taps of the 1609 Transformers since I don't yet know how to model them.

First image is a combined gain plot and schematic from the LTSpice Simulation.

Second image is the schematic in LTSpice.




Attached Images
File Type: jpg SVEN_6P1P_PP.jpg (99.3 KB, 1673 views)
File Type: jpg SVEN_6P1P_Schematic.jpg (179.3 KB, 1723 views)
  Reply With Quote
Old 28th September 2009, 04:38 AM   #2
diyAudio Member
Wavebourn's Avatar
Join Date: Sep 2006
Location: Pleasant Hill, CA
Send a message via Skype™ to Wavebourn
Shift one pole 5-10 times, you have all couplings 22n/470K
For example, 220n from driver stage to output toobs.

I would use 6N1P instead of 6N2P, and about 8 mA of current since you go with UL (higher Miller capacitance than when pentode).
One more option: 6N2P on input, 6N23P as a driver.
The Devil is not so terrible as his math model is!
  Reply With Quote
Old 28th September 2009, 12:38 PM   #3
diyAudio Member
Join Date: Mar 2008

hi, please let me know where i can get the 6n2p spice object from.
thank you in advance
  Reply With Quote
Old 28th September 2009, 04:56 PM   #4
TheGimp is offline TheGimp  United States
diyAudio Member
TheGimp's Avatar
Join Date: Aug 2009
Location: Johnson City, TN
I did a first order approximation on the spice model by substituting the models as follows:

6P1P - 6AQ5 - 6BQ5, change capacitance, Mu values
6N1P - 6SN7 change capacitance, Mu values
6N2p - 12AX7 change capacitance, Mu values
6N23P - 6DJ8 change capacitance, Mu values

I used Duncan Munro's file dmtriodep.icn from

I hope to refine the models as I learn more about PSpice.

I changed some of the resistor values to get U1 up to 1.8mA cathode current, U2a/b up to 2.64mA cathode current(combined) and measured (PSpice) the output cathode current at 51.5ma.

This gives plate dissipation for the tubes at:
6N23P - .335W
6N2P - .24W
6P1P 12.4W (12W limit)

The 6P1P is a tad hot and I'll need to increase the cathode resistor value to get it down a bit to play it safe.

Changing one of the coupling caps to 220nF got rid of the pronounced peak in frequency response.

Thanks all.

  Reply With Quote
Old 28th September 2009, 05:41 PM   #5
Arnulf is offline Arnulf  Europe
diyAudio Member
Join Date: Feb 2009
I like the straightforward no-nonsense schematic. You drew a pretty response picture too but alas one cannot listen to pictures. Therefore my suggestion is: less plotting, more soldering Let us know how it sounds !
  Reply With Quote
Old 29th September 2009, 04:35 AM   #6
TheGimp is offline TheGimp  United States
diyAudio Member
TheGimp's Avatar
Join Date: Aug 2009
Location: Johnson City, TN
While the "taste is in the Pudding" I prefer to do a bit of 'what if' with the simulator before hitting the hardware. It oft time saves a lot of 'Oh! SH**!'

I found I had a problem with the 6N2P model for some reason. I got better results with the 6N23P model, and made some tweaks.

I think the phase splitter is better in this iteration.

Attached Images
File Type: jpg Image3.jpg (185.9 KB, 1498 views)
  Reply With Quote
Old 30th September 2009, 02:24 AM   #7
TheGimp is offline TheGimp  United States
diyAudio Member
TheGimp's Avatar
Join Date: Aug 2009
Location: Johnson City, TN
I tried several 6N2P models and they all wanted to go into cut-off, where-as the 6N23P and 12AX7 models do not.

The problem I've seen with PSpice is that they all are approximations that will get you close, but you have to build the circuit and tweak it in to get it where you want it (I'm currently using LT Spice, but also have an old 16 bit version of ICAPS/4. I had hoped to get a free update to 32bit, but I don't think that will happen).

The current design is close, but the cathode current is wrong on the output tubes. This is wrong whether I use a 6AQ5, 6BQ5, or 6V6 model. I am comparing the results to the cathode current vs grid bias with 250V plate voltage and about -10.5V grid bias yielding a 210 Ohm cathode resistor. Those parameters should put me near 50mA cathode current, but the simulation says 30mA. This is way off, even ignoring that I should be able to drop it to 190 Ohm when Screen current is accounted for.

I've mapped the cathode currents, plate voltages and gain for all stages.

Time to stick a fork in it and call it done.

I'll start pulling the parts together when I get back from visiting my dad this weekend. I'll build a prototype before I lay out the final chassis. The prototype will be on an aluminum chassis from a Tek 545 oscilloscope I'm using for parts.

Attached Images
File Type: jpg SVEN_0.6.jpg (157.8 KB, 1273 views)
  Reply With Quote
Old 10th November 2009, 02:32 AM   #8
TheGimp is offline TheGimp  United States
diyAudio Member
TheGimp's Avatar
Join Date: Aug 2009
Location: Johnson City, TN
Well, I finally got off my keyster and started the prototype.

Now that I have a current limited 100-350V power supply built I have started building my first amp.

I was tempted to fiddle with the phase splitter, but thought "No, leave it alone and just build it!", so I have the first gain stage and phase splitter wired up along with the filaments, and a delay tube for the B+.

I've done a power test on the filaments and delay tube. All looks good so far.
Attached Images
File Type: jpg Sven_6P1P_Prototype.JPG (424.7 KB, 1176 views)
File Type: jpg Sven_Prototype_Schematic.JPG (95.6 KB, 509 views)
  Reply With Quote
Old 12th November 2009, 02:57 AM   #9
TheGimp is offline TheGimp  United States
diyAudio Member
TheGimp's Avatar
Join Date: Aug 2009
Location: Johnson City, TN
The Prototype Sven PP is up and running!


I ran power up tests for bias conditions and found I had a problem with my bench supply and could only get 220V out of it with the load I was running. So I did the initial tests at 220V B+ and 180V on the screens.
This is with an old output transformer that is supposed to be from a Lafayette amp. I have several different OPTs I want to try before switching to the Hammond 1609s I expect to run with it.

After confirming bias conditions (running a bit cool at 32mA on the output cathodes) I ran some gain measurements. I adjusted a signal generator to get 2.8Vrms out on a 8 ohm resistive load and then measured the input required to achieve this (1W out).

I got:

20Hz 0.432Vrms Waveform distorted at driver and output due to low B+
50Hz 0.254Vrms
100Hz 0.256Vrms
1KHz 0.236Vrms
10KHz 0.236Vrms
20KHz 0.208Vrms
30KHz 0.190Vrms
40KHz 0.180Vrms

Powered it down, disconnected the load and signal generator and hooked up a Kenwood KL5050 and CD player.

First up YES Fragile, Roundabout - muddy muted mids and guitar sounds strained. Volume won't go up above "background music level"

May need to add an additional gain stage since I haven't hooked up the GNFB yet.

So, I think "What if I switch out the front end tubes ?". This was with the 6N23P in the first stage and driver position, which it was designed for.

Switched both to 6N1P - Voices still a bit muddy, but the guitar sounds more 'real'.

6N6P - between 6N23P and 6N1P. not impressive.

6N2P - Damn! Volume up at least 9dB. Guitar is right there, Voices are clean and clear.

At this time I shut it off and ejected the CD, then went upstairs and got the wife to bring one of her CDs down and played Beatles "Let it Be".

After listening for 5 minutes she asked "Can I have one of those upstairs?"


After she went back up stairs I listened to Billy Joel Storm Front, Jefferson Airplane "Jefferson Airplane" (1989) and Bonnie Raitt "Sweet Forgiveness".

I'm really impressed with how well this little amp compares with the SS amps we have (Sony STR-5800, Philips 951 and Magnavox 931Pro - wife's system upstairs ).

I'm also surprised at the difference switching tubes made. I suspect this could be compensated for by changing the bias and load resistance.

I may have some time to play with that after I switch out the OPTs and try different ones. I also need to correct the bias on the screens (220V and should be 200V) and the cathode bias (32ma vs 50mA target with 70Ma max).

My power supply consists of an AnTek AN-1T200 with a KBU6K bridge rectifier followed by a 47uF 450V cap, 7H (Gap Core switching power supply) inductor, and a 220uF 450V final cap. The Plates of the output tubes connect here, then there are R-C filters for the driver and preamp tubes.

As my last test of the evening, I hooked up a 60uF 440Vac Motor Run capacitor across the 220uF 450Vdc cap in my power supply with a switch shorted by a 220K resistor. My reasoning is that once the cap is charged up to the average DC value, I can switch between the cap in and out of the circuit to see if it makes a 'sonic' difference.

I can see no discernible difference, and conclude that for this amp it makes no difference. It may in larger amps, or amps with a lower filter capacitance to load ratio.

Overall I'm really happy with this amp, and foresee building several stereo units once I'm finished tweaking it.

Attached Images
File Type: jpg Prototype_Sven_Running _A.JPG (363.7 KB, 440 views)
File Type: jpg Prototype_Sven_Running_B.JPG (427.9 KB, 373 views)
  Reply With Quote
Old 12th November 2009, 03:38 AM   #10
diyAudio Member
Join Date: Nov 2007
Location: Dallas
Originally Posted by Wavebourn View Post
Shift one pole 5-10 times, you have all couplings 22n/470K
Is that the same procedure they call "slugging the dominant pole"?

For frequencies highly attenuated, even the worst phase shift won't
make an oscillator. But at corner frequency of unity loop gain (or NFB)
you had better have 90deg phase shift or less... Thus the slope that
leads into your corner can't be any sharper than -6dB. or is it -3dB?

Or to put it another way: If you are gonna have a bunch of filters
rolling together at some higher rate (and phase shift). It had best
already have been attenuated below unity loop gain by a filter with
a lower corner frequency and lesser slope.

Am I telling the story straight, or do I got it messed up somehow?
I don't wanna be repeating nonsense, but it happens sometimes...

Last edited by kenpeter; 12th November 2009 at 03:41 AM.
  Reply With Quote


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

Similar Threads
Thread Thread Starter Forum Replies Last Post
g. pimm's rev3. and rev. 4 -6P1P/5687 ??? sparkle Tubes / Valves 3 20th July 2004 10:56 AM

New To Site? Need Help?

All times are GMT. The time now is 08:45 AM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2