PMsix61 Limiter DIY

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

alexc

Well-known member
Joined
Sep 21, 2004
Messages
2,571
Location
Hobart
Hi All

Starting to build this PMsix61 vari-gm limiter.

Based on the venerable original and using some design features and modules from Analag/Silent Arts PM660 limiter offerings.

Intended to step further towards the original 'brute force' approach with higher powered signal and sidechain amps.

The cathode bias will use either of the traditional arrangement for a 660 as well as a version of the PM660's flexible cathode-clamping. This feature was in my opinion a point of merit in the PM660 and as far as I am aware, novel to Analag's design.

Still done relatively 'poor' style that I have happily used a number of times - point-point on the top tray and pcb modules and other components inside. Standard 2RU rack suffices and the whole construction is about keeping to easily available stuff.

As you see it in the first pic, it is around 1000usd including all the tubes and everything with 200 of that in shipping down heeer in Hobart at the bottom of Australia.

I am replacing the little edcor with a utc shortly. Mostly for aesthetics  :) but also because sometimes I long for the luxuries in a transformer. 

-----

I like to use a  mix of Edcor/Hammond transformers and some exotic old old stock signal transformers where I can.

Mounted on an inexpensive 2RU, it is a single channel unit only, making construction somewhat more robust.
Tubes are also relatively low cost. Labelling panels is pretty basic stuff using Sifam printable film.

So - here is the start of the build. I hope describing it here motivates me to make a reasonably quick build which I can then use and observe whiles I optimises it.

A lot Nearly all of the information I'm talking about here is out there in the Poorman 660 Support Thread, Pimp our Poorman 660 and the myriad of  Vari-Mu threads in general.

Its all out there :)

--------

First pic of my layout which I am marking for drilling and all that.

Will take a few days and then a week or so for the front panel.  (I hate hand filing those meter slots to size  :mad: )

My style of construction is .. hasty at best and massively messy at worst.
Here I am keeping it 'simple' without too many bells and whistles in order to get reasonably robust and stable.

I'll be using it as a instrument limiter/compressor after a tube preamp channel of Orange 86+eq+drive.

The sorts of figures I want to achieve are improvements such as few dBs better noise floor, a few dBs higher output, a percent and a half better THD and certainly doing something with the high IMD issues, as discussed by Kingston in some detail in his postings. Also, a few more dBs GR available. And hopefully, achieving a slightly smoother, less grainy audio quality.

I've used all these parts in other projects and they represent the best mix of performance, cost and character I can come up with.

Time will tell!

--------------------

My reason for doing this one is that I very much admire the Poorman 660 design as well as the predessor designs too.

Apart from improved construction values, I am trying to get a stronger level from the signal amp and a more brutish level out of the sidechain amp for the option of original-like constants as well as some more general constants.

It will be interesting to note the effect of the transformer choices too.

All that in a context of applying some tube matching spells as well as ironing out some build mismatches which bring down my original PM670 a little bit.

Lowering THD under GR and examining issues relating to IMD are relatively new a concern for me.

-------

All that  IFF  it works!  ;D



 

Attachments

  • Alexc_Audio_PMsix60_pic1.jpg
    Alexc_Audio_PMsix60_pic1.jpg
    270.3 KB
Did the panel layouts - here's them papered up and centre punched.

Ready to drill then hack then file into submission.

 

Attachments

  • pm_six61_drill1.jpg
    pm_six61_drill1.jpg
    276.6 KB
And here's the panels ready to file and the top tray marked, stripped and ready for drilling.
 

Attachments

  • pm_six61_panels1.jpg
    pm_six61_panels1.jpg
    108.4 KB
And some other panels I did at the same time. 

Lots of tube stuff here  :)

Next I'll be drilling and filing the top tray as well as filing those meter slots.
 

Attachments

  • vari-mu-panels.jpg
    vari-mu-panels.jpg
    221.5 KB
Now the rack top (tubes and transformers tray) is drilled, fitted and sprayed clear to prevent rusting, and the front, rear panels all drilled and fitted, ready for the printed films.

While I'm doing that, I'm bringing the other vari mu and preamp panels to the same place  :)

 

Attachments

  • pm_six61_panels2.jpg
    pm_six61_panels2.jpg
    285 KB
Now, on the subject of the signal and control circuits I'm going to use - same topology as the original 660 with some differences.

Just about all of this is described by other forum members - I'm merely following and several years late!

- GR tubes are 4 6K4P remote cutoff  pentodes per signal phase. 8 tubes with 2.5A 6.3V regulated dc heater all up
- control amp has AX7, BH7, 2x6V6 tubes. 1.5A unregulated 6.3V dc heater all up

- PSU module for GR tubes is SS regulated and provides ~110Vdc for plates to idle at -1.5V bias and around 12mA or so.
  There is an additional regulated HV there which I may use for the GR tubes screen to allow messing with the pentode mode.
  Or  for sidechain pre+driver stages plates. Whichever seems to be best. It provides the reference -ve rail and utility +ve rail also.
  Total dc HV available is 200mA or so. Around 115mA for the signal amp.  85mA for the control amp. More or less.

- Additional rectifier+caps+choke filtered HV for sidechain pre+driver and finals stages.

Other things are :

- option of pnp transistor voltage clamping to GR tubes cathodes
- option of simple dc 'thumper' for tube balance
- option of 600R T-pad after the signal output transformer
- buffered VU meter with sensitivity switch
- 'standby' switch that disconnects HV
- seperate attack and release rotary switches

Things of interest retained from the original, in addition to the obvious beefy signal chain, brutish sidechain amp  are :

- GR bias scheme and adjustment
- GR meter with switch for monitoring plate current for signal amp +phase, both phases and -phase
- sidechain bias adjustment
- some time constants

OK - that sets the stage. Doubtless things will evolve/change as I progress.

Cheers
 
So now the panels are all done - about as good as I get them to come out usually.
One day I WILL have my chassis custom made!

I'm happy I also got three other projects' metal work to the same place.

So - tomorrow the fun begins -  I assemble the top tray and the panels.

Then the wiring begins    8)

Still drawing up the circuits - mostly got it together and working out the small parts.

 

Attachments

  • pmsix61_panels_final.jpg
    pmsix61_panels_final.jpg
    92.5 KB
Completed the basic assembly now to the 'rolling chassis' stage.

 

Attachments

  • pm_six61_front2.jpg
    pm_six61_front2.jpg
    304 KB
As well as completing the psu module and a spare    8)

I have to drop the reg voltage on these quite a bit, so I'll be doing that and the 'protection' mods described in the pm660 support thread.

Also need to do the vu meter pcb, so I'll get that started now too.

Maybe a couple of days before I have the chassis power wired up. I'll test the power, probably make up some led lights for the meters.
They are old MCI/Triplet meters. I have been fortunate to have come across these quite a few times. Anyone who used them would know what I mean.

Right then. Carry on.
 
looks impressive :eek: go go alexc.. i wish i could understand this tube marlarkey so i could build a compresser....


skal1
 
Some more progress on the wiring .. 

Still waiting on a heatsink to arrive. Rest of the wiring should be a couple of days.

I will triode connect the pentodes via a smallish screen resistor.
The rest is pretty much as per a channel of the 670.

Looking forward to starting testing and other fun stuff. Soon :)

 

Attachments

  • pm_six61_inside_1.jpg
    pm_six61_inside_1.jpg
    226.5 KB
Just been wiring up the sidechain amp - I notice the driver (12BH7) for the finals (6V6) are biased quite hot and very lean :)

Seems like it is about 2.6mA per tube with 430V dc B+ and 100K plate resistors.
And a bias voltage of 2.6V over the kathode 1K resistor.

Gain would be quite low - around x8 or so I think.  (around 1/2 of the tube mu)

So I wonder why bias  for such low current with a high plate resistor in a BH7?

Maybe there are good reasons for such low current? Save power maybe.
Those extra mA make a difference in a unit with a 12W sidechain  :)

Perhaps it will all become clearer when I finish, but I think I will go for my own op-point given I have only 300V of B+ for this tube.

I'm going to do :

300V B+, 5mA per triode, 100V plate quiescent
Plate resistors of around 40K or so with a bias of -5V on a 1K kathode resistance.


The pre-driver is about what I would expect, so I'll do that the same.







 
alexc said:
Just been wiring up the sidechain amp - I notice the driver (12BH7) for the finals (6V6) are biased quite hot and very lean :)

Seems like it is about 2.6mA per tube with 430V dc B+ and 100K plate resistors.
And a bias voltage of 2.6V over the kathode 1K resistor.

So I wonder why bias  for such low current with a high plate resistor in a BH7?

Alex,
I think the cathode resistor is set-up to get the correct feedback ratio with the tertiary winding.  I've experimented with this and its quite critical.  Not enough feedback and you don't get the 100 ohm output Z necessary for the attack speed, too much feedback and you get oscillation,  I would leave it stock myself.  The ECC83 is biased to cut-off 3.6V so only the peaks get amplified.  See Lolo's old post.
best
DaveP
 
Thanks Dave for jogging my memory.

I now do remember something about the vagaries of the first stage amplifying just the peaks.

I'll have to do some tweaking because  I only have around B+300V for the driver.
I'll have to get 'as near to it' as I can and figure out a stable set of op points from there.

Will be checking the output impedance when I finish wiring up s/c and get the psu going.

CHeers
 
Thanks!

Heeding Dave's advice, for my starting point., I'm going as close as I can figure for the sidechain driver BH7 to stay reasonably close to the ballpark :

- op-point of 70Vdc plate with 80K resistors for 2.7mA and 1K cathode resistors giving 2.7V bias with driver B+ 300Vdc
  ( I can go to 315Vdc max I think, but I like to have a bit of pi-filtering after the finals B+ of 320Vdc)

That's around where I think I can get. I can change if necessary later.

Now to improve my understanding of the dc bias control at the input of the pre-driver ax7.
Good thing I made a big doc of  much of the info contributed by all the members over the years  :)

 

Attachments

  • pmSix61_inside4.jpg
    pmSix61_inside4.jpg
    361.3 KB
Back
Top