the Poor Man 660 support thread

[Silvas]

Actually, the 6n5p seems to be cheaper than the 6bc8 !

So, it wil be cool if you guys can try different input transformer possibilities and report here. I cannot (nothing at hand !) but i was playing a bit with RV3 and RV6 values with the stock PM670, hoping to shape a bit the distortion. So i left RV6 at -4.5 and tweaked RV3 to have minimum 3rd order distortion, using a 1k tone, unity gain, no compression, (-14dB referenced to 0dBFS, interfaces calibrated at 0dBFS=+18. A bit hot for a stock PM i think?). It took off some of the unpleasing distortion and left me with around 1%...mostly 2nd order. I ended up with something like -3.5 volts for the left channel and -3 volts for the right one. I had to tweak my VU´s after (which are unstable ! have to look at this later).

Best regards.

 

[elektrovolt]

good job, Kingston!
I will get some 6n5p's then for my other 6bc8 box (drip 175) as I sold my poorman...

 

[radiance]

good job, Kingston!
I will get some 6n5p's then for my other 6bc8 box (drip 175) as I sold my poorman...

Why is that? Having problems with your drip 175 when using the 6bc8 tubes?

 

[elektrovolt]

well,
I had a lot of bad 6bc8's. bad - as in huge unbalance in the triode twins but different curves in one tube also.
Finally I found four good and somehow matching 6bz8's rca tubes for the poorman.
for the 175 I have two fine GE's, but I would like to experiment with it when it is finished.
Don't have the money for the Sowter transformers... some day...

 

[Kingston]

I have now done some comparing of the better matched 6bc8 vs. 6n5p tubes now that I had the time to go through all that I had. There's more to it than meets the eye, umm, ear.

For example, I cannot get the 6bc8 overdriving in pleasant ways. When overdriven, 6bc8 crackles in a nasty way, pointing at a sharp and more linear cutoff in the breakup curve. 6n5p break up is so soft it's hard to notice, but has an odd smearing effect. It would seem they both actually have an equal headroom, but 6n5p sounds like it has more range due to the very soft break up.

Furthermore, 6n5p curves always measure somewhat different shape than 6bc8, and this has an obvious effect on compression. 6n5p pumps in a more obvious way with same gain reduction as 6bc8, which is a good thing in my book, but others might not like it. 6bc8 vari-mu curve is more linear than 6n5p.

To summarise for a well matched set of tubes in poorman: 6bc8 is the more accurate but subtle "mastering tube", while 6n5p is a "vintage warmer".

I had to tweak my VU´s after (which are unstable ! have to look at this later).

With the way they are configured right now, measuring current draw of the vari-mu amp B+, there's no way to get them absolutely stable. Tubes are "living things" and can deviate a lot even during one day. Maybe after some months of usage they will finally stabilise in poorman, probably depends in a tube. The VU's of poorman are more eye candy than an accurate measurement.

 

[Silvas]

I have now done some comparing of the better matched 6bc8 vs. 6n5p tubes now that I had the time to go through all that I had. There's more to it than meets the eye, umm, ear.

For example, I cannot get the 6bc8 overdriving in pleasant ways. When overdriven, 6bc8 crackles in a nasty way, pointing at a sharp and more linear cutoff in the breakup curve. 6n5p break up is so soft it's hard to notice, but has an odd smearing effect. It would seem they both actually have an equal headroom, but 6n5p sounds like it has more range due to the very soft break up.

Furthermore, 6n5p curves always measure somewhat different shape than 6bc8, and this has an obvious effect on compression. 6n5p pumps in a more obvious way with same gain reduction as 6bc8, which is a good thing in my book, but others might not like it. 6bc8 vari-mu curve is more linear than 6n5p.

To summarise for a well matched set of tubes in poorman: 6bc8 is the more accurate but subtle "mastering tube", while 6n5p is a "vintage warmer".

But, in your oppinion, is there a better chance of getting a right pair (or quad) using the 6n5p, assuming you are matching them just by ear ?

I had to tweak my VU´s after (which are unstable ! have to look at this later).

With the way they are configured right now, measuring current draw of the vari-mu amp B+, there's no way to get them absolutely stable. Tubes are "living things" and can deviate a lot even during one day. Maybe after some months of usage they will finally stabilise in poorman, probably depends in a tube. The VU's of poorman are more eye candy than an accurate measurement.

BTW, my meters are tracking exactly 1/2 of the actual measured G.R ! i tried removing both the diode bridge and internal resistor of my nissei VU´s without luck. Any possible cure ?

Nice, nice work on the tubes thing, Kingston ! it encourages me to try some things on my PM.

 

[Kingston]

But, in your oppinion, is there a better chance of getting a right pair (or quad) using the 6n5p, assuming you are matching them just by ear ?

Whoah, I'm not matching these by ear! I'm using lolo's gain matching rig, and meticulously measuring each tube (both triode sides) with 7 grid voltage points.

And this far, it seems there's a better chance finding a matched 6n5p quad. They are cheaper tubes as well, so a better deal in every way.

[edit]

just to avoid confusion, the above 6bc8 vs. 6n5p post was about comparing how they sound as matched quads. They don't sound the same.

 

[Silvas]

But, in your oppinion, is there a better chance of getting a right pair (or quad) using the 6n5p, assuming you are matching them just by ear ?

Whoah, I'm not matching these by ear! I'm using lolo's gain matching rig, and meticulously measuring each tube (both triode sides) with 7 grid voltage points.

And this far, it seems there's a better chance finding a matched 6n5p quad. They are cheaper tubes as well, so a better deal in every way.

[edit]

just to avoid confusion, the above 6bc8 vs. 6n5p post was about comparing how they sound as matched quads. They don't sound the same.

Yep, i know you are tightly matching the tubes using lolo´s circuit. I´m matching them by ear ! but i dont have lots of 6BC8 so i want to give the 6N5P a shot...please excuse my crappy english !

 

[lolo-m]

BTW, my meters are tracking exactly 1/2 of the actual measured G.R ! i tried removing both the diode bridge and internal resistor of my nissei VU´s without luck. Any possible cure ?

It seams that your Vumeter isn't sensitive enough. Did you try to remove the vumeter resistor only ? I don't know the nissei specs, but most of the time, resistor act as a voltage divider in the vu (in our case).

With the way they are configured right now, measuring current draw of the vari-mu amp B+, there's no way to get them absolutely stable. Tubes are "living things" and can deviate a lot even during one day. Maybe after some months of usage they will finally stabilise in poorman, probably depends in a tube. The VU's of poorman are more eye candy than an accurate measurement.

The vu meter topology is exactly the same than Fairchild's. And to me it is a clever position. It isn't really accurate most of the time, but who cares ? A varimu must be used with ears only as there's no make up gain pot. The vu helps you to know if you're compressing a lot or not. That's all.
PS: I bought some 6N5P 6 months ago, I should give them a try too !

 

[Kingston]

lolo-m, I have forgotten to thank you. It's pretty much you who turned around the whole "poorman" experience for us. The tube matcher was an absolute necessity since there aren't four tube's in parallel to even out the responses like in the "richman".

thanks!

 

[jackies]

nevermind.

Great effort on your part!
Now, to solid state sidechain - that has to be pretty simple, a 3886 chip would probably do.
What kind of output voltages we looking for though?
Like, would 3886 on +/- 17v be enough? Or would it be better with +/- 35v?

 

[lolo-m]

lolo-m, I have forgotten to thank you. It's pretty much you who turned around the whole "poorman" experience for us. The tube matcher was an absolute necessity since there aren't four tube's in parallel to even out the responses like in the "richman".

thanks!

You welcome ! Sharing is one of the reason I'm here for. Learning is the other.

nevermind.

Great effort on your part!
Now, to solid state sidechain - that has to be pretty simple, a 3886 chip would probably do.
What kind of output voltages we looking for though?
Like, would 3886 on +/- 17v be enough? Or would it be better with +/- 35v?

-/-17 is enough... You can also make a kind of copy of the PRR threshold detection followed by my SCAB... easy and cheap... but again this should be in the Pimp Thread  !

 

[Kingston]

Now, to solid state sidechain - that has to be pretty simple, a 3886 chip would probably do.

Can't say I'm not excited of the prospect. I mean, take out the inefficient tube power amp, replace it with over-designed chip amp side-chain, and drive 6-8 6K4P's per channel with a wide voltage swing. Maybe later this year...

 

[radiance]

Quick question on the slow start heater psu boards...

As has been said before, you'll get (and I'm getting) nowhere near DC with the 1000uF caps.
Actually I measured something like 0,5V AC.

What are the benefits of having true DC heater voltages? Less shielding needed maybe?

Man, my build is going sooo slow, but I like it. Along the way this project is changing and it keeps me from making final decisions on component placement and metal work. There was a point were I wanted to sell my poorman but I'm glad noone was interested cause I like the new direction this project is heading. Solid state side chain, more tube choices, transformer replacements...

I've some Haufe 5223 5:1 transformers which seem like a perfect input (T1) replacement. Someone more wise than me would like to comment on this choice? Info can be found here http://audio.kubarth.com/rundfunk/index.cgi#trafos

Thanks!

 

[Kingston]

Solid state side chain

that's a new project already. But Scamp extension makes the poorman sidechain partly solidstate.

I've some Haufe 5223 5:1 transformers which seem like a perfect input (T1) replacement. Someone more wise than me would like to comment on this choice? Info can be found here http://audio.kubarth.com/rundfunk/index.cgi#trafos

Why not. It handles plenty of level. But no dual windings so you can't play mid/side tricks with these.

What are the benefits of having true DC heater voltages? Less shielding needed maybe?

If you manage it *clean* DC, it makes the unit less subject to hum. AC heater wiring is a very critical thing to get just right.

But it really doesn't have to be regulated at all. It's actually a massive source of heat and really inefficient, and mostly unnecessary. The greatest benefit is absolutely stable voltage that doesn't deviate with mains, but that's not a problem in the modern world.

 

[Silvas]

Hello. Turned out that my meter (and control interaction and weirdness) problems were related to too much current being pulled from the 5687´s....so, for now, i took khstudio´s route on pulling one tube from the sidechain. Then, i adjusted the heaters again (6.5 was the closest i could get with the power resistors,  a bit on the high side but don´t mind) and voila ! Stable meters and no more interaction ! My tracking isn´t exact, tho. I´m reading 6dB´s on the VU for 10dB measured G.R. I tweaked rv3 again for minumum distortion at -14dBFS (converters at +18), voltages ended up -2,54 left and -3 right...and i´m very happy now. No ugly distortion for me, unless i slam the input. Mostly 2nd order...trafo exchange somewhere in the future ! I´m trying the other tubes and see (hear) what happens. Not sure if i´ll use this baby on the 2mix, maybe when i put the 6N5P. Need new 5687´s to try, but i wouldn´t mind running the beast with only one tube. To my ears, the difference is so small and the benefits are many (MUCH less heat everywhere !). My 5687´s are JAN Phillips, date code 8639, blue lettering, and i´m using edcor power TX, which was getting TOO hot. maybei have same tubes as khstudio´s problematic 5687 tubes.

I just need to put on the bypass boards and i´m done. Very very nice comp, indeed. I´m loving it on acoustic gtrs. Not so nice on drums, cool on bass and vocals. Tried it on a grand piano, nice, but you have to be picky with your levels. I found it quite transparent on trumpet and trombones. Tried it on a string section and it ruled as a "character" compressor.

I´ll put lots of pics as soon as i´m done with the bypass boards and i need to tidy it up.

Best regards.

 

[Silvas]

BTW, my meters are tracking exactly 1/2 of the actual measured G.R ! i tried removing both the diode bridge and internal resistor of my nissei VU´s without luck. Any possible cure ?

It seams that your Vumeter isn't sensitive enough. Did you try to remove the vumeter resistor only ? I don't know the nissei specs, but most of the time, resistor act as a voltage divider in the vu (in our case).

With the way they are configured right now, measuring current draw of the vari-mu amp B+, there's no way to get them absolutely stable. Tubes are "living things" and can deviate a lot even during one day. Maybe after some months of usage they will finally stabilise in poorman, probably depends in a tube. The VU's of poorman are more eye candy than an accurate measurement.

Yes, i tried first to remove the resistor only. A bit better....then, i noticed some kind of weird induction over the VU cables, which was causing the VU´s to move as i moved the cables around the case...so i removed the bridge and that solved it.

Maybe removing or replacing the 22K and just left the VU with the 5k trimmer ? have to try that. Anyway, i can live with the VU´s behaving like that...this is a comp you want to really use with your ears instead of yor eyes !

 

[envelope]

Hi,
here´s a chart with an English translation of the PM670 power transformers wire colors from the European groupbuy:

• 
  R2-0134-002  Rondo 134VA
  Prim: 2x115V  yellow-white / yellow-white
  Sek.1: 250V/0,2A black/black
  Sek.2: 9V/6,0A grey/grey
  Sek.3: 2x15V/2x1,00A red-green/white-blue
  tx shielding:  green yellow (connect it to ground)

Volker, would you please set a link to this post on 1st page?
regards
Bernd

Hello, can anyone comment on the use of this power transformer over the EDCOR version, which I have read is under spec'd
thank you, Stuart

 

[Silvas]

Hi,
here´s a chart with an English translation of the PM670 power transformers wire colors from the European groupbuy:

• 
  R2-0134-002  Rondo 134VA
  Prim: 2x115V  yellow-white / yellow-white
  Sek.1: 250V/0,2A black/black
  Sek.2: 9V/6,0A grey/grey
  Sek.3: 2x15V/2x1,00A red-green/white-blue
  tx shielding:  green yellow (connect it to ground)

Volker, would you please set a link to this post on 1st page?
regards
Bernd

Hello, can anyone comment on the use of this power transformer over the EDCOR version, which I have read is under spec'd
thank you, Stuart

AFAIK, The Rondo heaters secondary is 6A. Same as the Edcor....is this right?

 

[radiance]

yep.

I use this rondo to power the heater of 1 pm660 board. The other heater has it's own (8A) rondo (toroid).
None of them gets hot, not even mildly warm....