Measurement Software - 'Room EQ Wizard'

[alexc]

While I'm doing a bit of tidying up ..  only a bit  ...  I have nfb on a front panel control - 5K from speaker output to 6SN7 cathode (460ohm cathode resistor).

It's a must have to tame the raw power and also the hum, and the extension of the freq response. Thge amp is still pretty good with minimal nfb, but it really does deepen the sonic palette to dial her in .. or .. out.

Along with the 'instrument' input's tone-stack-bypass switch which gives a preamp boost, pretty much any basic THD mode is .. .. over-represented perhaps.

.....

Listening to hifi thru the aux's is pretty good. Balance with the instrument input is good too

Power to burn, no doubt about that. 15Wrms into a load box sure, but into a 90+dB sensitivity speaker with some weight in the box and ... whammo.

It's big. In that mode, I have 25dB + ear muffs and it's still impressive. 

These are the no-transducer version of the Iso Maxx headphones I use for drum recording.

.....

I have a bunch more things to sort out before deciding on which finals to use next ...

 

[alexc]

Some more adjusticating :

- set finals up for 80mA at the cathodes for each of the 6550s
- added nfb, a simple 5K pot into the first 6SN7 cathode (490R cathode resistor to ground)

As expected, the nfb tames the wild beast - I have a minimum of 6dB and a max of 12dB on the front panel pot

At 12 dB nfb, I get my first baseline nfb setup :

freq response is      -0.8dB at 20Hz    and      -1.2dB at 20KHz    with the  -3dB points at  12.7Hz and    26.3KHz

Here's the response plots ..

 

[alexc]

So that's where this amp will stay unitl the next occasion it's on my bench 

Set up for  a max 'low THD' of 4.5%  at 20Wrms into 4ohm  for use with big, sensitive speaker 
For pentode-UL mode, that is. I forgot to do the triode mode 

Seriously, I did. Sounds good though, and is probably the setting I will use most.

----

Of course, down at lower listening levels is where I usually sit - on the 3way Sansuis, it typically was < 3Wrms.
Even less for the big Jensen 15" box. At those levels, the sine wave THD is around 1.5% or so or less.

Hum voltage is now down around 20mVpp at the speaker terminal for max power with 12dB nfb, which is around -62dB on the fundamental.

So - all working pretty much as expected. From here, just more THD measures and button up - looking at the spectral distribution and such.

---

As usual, better layout means lower noise - mine in now ready for a 'tear down' and 'rebuild' - I have the basic amp structure worked out  ;D ;D 

Not likely right now - I'm ready to apply some of this 'new knowledge' of thd measures to my next so called 'design'  .

There I will do the measures as I prototype the stages and 'build right' from the onset. Or thereabouts.

 

[alexc]

Looking at the above THD spectra for max power with THD 4.5% ...

- H3 at 3.32% dominates the THD - I have to check back now and look at each stage again.

---

I did some 'tube subs' while watching the effect on main output THD ...

- up to 3.5% THD, H2 dominates 3:1
- from 3.5% to 4.5% THD, H3 equalises then begins to dominate

Tube 'subs' of the same type, alter that outcome a little.  Tube 'subs' of similar but different types, alters the THD at which that situation comes about, either earlier 'breakup' or 'later.

You can hear the transitions by ear when when connected to the load box.

--

So it appears, the amp has a basic spectral character in each of the main THD bands. Tube changing varies the measures but not the shape  ..  too much.

The main defference between how it was and this baseline, is that each stage is a lot more .. center biased 

Before, the stages were not well setup, clipping early and with no nfb -  BUT THD was almost entirely H2, in the preamp.
Now its more 3:1 until THD rises strongly and that all comes out to the speaker output.

I need to review those early figures  now  to make sure 

ONE thing for sure, when operated with the 2x Sansui 3ways- 8ohms, in parallel for 4ohms load - reflecting 1600ohms to the parallel finals, is MUCH better - the bass kicks up a notch or two and the impact is major for the same power. (adjusted, more or less)

In both pentode-ul and triode mode the improvement with the lower Z is big.

I did target that speaker match originally 1600ohms for triode mode.

 

[alexc]

Also as a result of this THD 'tuning', I'm going to invest in some more test toolage :

- an 'active' HV probe;  I've not seen too much on this, and I'm not sure if there's much benefit beyond my simple passive probe.  But I'm doing some researchin' ....

and

- tube test fixture with parallel'd sockets and easily accessable plate, cathode resistances
- regulated bias source, for grounded-cathode variable 'fixed-bias' characterisation
- variac for varying HV easily
- separate filament transformer

And, hand in hand with THD measures, is the abandonment of the importance of 'time spent' on a build

 

[alexc]

NOw, back to the Sansui SP2500s ...

I did a recap of the original Nichicon biopolars in the crossover - they may not have needed it, but they were 40+yrs old.

I'm too new into crossovers and speaker measurements to  know if they are not .. better - all I can say is they sound as I expect and very detailed.

Here's the crossover units - 'tweeter' and 'squawker' each have a 3 position rotary, selecting taps on each's seperate iron-cored inductor. No air-cored stuff here!

The settings are 'Soft', 'Natural' and 'Clear'  in order increasing treble or mid.

 

[alexc]

So - all buttoned up, auditioning on the recapped speakers and have the measurement mic setup and feeding into the Motu and Room EQ Wizard.

I did the 'spl' calibration for REW's suite of test sweeps. Easy, no probs.  24bit, 96KHz remains all good.

I am yet to download from Dayton the 'cal' file for the mic - I'll do that and see how it works. The mic is plenty flat, according the supplied test curves.

Here's the first view of the speakers response, magnitude spectra.

I should point out, the room is not much good - a wooden box of rectangular  dimension .. but worse!

In time, I'll set myself for portable testing, and take it out in my fields and try there ...  with a good long power extension cable

 

[alexc]

Back to my EL34s - I have wired for the eh 6922 type in the first stage position now.

It's a hell of a tube, along with the 6N1P and 6N23P. 

With 10dB nfb reconnected,  and using my 'worst' tube in the 6922 position :

-  I'm getting  30Wrms into 8ohms load box, with 0.345% THD at the speaker and  0.85% THD at the 6922 plate

And I haven't yet set it right for a 6DJ8 which is my current choice. 

At normal listening levels of <10Wrms, it's more like 0.25% THD at the plate and  <0.2% at the speaker outputs.

So - I'm on the right track! I imagine a series pair of 6922 has got to be the lowest high-gain THD possible. I'll find out soon enough!

---

As usual, waiting on parts - so more speaker spectral measurements AND trying to learn to make sense of the meaning    ;D

My early opinion is that the bigSE amp I've finished tuning, sounds better on the same speakers and at the same levels.

That SE has quite a bit higher in THD  than the PP but it has a lot more bass and impact.

Look like my next hifi effort will be another SE jobby, but I'm still thinking...

The only firm decision is that the finals are going to be a 7915a  and the first stage will be a 6DJ8.

 

[alexc]

Next up for a tune and measure ....  the smallSE Amp 

This is what I most often use in my 'listening room' (such as it is)

It is stereo, each side  has a single final, currently 6L6 in SE  into a fairly cheap sino output traffo.

 

[alexc]

And here's the inside pic

 

[alexc]

Setup is  'fixed bias', variable on a pot, currently set to 6L6 -26.3V for 57mA primary current.

Running conservatively, I have have available only 65mA per channel of HV, leaving a 15% reserve of power.

Makes  3.08Wrms into 8ohms, no nfb and with 5.5% THD 1KHz at the speaker terminals

Using 6SN7 stage1 as VA, low sensitivity. With 23K5 plate load and 5.5Vbias for 6.8mA and gain x11.9.
It gives 1% THD at 46Vpp which drives the final as described above.

Here's the response of the 6SN7

 

[alexc]

and the speaker outputs response ..

 

[alexc]

and the THD at the speaker outputs ..

Noise voltage is 7.5mVpp with max gain. Nothing needed there 

Typical listening levels are around 1.9Wrms of music - down there, the sine wave THD is more like 1% at the speaker terminals.

Add an active sub to these and nothing much more needed to enjoy at strong listening levels.
Very unfatiguing to listen to for long periods - top end is notable for the lack of harshness.

(the sub is another matter!)

 

[alexc]

and the test song ....  Carl Carlton  'I Won't Let That Chump Break Your Heart'  ;D

Fantastic recording of that genre  ... superb recording.

Its sounds absolutely wonderful into the Sansui speakers, even though they are a little less sensitive than my regular critical listen speakers.

 

[alexc]

So this ultra simple smallSE Amp has a big, vibrant sound with sensitive 'old style' speakers. 

Of course, won't cut it with todays low sensitivity, 200Wrms 3way jobbies 

5.5%THD, 1KHz at max output isn't too bad, with no nfb. 

I expected worse, but as I said, setup conservatively and for low sensitivity. The THD at < 100Hz shoots up, of course. I may wind up replacing these traffos with something of less THD at the bottom .... maybe 

Caps are good quality, with some big ol Russian greens in there, rest wima and nichicon. PSU traffo and choke are super quiet.

Has a very good low  hum level, around -65dB down from fundamental wide open at 3Wrms into 8ohms.
(ie. 7.5mVpp/14Vpp).  Now, with no nfb, thats pretty impressive. It has nice wiring for the most part.

Gain is around x4 or +12dB from input to speaker terminals.

Freq response is pretty amazing 12 .. 25K    These traffos cost very, very little ...  around 10years ago

-----

I've never tried it with nfb - I keep it that way as a 'reality check' on what enjoyable, live music sounds like

Oh yes - it has early 70s RCA black plate 6L6s and Sylvania 50s 6SN7s.

I really lucked out on the 6SN7s - these are superb new old stock (really), low THD (for a 6SN7) and amazing match.

All for a reasonable price, not much more than my 'standard' modern TungSol units.

This amp is also pretty good with 6SL7s - more sensitivity of course, and I run the second half as a cathode follower.
That would be the ticket for running nfb, if wanted.

Sylvania 50s 6SL7 again - what a liquid, thick sound with those   

I have to resist the temptation to 'hype'  up the sound with these and stay with the 6SN7s minimal path  setup 

 

[alexc]

Next up in the amp testing series ....  the 'Encel MFB Stereophonic'

A mid sixties japan hifi, completely original and made by Sansui or Toshiba, I think, rebranded by oem for an Australian hifi distributor  back in the day.

When this was made, the Beatles were making busy creating Sgt Pepper's LHCB

It's pretty good sounding, and I've never measured it - I replaced a couple of faulty electros, a few burnt out resistors .. but that's all.  It needs a full recap, including the HV resevoir caps. I'll do that one day!

I used this a great deal in the past 8yrs but not so much lately.

----

ax7s and 6bq5 (el84)  pp with (lots prolly) nfb and  rated at 20Wrms per side.  It has philips miniwatt early 60s tubes all round. I think two of the 6bq5 are later .. 'toshiba'

Output traffos are satisfyingly bulky for 6bq5s and bass response I know is massive with my regular sensitivity 2-way 8" vifa boxes. This one drives all my speakers with no trouble - just like a more modern solid state amp.

And it plays all day long without a hiccup. Down here, its generally quite cool temperature wise, except for high summer. In the winter, the heat is a welcome addition

----

It has variable damping and 'motion feed back' input - used on speakers with a feedback coil (which I don't have!).
Wish I had the schemo - I'm not going to draw it up unless I have to

And finally, check the nicely lighted meter  8)  It shows L or R or L+R or off  ....  far out man!

 

[alexc]

Fantastic chassis - simple, classic and no nonense   

 

[alexc]

And the inside underside  !

 

[alexc]

Side shot showing the extensively folded inner chassis with the sit-in psu traffo, chunky 6BQ4 pp traffos and the big ol' can capses.

 

[alexc]

Wiring is all a work of Japanese art!

The AX7s in the phono preamp tube has a shield that is actually pressed around the tube and soldered tight. Like a personal body shield for an AX7.

Nicely shielded jackbay with 'crystal' or 'mag'  phono inputs, mic, 'tape headamp' as well as 'aux', 'tuner', 'tape playback' and 'tape rec out'  on two seperate jack sets. Phew! ANd it's all point-point wiring in large guage hookup and shielded wire.

These guys knew how to rock it, build wise.  That's just 'how they rolled'.  Those were the days ....

-----

All those inputs are independently selectable, per channel,  on the front panel rotary switches  

There's also several stereo field modes :  'mono' from either L or R source,  or L+R  *and*  'stereo' with 2 settings.

All I know is you can set it to hear the 'mono' beatles records, released with L and R tracks  really well.  All that 'band in one speaker' and 'main vocal arrangement in other speaker' stuff works brilliantly well.

You can run the left and right channels independently : like 'crystal' on one, and 'mic' on the other. Crzay!

As well as the 'Damping' 10-pos rotary switch, which functions effectively as a 'sub bass' control - can get huge bass out of it.  The main 'Bass' control stays neutral with the 'Damping' in the middle pos5. But with really big bass!

---

AND the power - it seems like a really, really loud 20Wrms into 8ohms (has 4 and 16 taps).

It appears to be pentode -mode (no ul taps), ac balance on the phase inverter plates.

I'm sure it will be very interesting - this is my fav 'high sensitivity stereo hifi amp with tone controls'    I use it instead of the Pioneer AV power amp for general purpose lounge hifi stuff - you know, like with Cubase in the TV room.

Right then - this one's all hooked in and juiced up ready to test - I've been really wanting to know the THD and gain measures for this unit. And of course the sensitivities of the various inputs and outputs.

I know it performs very well. Now it will be time to 'see' how well.

50year old tubes! 'Made in Australia Miniwatt'  (que?)