Frank's Pultec Support

[radiance]

I've tried without the 10K resistor. Doesn't seem to make any difference.
I'm doing some more tests...

Have a look at Steffens schematic of the Pultec which uses an opamp debalancing stage as well.
No 10K (or 620R) resistor there...

 

[Nele]

the 10k was probably taken from de gyraf-pultec. it's used there for loading the 5402 input transformer.

 

[radiance]

AFAIK, it's not needed when using an opamp debalancing stage...

 

[Hank Dussen]

OK here we go.
In short I'm building the pultec filter in 2 different ways: the classic (transformer de-/balancing and 2520 opamp for gain make-up) and Franks way (5532 based de-/balancing and gain make up). I want to be able to switch between these two way with relays. (See the previous page)
The classic circuits seems to work correctly. Franks circuit doesn't. The low boost is not enough, the low cut too much. (just like others reported)

I've compared the filter to the different schematics online.
This one seems to be the original schematic:http://home.earthlink.net/~cayocosta/diy_files/eqp_1a/eqp1a_filter_schematic.jpg
This one is a clearer copy of this: http://www.gearslutz.com/board/attachments/geekslutz-forum/268531d1324562998-pultec-schematic-discrepancies-eqporiginal.jpg
And this one is called ultra enhanced (drawn by lolo-m): http://nbns-studio.pagesperso-orange.fr/ULTRAPULTEC/ULTRAENHANCEDPULTEC.gif
These are the differences I've found between the last 2:
- The 100K resistor near the output is originally a 10K resistor.
- There should be an extra 1K resistor between the low cut pot and the output.
- There's an extra 10K resistor to ground at the input, probably for some loading purposes. I don't think it's necessary. I've tried with and without. Doesn't seem to make a difference. What's more the input of the filter is always connected to ground from the moment I turn the power on. I seems like there is some sort of connection to ground inside the 5532 opamp. But, it doesn't seem to be a problem. (see further)

If you follow the mods posted earlier by briomusic it will get you close to the original pultec filter.
Only difference I've found is that the connection between 'C' of the LF switch and the low cut pot should be on the other side of the added 1K resistor.
And of course the added 10K resistor to ground at the input.

None of all these changes seem to solve the problems I have with low frequency boost and cut.
The only thing that changed the low filtering behavior is lowering the 10K (or 100K) resistor near the output to, for example, 1K. This gives me more low boost and less cut which is exactly what I want. But it also lowered the output, changed the frequency points and changed the low frequency behavior of the 2520 version which I don't want.

I've also checked the mods described by sangeet here: http://www.nrgrecording.de/nrg/viewtopic.php?t=91&start=60
I see how changing the low cut pot from 100K to 20k would lower the maximum cut. I'm sure it works but that's not what I want to do, because the 100K works perfectly with the 2520 amp and I want to make it switchable.
I tried changing the 1K resistor (at the LF cut) to 10K as you described to change the low boost. It didn't make a lot of difference and looking at the schematic I personally don't see how it will. Are you sure that's the only thing you changed to get more low boost?
You also changed some resistors to raise the gain to get unity gain. I did this in a similar way but after the filter instead of before.

I've done some more tests (without any of the relays):
Franks debalancing -> filter -> Franks make-up amp and balancing ->wrong LF boost and cut (as mentioned before)
transformer debalancing -> filter -> 2520 based make-up and transformer -> correct LF boost and cut (as mentioned before)
transformer debalancing -> filter -> Franks make-up amp and balancing -> wrong LF boost and cut
Franks debalancing -> filter -> 2520 based make-up and transformer -> correct LF boost and cut

This made me draw the same conclusion as I did before: the filter doesn't like the 5532 make-up amp/balancing stage.

Now, enter Steffens schematic (above). I have seen this before but never really analyzed it.
I don't think there's a schematic of Franks de-/balancing/amp section available but it seems to be exactly the same as in Steffens schematic. (I haven't compared the filter section though)
BUT in Steffens schematic there an big section between the filter and the amp section which is completely left out in Franks version!
What is this section and why is it left out??
Could it be that this is some sort of buffer that changes the way the filter "looks" at the amp section?
In short is this the thing that should solve the problems I'm facing with?

 

[radiance]

That section in steffens schematic is for gain make up and is based on the neve Ba283 if I remember correctly.

As I see it, you should always use the 2520 for gain make up and just switch between transformer and opamp balancing/de-balancing.

Have you tried 5532>filter>2520>5532  ?

 

[Hank Dussen]

That section in steffens schematic is for gain make up and is based on the neve Ba283 if I remember correctly.

As I see it, you should always use the 2520 for gain make up and just switch between transformer and opamp balancing/de-balancing.

Have you tried 5532>filter>2520>5532  ?

I haven't tried it but I think this will work but as it's designed right now the 5532 does the amplifying and balancing at once.
I could, of course, lower the gain make up of the 5532 and leave this to the 2520, and have the one after the other.
But this was not my initial plan as I wanted the 2520 DOA only in the 'classic' circuit and keep Franks circuit as it is for a 'clean' version. But could definitely be a solution.

Also, the 5532 output-section of Steffens schematic is exactly the same in Franks version. So they should have the same amount of gain. I really doubt if the big section before the 5532 is just there for gain make-up as that would make 2 make-up amp in series. with a lot of gain)

Check out PRR's post is this threat about Steffens pultec: http://www.groupdiy.com/index.php?topic=6501.0
People were presuming this extra section was just there for the sound but may be they were wrong and Steffen put it there for an actual reason?

 

[radiance]

Also, the 5532 output-section of Steffens schematic is exactly the same in Franks version. So they should have the same amount of gain.

If  the 5532 output-section of Steffens schematic is exactly the same in Franks version than they both have unity gain...not the needed 20dB.

I really doubt if the big section before the 5532 is just there for gain make-up as that would make 2 make-up amp in series. with a lot of gain)

Thrust me, it is there for gain make up and the 5532 is just for balancing and does not add extra gain.

Check out PRR's post is this threat about Steffens pultec: http://www.groupdiy.com/index.php?topic=6501.0
People were presuming this extra section was just there for the sound but may be they were wrong and Steffen put it there for an actual reason?

What PRR is sugesting is maybe what Frank has done (but we never know until we see a schematic)...but that involves lowering the 20K resistor (in Steffens schematic straight after the bypass switch) to 2K. 
Either way, the Neve gain make up stage Steffen's using DOES have a sound but it might be possible that he run into the same problems you're having when he tried to use the 5532 for gain make up AND de-balancing. So  apart from the sound, that might be the reason he opted for the Neve gain make up stage.
Note that Steffen also has a 50K gain pot straight after the filter section....this might change things as well compared to your configuration.

Another thing you might want to try is this:

5532>filter>2520

AFAIK, the 2520 is capable of delivering gain make up AND de-balancing  (check Pablo's schematic)

 

[Nele]

5532>filter>2520

I think he already did...

Franks debalancing -> filter -> 2520 based make-up and transformer -> correct LF boost and cut

 

[radiance]

5532>filter>2520

I think he already did...

Franks debalancing -> filter -> 2520 based make-up and transformer -> correct LF boost and cut

No, I ment 2520 without transformer...

 

[audiophreak]

Hi ,
    I'm just finishing up a pair of these and was wondering how to implement a Bypass . I'm building these with the Gyraf SRPP  tube gain stage ,  I looked at the G-Pultec for bypass but couldnt figure out how to implement it on these PCBs , I looked at NewYorkDaves Passive EQ , and just bypasses the filter section - I think by just connecting filter signal IN to Filter signal Out with a Resistor in line - can that be done in this case ??  and how would I determine that Resistor value ??

  Thanks,
                Chip

 

[MatthisD]

My understanding of the original is that, in bypass there is a voltage divider of 10k and 909ohms to give 21.5dB attenuation. I didn't see a schematic for this version but if the passive attenuation is the same then a divider of 10k and 910ohms(or 10k+1k in parallel) would mimic the attenuation. Ruffrecords' pultec/gain stage uses a 10k, a trimmer and a 470ohm resistor as a pad with a DPDT switch as bypass. Handy if you need to fine-tune gain in bypass.

 

[audiophreak]

ahhh  ... I will look at the Ruffrecords' pultec schematic , Thanks !!!

 

[audiophreak]

Hi All ,
        Well , I've finished ( sort of )  my Franks Enhanced Pultec with Gyraf SRPP  ( Thanks briomusic ).
Its kind of a Russian version ala Kingston style with inductors from ChrioN , Gyraf G9 HT circuit slightly modified with help from ruffrecords ( Thanks Ian ) and perf-boarded G-pultec heater circuit.
    It doesn't have the HUGE bottom end like the G-pultec , but its tight and right 
I put in Igors relay bypass boards , but dont like that they are energized when circuit is IN , so with EQ IN ,  the relays are constantly powered ON ,  i just think thats a waste of power and the relay spec sheet warns about being powered for extended periods of time , would like it the other way round, I want to bypass for comparisons , if I dont like what I hear, I'd just take the EQ  out of the signal chain .. the D-LA2a is the same way , relays ON  all the time ( I used the 5v circuit from the D-LA2A ) 
... anyway , just got to figure that out and then put the lid on it.

Thanks to All and this wonderful place -  I have learned so much ... and have so much more to learn.

  here are some pics -  https://picasaweb.google.com/audiophreeek/EnhancedPultec?authuser=0&authkey=Gv1sRgCO3-1OzTr5z_dg&feat=directlink

Happy and Safe Holidays  to All 

 

[audiophreak]

Hi All,
      I know its been a while , just got back to messing around with this. In my build, I hear a low end drop off and slight level drop when EQ is switched IN  ( using Igors relay bypass PCBs .)
    I have swapped three different tubes ( ECC88 - Gyraff SRPP gain stage with briomusic hook up ) and three different output caps,I hear very slight changes but nothing that compensates with all boost and cut pots full ccw ( no boost or cut ) when relay switching circuit IN and Out .
   
  any thoughts ???  or is that normal ( how the unit sounds " In Circuit " with no boost or cut.)

 

[mitsos]

Have you swept it?  What's driving the pultec input?  Someone else had an issue with the low end using a gyraf amp, turned out it was a wrong resistor in the amp, creating a HPF. Don't remember the specifics, but look in the gyraf thread, it was fairly recently. 

good luck

 

[audiophreak]

Thank you mitsos  !!!

    I dont have the equipment to sweep it, but all boost and cuts work and all freq's change and seem close enough. Its just with my G-Pultecs when I switch between bypass and IN - there's no noticeable boost or cut in signal level or frequency change with all pots full CCW.  With this one there seems to be a slight drop in signal level and a slight bass rolloff - with all pots at full CCW ... otherwise it boosts and cuts fine and sounds pretty good too.

  I will check the other thread , Thanks again 

     

 

[r2d2]

Nice eq project ,

better more if with the Low-mid section  ,  Hi-mid secfion controls,
and in-out transformer options  (as well) ,

r.