the Poor Man 660 support thread

[prescott]

I almost finished with reading all the pages, and yes all my questions are answered.
Getting closer and closer...

Just an other one:

What about using 5670/6n3p tubes?
Different pins but also for vari-mu comps, and much more easier to source (50 USD for 50 pcs).
May be its easier to match them...(may be)
Because i like more the 6bc8 sound over the 6n5ps (more space, cleaner to me, i dont like the 6n5p sound)
I listened it without the sidechain section, 8 pcs 6n5ps and 2 different brand of 6bc8, and I liked better the 6cb8s, even they aren't balanced.

If I connect the pins correctly, would it work (5670)?

Thanks!
Peter

 

[radiance]

Sorry, I'm not a tube expert so I will leave your question for someone else to answer...

Because i like more the 6bc8 sound over the 6n5ps (more space, cleaner to me, i dont like the 6n5p sound)
I listened it without the sidechain section, 8 pcs 6n5ps and 2 different brand of 6bc8, and I liked better the 6cb8s, even they aren't balanced.

So, your saying that even with unmatched tubes you like the 6bc8 better than the 6n5ps?
This surprises me and leads me to believe that your 6bc8 tubes are somehow matched ...i.e. you got lucky.
Thing is, I never was able to judge the sound of the 6bc8's cause I never got a matched pair. When unmatched they sound like crap.
6n5p's OTOH, are way more close specced and you'll find a matched pair easily. They're also not very expensive...

 

[Kingston]

I've plotted out about 40 6BC8/6BZ8 and 6N5P tubes on a very accurate matching system and then made visual charts on excel.

On 130-150VDC B+ and the grid control voltage range that poorman uses for compression (about -1VDC to -10VDC depending on your set trimmers) 6N5P works out a nearly linear scale for compression. 6BC8 is non linear. These are averages from a decent selection. This might explain preferences, but the difference is small enough that time constant networks and attack/release settings will have a more pronounced effect on sound.

On the other hand if you don't match the tubes there will be quite horrible cross-over distortion that is not level/headroom dependent and sounds absolutely terrible. This the worst short coming of the poorman design. It needs absolutely perfectly matched tubes to not sound like a fart machine. It's of course tremendously difficult to find a perfect quad from even 40 of the aforementioned tubes. The design could be improved with individual biasing of the triode halves but then you better build another project and not be confined to the existing PCB.

 

[creal]

hi all,
Volker can you send me the schematic of the beast?
it's time to finish my PM670, the boards are populated since two or three years but i never find time for finish this compressor.
I started the wiring this afternoon, but i lost the schematic and i like to understand what's happen and how does it works.
thanks.

Cyril

 

[duantro]

I'm in the same boat as Creal- about to wire it up and don't have a schematic. I picked mine up in the BLack Market from Sintech. Could I get a copy also Volker?

 

[Kingston]

I know this has been discussed in the past but maybe it would be a good time to release the poorman schematics out in the open. It's not exactly a commercial proof design and I don't understand why it needs any special protection.

It might spark healthy discussion, revisions, optimisations and other improvement ideas.

 

[telefunk]

On the other hand if you don't match the tubes there will be quite horrible cross-over distortion that is not level/headroom dependent and sounds absolutely terrible. This the worst short coming of the poorman design. It needs absolutely perfectly matched tubes to not sound like a fart machine. It's of course tremendously difficult to find a perfect quad from even 40 of the aforementioned tubes. The design could be improved with individual biasing of the triode halves but then you better build another project and not be confined to the existing PCB.

The tube matching is absolutely as important in the original Fairchild design, only the SC tubes are more sensitive here because of the less powerfull amp. But regarding the distortion/headroom issue, the main concern here is to have perfectly matched input amp tubes for each channel, because if the matching fails there, you have phase problems from hell.  SC tubes on the other hand needed to be well matched together for the stereo link to really work.

To route the biasing to the back panel is no big deal, will save time later.

 

[Moby]

I know this has been discussed in the past but maybe it would be a good time to release the poorman schematics out in the open. It's not exactly a commercial proof design and I don't understand why it needs any special protection.

It might spark healthy discussion, revisions, optimisations and other improvement ideas.

Thumb up 

 

[[silent:arts]]

I know this has been discussed in the past but maybe it would be a good time to release the poorman schematics out in the open.

I would like to release it but need the permission from analag.

 

[Moby]

I know this has been discussed in the past but maybe it would be a good time to release the poorman schematics out in the open.

I would like to release it but need the permission from analag.

Yes, of course. But it's complete new beast if you include PSU, sidechain amp , transformers,  Tubes, etc. modifications we all published here.

 

[alexc]

I finally finished my poorman 660 channel strip

Orange 86 pre + eq with drive + poorman 660 limiter

Sounds really good - the eq and limiter are each hardware bypassable.

 

[alexc]

The eq+ drive is a fender style tone stack with an av7 type tube (ge 6bz7) at the front for eq makeup and a gain stage, feeding a 6as6 pentode with pots pre and post as well as a variable cathode bias pot. All that feeds a white cathode follower output stage with a 6N1P tube to edcor 15K:600 traffo.

I used the same tube and traffo in the Orange 86 output section.
Found I liked it better than the au7 or 6922 type.

The poorman 660 limiter users 6n5p tubes in the gr position and 5687 in sidechain.

So - 9 tubes and 9 signal transformers   

I have Kingston to thank for my respect of these russo tubes.
The more I use, the more I like.

No special matching of the gr tubes - sound of the limiter is pretty clean - no undue distortion there.
Some nice warmth but no faulty type distortion. Or thump.

Uses an Edcor psu traffo and the standard poorman 660 psu to which I added heater supply caps for a total of 50K uF - around 10K uF per amp used. The stock heater supply ripple was very high and the limiting factor for noise. The major cap-age knocked down the ripple to tiny level and noise floor dropped super dramatically.

Noise ends up being very good for the pre + limiter each and together. Something like -80dBu spl (20Hz-44KHz) for line input at unity gain (no GR) on my RTA where loopback noise level is -84dBu. The RTA is calibrated for absolute voltage on the dBu standard to the limit of my CROs ac resolution on the 2mV/div range.

The noise of the EQ+drive is a lot higher - I'm just doing some optimisation of that now.
But it looks like it will wind up as I expect it to.

Wot a mess eh?
Pretty typical for another of my franken-builds.
I like to prototype interesting combos of diy projects available here.

A nice addition to my set of crazy ass channel strip builds.

Happy christmas to all

 

[Kingston]

 

[alexc]

Thanks man - I just laffed my head off  ;D ;D

 

[shabtek]

originality -DIY excellence

 

[alexc]

Thanks Shabtek.

I am doing this concept again a bit more .. properly   ie. seperately

I like the poorman limiter so well that I am doing a v2.0 of it.
On it's own, one channel, in a 2RU enclosure (+ 3ru on top) for a total of 5ru.

The same PSU for the GR and sidechain pre-driver heater and hv.
The sidechain amp finals hv is seperate and unregulated. 
Done point-point and using the heroic sidechain amp of the original. 
And using the 8x 6K4P GR pentodes.
UTC (the cheaper ones!) signal traffos and an Edcor 10W 8K:600 for the main output.
Edcor PSU traffo and choke.
JLM vu meter buffer amp for the main outs to minimize 'collateral distortion'

The Preamp and EQ+drive will be in a seperate 2RU enclosure.
Pre will be again the Orange 86 which sounds just very, very, very good. Pretty much stock.
Haufe traffos for the mic and line inputs.

EQ will be a fancier version of the fender style preamp - using switched freqs and a 'lift' pot.

Again with a pentode 'drive' stage (inspired by Bernbrue's Rude Tube ) this time feeding a single ended mega-triode (8W plate) output stage with Carnhill 9600:600 traffo.
PSU traffo is a locally available sino one with hv and dual +/- rails feeding a jlm psu pcb for +48V and regulated heater.
A led VU meter as well.

-----

So - same concept, but 2x 2RU enclosures (7ru total space) for 1 channel of total tube goodness.
16 tubes, 9 signal transformers.
Parts around 2Kusd.
eta mid 2013

Life's short - gotta live it .. to the max!
These things are the biz for exacting intruments - like a semi acoustic jazz guitar.
Clean but with some time spent 'in the can'.

I'm trying to get the sound of my (diy) fender super amp+ross comp in a direct balanced, studio hifi context.
But without the room shaking

It's all Analag, Silent Arts, Charlie and Bernbrue 's fault.
And Leo, of course. So very much so.

I am merely a .. conduit. With a taste for Christmas brandy!

 

[bernbrue]

It's all Analag, Silent Arts, Charlie and Bernbrue 's fault.
And Leo, of course. So very much so.

Honestly, I´ve never expected my Rude Tube to be the left leg of your Frankenstein Monster. Alex, you are doing so crazy things no one ever thought of. I like that very much. So much to learn from. Go on!!
regards
Bernd

 

[prescott]

I asked in the pimp thread with no answers, hope more luck here

My 6k4p tubes just arrived
Please help me how to connect the pins to triode mode!
I've read the post about it, but not 100% clear...
Pin 1 is for grid, 3-4 heater, but dont know the others...

Thanks!
Peter

 

[alexc]

There's a few ways you can connect a pentode such as the 6K4P into triode mode.

- connect grid 2 (screen) to plate with no screen stopper resistor

or

- connect grid 2 (screen) to plate with screen stopper resistor

or

- connect grid 2 (screen) to cathode

Note in the 6K4P, the grid 3 (supressor) is hard wired internally to cathode


For use as GR tubes in a limiter, you can also use them in pentode mode with a suitable screen supply circuit.

There's a lot of info around relating to this - search for the usual suspect terms - 'vari mu', 'fairchild', 6386 etc.

It is not a sub for the 6BC8 - you would need an adapter of some kind. I also posted an adapter schematic for pentode conversion to triode in CJ's fairchild build thread.

You can get the pinout from the data sheet from this site http://www.mif.pg.gda.pl/homepages/frank/vs.html

 

[Freddy G]

Hi guys! Been a while...
I have pulled my 670 project off the shelf and dusted it off...it was 2010 whan I last looked at it and had it working with no problems. I shelved it just before I was about to add Lomo's? side chain amp booster.
So I plugged it in today and forgot that there was voltage on the meters. I accidentally shorted them against a transformer...big spark! Now I'm seeing that the rail that should read 136V is reading 245V. Did I blow Q5?
Are there still guys working and supporting this thing?