Customising Gpultec freq bands

[atticmike]

Hey there,

I'd like to know what capacitor I'd have to swap the 90 hz for in order to go 70 hz for that band in the LF?

Otherwise, can someone give me an understanding of what I have to do math-wise to figure it out?

Mike

 

[gyraf]

Find the capacitor(s) that are switched in at the wanted position, and change those - lower value = higher freq.

 

[atticmike]

Find the capacitor(s) that are switched in at the wanted position, and change those - lower value = higher freq.

Is it a trial and error procedure or can I predetermine the value?

 

[warpie]

C= 1 / (pow(2pi*F) * L)

Bandwidth will change but shouldn't be a problem for only 20Hz difference.
I'm not familiar with the pulteq circuit but maybe you also need to change the series resistor (if there's any) to get the Q you want.
Q= (2pi*F)*L / R

 

[gyraf]

Pultec lows are simple R/C shelving filters

Find value of capacitor in question

Calculate load resistance from your frequency-label and this capacitor value - http://www.muzique.com/schem/filter.htm

Then calculate new capacitor value based on same load resistance and wanted-frequency.

Jakob E.

 

[warpie]

Pultec lows are simple R/C shelving filters

Ah OK, sorry then. What I said doesn't apply here 

 

[atticmike]

Pultec lows are simple R/C shelving filters

Find value of capacitor in question

Calculate load resistance from your frequency-label and this capacitor value - http://www.muzique.com/schem/filter.htm

Then calculate new capacitor value based on same load resistance and wanted-frequency.

Jakob E.

hm, just figured out it must have 5650 ohm to match the rest sort of.

However, I couldn't make this out from the board itself:

I got the pcbs from gustav but will be using this filter board:

http://www.diy-racked.com/storage/EQP1FilterPCB.jpg

Can you help me out Jakob?

And as I said, with the 5650 ohms at 60 hz, it fits the boards value but if I go 120, it'll not match (0.235 instead of 0.220).

Is it because of the 10% tolerance?

 

[gyraf]

It could be.

Also note that there are two capacitors for low, one for boost and one for cut. They are loaded differently (and hence not same value capacitor for a given frequency)

Jakob E.

 

[atticmike]

It could be.

Also note that there are two capacitors for low, one for boost and one for cut. They are loaded differently (and hence not same value capacitor for a given frequency)

Jakob E.

I mean, the filterboard is just an extended version of your build with a broader range at the HF but should be based on the same resistor values or do you something different?

 

[gyraf]

I don't know purusha's board at all

What I mean is that at 60Hz selected, there is both a 470nF (low boost) AND a 22nf (low cut) capacitor in action. Both needs changing to new value in order to get a different frequency setting, unless you want boosting at a different freq than cutting.

Jakob E.

 

[atticmike]

I don't know purusha's board at all

What I mean is that at 60Hz selected, there is both a 470nF (low boost) AND a 22nf (low cut) capacitor in action. Both needs changing to new value in order to get a different frequency setting, unless you want boosting at a different freq than cutting.

Jakob E.

yeah I got that but it looks like the values from your filter board cover the ones from prusha's since from what I know it is just as yours with a little more options at the HF:

Weirdly, at the bottom he indicates either 1 or 10k resistors where I don't know in what relation that choice might be?

 

[lolo-m]

Cut and Boost at 90Hz can give you a shef around 70... Pink noise and analyser helps you to understand what's happening.

 

[mitsos]

Either stick a smaller cap in parallel with each the 330nF and the 15nF in the low sections.  100nF for the boost and about 3n3 for the low cut, soldered under the board.

Or, add the extra frequency. Should be pretty easy with this version, since the switches are wired in.  Just move the two 120Hz wires to the 6th switch positions, the 90Hz wires to the 5th, and add capacitors that correspond to 70Hz directly to the switch at the fourth positions, and wire the other side of the caps to the PCB. Pretty clear in the images you posted where the wires should go.

For the 1K and 10K resistors, do it like in the Gpultec schemo.

 

[atticmike]

Either stick a smaller cap in parallel with each the 330nF and the 15nF in the low sections.  100nF for the boost and about 3n3 for the low cut, soldered under the board.

Or, add the extra frequency. Should be pretty easy with this version, since the switches are wired in.  Just move the two 120Hz wires to the 6th switch positions, the 90Hz wires to the 5th, and add capacitors that correspond to 70Hz directly to the switch at the fourth positions, and wire the other side of the caps to the PCB. Pretty clear in the images you posted where the wires should go.

For the 1K and 10K resistors, do it like in the Gpultec schemo.

so you are implying that the change from 90 to 70 hz will change all the other caps as well? I just don't understand the theory behind calculating the caps for 70 hz. I just want to swap the caps of the 90 hz to a value that it equals 70 hz.

 

[ruffrecords]

so you are implying that the change from 90 to 70 hz will change all the other caps as well? I just don't understand the theory behind calculating the caps for 70 hz. I just want to swap the caps of the 90 hz to a value that it equals 70 hz.

I think he was just implying you need to change two caps not just one.

The thing about the Pultec EQP1A lo boost/cut is that you can boost AND cut at the same time. So these two don't just cancel out, the frequency at which boost begins is different than it is for cut. Note that these frequencies are NOT the frequencies on the front panel switch but between 5 and ten times greater. The boost/cut happens at a rate of about 20dB per decade so to get 20dB boost at 20Hz you have to start the process of boosting at about 200Hz.

Now, in the original Pultec the ratio between the frequency on the switch and the frequencies at which boost or cut starts varies with frequency so there is no hard and fast rule. I am not sure why this is but it is probably just so they coupld use readily available component values.

Anyway, to answer your question, the boost cut is a simple RC filter so to change the frequency by 10% you simply change the capacitor by 10%. To go lower make the capacitor bigger and vice versa.

Cheers

Ian

 

[atticmike]

so you are implying that the change from 90 to 70 hz will change all the other caps as well? I just don't understand the theory behind calculating the caps for 70 hz. I just want to swap the caps of the 90 hz to a value that it equals 70 hz.

I think he was just implying you need to change two caps not just one.

The thing about the Pultec EQP1A lo boost/cut is that you can boost AND cut at the same time. So these two don't just cancel out, the frequency at which boost begins is different than it is for cut. Note that these frequencies are NOT the frequencies on the front panel switch but between 5 and ten times greater. The boost/cut happens at a rate of about 20dB per decade so to get 20dB boost at 20Hz you have to start the process of boosting at about 200Hz.

Now, in the original Pultec the ratio between the frequency on the switch and the frequencies at which boost or cut starts varies with frequency so there is no hard and fast rule. I am not sure why this is but it is probably just so they coupld use readily available component values.

Anyway, to answer your question, the boost cut is a simple RC filter so to change the frequency by 10% you simply change the capacitor by 10%. To go lower make the capacitor bigger and vice versa.

Cheers

Ian

I know this might sound like a cancer in your ears but I really can't climb behind how the theory actually works and how I can determine that 70 Hz / alter the cap to get the respective band.
I appreciate all your input to help me figure it out but I was given yet only pieces of information and not a solution / approach to the problem. The only thing I could do was to approximate the issue:

 

[mitsos]

I think you are overthinking it.  Google RC filter and reactance and let it sink in for a while... most people have a-ha moments with pultec circuits, so don't worry if it doesn't make sense now, at some point it just does. The math for the low shelves is like an RC filter, that why the link to the RC filter calculator up above.

High school algebra.  You have 2 variables in your equation, C, and F; R is a constant.

You have C and F for 60 Hz, so you can solve for R. Then use this R and 70Hz in the same equation (the "musique" calculator actually) and solve for C.  That's the cap that will give you 70Hz in that circuit.

Do the same for the low cut and you're done.

Since your resistance stays the same, you just change the cap to change the Frequency. 

Anyway, to answer your question, the boost cut is a simple RC filter so to change the frequency by 10% you simply change the capacitor by 10%. To go lower make the capacitor bigger and vice versa.

 

[ruffrecords]

I know this might sound like a cancer in your ears but I really can't climb behind how the theory actually works and how I can determine that 70 Hz / alter the cap to get the respective band.
I appreciate all your input to help me figure it out but I was given yet only pieces of information and not a solution / approach to the problem. The only thing I could do was to approximate the issue:

No problem. The 90Hz caps are currently 330nF and 15nF. To change it to 70Hz you need to add 100nF in parallel with the 330nF and 3.9nF in parallel with the 15nF.

Cheers

Ian

 

[atticmike]

thanks guys, most appreciated

am I actually good with 63v for the caps?

Mike

PS: Would you guys actually prefer metalized caps over wimas?

 

[atticmike]

Or what about the Evox Rifa caps?
Material: Metalized Polypropylene Sulfide capacitor
Long life (>200000 hours)
Leg spacing: 5mm
Voltage Rating: 100VDC/63VAC
Tolerance: +/- 10%
IEC Climate Category: 55/125/56
Temperature range: -55...+125 degrees Celsius