the Poor Man 660 support thread

[telefunk]

For being such a poor man, he tends to get quite an amazing sound out of himself, dont you think? Besides, being poor and hot is a rare combination.

 

[Kingston]

For being such a poor man, he tends to get quite an amazing sound out of himself, dont you think?

no, I didn't think so. took me years to get a decent sound out of the original design, nearly a redesign.

 

[telefunk]

For being such a poor man, he tends to get quite an amazing sound out of himself, dont you think?

no, I didn't think so. took me years to get a decent sound out of the original design, nearly a redesign.

I hear you.

Because I started building mine just late last year, I had the advantage of reading all the experiences of you guys trying out things. I built mine with "Bluebird Time Constants", modded PSU, overkill power transformers and did all the intensive tube matching etc. I actually tried it once with stock TC circuit but found it utterly useless.

So, mine isn't really stock either and has never been. But it sounds really cool now.

 

[radiance]

Mine took an awe full time to build as well.
Thanks mostly to Kingston's persistence with finding the cause of the distortion and lolo's side chain booster it does sound very good now...

But I can understand Kingston's concern...it's too much time for this project and the fact that it could have been soo much easier and probably better designed...

 

[lolo-m]

I wouldn't say that the PM670 isn't well designed. I would say that the project has some lacks :
- First : it isn't a newbie project.
- Second : the name given to  the project has been a HUGE error... Most of us have never heard a real Fairchild and I'm sure most of us would be disapointed by the Stock Time Constants. The Fairchild is a legend. Its legend comes more for the fact it had the quickest attack and less distortion at that. When you name a compressor 670 or Fairchild, poeple are teased by their fantasy and expect something that will change their crapy mix or take into perfect ones by magic. Nothing's magic.

I build mine as a reseach and learning tool. I learned more with that project than with the 8 1176 clones I did without trying to understand how they work because they did work like 1176. I'm sure that Analag did this design that way.

Stock TC works. Stock TC may please some poeple and some not. Any compressor may please some and disapoint some others.

To have a really well working PM670, you have to understand how it works and how and why it is designed that way. Then, you'll be able to correct the errors you made or the things you don't like. Maybe the worst error made on this project was to design a PCB and to sell it  :-\... People thought it was a build and play DIY project and it isn't.

To conclude, I would say :
Thanks Analag for this exiting, inspiring learning tool/project.
Thanks Silent Arts for a really well designed PCB and investment in this project. Without the PCB selling, I certainly would have never learned so much because I would have never start a project like this one.

PM670 please some but it doesn't sound as a SSL or a UREI or a Fairchild 670... Is this a crime ?
Never forget there's no magic in electronics...

 

[skal1]

THis is a stock pm660

ok lads, i have one channel working now but i  seem to have an issue in the  sc or maybe it is me.



so their are a few ? that i need  answering ,thanks



1}what should be the current draw between ct and pin 1 and 4 of TX4

2}if 1 side of the tx 4 is drawing to much current what would you contribute this to {bad tx bad tubes) or a cathode rs


The reason i ask these ?  is that 1 side of the tx is drawing over 31ma and the other draws about 10ma.

3) what  type of tube circuit is in use for the side chain amp , maybe this can explian why there is more current on 1 side of the tx4

thanks in advance

skal1

 

[alexc]

I used my poorman for recording a guitar track recently. It's been in my studio rack for 18mths or so, no probs.  8)

guitar -> valve pre -> vca compressor -> poorman limiter (fast time constant) -> daw

Was very happy indeed. The poorman limiter added a whole truck load of charasma.

Apart from the simple modification of the time constant network, I don't see so much wrong at all with the design or the implementation, considering it is targeted at the 'diy' community.

I would probably recommend for the less adventurous persons, to make a pair of  'one channel in one box' units

This would avoid reported issues relating to power transformer capacity (heater supply), heat buildup, inter-channel interactions, overall wiring messiness and so on.

The PPA mods to the psu pcb worked well for me too and I'd recommend them for improved psu pcb 'mean time between failures'

That would leave basically just tube matching and wiring of the 2x dual switched attenuators as the main challenges. 
Some patience and care goes a long way here. Tube matching is the main determinant of 'audio quality' imho.
Meaning 'buy lots' and 'try lots' of tubes.

Beyond that, I would use some shielded transformers - probably trial and error to find some that would be 'better' than the specified edcors. That and a toroidal psu transformer if possible. And wait until testing to find the quietest location for the signal/control transformers (in addition to the usual toroid rotation to find quiet spot).

Makes for a markedly quiet unit. If the basic thing isn't working, you're doing some thing(s) wrong!


Do I think it deserves a 'Mk II' revision?

Sure do!

Maybe something in the way of a 'drip' treatment to make it really easy.
A little less cramped on the psu pcb, maybe with one pcb for the whole single channel and possibly sowter traffos onboard.

Sound familiar? A 'poorman' no more 

I found the poorman 'sound' desirable enough to go to the next level, inspired by Analag/SilentArts, Lolo and DaveP, not to mention CJ, PRR, EMRR and a number of notable others. Should be done in another year or so  ;D

One thing  I say - in cold weather, the poorman 670 makes a welcome heater. Serious!

PS    - for 'mastering', I don't think of the poorman 670. I use a TC Finalizer!

PPS  - one thing I'd *really* like to see - an Analag/Silent Arts 'channel strip' using this configuration.
          That would 'rawk'. In a 3Ru. None of this little 500 series stuff 

Cheers 

 

[skal1]

Ok lads just bumping the thread to see if some 1 could answer my ? about the plate current draw of amp and side chain amp.

cheers

skal1

 

[alexc]

My dual PM660 measured :

- 1 channel only connected, Edcor PSU traffo (250-0-250VAC secondary, 2x6BC8, 2x5687 tubes
- gain switch at max
- signal 1Khz, around 0dBu ( I think - was high enough to give around 12 to 14 dB GR)


245+ dc supply (sidechain amp) no signal        : min threshhold : 48mA, max threshold : 37mA, typical 48mA
245+ dc supply (sidechain amp)  signal present : same

max 48mA

135+ dc supply (signal amp) no signal        : min threshhold : 9.3mA, max threshold : 10.0mA
135+ dc supply (signal amp)  signal present : min threshhold : 7.8mA, max threshold :2.7mA

max 10mA

and

250VAC supply  no signal        : min threshhold 79mA, max threshold : 59mA
250VAC supply  signal present :  min threshhold 75mA, max threshold : 50mA

max 79mA AC HV

This is for one channel, so double it for a dual unit

Cheers

 

[emrr]

I don't have the schematic in front of me, but the current swing with no signal, only difference being threshold setting, is curious.  This suggests threshold varies SC amp grid bias, since current swings 20%.

 

[alexc]

Yep - interesting results.

I'm fairly sure they are correct measurements for my unit.
I remember trying all the knobs when measuring the currents and that's what I found.

The performance was generally as I expected so I don't think there are any gross problems with the build.

I did note that the sidechain results gave 37ma only at max setting and 48mA at all except max.
Possibly I could have a threshhold switch error on that max setting.
Safe to read 48mA for all settings, I guess.

The signal amp with no signal present definately did vary with threshold by around 7%

 

[emrr]

Hmm, trying to remember what interactions might be more active with threshold at the top of the voltage divider.  T3 sec CT connected to a bias?  If so, more directly connected to grids at atten max gain. 

 

[alexc]

T3 secondary is not connected to bias.

 

[emrr]

Not that then. 

Someone can quote my old post better than I remember probably; I recall measuring 36-38 mA for SC amp with no signal and atten equivalent grid to ground resistors. 

 

[skal1]

Thanks alex for the test result , looks like my meter is shot will have to get a new 1 today and measure again.

cheers

skal1

 

[alexc]

I have another channel on my bench at the moment - going to check it and I'll post those figures for comparison.
I'm interested in taking another look.

Cheers

 

[Holger]

I've updated my 'temperature' post. I think it's better to keep all my measurement and findings there.
http://www.groupdiy.com/index.php?topic=28274.msg613484#msg613484

 

[living sounds]

I've finally started working on my PM670 again...

Got one channel connected, and the voltages are good, it's passing audio, but what comes out is at least 20db below unity gain, and changing the settings of the gain or threshold switches only affects the amount of distortion I'm getting. Clearly something is wrong. But what's really worrying me are the two 5687 tubes, I've measure their temperature, it exceeds 100° C... wrong resistors somewhere maybe?

Thanks!

 

[alexc]

Are you getting the heater voltage you expect?
One channel of poorman should be around 2.5A

The heater circuit is pretty straight forward - assuming you have soldered the sockets without shorts,
and your heater dc wires are not shorted, you would expect excessive current to cause excessive heat.

I would expect the 5687 get close to 100C at the surface of the glass.
I haven't measured the temp, but it certainly ain't cool!

So - measure the voltage at the poorman pcb with tubes and if you can safely, the heater current.

With respect to the signal side of things, you say the voltages are OK. so I assume your (approx) 245V, 135V and -17V are all in the ballpark.

That being the case and assuming you have the rather simple pcb with tubes populated correctly,

check and check again the transformer and switch wiring.

Those edcors can be pretty delicate with the magnet wire to lug solder, so if possible trace signal from onto the primary of T1 and then from the secondary of T1 thru to the Level switch. Check the Level switch does what you expect - attenuates the input signal.

Get that going right before moving on to the secondary of T2, checking for audio signal.

Obvioiusly take care of high voltages, especially on the primary of T2 and primary of T4. These should all be heat shrink insulated.

 

[living sounds]

Are you getting the heater voltage you expect?
One channel of poorman should be around 2.5A

The heater circuit is pretty straight forward - assuming you have soldered the sockets without shorts,
and your heater dc wires are not shorted, you would expect excessive current to cause excessive heat.

I would expect the 5687 get close to 100C at the surface of the glass.
I haven't measured the temp, but it certainly ain't cool!

So - measure the voltage at the poorman pcb with tubes and if you can safely, the heater current.

With respect to the signal side of things, you say the voltages are OK. so I assume your (approx) 245V, 135V and -17V are all in the ballpark.

That being the case and assuming you have the rather simple pcb with tubes populated correctly,

check and check again the transformer and switch wiring.

Those edcors can be pretty delicate with the magnet wire to lug solder, so if possible trace signal from onto the primary of T1 and then from the secondary of T1 thru to the Level switch. Check the Level switch does what you expect - attenuates the input signal.

Get that going right before moving on to the secondary of T2, checking for audio signal.

Obvioiusly take care of high voltages, especially on the primary of T2 and primary of T4. These should all be heat shrink insulated.

Thanks alot, I'll tackle this one more time this week!