the Poor Man 660 support thread

[Kingston]

This is the first unit where I can really hear the edcor XSM 600/10k step up transformer at work. I can't say I like it much. I was surprised how grainy it sounded and how it ***** all over "s" frequencies. No good. I'll start looking for alternatives. Lundahl certainly has a few that fit my mid-side plan.

Yes, EDCORS are really poormans transformers   "****" in the high frequency is caused by huge nonlinearities in the 60k region. Firstthing I did when I recieved them was measuring and I was really dissapointed, but they are OK for prototyping  . Lundahl are really transparent and no way it's comparable to the EDCORS. If you want clean sound that's the way to go. If you ask me I will go with Jensen.
Regardind crossover distortion it's probably caused  by EDCORStoo because they are not symetrical (try to measure the Inductance). In the puh-pull arrangement it's really important. Of course I read that you matched the tubes. Hope you did that with various grid voltages  
BTW, do you have hum problem?

I would prefer jensen and cinemag to lundahl, but neither of those seem to have a compatible line 1:4 ratio step up available with dual, separate windings (or even center tap, but maybe I'm just blind). I need the dual windings for mid-side encoding. This can be hacked into the edcors as well, but now I'm thinking it's probably not a good idea due to the imbalances between the windings.

I measured the cross-over distortion further. I don't think I have matched the 6BC8's well enough. this time I indeed need to measure them with various grid voltages, but also taking into account the imbalance of the edcor windings. Lot of work, but I have a lot of 6BC8's to try at least. I can see (and hear) the cross-over distortion even if I completely take out the sidechain off the signal path. Nasty wobble near the A/B gain stage zero crossing, a rounded off discontinuity and sounds like crap.

I have no hum problem at all. I see the usual 50hz transformer bump at about -90dBu. No rectifier harmonics at all for one channel (general noise floor at about -105dBu). The other channel has negligible rectifier harmonics at about -100dBu. I have extensive (read: paranoid) shielding where ever it's possible for all wiring, and an unnecessarily thick (and heavy) steel cage for the PSU transformer (which does surprisingly little, I might remove it completely).

But really the lack of hum is probably due to the correctly over-specified and high quality PSU transformer. It doesn't get hot, and never reaches even close to core saturation.

 

[Moby]

I measured the cross-over distortion further. I don't think I have matched the 6BC8's well enough. this time I indeed need to measure them with various grid voltages, but also taking into account the imbalance of the edcor windings.

No way you can match the tubes and compensate transformer imballance, but dynamic tubes match worked for me. I have no crossover, but "ugly" charachter is present of course  It works nice as channel strip compressor, tyracking guitars,  maybe good for squashing drums but no way to use it on master material. The  F670 replica transformers can do the job too, but I will keep them for some more sophisticated varimu project. (I already started that )
Regarding hum , I had that issue, not the terrible one (sorry -105db noise sounds impossible to me, I measure with input max and no GR) so I decided to put power transformer in separate case. Now it works like a charm and I measure -80db with +4db scale. I have no problem with slow start fillament or other cables, just input transformer was picking the PSU tx magnetics.

 

[inputoutput]

The  F670 replica transformers can do the job too, but I will keep them for some more sophisticated varimu project. (I already started that  )

Moby, Do you mean the replicas that sowter makes, would they work in the PM?


---

Would be interesting to emailing Sowter + cinemag and Lundahl, to hear how much they would charge for making audio-transformers for this project.  Can anyone help me get the transformer-spesifications, and I'll email them.

 

[Moby]

I think we have to move this discussion to the "pimp" thread. Yes It's possible to change EDCORS to something more transparent but why? It works  ;D

 

[[silent:arts]]

...
Regarding hum , I had that issue, not the terrible one (sorry -105db noise sounds impossible to me, I measure with input max and no GR) so I decided to put power transformer in separate case. Now it works like a charm and I measure -80db with +4db scale...

the main problem of software analyzers. - 105dB is something like -105dBfs.
without knowing the headroom of the system it says nothing.
your measuring of -80dB referenced to +4dB results in a [realistic] noisfloor of -76dBu.

 

[Kingston]

sorry -105db noise sounds impossible to me, I measure with input max and no GR

I mean the general noise floor viewed in spectrum analyser (white noise ie. flat line). Above this you can see the peaks from transformer and rectifiers, which are the major contributors for the RMS level dBU noise (hum) measurements, about -90dBu for me (not looking at the analyser, but a VAC RMS meter).

Would be interesting to emailing Sowter + cinemag and Lundahl, to hear how much they would charge for making audio-transformers for this project.  Can anyone help me get the transformer-spesifications, and I'll email them.

At least lundahl has a variety of off-the-shelf models which can be configured as 1:4 step up input and 4:1 stepdown outputs, with dual (and even quad) windings. For each model they have a spec paper with recommended usage. Currently I'm thinking a general purpose model LL1545a would work for both output and input. LL1676 would be actually even more ideal for input (but way more expensive).

Higher mumetal shielding won't hurt either (speaking of hum removal).

I think we have to move this discussion to the "pimp" thread. Yes It's possible to change EDCORS to something more transparent but why? It works  ;D

It's not really pimping when were looking for alternatives for recommended model that works just barely, and even then is basically the lowest of lofi options. I think it's a waste of a perfectly good high end compressor not to look for better options.

count the beans and you have a 1000e unit with lofi I/O wasting much of the good the tube stages are doing. There's something wrong with that.

Remember, for sidechain use those edcors might actually do only good. They slightly shape the audio going through and this might be a highly positive thing for a time constant network.

 

[inputoutput]

It's not really pimping when were looking for alternatives for recommended model that works just barely, and even then is basically the lowest of lofi options. I think it's a waste of a perfectly good high end compressor not to look for better options.

double that!  To keep the edcores in the SC makes sense.  Some new kinds of in and out trannies would be really exiting to try out.. 

 

[Gus]

So why doesn't someone design a new transformer(s) for the amp section and have it built?  Use the same size core but change the way it is wound.  What are the Edcors pri sec pri?  

 

[Moby]

So why doesn't someone design a new transformer(s) for the amp section and have it built?  Use the same size core but change the way it is wound.  What are the Edcors pri sec pri?  

Because EDCORS can't be beaten with that price. There is a lot of drop in better or similar transformers but much more expensive and no need to design another one poorman TX 

 

[inputoutput]

Having spent this much time and money on this box, I don't see any problems with some of us going that extra mile to see if that distortion can go away.

Could anyone help me by posting the info I need for ordering custom I/O transformers?

 

[emrr]

there must be a dozen off the shelf solutions that someone could try. 

 

[1954U1]

About the Edcors, I'm waiting from Edcorusa for some WSMs, maybe they're better than the XSM TXs I too have.

 

[Holger]

My version:
3 U, Frontpanel made by Frank/NRGrecording (thank you very much, Frank).
Meters are old Sifams.
I built it the original way, no modfifications.

Very special thanks to Rowan/Analag for sharing.
I really like his design ideas, I really like my D-AOCs, I will build a second PM 670.

Thanks to Volker for the PCBs.

 

[[silent:arts]]

Holger, nice and impressive work - as always

 

[elektrovolt]

nice, it looks like a real old machine. beautiful work !

all those transformers are so close together. isn't that giving you some interference or noise ?

 

[maxheadroom]

really nice build! i like this vintage flavor!
how did you made the frontpanel color?

 

[Holger]

...all those transformers are so close together. isn't that giving you some interference or noise ?

I don't know  I assembled the unit, did a simple check like 'does it pass audio and compress?', then I took it to my new studio which isn't functional because a lot of wiring still needs to be done. Didn't had a chance to listen to the compressor under studio conditions 

...how did you made the frontpanel color?

The panel is custom painted by a local paint shop. Look here for more stuff:
http://www.groupdiy.com/index.php?topic=32677.msg424504
http://www.groupdiy.com/index.php?topic=32341.0

 

[lewilson]

Ive taken a little brake and re-tested my poorman. I should have written the results the first time.  Anyways the response is very flat from 25 cps up to 22k cps where it starts to drop . It has a little distortion at 25 cps but cleans up well at 50cps. Also I have no crossover distortion even with heavy compression. I can do  a max input level later but I think its not an issue.
Im wondering why some are having trouble with this comp? Has anyone else tested there comp with good results?  

 

[emrr]

your distortion comment sounds like what one would expect out of any vintage comp in good working order.    We still haven't seen a max input or max output #.  I'd check level with pink noise, and also with something like 30 Hz and 40 Hz sine waves, to really evaluate the bottom. 

 

[elektrovolt]

Im wondering why some are having trouble with this comp? Has anyone else tested there comp with good results?  

here still having the same problems. first, the attack / release control which did not do anything, just very slow, and when turning the pot it only gives a deeper reduction. bad frequency response, especially in the high freqs and too loud and muddy low end. (edcor?)

now when I fire it up, meter and sound comes up, and goes down again. still nothing, can't find it. quite **** actually