I obtained a pair of soviet era inductors that I was hoping to use for the high boost in the Pultec circuit but I can't figure out how to achieve the same curves. Is it possible to use these values in the circuit;
340mH, 910mH, 1350mH, 3210mH.
I've managed to design a circuit involving these values however there's a volume drop of about 20dB compared to the original design. Is there any way to use these inductors without a volume drop or is it inevitable?
340mH, 910mH, 1350mH, 3210mH.
I've managed to design a circuit involving these values however there's a volume drop of about 20dB compared to the original design. Is there any way to use these inductors without a volume drop or is it inevitable?
I understand that but I'm getting a volume drop of 50dB compared to the 20-30dB of the original.
Okay, I've fixed the problem by multiplying the value of all resistors and inductors by 14 and dividing the value of each capacitor by 14 and the curves match perfectly with the inductors I have in LT spice. However, in practice will I need some sort of buffer at the end of the passive eq section because of these changes?
Most passive eq's have a buffer at the end. I like my tubes to distort audible a bit, depending on how much you drive them. (Otherwise whats the point of using tubes.)
The values of the caps and inductors are a balancing act, as you've figured out.
Great eq's for general shaping of sounds.
The values of the caps and inductors are a balancing act, as you've figured out.
Great eq's for general shaping of sounds.
Thanks for your reply. I would like to build the eq unit without any transformers if possible. Would it be sufficient to add the follower (buffer) circuit from an api 550a before and after the passive eq section or will this not provide the correct impedance?
I'm not really sure what I'm doing. Will the input impedance or output impedance affect the eq curves or will both?
http://www.gyraf.dk/gy_pd/pultec/pultech.gif
http://www.gyraf.dk/gy_pd/pultec/meq5.gif
http://www.gyraf.dk/gy_pd/pultec/gy_pd_sch.gif
We see that the original expects a source Z under 300 Ohms and a load greater than 10k. Anything more/less will mess the shape of the EQ.
If you have really multiplied all inductances by 14 (means pretty big chokes) then you need a source less than 4k and a load greater than 140k.
The 4k can be "any studio source". The 140k can't go out of the box, and you need a gain of 10 to make-up losses, so you put any decent gain of 10 Zin>140k amp after the L-R-C network. if you like tubes, the Gyraf booster is fine (the original yu can't get >140k from a grid transformer). Or use a good audio opamp.
http://www.gyraf.dk/gy_pd/pultec/meq5.gif
http://www.gyraf.dk/gy_pd/pultec/gy_pd_sch.gif
We see that the original expects a source Z under 300 Ohms and a load greater than 10k. Anything more/less will mess the shape of the EQ.
If you have really multiplied all inductances by 14 (means pretty big chokes) then you need a source less than 4k and a load greater than 140k.
The 4k can be "any studio source". The 140k can't go out of the box, and you need a gain of 10 to make-up losses, so you put any decent gain of 10 Zin>140k amp after the L-R-C network. if you like tubes, the Gyraf booster is fine (the original yu can't get >140k from a grid transformer). Or use a good audio opamp.
API: API 550A 3transistor buffer : Free Download, Borrow, and Streaming : Internet Archive
Okay thank you. So the input should be perfectly fine without the transformer. I would like the amplifier circuit to not use any tubes because of the high voltages. Would the buffer in the above link be satisfactory to put at the end of the passive eq to feed into a mic preamp or audio interface?
Okay thank you. So the input should be perfectly fine without the transformer. I would like the amplifier circuit to not use any tubes because of the high voltages. Would the buffer in the above link be satisfactory to put at the end of the passive eq to feed into a mic preamp or audio interface?
A filter ideally needs zero ohms driving it and the output drives into infinite ohms.
But you can compromise a bit.
But you can compromise a bit.
Or put another way the source and load impedances are part of the filter design. For RF filters both would commonly be 50 ohms and filters are designed assuming this.
For audio the typical assumption is low source impedance, high load impedance. This allows simple passive filters to not really require buffering, if they are at an intermediate impedance, say 1k ish. Complicated passive filters though will generally require buffering to behave repeatably. (whereas RF RLC filters can be very complex as the 50 ohm thing can be relied on).
For audio the typical assumption is low source impedance, high load impedance. This allows simple passive filters to not really require buffering, if they are at an intermediate impedance, say 1k ish. Complicated passive filters though will generally require buffering to behave repeatably. (whereas RF RLC filters can be very complex as the 50 ohm thing can be relied on).
Okay so my source has an impedance of 600 ohms so I'm guessing I can avoid using a transformer on the input. Can I use the API buffer on the output to run into a microphone preamp to make up for the losses? Will I have to change any of the resistor values on the API buffer?
> use the API buffer on the output to run into a microphone preamp to make up for the losses?
Losses are about 20dB. A 3-transistor buffer into a high-gain microphone preamp is clumsy and costly. Many mike preamps do not do well at low gains.
Three transistors can do a 20dB make-up just fine. You can even use API's transistor types rather than the jellybeans I simmed. Sim says 0.1% THD at 5Vrms or +16dBu output in 10k load, which is fine for a "vintage" effect.
Losses are about 20dB. A 3-transistor buffer into a high-gain microphone preamp is clumsy and costly. Many mike preamps do not do well at low gains.
Three transistors can do a 20dB make-up just fine. You can even use API's transistor types rather than the jellybeans I simmed. Sim says 0.1% THD at 5Vrms or +16dBu output in 10k load, which is fine for a "vintage" effect.
Attachments
Thanks. Just to clarify, can I feed the signal directly from the passive eq stage into that amplifier? I'm thinking to maybe just to build the API 2520 amplifier that were used in the SS Pultecs.
Do R4 and R1 in the API 550 Buffer Amplifier control the input and output impedance respectively?
So in theory could I have two buffers, one before and one after the passive eq section to obtain the correct curves? So expecting a 600 ohms input, would these values for the buffer resistors be ideal?
INPUT: R4=600, R1=4k
OUTPUT: R4=140k, R1=600.
My reasons for this is because I would like to experiment with which output amplifier I prefer the sound of.
So in theory could I have two buffers, one before and one after the passive eq section to obtain the correct curves? So expecting a 600 ohms input, would these values for the buffer resistors be ideal?
INPUT: R4=600, R1=4k
OUTPUT: R4=140k, R1=600.
My reasons for this is because I would like to experiment with which output amplifier I prefer the sound of.