the Poor Man 660 support thread

[[silent:arts]]

Lolo, I don't feel offended, and I think I do understand what's happening.
And I'm thankful for you, Moby, EMRR, Kingston & and all the others did trouble shooting, modifications, suggestions etc.

Evidence is that you'll have disapointed customers. Man, don't forget you sold boards at a price so poeple want to have something working...

understandable. now I understand several disclaimers I read from other PCB sellers, and will add them to any of my PCBs in the future:

I have no idea. I make PCBs, but I am a beginner when it comes to electronics. Use the search functions and the Meta-Meta.

No individual technical or other support is included in the price.

I like this one best:

XXX designs and creates circuit boards which are primarily ART but can function as electronics. As such, XXX is held harmless from any harm or damage which might occur to individuals in the use or function of any XXX product.

Kingston, for sure it would be interesting what other tube guru's would say.
but the decision to publish the circuit is up to analag, he mentioned before any PCB was sold he wouldn't be happy with it in public. the decision how it is handled now was a decision from this group, and everybody was aware of this buying the PCBs (and the schematic, that is).

 

[livingnote]

Funny this is not my project - but it does strike a chord. Considering all the work that goes
into making a working rev like the PM (and I haven't built mine yet so I don't really have too
much to say about the specific project) it's pretty amazing people sometimes misunderstand
what the forum is about - if I may say so. "sold for profit or not" is this much-trampled topic,
but what really is bugging is the flip side, people who do this idealist "DIY is nonprofit" thing
but in translation just want to pay close to nothing for what they get, don't want to spend weeks
understanding the device itself, but are very interested in getting sound worth 1000's of $ for
a few hundred - I'd say that's a profit-oriented mindset front and center, but on the side of the
buyers, not the poor developers who tear up their heini designing and supporting and distributing
and hope to get at least their costs covered with a measly margin of 20 or so € a pop. Wanting
to save 2000$ on something = making 2000$ profit. If you build 10 great-sounding
GSSLs for your studio diy-style, that's a 20'000€ profit you made right there.

So please guys don't feel offended, I don't mean to tear up a bush I'm not part of, it's just
this general undercurrent I don't like of people wanting to pay low profit prices and then take
developers to town if it doesn't just light up the sky right away, with little blood, sweat and
tears attached. Being part of a debugging/improving process is the real price that's on the
privilege of building something new, and landing in corners where things don't quite sound
like they should is just a toll gate you have to pass. If you ain't up for it, you should probably
get an ADL 670 - or how about a Gyratec?

Or alternatively be willing to pay 250€ for a PCB because that's just what development work
costs if it's been through enough tears to really sing.

The money Volker and Analag got is probably not even a third of what they put in it just to
make it happen. We should be supporting them.

 

[radiance]

Hey hey...we should move this thread to the brewery 

...& maybe move it back to the LAB when schematics are posted.

 

[emrr]

Actually plenty of successful tube comps use the 6BC8.  UA 175/176, Langevin Leveline, Altec 436; all well-loved devices - clearly not by all. 

 

[Moby]

"sold for profit or not"

Who's talking about profit? I searched the thread and can't find comment about it. To be clear I don't care if somebody is making a profit from any project here ore anywhere else (I'm the 1. one who's living from music projects ) but I don't like unfinished business. That's all folks 

 

[radiance]

this build has been a terrible pain in the ass for me, however it is working now but only not finished.

Would be nice to know what caused your problems...

 

[elektrovolt]

well, I don't know for sure, but I do have an idea: rebadged components, such as the MPSAs and IRFs I guess.
I rebuilt the PSU several times and rebuilt one of the poorman boards too.

 

[[silent:arts]]

well, I don't know for sure, but I do have an idea: rebadged components, such as the MPSAs and IRFs I guess.

it didn't happen you ordered them from Reichelt in Germany?

 

[elektrovolt]

no, I got them from a few local shops. (stuut & bruin, radio twenthe)

 

[Moby]

well, I don't know for sure, but I do have an idea: rebadged components, such as the MPSAs and IRFs I guess.

PSU design is 100% clean and working. At least that's a known an proved design so only part source or human mistake can be a reason of your problem. I bought mine parts locally too and it works as a charm.
I'm 99% sure that a lot of people has a problem with poor mains transformers. Constant 5A for heaters can saturate the TX easilly so for people with low voltages that will be the 1. thing to check.
One more thing...
Placing mains TX in the box is definitely causing the Hum on the input stage transformers (well it gaussing the output too but it's 4:1). I removed the damned "gaussator" out of the PM's box and now it's silent , but you must have a  heavy cabling to extend the 5A heating windings and some more voltage to feed slow start PSU because of possible voltage loss in the 1m cable.

 

[Holger]

I did some testing today. A couple of people mentioned that their poor man is a sort of low fi distortion box. My box is fine.

My procedure for this file: sine sweep from 20Hz to 20kHz over 30 seconds. left is bypassed, right is through PM.
In spectral view, it looks like this:

When the channels are layered, you can see the level differences:

Level wise, the channels aren't perfectly matched, but it illustrates the 'shape' of the frequency response.

And now for a large audio files (48kHz, 24bit). The source is mono (L=R). In the first 1/3 both channels are bypassed, the second third is left bypass, right compressed (tried to match the peak levels), the last third is both channels active.
I set the compressor to max threshold, used time constant 2 or 3 IIRC.
Download the files and watch/listen in your DAW.

link to drum file

And here some Apple piano loops, same procedure.

link to piano file

A fine compressor, it certainly doesn't do any harm to my audio material 
The fun really starts when pumping up the input gain, unfortunately the signal needs to be attenuated later.
Maybe I find more descriptive audio examples one day.

 

[radiance]

Thanks for posting this!

As I read this correctly your unit has a very broad 3db dip centered at 4K right? Certainly no low end roll off....

 

[Holger]

Yep...

 

[mrclunk]

nice one Holger, good news what programs that btw?

So i've had mine nearly finished for a a few months now. When i first powered it up i had the common psu problems. (my fault)
Then i heard more mains hum on the right channel than i was prepared to stand for.
Fixed that with some steel but i've now got a small distortion problem.

I'm pretty sure it is to do with tube mismatch as i've tried swapping them around for the least distortion and it certainly gets worse but no better than (annoyingly) the sets i fitted initially...
At the moment i only have six 6bz8's to play with but i'll buy some more soon. Hopefully i can find some better matches.

On most material i've tried it sounds great and i can't detect any distortion but then i tried a piece with a midi piano solo and its really quite noticeable.


original file
http://sonicamusic.co.uk/old_files/test/test%20piano.wav
through the pm670, approx 5dB of GR.  
http://sonicamusic.co.uk/old_files/test/pm670%20piano%20-5db.wav

I can hardly hear it at home but in the studio there's very audible distortion on the piano hits.

So Lomo i would be very interested in your 6BC8 matching circuit if you care to post it! please
Also have one of your SCAMP pcb's to go in, but want to get rid of the distortion first.

Some pics. I've tidied it up since i took these and added a sheet of metal between the right channel and psu. I still may move the power transformer out of the case tho.
When i've sorted out the gremlins and saved the cash i'll order a nice engraved panel.

Valve wise i've used NOS RCA 6BZ8's and GE ***** 5687's.

many thanks to Volker and Analag  ;D

 

[jackies]

So, umm, what exactly are we trying to match here?
Transconductance? Plate current? Both??
I have 2 tube testers here and wondering which parameter is the important one for 6bc8.
:

 

[Holger]

nice one Holger, good news what programs that btw?...

The screenshots were taken from Soundbooth which is part of Adobe's Production Suite.

 

[Kingston]

Fixed that with some steel but i've now got a small distortion problem.

I'm pretty sure it is to do with tube mismatch as i've tried swapping them around for the least distortion and it certainly gets worse but no better than (annoyingly) the sets i fitted initially...
At the moment i only have six 6bz8's to play with but i'll buy some more soon. Hopefully i can find some better matches.

On most material i've tried it sounds great and i can't detect any distortion but then i tried a piece with a midi piano solo and its really quite noticeable.

That's the exact same distortion me and everybody else here have, even the ones claiming they don't.

There are two significant factors, of which only one can be fixed:

1. The edcor transformers around the vari-mu tubes are not symmetric enough. It's not just the lack of quality control, they simple aren't cut out for the task. This can be easily fixed, but it gets expensive.

2. The 6BC8 vari-mu gain reduction stage will not handle "modern levels". Don't expect to be able to blast normal fully mastered 0dbFS (usually around +10 to +15dBu depending on your converter) music through it. You have to attenuate it significantly before it hits the poorman.

 

[[silent:arts]]

Thanks Holger. Have to give Soundbooth a try, looks interesting.

mrclunk, congrats, nice work.
I disagree with an engraved front-panel, the Dymo just looks way cooler ;D

 

[[silent:arts]]

...
That's the exact same distortion me and everybody else here have, even the ones claiming they don't.

There are two significant factors, of which only one can be fixed:

1. The edcor transformers around the vari-mu tubes are not symmetric enough. It's not just the lack of quality control, they simple aren't cut out for the task. This can be easily fixed, but it gets expensive.

2. The 6BC8 vari-mu gain reduction stage will not handle "modern levels". Don't expect to be able to blast normal fully mastered 0dbFS (usually around +10 to +15dBu depending on your converter) music through it. You have to attenuate it significantly before it hits the poorman.

why only one can be fixed?
if the edcors are the factor, a replacement to another brand / transformer would fix it. but this is an expensive fix.
if "modern levels" are the cause, use lower levels. you wouldn't feed an "original" with +15dBu either ...

 

[jackies]

Apparently, a set of well matched tubes would be beneficial...

So, umm, what exactly are we trying to match here?

Yeah, let's talk more about matching tubes for vari-mu devices.
I, naturally, don't expect you to just spill the info and was doing some research myself.
The ever-helpful meta even has a link:
http://www.groupdiy.com/index.php?topic=6352
Vari-mu tube matching criteria
Unfortunately, not much in terms of practical advice on there...
???
So I gather we're not trying to match Gm or current, but rather need similar gain curves?