Poor Man's Pultec EQP1-A Build Support Thread

[Humner]

That's interesting as I have had similar ringing in the same spot - 16-17kHz - when testing an EQ(not ian's pultec, but a different style of inductor EQ) and I'm using an RME Multiface with a TL071 as the make up gain/output buffer and going straight to my RME unbalanced.

I wonder if the RME has caused some issue here.  I still haven't fully tested my poorman's pultecs since I received prescotts inductors, when I get some time I might do some similar tests and see if there any relation to the RME etc.

 

[ruffrecords]

Hi
I'm building the same 3 band Pultec  with inductors.
I want to use stepped pot instead regular pot.
How I can realize 1 db for step (lorlin switch 1 x 12)?
Suggestions?
Thanks!

You need to work out the resistor values for each step. You realise the Pultec can give as much as 20dB cut or boost but with a 12 position switch you will only be able to go from 0 to 11dB in 1dB steps. I can work out the values for you if you wish but it will take a little time.

[/quote]

 

[ilfungo]

Many Thanks Ruffrecords!!!
11 db for me it's perfect!!!
If you will help me I will wait...
THANKS!!!


there is somewhere a wiring istructions on forum? ( 3 band Pultec)

 

[jonasmaks]

I wonder if the RME has caused some issue here.

Hi Humner - unfortunately no, it is not. I've made more measurements today with different interfaces - same result.

Is the 26K bump there with the controls flat?

Hi Ian. There is no bump from 26K with flat controls, flat response was included in first graph. Anyway, today I removed "WTF" 9042 from the PCB and installed stand alone 100 mH Toko inductor in S-A place.
Guess what? No ringing whatsoever. Graph with different Q values attached. Green & Violet - 100 mH Toko, Red & Blue - first winding of "A high performance, professional audio, multi-tapped inductor, intended for line level inductor based equalisation applications."

Could it be possible that whole inductor resonates at some frequency depending on elements around? In my case brake points are ~26K for 9042 and ~16K for 9050.

Ian, I've looked at your "3 Band Pultec Test Results" again. Maybe I've become paranoid about everything, but HiBoost@10K (page 6) does not look very symetrical to me

 

[ruffrecords]

Could it be possible that whole inductor resonates at some frequency depending on elements around? In my case brake points are ~26K for 9042 and ~16K for 9050.

Quite possible. We both already know that the 10K and 16K calculated capacitance for 100mH actually result in frequencies that are too low. I had put this down to a parasitic parallel capacitance in the inductor which appears to be about 320pF. The theoretical resonant frequency of 320pF with 100mH is, surprise, surprise, 28KHz. This may well simply be the self resonance of the inductor but why it is not simply swamped by the parallel 680pF capacitor I do not know. Very odd.

Ian, I've looked at your "3 Band Pultec Test Results" again. Maybe I've become paranoid about everything, but HiBoost@10K (page 6) does not look very symmetrical to me

What makes you think it should be symmetrical?

Cheers

Ian

 

[jonasmaks]

What makes you think it should be symmetrical?

Graph in my previous message "100mH.png". Peaking with single inductor gives (at least shows) symmetrical bell shape frequency response, doesn't it? Peaking with VTB 9042 shows gausian rise but somehow exponentional decline because of new resonance nearby. Previous graph "HiBoostLoQ.png" shows that the closer new rise is - the  steeper the decline is.

 

[ruffrecords]

What makes you think it should be symmetrical?

Graph in my previous message "100mH.png". Peaking with single inductor gives (at least shows) symmetrical bell shape frequency response, doesn't it? Peaking with VTB 9042 shows gausian rise but somehow exponentional decline because of new resonance nearby. Previous graph "HiBoostLoQ.png" shows that the closer new rise is - the  steeper the decline is.

Ah, now I understand what you are getting at. Looking at the 100mH graph I take it the green and purple lines are the single 100mH at different boosts and the red and green ones are with the VTB9042. The more I look at this the more I become convinced there are two additional resonances. If you look at your original set of graphs for the high boost, in every case there is a dip somewhere between 20KHz and 30KHz followed by a peak at at 40 something KHz.

The dip implies either a parallel resonance in series with the wanted series resonance or a series resonance to ground, neither of which I can see an obvious mechanism for. The peak implies a parasitic series resonance which again is hard to visualise but both must have something to do with the construction of the inductor.

Can you do an experiment for me? Can you wire just the 100mH part of the VTB9042 into the circuit and repeat the measurements. I am just wondering if it is anything to do with parasitic resonance of the other windings in the inductor possibly via the capacitance across switch contacts.

Cheers

Ian

 

[jonasmaks]

Hi Ian,
Graph shows frequency response of VTB9042 (100 mH part) connected directly between QMAX resistor and filter input (via switch) with Max/Min Q value. Red curve represents frequency response of stand alone 100 mH inductor with 1nF condenser resonating @ ~15.9kHz. So no big strays around switch or leads or PCB…

Anyway, would be so great if someone who owns Pultec based on whose VTB 9042/9050 inductors could repeat measurements of HiBoost (and MidBoost if any) frequency response above 20K.

 

[ruffrecords]

Hi Ian,
Graph shows frequency response of VTB9042 (100 mH part) connected directly between QMAX resistor and filter input (via switch) with Max/Min Q value.

So just to be clear there is no capacitor in series with the VTB9042 so these are purely self resonances.

Cheers

Ian

 

[jonasmaks]

So just to be clear there is no capacitor in series with the VTB9042 so these are purely self resonances.

No capacitor. Switch and few centimeters of lead to inductor. I've updated graph with all taps measured. With growing inductance (resistance) quality gets lower. Black line represents VTB9050 self resonance (~18.7 kHz) with 100 mH tap connected same way.

Ian, I think you definitely should know Colin from Audio Maintenance? Maybe he could comment this? Something went wrong with particular batch of current inductors? Or this is determined by design?

I extended experiment connecting VTB9042 and 9050 directly between sound interface (FF400) mic input and output. Lines in magenta represent corresponding frequency responses in this configuration. Dips slightly moved to ~19.8 and ~30.9 kHz.

And probably I should stop looking what is wrong with whose Carnhill inductors - maybe they just not intended to work with such high (higher than, say 8-10 kHz) frequencies? As far in 1073 EQ VTB9050 works only up to 7.2 kHz?

 

[ruffrecords]

Yes, I know Colin very well.  I will get in touch with him.

Self resonance is a fact of life in inductors:

https://wiki.analog.com/university/courses/electronics/comms-lab-isr

but I had not expected it to occur so close to the audio band.

Cheers

Ian

 

[ruffrecords]

Many Thanks Ruffrecords!!!
11 db for me it's perfect!!!
If you will help me I will wait...
THANKS!!!


there is somewhere a wiring istructions on forum? ( 3 band Pultec)

It has taken a while but I have finally managed to work out all the resistor values required to get up to 11dB boost and cut in 1dB steps. I was a bit concerned that it might not be possible with the mid boost/cut because this circuit was always a bit of a kludge but it turns out not to be so hard in the end. Attached is a document describing all you should need to know.

Cheers

Ian

 

[ilfungo]

THANKS IAN!!!

 

[jonasmaks]

Hi Ian,
My MidBoost gain stays somehow high in 0dB position so I thought that adding 300 K resistor mentioned in SteppedPotsfor3BandPultec.pdf would solve this issue.
Frequency response with MidBoost enabled drops practically to flat when resistance is changed (yellow line in graph - flat is in green) - it was ~5 dB higer with 4.7K resistor (blue line) - but no boost occurs when resistors network is low (OK, I'm cheating and still using 50K pot). The way to get boost is shorten that 300 K resistor and pot (red line). How sensitive your design is for 5K/50K pots use instead of 4.7K/47K (in one post you wrote that not much, but still)? One more thing - I'm using 10K:10K Canhill input trafo with 12K load resistor instead of nominal 100K - am I right with this value? Output to tube is loaded with 470K.

 

[ruffrecords]

Hi Ian,
My MidBoost gain stays somehow high in 0dB position so I thought that adding 300 K resistor mentioned in SteppedPotsfor3BandPultec.pdf would solve this issue.

The mid boost/cut was always a compromise design. I could not find a way of modifying the original poor man's Pultec to include a proper mid section. But so many people wanted a simple to add mid section that I was forced to come up with this compromise design. One of the compromises is that using pots, the  minimum boost/cut is not zero - that's why there is a three way switch in series with the pot so it can be switched completely out. The only way to get zero boost/cut is with an open circuit. The boost pot + series resistor is in parallel with the 47K pot of the high boost, so in the nominal minimum boost position there is just under 6dB of boost. Changing the series resistor to 300K will reduce this to about dB.

Cheers

Ian

 

[jonasmaks]

Hi Ian, thank you for replay. I clearly understand that MidBoost is a compromise, I'm trying to figure out why Mid Boost stopped working (boosting) at all after changing resistor from 4.7K to 300K.

 

[ruffrecords]

Hi Ian, thank you for replay. I clearly understand that MidBoost is a compromise, I'm trying to figure out why Mid Boost stopped working (boosting) at all after changing resistor from 4.7K to 300K.

The nominal loss in the EQ is due to the 47K pot and the 4K7 pot and is  20.82dB. With 300K in parallel the 47K drops to 40.63K and the attenuatioin is reduced to  19.68dB which is a boost of 1.13dB. To get more boost the 300K needs to be reduced. Clearly my conclusion that the boost in 1dB steps to 11dB can be achieved using the 1dB cut values in sereis with 300K is completely wrong! I thought it was too good to be true! Back to the drawing board.

Cheers

Ian

 

[ruffrecords]

I found the error in my spreadsheet that led me to belive you could use one set of resistor values for mid boost and cut. I have corrected this and it is now clear that the cut resistor values are approximately one tenth the value of the boost ones. So there is no single set of resistors you can use for boost and cut. Instead a two pole 12 way switch will have to be used with cut values fitted to one pole and boost values to the other. I have updated the document to reflect these changes. The file is called SteppedPotsfor3BandPultecv0.2.pdf and you can find it in the 3BandPultec folder of the DIY tab of my web site:

http://www.customtubeconsoles.com/diy

Cheers

Ian

 

[salomonander]

hey guys,
i need to step in here as well since i have a bunch of these exact inductors that i was just about to solder onto the pcb. im really not certain after reading about the high end problems. i can do some measurements as well - but they are not soldered in yet. i have 8 of them though so if you tell me what to do ill try my best to hepl sorting this out.

 

[jonasmaks]

Hi salomonander,
If you would run your sound interface at regular 44.1K sampling rate you probably would not notice any HF problems (or at least they be minor). So you  should run tests on as high sampling frequency as you sound card allows (I've done at 96K).
I connected first tap of inductor (S-A) in series between my sound interface input and output and ran FuzzMeasure software sweeps.
You can get demo (with no much limitations) from http://supermegaultragroovy.com/products/fuzzmeasure/ or alternatively use http://audio.rightmark.org/products/rmaa.shtml if you are on PC. Now it depends on your sound interface impedance what type of curves you will get (like red in this page - https://wiki.analog.com/university/courses/electronics/comms-lab-isr if impedance is low (Mic IN) or blue if high (Instrument IN). Or just connect selected inductor to Ian's PCB on leads and run the sweeps. The place to start - http://supermegaultragroovy.com/products/fuzzmeasure/gs/

I've added primitive drawing of my test setup.
Looking forward for your results!!!!!!