the Poor Man 660 support thread

[[silent:arts]]

thanks Kingston, I nearly missed that beauty

 

[phishman13]

Not sure if this has been definitively answered in the many pages of this post, but these ceramic tube sockets from tubedepot.com fit these boards perfectly:

http://www.tubedepot.com/sk-9pinpcf.html

Just an fyi.

 

[mitsos]

thanks. wish I had seen that before I bought mine... I ended up soldering pieces of resistor leads like lolo-m suggested above. Works perfectly so far, but I~ve only tested connectivity, still have to wire this beast together...

I'm not 100% sure of the resistor values for the switches, but I guess the values Kingston posted should work fine? Is that what most people are using?

cheers!

 

[Kingston]

Not sure if this has been definitively answered in the many pages of this post, but these ceramic tube sockets from tubedepot.com fit these boards perfectly:

http://www.tubedepot.com/sk-9pinpcf.html

Just an fyi.

Same sockets from a location that might work better for some:

http://uraltone.com/kauppa/product_info.php?info=p681_Ceramic-noval-tube-base--PCB-1-.html&XTCsid=9p6hpnrpemmkdqarshmcdklh41

or if you're feeling fancy:

http://uraltone.com/kauppa/product_info.php?info=p2146_Ceramic-noval-tube-base--PCB-1-G--with-gold-plated-pins.html&XTCsid=9p6hpnrpemmkdqarshmcdklh41

I'm not 100% sure of the resistor values for the switches, but I guess the values Kingston posted should work fine? Is that what most people are using?

I don't know what everybody uses, but I strongly suggest a logarithmic range from no attenuation to about -24dB for both the gain and compress rotary switches. Number of steps is up to you, but 1-2dB steps are ok. Even just a 12-step rotary would provide plenty of accuracy for 24dB range.

A note about the -24dB number: it's very non-critical, but in this particular design seems like the natural range. Use less than that, and you risk distorting the gain stages all the time. More attenuation than about 24dB will not really drive the stages enough, and would have no real world use.

I used this calculator: http://homepages.tcp.co.uk/~nroberts/atten.html

Building and wiring them was a *****! Worst DIY chore ever for me.

 

[mitsos]

A bit of a chore... I try to see if as relaxing.

So, they're supposed to be 15K, right? So basically you're saying build two of the same switch, right? I'm using 24P switches from fleabay that I got when this project first got going (along with the rest of the parts) : 

cheers!

 

[Kingston]

So, they're supposed to be 15K, right? So basically you're saying build two of the same switch, right?

Yes, 15k decks, all exactly alike. Two per balanced rotary switch. Total of 8 equally tedious decks for a stereo unit. I have gone "zen" with many DIY tasks, but unfortunately not on this one.

Next time I will shortcut with a 15-20k log dual pot if I manage to find such a thing. Actually for gain control you could just use a 600 ohm T-attenuator before the input transformer, just like with 1176 for example. Might actually be a stellar idea with the hairball version. Then just wire CV directly to input transformer secondary center tap.

 

[mitsos]

....very interesting. and time-saving as well!  So you'd just wire PL1 (1,2,3) on the schemo direct to the trafo secondary?  From what I can see, that's how the fairchild does it right? 

 

[Kingston]

From what I can see, that's how the fairchild does it right?

Yes. It will work perfectly with poorman as well. Performance will be (at least theoretically) improved since CV is balanced better to the grids through the transformer secondary.

 

[mitsos]

Next time I will shortcut with a 15-20k log dual pot if I manage to find such a thing. Actually for gain control you could just use a 600 ohm T-attenuator before the input transformer, just like with 1176 for example. Might actually be a stellar idea with the hairball version. Then just wire CV directly to input transformer secondary center tap.

I was thinking about the pot thing.. You might be able to get away with getting a custom pot from Omeg (they might even let you sample). If they won't do 15K, you could get a custom 4Gang 10K,4K7,10K,4K7, and use one 10K and one 4K7 in series for each 15K.  Would be a slightly pricey pot in small quantities, but (without looking at the schemo) I think it would work.

 

[[silent:arts]]

while prototyping I used a 10K pot in series with a 4K7 resistor.
worked fine.

 

[emrr]

A dual 25K or 50K will work fine.  Any way you do it, the DC resistance of the secondary windings will swamp the pot resistances and dictate a lot of the CV action.  With a larger pot value, the extra resistance will possibly give slightly slower minimum attack time. 

Someone needs to report on any differences with the CT of the sec tied to ground with the center of the pot/CV tie point, versus the stock condition. 

Look at the Gates Stalevel input.  Dual 100K with a 200K parallel shunt. 

 

[phishman13]

I used this calculator: http://homepages.tcp.co.uk/~nroberts/atten.html

Yeah me too, with no attenuation going all the way down to 23db attenuation.  Haven't wired it up to know its gonna be good, but I'm hoping it is since I really dont want to solder 4 of those switches more than the 1 time that i have to.  Its nice to see some words/experience that confirm my decision.

 

[Kingston]

Someone needs to report on any differences with the CT of the sec tied to ground with the center of the pot/CV tie point, versus the stock condition.

while prototyping I used a 10K pot in series with a 4K7 resistor.
worked fine.

Works, sure. But it doesn't go quiet enough (at least for my use it didn't). A better ratio would be more like 1 to 10, ie. 10k in series with 1k. But seeing what emrr writes above (that the pot resistance is not critical) maybe 25k in series with about 3k would be great. In fact the series resistor for either deck is only needed to keep the the attenuator completely in the useful log range, because -60dB attenuation is pointless.

 

[rrs]

Been a while since Ive been back here and good to see many new poor men.

Looking back over recent posts I see some progress is being made on experimenting with different transformers and tubes. I have just purchased some 6N5P tubes but may have to settle for using my ear to see if they match well. I did see lomo's test circuit but not too confident in what I'm doing to build one.

Also I noticed Lundahl LL1676's have been tried. These ave very expensive here in AUS ($180 a pop) so just wondering other alternatives. The Sowter 4383 or Cinimag CM-9589 maybe for in and out.

I have used mine for around a year now and love it. It does compress nicely with Lomo's CS but I did feel it could do with some improvement as I feel to get this thing sounding it's best requires a LOT of nursing levels and fine tuning but when it's right it sound great.

Also I notice a slight lift in the high's which I'm not that fond of though. So I felt better trannies and more attention to the tubes (quality/ matching) was in order. I did get some CDS 6BZ8 which didn't seem to exhibit the issues some here had but one tube has a blue discoloration and is distorting ever so slightly in that channel.

 

[emrr]

Today's bench test of the side chain amp; I'd show you a picture but you'd be scared.  Pile of loose parts and test clips, bench PSU. 

247VDC = 38 mA
13.7VDC at top of R10

-15dBm into T3 = -3.2dBm on sec
20.5dBm T4 pri = 8.3dBm sec
23.3dB gain through SC amp

Then disconnected T4 sec from rectifier, put 5K load on it, and did level listening tests. 
By ear, looks like approximately 10 dBm max into T3 pri with threshold control fully open, which matches T4's max rating of 2.5W. 

Worth considering when planning threshold attenuators. 

 

[rrs]

Ok brought a couple of Lundahl LL1676's for input transformers to replace my Edcor's.

So wondering there is a diagram on Lundahl site showing wiring for 1:2+2. Is this another was of referring to 1:4??
HERE
would this be how to wire for the poorman input.

 

[Kingston]

would this be how to wire for the poorman input.

No.

Input: only pin 8 (can & core) connects to ground. Pins 6 and 9 are joined together (creating a center tap), but not connected to anything. Primary in parallel, secondary in series basically.

Output: The whole transformer is reversed of course (secondary is now your primary). pin 8 (can & core) connects to ground, again, but Pins 6 and 9 are joined together and this center tap is connected to B+.

These are probably the best transformers one could possibly have for this task.

 

[rrs]

Is this better

Not sure what you mean by B+. Is that what I have referred to as output +

 

[Kingston]

The input above is correct.

Output almost as well.

B+ is the high voltage supply for your 6bc8 tubes, I think marked as 136V on the poorman, but can be anything from 110-150V really.

Look in the schematic. See what's connected to output transformer primary center tap. You need that connected to your Lundahl center tap as well (pins 9/6).

 

[rrs]

Good I understand more now.
Thanks.

Oh and also does anything need to be modified before installing 6N5P tubes replacing the 6BC8.