the Poor Man 660 support thread

[lolo-m]

Also, what is the best way to disable sidechain amp completely, and cleanly take it out of the signal path? I'd like to isolate this. I'm quite positive the problem is with the 6BC8 amp, but just to be sure...

Disconnect the "FROM T4"
Disconnect the "TO TRANSF.3" header.
It will eliminate Side Chain electronics of the circuit and should be safe...

Make a report, your 6BC8 stage has got a weird gain, really. Mine is about 16dB with quite the same vumeter wiring...

 

[emrr]

I have yet to see anyone post max output numbers.   6.9dBu sounds like a pretty hot input (24.9dBu expected output of PM), and if I'm reading you correctly, you have no input attenuation at the PM?   Just to get it off the list, what happens when you put a much lower input into the unit?  Or turn the input attenuator down?  I'm sure you must have done this, but I can't tell.  It sounds like you are overdriving the unit, and have another problem as shown by your gain figure.    You mention unit setting, and modern output levels; define modern output level, and is your unity setting giving 6.9dBu output, or are you operating at different levels?

 

[Kingston]

Disconnect the "FROM T4"
Disconnect the "TO TRANSF.3" header.
It will eliminate Side Chain electronics of the circuit and should be safe.

damn that did nothing. but at least it's isolated now. I won't stop working on the 6bc8 amp until I have 18dB of clean gain. Oddly, now the VU meters pump in the opposite direction, when they should be still. I can affect it's movement somewhat by tweaking RV3 and RV6 (and turn the pumping direction!).

I have yet to see anyone post max output numbers.   6.9dBu sounds like a pretty hot input (24.9dBu expected output of PM), and if I'm reading you correctly, you have no input attenuation at the PM?   Just to get it off the list, what happens when you put a much lower input into the unit?  Or turn the input attenuator down?  I'm sure you must have done this, but I can't tell.

6.9dBu is nowhere even close to hot input. some 12dBu is pretty standard (even low these days) and I have yet to own a piece of gear that would not handle it with grace.

If I feed a gradually lower input (or turn down the input attenuator), the distortion starts to go away. Oh by the way, everything up to the 6BC6 grids seem to work fine. I can measure correct VAC RMS levels around input transformer. secondary is always 4X the VAC RMS of primary, and I can't seem to saturate the transformer itself.

It sounds like you are overdriving the unit, and have another problem as shown by your gain figure.    You mention unit setting, and modern output levels; define modern output level, and is your unity setting giving 6.9dBu output, or are you operating at different levels?

"modern level" to me is RME soundcard (or similar) at full blast, which equals to exactly 12.5dBu 50hz sine. That's the headroom I expect any normal piece of studio gear has.

And when I set *that* to unity, I get a gradual overdrive, pictured above. Also mentioned earlier, I don't see 18dB of clean gain with *any* material, any level. I get about 12dB.

What exactly is the headroom of the 6BC8 amp? What is the max level I can feed the unit, and expect clean gain when not attenuating the amp at all. Where was the 18dB amplification calculated from in the first place?

What do the RV3 and RV6 actually do? They seem to have a massive effect on gain. How can I actually measure or mathematically calculate the optimal setting for them?




To recap what *does* work,
set attenuator to "unity", then compression to full, and I get about 1-3% THD distorted compression, with what I call "modern level" material above. The distortion is audible, but not completely nasty.

But at this setting the output level is too low already (from unity). Normally at this point I'd compensate, but now we start entering the "nasty" territory.

 

[Kingston]

I now learned how to fine tune the gain, and actually how the whole 'remote cutoff triode' (ie. vari-mu) idea works. Looking at the 6BC8 data sheet, I have 130-138 of B+, and grids are biased down to -4.5V. In the tube data sheet this lands me into the amplification factor (dB) of about 18 (low resolution data sheet, hard to say exactly, depends on tube as well).

RV6 is our grid bias, and therefore the fine tune of amplification factor as well.

And this amplification factor slope is also where the time constants pump the gain (by altering grid bias), and we get compression. Why oh why didn't I see this earlier.

I now also see -4.5V was picked as the recommended value because it sits smack in the middle of the amplification factor slope, ie. balanced compression performance. I can also see the exact value is not awfully critical, as long as channels match.



Still have the same distortion, but at least I learned something.

[edit]

thinking this further, I might have got it completely wrong. maybe someone could shed a light on this.

 

[emrr]

6.9dBu is nowhere even close to hot input. some 12dBu is pretty standard (even low these days) and I have yet to own a piece of gear that would not handle it with grace.

it's an extremely hot input for any vintage tube circuit; I don't own any vintage tube limiter that can take that sort of input level.  My SA-39 is usually set for -28 to -34 dB at the input control.  Same with a Collins 26U or a BA-6A.

If I feed a gradually lower input (or turn down the input attenuator), the distortion starts to go away. Oh by the way, everything up to the 6BC6 grids seem to work fine. I can measure correct VAC RMS levels around input transformer. secondary is always 4X the VAC RMS of primary, and I can't seem to saturate the transformer itself.

"modern level" to me is RME soundcard (or similar) at full blast, which equals to exactly 12.5dBu 50hz sine. That's the headroom I expect any normal piece of studio gear has.

...but you can't put half that level into an input transformer with gain, then add all of the tube gain, without attenuation somewhere.  If you want unity throughput, you have to attenuate by the gain of the amp in front of the amp.  In a tube amp, headroom is the output max. In this unit, the attenuator knocks down level after the input transformer has added roughly 12 dB of gain, and if built to attenuate by only 18 dB, will still have a point at which it can be overdriven.   Clean output is the dB level we need a quote on; a known good unit, and yours for comparison.  If max output is below the max input of your system, it's possible you could be overdriving it to the point that it's self limiting, and reducing gain through simple overload.    For example, I have some tube preamps that have about a +8dBm output rating, and I can overdrive them to the point that they put out a listenable, compressed, +16 dBm.  You can tell they are distorted, but you would never guess the amount due to the nature of the distortion components.        

I don't see 18dB of clean gain with *any* material, any level. I get about 12dB.

right, I think that's a different problem.

What exactly is the headroom of the 6BC8 amp? What is the max level I can feed the unit, and expect clean gain when not attenuating the amp at all. Where was the 18dB amplification calculated from in the first place?

18 dB gain came from Analag measuring a signal at input and at output, and comparing the difference.   We haven't seen anyone with a well working unit comment on max output level; take that dB number and subtract 18 from it.   The actual Fairchild 670 is rated for a +8 dBm output, and has 7 dB of gain, if I recall correctly....which would be a max input of +1 dBm.   Whatever max output turns out to be, adding back the loss of the output (12ish dB?) is the 6BC8 headroom number.  All of that is without compression.  Add bias drive from the side chain amp, and the gain is reduced.  That in turn equates to greater headroom, at least in regard to the amount that can be crammed into the input before the output blows up.  

To recap what *does* work,
set attenuator to "unity", then compression to full, and I get about 1-3% THD distorted compression, with what I call "modern level" material above. The distortion is audible, but not completely nasty.

1-3% for a tube comp is pretty normal

But at this setting the output level is too low already (from unity). Normally at this point I'd compensate, but now we start entering the "nasty" territory.

Something else is going on there.

 

[lolo-m]

Disconnect the "FROM T4"
Disconnect the "TO TRANSF.3" header.
It will eliminate Side Chain electronics of the circuit and should be safe.

damn that did nothing. but at least it's isolated now. I won't stop working on the 6bc8 amp until I have 18dB of clean gain. Oddly, now the VU meters pump in the opposite direction, when they should be still. I can affect it's movement somewhat by tweaking RV3 and RV6 (and turn the pumping direction!).

When the 6BC8 output saturate the vu is mooving oddly arround 0. This is normal. And from what I recall from my old tests My TC Konnekt soundcard output couldn't be set to full on but half way up not to saturate the output without compression...

 

[Moby]

Oddly, now the VU meters pump in the opposite direction, when they should be still. I can affect it's movement somewhat by tweaking RV3 and RV6 (and turn the pumping direction!).

It's normal because it reacts on the 6bc8's current . For max gain current is minimum, and you trim that to 0. When GR starts current start increasing. Opposite, if you go more than preset gain current go lover so VU goes in "minus" or if you like it better tubes are sucking less current 

 

[andre tchmil]

I'm past page 97 now !!!

How does it sound ??

 

[inputoutput]

I'm past page 97 now !!!

How does it sound ??

DROP DEAD GORGEOUS !!!!!!!

Could someone else simply feed about 7dBu 100hz sine to their poorman, and tell me if there's any distortion?

The measurements comes out clean for the lower sinus tones. Nothing like your photos... The second and third harmonics grow larger from 400Hz and up

 

[inputoutput]

and heres a 500Hz sinuswave driving the Poorman pretty hard. Meter all the way down. (please let me add that I am absolutely no expert on how to measure distortion).  Sound goes through a Lavry DA/AD

 

[emrr]

Another assumption I neglected to mention, and a simplification; if there is something wrong with your amp stage, the headroom and output max may be affected.  As a benchmark in that situation, I'd be measuring to determine the headroom at which distortion appears acceptable.  Especially since I don't know what the expected headroom should be.

 

[lewilson]

I'll test max input of my unit and report. A badly matches set of tubes will cause big problems from my experience.

 

[Kingston]

Could someone else simply feed about 7dBu 100hz sine to their poorman, and tell me if there's any distortion?

The measurements comes out clean for the lower sinus tones. Nothing like your photos... The second and third harmonics grow larger from 400Hz and up

Thanks for your measurements.

Was that really a 7dBu sine? I can get a reasonably clean sine with much lower levels. I'll have to check the "breaking point" but I seem to recall something like -5dBu to get FFT like yours.

That 500hz wave for looks very much like my unit when driven "clean" at levels much too low compared to the rest of my set up. Some 30dB difference between the sine and its first harmonic, something like 1-2% THD. Quite a bit in my book, although emrr says this is reasonably normal.

I'll test max input of my unit and report. A badly matches set of tubes will cause big problems from my experience.

How did you match yours? I think mine are pretty strongly matched set already, but the only way I can measure is by checking the voltage drop on those 330ohm resistors, and by feeding various voltages of sines in (various grid VAC RMS voltages basically).

When the 6BC8 output saturate the vu is mooving oddly arround 0. This is normal. And from what I recall from my old tests My TC Konnekt soundcard output couldn't be set to full on but half way up not to saturate the output without compression.

It would seem I really am simply saturating the 6BC8 tubes.

it's an extremely hot input for any vintage tube circuit; I don't own any vintage tube limiter that can take that sort of input level.  My SA-39 is usually set for -28 to -34 dB at the input control.  Same with a Collins 26U or a BA-6A

Could be I simply expect too much from the headroom. Will be interesting to see more working unit measurements.

It's too bad the original designers never released such info. This thread from the very beginning has been a big guessing game when it comes to that. There's a saying in my country, "don't buy a pig in a sack". I did anyway.

 

[benlindell]

Just got my elma's, want to double check my switch theory before I put them together.

Took these values from Soren:

Step 1, Attenuation = 23 dB, Rx = 13938 ohms, Ry = 1062 ohms, Resistor = 1062 ohms.
Step 2, Attenuation = 22 dB, Rx = 13809 ohms, Ry = 1191 ohms, Resistor = 129 ohms.
Step 3, Attenuation = 21 dB, Rx = 13663 ohms, Ry = 1337 ohms, Resistor = 146 ohms.
Step 4, Attenuation = 20 dB, Rx = 13500 ohms, Ry = 1500 ohms, Resistor = 163 ohms.
Step 5, Attenuation = 19 dB, Rx = 13317 ohms, Ry = 1683 ohms, Resistor = 183 ohms.
Step 6, Attenuation = 18 dB, Rx = 13112 ohms, Ry = 1888 ohms, Resistor = 205 ohms.
Step 7, Attenuation = 17 dB, Rx = 12881 ohms, Ry = 2119 ohms, Resistor = 231 ohms.
Step 8, Attenuation = 16 dB, Rx = 12623 ohms, Ry = 2377 ohms, Resistor = 258 ohms.
Step 9, Attenuation = 15 dB, Rx = 12333 ohms, Ry = 2667 ohms, Resistor = 290 ohms.
Step 10, Attenuation = 14 dB, Rx = 12007 ohms, Ry = 2993 ohms, Resistor = 326 ohms.
Step 11, Attenuation = 13 dB, Rx = 11642 ohms, Ry = 3358 ohms, Resistor = 365 ohms.
Step 12, Attenuation = 12 dB, Rx = 11232 ohms, Ry = 3768 ohms, Resistor = 410 ohms.
Step 13, Attenuation = 11 dB, Rx = 10772 ohms, Ry = 4228 ohms, Resistor = 460 ohms.
Step 14, Attenuation = 10 dB, Rx = 10257 ohms, Ry = 4743 ohms, Resistor = 515 ohms.
Step 15, Attenuation = 9 dB, Rx = 9678 ohms, Ry = 5322 ohms, Resistor = 579 ohms.
Step 16, Attenuation = 8 dB, Rx = 9028 ohms, Ry = 5972 ohms, Resistor = 650 ohms.
Step 17, Attenuation = 7 dB, Rx = 8300 ohms, Ry = 6700 ohms, Resistor = 728 ohms.
Step 18, Attenuation = 6 dB, Rx = 7482 ohms, Ry = 7518 ohms, Resistor = 818 ohms.
Step 19, Attenuation = 5 dB, Rx = 6565 ohms, Ry = 8435 ohms, Resistor = 917 ohms.
Step 20, Attenuation = 4 dB, Rx = 5536 ohms, Ry = 9464 ohms, Resistor = 1029 ohms.
Step 21, Attenuation = 3 dB, Rx = 4381 ohms, Ry = 10619 ohms, Resistor = 1155 ohms.
Step 22, Attenuation = 2 dB, Rx = 3085 ohms, Ry = 11915 ohms, Resistor = 1296 ohms.
Step 23, Attenuation = 1 dB, Rx = 1631 ohms, Ry = 13369 ohms, Resistor = 1454 ohms.
Step 24, Attenuation = 0 dB, Rx = 0 ohms, Ry = 15000 ohms, Resistor = 1631 ohms.

here's a little diagram I threw together:

For the 2nd deck should I put the resistors in the same order and just change where 1 (out) and 3 (in) are? The other question from the schemo is which way the poles move, both down or in opposite directions? my guess is both down.

Hopefully I'm on the right track with this, I've done rotaries before but this one is just a little less set in stone.

Thanks.

 

[lolo-m]

Just got my elma's, want to double check my switch theory before I put them together.

here's a little diagram I threw together:

For the 2nd deck should I put the resistors in the same order and just change where 1 (out) and 3 (in) are? The other question from the schemo is which way the poles move, both down or in opposite directions? my guess is both down.

Hopefully I'm on the right track with this, I've done rotaries before but this one is just a little less set in stone.

Thanks.

The 1rst and 2nd deck are wired exactly the same way.

You wire an attenuator before a push pull amp. The output transformer is making this operation:
Push = Va
Pull = Vb = -Va
Va - Vb = Va - (-Va) = 2Va

The "pots" are moving the oposite way in you second drawing.

 

[Moby]

Just got my elma's, want to double check my switch theory before I put them together.

I still can't figure what's so confusing with att. switches  ??? Just imagine dual pot and if you can't grab one and "copy the situation" 

 

[benlindell]

So they are wired the same way and the resistors are arranged the same way as well?

 

[[silent:arts]]

the resistors are identical on each position.
than just follow the wiring guide

 

[benlindell]

So everything between each deck is identical. according to the table there needs to be 1062 for Ry when at pos 1 so I assume I have it in the correct place.

 

[elektrovolt]

just built another board, a newer white pcb, new parts...

but still the same problem. when I connect the board to the PSu and power up, the high voltages and meter come up and then drop again after several seconds, until they are at around 20V (136rail) and 40V (245rail) the meter gives zero (well, -20 then)
with the 5687's removed, the voltages are higher, but only half of what they should be. meter goes all the way to the right.
removing the powercord and the voltages go up!

with the 6bz8's removed too, voltages are OK.

to me, this seems the PSU can't deliver enough power - but I don't have a clue !

this just pisses me off! the same problem with the other pcb. tried different tubes, no difference.

could someone please help me with this ?