the Poor Man 660 support thread

[[silent:arts]]

would be nice if some other people post some measuring for comparison.

I would like to do this, but
a) I would need to buy a gun to get mine out of the studio where it is a the moment
b) too busy at the moment with the 51xAudio web shop
c) very busy since some friends are booked out for a three years world tour and need some technical help
d) very busy at my daytime job too (new studios to build etc)

 

[lolo-m]

Really... the increased Compression from 1k up is normal?
I'm aware of filtering side-chains but I find it strange that the Side-chain Amp ONLY test doesn't show the same response I get at the output. ???

The "Time Constant" components obviously effect the response which I find VERY interesting.
This must be where the sensitivity to the higher frequencies occurs.
It's definitely NOT coming from the "Side-Chain Amp" itself from the tests I've done.
Could someone please explain how this happens? I'd like to understand this better.

I'd love to see some measurements from others units.

The Distortion is probably my biggest concern at this point.
Look at the "Harmonic Distortion" pic I posted... now that's "Character" & probably why I really like it on some things like Bass & some guitars & Vocals... but a little much on most mixes.

Kevin, from what I learned about another varimu project I'm working on, the varimu tubes must be really well matched (6BC8 in the PM660/670). If they aren't, the frequency response is affected aswell as distortion figures when compressing. My too poor english doesn't allow me to explain this really well but what I can tell you clearly is that mismatched tubes in a varimu when compressing means :
- Longer attack times  
- Distortion  
- Bad frequency response  .
Looks like this is your main problem now.

You must check the plate current/grid tension of the 2 triodes of the 6BC8. Try to find 2 pairs of tubes which added plate current of  both triode1 and both triode2 (at several grid voltages) are the same... I'm quite sure your main problem will be solved...

 

[Moby]

I agree with Lolo. Dynamic tube matching is the crucial for Varimu performance. Before you continue to post responses please do that. Regarding 5687 , I ordered 25 tubes and they are hard to be matched. But I ended with +/- 1 ma tolerance. Mismatch in both stages are afecting the transformer performance too, so you will probably end with poor f.response and distortion.

EDIT: and yes, if you really want to improve the performance of PM you should consider exchanging T1,T2 with some "richer" option, replace 6bc8 with 6386 and add powerful scamp. Than you can erase PM in front of 670 . To be honest I don't expect clean sound from 10$. I don't know the phrase in English but in my country people like to say " As much as you pay that much music you get".

 

[1954U1]

Yes your curves arent bad at all.
About the curve of the SCamp [didnt noticed you measured it with T4 disconnected from circuit] different with 1 and 2 tubes in,
yes you definitely should match the 5687s.. [and the 6BC8s too of course!].

I'm still far from having my pm670 finished..  thanks, your posts and subsequent discussion/hints are really informative 

 

[Kingston]

Regarding 5687 , I ordered 25 tubes and they are hard to be matched. But I ended with +/- 1 ma tolerance.

How did you match them? Is the procedure simple?

Is the procedure the same for both 5687 and 6BC8?

 

[lolo-m]

Regarding 5687 , I ordered 25 tubes and they are hard to be matched. But I ended with +/- 1 ma tolerance.

How did you match them? Is the procedure simple?

Is the procedure the same for both 5687 and 6BC8?

Yes, the procedure is the same. Measure plate current at different grid voltages. Find 2 tubes where the added plate current of both triode match and you're done. This is why there's 2 tubes for each part in the design : this is a lot more simple to have well matched double triodes when you add currents.
Match should be with 136V or so for the 6BC8 and 245V or so for the 5687.

 

[Moby]

Yes Lolo with addition that 5687 can be matched only in static mode. I cant remember now exact value in the SC amp but you can measure that. Anyway, I measured all the tubes with tube tester before I dropped them in to sockets. From 25 5687  tubes (ALL JAN NOS Philips) 2 of them were too microfonic, 5 were out of range based from vademecum (5ma, 20ma), so I ended with 18 possible . I matched them to compensate current (1. dual triode one say 14ma other 12ma, then 2. 12ma/ 14ma) . I ended with 1ma tolerance. Hope you understand.  :

 

[emrr]

Or for those having only response testing software, match the tubes while looking at that, do it for both amps. Best match will be best response on the bottom end.  Remember you are balancing current in the output transformers, and the transformers themselves will contribute due to any imperfections in the individual pieces.  Tubes matched perfectly in one channel may not match perfectly in the other. 

To really look at transformer harmonic distortion, you need to inject low frequencies.  Testing at 1K misses most of what you'd hear from transformer contribution.  Test at 20 Hz, 50 Hz and 100Hz if you can; that's where the real action is.  Without compression too; just to observe best case as an audio amp. 

 

[khstudio]

Thanks everyone for your help. 

Just a few thoughts:
#1 - Both of my channels show exactly the same results... even when flipping tubes around but I haven't tried "Matching" yet.
I'm still getting a grip on how best to do it.  :-[  IF the tubes are mis-matched in BOTH channels, wouldn't they show "Different results"???

#2 - I know you guys have mentioned "Matching tubes" but I also remember reading that in "This" design it wasn't crucial  ???
... I think Rowan even said it.

Moby, Wow... you really did that much work to match them?    crazy.

Doug, good point. The high voltage feeding the plates is not exactly "Balanced" itself so.. ?
ABOUT MATCHING - Let me get this right.
Should I measure EACH side of the Primaries Resistance (from center tap out) then use that # & the voltage reading to figure the "current in each side of the output transformers" ?

Doug, I did measure other (lower) Frequencies also (without compressing) & they did not exhibit the high Harmonic Distortion I got when compressing - shown in the pic I posted. I wasn't that "deep" in the curve either so this tube matching has got my attention.

 

[Moby]

Moby, Wow... you really did that much work to match them?  Shocked  crazy.

Yes, but that takes few hours to do. I always check the resistors and capacitors before I solder them because it's much easier that way than finding the "guilty" one at the end of project . I also measure all Edcors when I received them and they are really good iron for 10$ but not good as old UTC or some Jensens. I see that people are posting responses up to 20K here but real mess is happening in range of 40-60khz what can really ruin good phase and square response. I have to fine tune that and I will post something about that under PM pimp.

ABOUT MATCHING - Let me get this right.
Should I measure EACH side of the Primaries Resistance (from center tap out) then use that # & the voltage reading to figure the "current in each side of the output transformers" ?

You can measure voltage across R1, 2, 3, 4, and using ohm law determine current for each triode. Than you will know are they matched.

 

[khstudio]

I meant the tubes. I always measure every resistor & cap (especially on this project)

You can measure voltage across R1, 2, 3, 4, and using ohm law determine current for each triode. Than you will know are they matched.

Funny... I'm wiring up test points right now to do just that.

 

[khstudio]

OK. I have test points wired for ch1 only right now.

I measured the "Resistance" of the Edcor Primary (10k:600):

222 ohms (Pins 1-2)
246 ohms (Pins 2-4)

Both channels measure the same resistance but they're not balanced side by any means.
I think this is what Doug was referring to?  If so, how does this play into it.

I'm sorry but I'm trying to get a grip on this so I can follow thru with this "Matching" to see if it improves my unit.

Thanks again. ;D

Hey... just a thought but can I add a 24 ohm resistor to balance the Primary of the Edcor?

 

[Moby]

I meant the tubes. I always measure every resistor & cap (especially on this project)

Sorry I mentioned resistors and caps but forget to add "even". Yes, I measured tubes as I described before. Static and dynamically.
Regarding DCR of TX I don't think you should bother with that. My main problem with EDCOR was inductance dis-balance but that's another story. Kevin, you own me English version of phrase I mentioned in earlier post. Money-Music 

 

[khstudio]

Kevin, you own me English version of phrase I mentioned in earlier post. Money-Music 

I tried to understand what you're trying to say but couldn't... I'll take it as a joke.  ;D

well, just finished & found one matched set & a second pretty close pair so we'll see.
These tubes can be pretty far off but I must have been lucky because the 4x 6bc8's I had in there were the best out of the batch.
I just had to swap one from each side to balance a little better but they weren't too far off.

I also measured the Voltage of the Primary of the Edcors & even though the Resistance is off the Voltage is almost dead on:
135vDC on the 222 ohm side
134vDC on the 246 ohm side
No need to focus on this I don't think.

Off to testing with my better matched pairs.
I'm skeptical because both of the channels exhibited the same high end loss when compressing but we'll see.

I wonder if the conversion from AC to DC is more sensitive to certain Frequencies???
... which in turn causes the high end to compress more?  Just a thought.

 

[Moby]

I tried to understand what you're trying to say but couldn't... I'll take it as a joke.  Grin

yee, that's a joke but true one 

BTW, I expect better bass response with matched tubes , hf roll is caused by something else.

 

[khstudio]

EDIT: and yes, if you really want to improve the performance of PM you should consider exchanging T1,T2 with some "richer" option, replace 6bc8 with 6386 and add powerful scamp. Than you can erase PM in front of 670 . To be honest I don't expect clean sound from 10$. I don't know the phrase in English but in my country people like to say " As much as you pay that much music you get".

ahhh... I missed your edited post.
Yes, you get what you pay for... That is true.  ;D

Well, after hours of measuring, matching & testing the distortion is still the same & when compressing the frequencies from about 1k on up get crushed more... just like the original pics I posted. 

At this point I give up... for now at least or until other post some measurements.
Tubes, side-chain, Transformers??? ... I have no idea & I'm beat up.

I do like this comp very much on certain things so that's where it will be used.

Could it really be the Transformers still? I wouldn't mind trying some upgrades... any recommendations of what to look for?
Cinemag may be a good choice at a fair price but I'm afraid to dump too much more time & $$$ out.

 

[tommypiper]

Sorry to hear you're still having problems, Kevin.  I've been reading along and I'm a bit lost with everything since I haven't built mine yet.  For those of us who have not built yet but are trying to understand your issues, is it basically just distortion during compression that is the problem now?

Sometime back you were saying the tubes were the fault and you got good tubes and that fixed the problem.  I asked at that time but didn't quite understand the answers, sorry.  What was wrong with the tubes and what did new tubes fix?

Is the distortion something new since you replaced the tubes?

Sorry, I'm just trying to understand what's going on.  And other guys have the distortion you have too?  But yet others don't?  Is that correct?

thanks.  (anyone else can answer too, of course...)

 

[khstudio]

Hey Tommy,

#1 - My unit DOES WORK but I'm trying to fine tune it... the problem is I'm not sure if some of things I'm trying to fix are the way it's supposed to be or not? Look at my post with the sweeps, etc... when compressing, the higher bands get compressed more (you'll have to read back to get the whole picture)

#2 - Yes I originally had problems with the Philips tubes & ordered Tung-Sols which finally worked as a pair.
I could only run 1 of the Philips in the side chain... this is what I'm doing now which works pretty well (1 tube in SC)
This helps my PT run cooler + the heat of the entire unit is lowered which I also like.

Please Note: All the recent testing I've done was with 1 & 2 tubes in the SC just to make sure THAT wasn't causing ANY of my current issues.

#3 - Yes, I'm getting a little too much distortion while compressing. I've spoken with a few other builders on the phone that have the same problem.

This compressor IS COOL... it's just a pain in the ass for me
It's going back in my rack tonight right or wrong because I like it on Bass, some vocals & some Guitars.
The BIGGEST reason I'm still working on it is I REALLY like what it does to mixes BUT the Distortion is too much.

Hopefully others will get something out of all my digging, posting, BS... whatever.

 

[idylldon]

This compressor IS COOL... it's just a pain in the ass for me
It's going back in my rack tonight right or wrong because I like it on Bass, some vocals & some Guitars.
The BIGGEST reason I'm still working on it is I REALLY like what it does to mixes BUT the Distortion is too much.

This has been my experience with it as well.  It distorts too much to use on the mix buss, which is a drag.  Unfortunately, I'm up to my neck in alligators right now with a new job and won't have time in the foreseeable future to do any more digging around or testing of my unit.  It will most likely be used in my studio in the same way you're using yours, Kevin.

Hopefully others will get something out of all my digging, posting, BS... whatever.

You've added a LOT to the knowledge about this design and I appreciate your efforts a great deal.  Thanks!

Cheers,
--
Don

 

[lewilson]

My unit does not distort when compressing.  Maybe Your starving the  varimu tubes for current. I have 100 ohm trimmers and some old vu meters that seem to work well together.
   When I first finished the pm670 I did a ruff demo mix  using about max 15 db compression  on the 2 buss.  Might not be any use because there are no before-after  to compare.
  Heres the link
http://www.thesixtyone.com/#/greenarmy/song/girl+from+the+underworld/62615/