Poor Man's Pultec EQP1-A

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I'm working on making one tube stages work with 1U of 51x so that would make an awesome project. But I'll say no more :)
 
I did not mean to imply that you could not do gain make up without an interstage transformer I was just answering a specific query about the Gyraf tube gain make up stage which does use one. Of course there are plenty of ways of achieving the necessary gain make up. My personal preference would be a mu follower stage based on a 6CG7.

Cheers

Ian
 
OK, I have done a few more sims and worked out some initial capacitor values. Looks like it is going to be very hard to avoid the need to parallel up caps in a few cases e.g. 8nF = 4n7 + 3n3 and we might need to do 15nF = 10nF + 4n7 and 150nF = 100nF + 47nF but apart from that the values have so far turned out reasonably OK. For experimental purposes I have added 200Hz lo boost/cut frequency and a 30KHz hi boost/cut frequency.

The key to creating the Pultec 'bump' is that the boost and cut turnover frequencies are NOT identical for a given frequency shown on the switch. In fact they are an octave apart although by the time you reach the switch frequency the boost or cut is pretty much fully established ( you can see these turnover frequencies quite clearly on the curves published by Pultec).  If the turnover frequencies were identical then the boost and cut would simply cancel out. So, with the values in the circuit below, with max lo boost and cut together you get a 5dB dip at the following frequencies

Code:
Switch f    Dip f
20            50
30           100
60           240
100          350
200          750

Similarly with max hi boost and cut together you get a 5dB peak at the following frequencies:

Code:
Switch f    Peak f
5K            1K6
10K           3K5
20K           5K
30K           10K

PMEQP1-A.png



Cheers

Ian
 
I have been doing some further sims and research concerning the Pultec 'bump'. It seems the little 5dB dip at the lo end is supposed to be followed by full boost at lower frequencies but the current circuit does not do this. I then realised that Pultc have the lo cut pot outside the main pot divider that the EQ is based on and has its own separate load to form its own separate pot divider for lo cut. This load could be returned to ground as you would expect but is in fact returned to the top of the lo boost pot (this is NOT immediately apparent from the way the EQP1-A circuit is normally drawn). The effect is to reduce the lo cut if the low boost is used at the same time and neatly produces the Pultec dip followed by full lo boost.

So I have redrawn the circuit to reflect this and called it the MKII. I also changed the way the hi cut/boost shelving is achieved by moving the series resistors to the wipers of the associated pots. This reduces the their interaction with the lo cut/boost pots. With the same resistor values as before you get the little 5dB upward bump reported earlier followed by a flat response. I wondered if it was possible to get the upwards bump followed by full cut (to mirror the lo bump and also be more like the peaking hi boost in the real EQP1-A) so I started by reducing the hi cut series resistor to zero which essentially gives infinite cut eventually but this destroyed the bump completely. After trying several different values I settled on 100 ohms as a good compromise and this is shown on the MKII schematic below. I'll run a couple of sims of the hi and lo bumps and post them so you can see what they  look like.

Cheers

Ian


PMEQP1-AMKII.png


 
Here are the plots as promised. First one is 60Hz with full boost and full cut. This clearly shows the dip at about 800Hz followed by the regular boost. The second shows maximum hi boost and cut at 5 KHz which produces a nice peak at about 1.5 KHz then falls away. I found an article somewhere on the net that had the lo plot from a real Pultec that we could use for comparison but I can't find it right now.

Cheers

Ian


60lobump.jpg



5Khibump.jpg



 
very nice Ian


just in these days I am developing(near to be finished) an input and output board using DOA's for pulteq eq because some my friend wanted it , one for input as sbalancer and one in output as balancer in trasformerless or trasformer output mode. It's possible using a integrated opamp DIL8 instead of the DOA. The board has two servo DC stages and does not use electrolytic caps in the signal path. Just for who wants use DOA's in a pulteq eq and no trasformers.
 


Pier Paolo
 
ppa said:
very nice Ian


just in these days I am developing(near to be finished) an input and output board using DOA's for pulteq eq because some my friend wanted it

Hi Pier,

Forgive my ignorance but what is DOA?

Cheers

Ian
 
ruffrecords said:
ppa said:
very nice Ian


just in these days I am developing(near to be finished) an input and output board using DOA's for pulteq eq because some my friend wanted it

Hi Pier,

Forgive my ignorance but what is DOA?


Cheers

Ian

DOA is discrete op amp.  

I have designed this board for the classic pulteq eq but it is good also for your eq.
my board can be used with a lot of DOA's as my doa's , JE-990 , 2520 etc..  
actually I think that with the Poor Man's Pultec EQP1-A with my board it is possible to make a very cheap
Pulteq eq but with very high performances.

Pier Paolo
 
This is shaping up to be a great project. With DOA's, should be able to squeeze 2 into a 1U case. Keep up the good work  :)
 
ppa said:
DOA is discrete op amp.  

I have designed this board for the classic pulteq eq but it is good also for your eq.
my board can be used with a lot of DOA's as my doa's , JE-990 , 2520 etc..  
actually I think that with the Poor Man's Pultec EQP1-A with my board it is possible to make a very cheap
Pulteq eq but with very high performances.

Pier Paolo

Thanks Pier, I should have guessed DOA meant discrete op amp.

What is the input impedance of your DOA? As you know I have increased the impedance of this EQ to use a 10K:10K input transformer. This means the output should preferably loaded with not less than 470K.

Cheers

Ian
 
chrispbass said:
This is shaping up to be a great project. With DOA's, should be able to squeeze 2 into a 1U case. Keep up the good work  :)

thanks

the board enters in 1U case and it is just 170x100 mm so it's possible making a stereo eq.

Pier Paolo
 
the board can use or a IC opamp on input stage and a DOA in the output or two DOA's in these ones.
The input stage has both sockets 2520 style and DIP8.
 
this is the schematic

this is the scheamtic for the traditional pulteq eq , for the Ian's eq R14 = 1M and DOA2 is APP2055,  APP99F, FT992 ONLY.
the first DOA1 can be API2520 , JE990 and others.


EDIT: new schematic for one error in the previus
 

Attachments

  • IO section pulteq eq.pdf
    239.1 KB · Views: 270
ppa said:
this is the schematic

That's a lot of "gnd" for one design  ;)
BTW, there is still a mistake I think, the power supply of the upper left OPA604 differs from the others.
 
Chris_V said:
ppa said:
this is the schematic

That's a lot of "gnd" for one design  ;)
BTW, there is still a mistake I think, the power supply of the upper left OPA604 differs from the others.

I think you will need to change the value of C250 as well if you want any HF response!

Cheers

Ian
 

Latest posts

Back
Top