• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

845 PP Prototype

artosalo

Member
2010-02-16 9:00 am
Designing and constructing a 845 PP prototype for a friend.
Williamson style front end, AC-filaments, Toroidy transformers, separate power supply for each channel etc.
 

Attachments

  • IMG_20190414_194214 (Medium).jpg
    IMG_20190414_194214 (Medium).jpg
    230.3 KB · Views: 981

multi

R.I.P.
2011-01-28 2:52 am
artosalo Designing and constructing a 845 PP prototype for a friend.
Williamson style front end, AC-filaments, Toroidy transformers, separate power supply for each channel etc.

Hope to see more of your design; hope to make similar; may be GM 70
Phil
 

artosalo

Member
2010-02-16 9:00 am
Here are the schematics. At the moment supply voltage for output tubes is 660...670 V. With this voltage the max. output power is some 30 W.
As the driver, now 6BL7 is used, because supply voltage (+Ub2) is less than I planned in the beginning.
With 380 V or more 6SN7 can be used too. Driver must produce clean 180 Vpp + some headroom.

There is one GZ34 in the power supply to fulfill "customer desires", not any technical aspect.

What is a small surprise to me, is that hum level at the 8 ohms output is only 1,3 mV (rms).
This is quite good value with AC-heaters.
 

Attachments

  • 845PP_power.JPG
    845PP_power.JPG
    121.5 KB · Views: 733
  • 845_PP_skema.JPG
    845_PP_skema.JPG
    232.8 KB · Views: 755
Last edited:

6A3sUMMER

Member
2016-06-07 6:50 am
artosalo,

Nice circuit! I like the potentiometer in the concertina, allows null of the sum total of the 2nd harmonic distortion of all the stages through the amp. Can leave a little 2nd harmonic in, if that sound pleases.

I believe you have 100Hz power mains. If you get a chance, put in a 1kHz tone, adjust for 1/4 of max power out (1/4 full voltage), and see how far down the 900Hz and 1,100Hz intermods are due to the AC filaments. It will be interesting to see if the push pull circuit cancels that effect (single ended has lots of that intermod).

Thanks!
 

artosalo

Member
2010-02-16 9:00 am
It will be interesting to see if the push pull circuit cancels that effect (single ended has lots of that intermod).

Good idea. I try to do IM-test next week.

What kind of power supply are you using for testing?
I mean the one on the background.

That power supply is small Heathkit IP-2717, only capable for 400 V / 150 mA.
But just fine for Williamson stages (6SL7 and 6BL7)
For output tubes I used in the beginning a pair of Soviet made UIP-1 (600 V / 600 mA).
Universal-Netzanschlussgerat- Deutsch Power-S Unknown - CUSTO

What are you using for output transformers?

This: TTG-6AS7PP - Tube output transformer [5kOhm] 2x6AS7G / 2x6N13S Push-pull or similar - Shop Toroidy.pl
 

artosalo

Member
2010-02-16 9:00 am
If you get a chance, put in a 1kHz tone, adjust for 1/4 of max power out (1/4 full voltage), and see how far down the 900Hz and 1,100Hz intermods are due to the AC filaments.
It will be interesting to see if the push pull circuit cancels that effect (single ended has lots of that intermod).


100 Hz IMD-test done. The output power under test was 8 W (to 8 ohms).
The IMD level at 900 Hz and 1100 Hz is -72 dBc which is about 0.025 %. I think this is not a bad result at all.

At the same time 2nd. harmonic is -68 dBc (0.04 %) and 3rd. is -57 dBc (0.14 %).

The hum-nulling potentiometer at the filament of 845 has no effect to +100 Hz IMD.
 

Attachments

  • 845_PP_IMD+THD.jpg
    845_PP_IMD+THD.jpg
    92.5 KB · Views: 200
  • 845_PP_IMD_100Hz.jpg
    845_PP_IMD_100Hz.jpg
    77.5 KB · Views: 187
Last edited:

6A3sUMMER

Member
2016-06-07 6:50 am
artosalo,

That was a good test.
And the results are excellent.
The amp must sound wonderful.

It is nice to know that push pull is so much better in cancelling that effect that can be such a problem for single ended amps that use AC filaments.
It makes me want to try a push pull 45, 2A3, or 300B with AC filaments.
 
Last edited:

LKA

Member
2014-12-28 5:31 pm
100 Hz IMD-test done. The output power under test was 8 W (to 8 ohms).
The IMD level at 900 Hz and 1100 Hz is -72 dBc which is about 0.025 %. I think this is not a bad result at all.

At the same time 2nd. harmonic is -68 dBc (0.04 %) and 3rd. is -57 dBc (0.14 %).

The hum-nulling potentiometer at the filament of 845 has no effect to +100 Hz IMD.

It is not AC heating related IMD, but output stage power supply ripple interaction with input signal.
 

artosalo

Member
2010-02-16 9:00 am
Let's see. I do the same test with heaters powered with DC.

It is not AC heating related IMD, but output stage power supply ripple interaction with input signal.

Below is the IMD-test result with DC-heaters. Now the IMD level is -78 dBc, which is 6 dB lower than with AC. So it is seems that 100 Hz IMD due to AC-heaters is dominant.
During the test the 100 Hz ripple voltage at the plate of 845 was 0.9 Vrms only.

I made an additional test.
I increased the +Ub filtering to 0.42 Vrms ripple.
Now with DC-heaters the 100 Hz IMD dropped to -82 dBc. The I switched to AC-heating. 100 Hz IMD jumped to original value -72 dBc.
It is obvious that in this case the 100 Hz IMD due AC-heaters is dominant, not 100 Hz ripple at +Ub.
 

Attachments

  • 845_PP_IMD_DC_heat.jpg
    845_PP_IMD_DC_heat.jpg
    89 KB · Views: 723

artosalo

Member
2010-02-16 9:00 am
HIFI amplifiers require clean supply voltages, this is self-evident.

In that document one detail is worthy of remark:

...After rectification the voltage is buffered in a large capacitor of 330uF/450V.
The resulting high voltage has a ripple of about 14Vpp when the total current demand equals 4 x 80mA.

My filtering circuit is below with ripple levels as printed in the schematic.
In my case the current draw is 4 dB less but ripple level 15 dB lower.
 

Attachments

  • 845PP_poweri_ripple.JPG
    845PP_poweri_ripple.JPG
    47.6 KB · Views: 656

artosalo

Member
2010-02-16 9:00 am
As already shown by my tests the 2.5 Vpp ripple at the plate voltage of the 845 output tubes generate 0.013 % (-78 dBc) 100 Hz IMD.

This is 6 dB below the IMD caused by the AC-heaters. I would like to know why would anybody need as low as 1-10 mVpp ripple (at PP-output stage) ?