PE 300-794 plate amp rumble filter (1 Viewer)

[Brian Bunge]

I've seen the question come up several times without any answer given concerning a rumble filter when the amp is modified to flat response. I also had a customer asking about this, so I called PE yesterday evening to get an answer. Apparently there is still a 2nd order filter centered at 15Hz.

Just thought everyone would like to know!

Brian

 

[Patrick Sun]

PE Paul told me (here's the PE Parts tech board posting) the F3 was 20Hz with the "no boost" mod that is now shipping with the 300-794 plate amps for the past few months.
I haven't given my 300-794 a workout yet (I wanted to see if I could clip it on the really stressful bassy DVDs, because I can't with my old, fan-noisy, pro Peavey CS-800X amp). One more thing on the list to do...

 

[Brian Bunge]

Thanks Pat. I'd tend to take Paul's word over the tech on the phone. An F3 of 20Hz plus the 17.5Hz tuning of the sub I just finished should be giving me plenty of protection for the DVC 15.

BTW, I keep forgetting to thank you for sending me those resistors. Much appreciated!

Brian

 

[Ryan Schnacke]

I've created a speadsheet that calculates the boost, boost frequency and filter center frequency of the 2nd order Sallen-Key filter used in this plate amp. With R26=56k, R27=120k I get Fc = 19.4Hz not 15Hz.

The equations I'm using are from my Microelectronics Circuits text (Sedra, Smith). And I've simulated the default case in HSPICE and found F3 = 19Hz. By 15Hz its down 5.5dB.

I'd share the spreadsheet file but I don't have any web space to put it. And its in UNIX StarCalc so I'd have to convert it to Excel and make sure it still works. But if anyone wants it I can convert and email it to them. Or maybe somebody with web space could make it available to everyone.

Some more examples of zero boost configurations:
R27=130k, R26=62k, Fc=17.7Hz
R27=150k, R26=75k, Fc=15.0Hz
R27=180k, R26=91k, Fc=12.5Hz

 

[Patrick Sun]

OMG! Sedra and Smith! The Horror! The Horror!

Never was too crazy about Transistors I or Opamps I in college...

 

[Ryan Schnacke]

Eh, they're not so bad once you get to know them. Ideal op-amp theory is easy enough to make it fun (gasp!).

Hey, give me back my pocket protector! I'll tell the teacher.

 

[David_Rivshin]

If this helps the discussion I believe I've found a good schematic for this popular amp, which is fortunate because otherwise I was going to have to reverse-engineer it and that's just a pain This is actually listed for the Apex Sr. 350W amp, but the pre-amp section seems identical, and even says 'SUB-250' on it. Here's a link: http://www.apexjr.com/Apexsenior.htm . The left half is obviously the input section, which feeds into the volume and EQ section (top-right), and the auto-power section (bottom right).
Now I'm not a EE by any means, but here is my interpretation of how to get rid of all the EQ, rumble filter and boost: remove C14 and C18, and then add a wire between the input side of C14 and the output side of C18. I'm assuming that the pull-down and series resistors (R34 and R35) are necessary, otherwise you could remove R35 instead of C18. Either way this will leave U3B and U4 hanging, so it might not be a bad idea to put in a pull-down of somesort to force the output of U3B to ground, and hopefully save alittle electricity.
Of course that only works if you want to remove the EQ wholesale; including boost, filter, and phase-adjustment. To modify it to desired specs would require much more knowledge than I even pretend to have Hopefully Ryan can use this to verify his work. Personally, since I have a 2039CS, I'm looking for as little boost as possible and an F3 of around 15Hz for a rumble filter.
-- Dave

 

[Ryan Schnacke]

Wow, you can read the numbers on that schematic? Okay, I suppose with a little effort I can.

This is good stuff. Thanks for bringing my attention to it, Dave. When you start cascading op-amp circuits together like this you want to think of them in stages. Each stage consists of the op-amp itself plus all the resistors/capacitors between the output of the previous op-amp and the output of the current op-amp. And resistors/capacitors placed after the output of the op-amp are a part of the next stage.

So we do see two input stages on the left (good call, Dave). The one at the top is obviously for the right input and the one on the bottom is for the left input. We can also see the pre-amp level output stages for the right and left channels in the middle of the left side of the schematic. Its interesting to note that you could use the preamp level outputs even if you're inputting only a high-level signal. If you input both high-level and pre-amp level then the result will be a weighted sum of the two.

The two resistors in the middle (I can't read their names but the values are 1k) sum the right and left signals together for the following stages.

The next stage (with op-amp U3A)looks like it implements the gain control. You can see a potentiometer (variable resistor) is used in a voltage divider on the input. Not sure what the capacitor in the feedback is doing.

The next stage (with op-amp U3B) is clearly the 2nd order Sallen-Key high-pass filter that is often modified. You can see that the resistors are labelled R26 and R27 just like the PE amps. And the values indicate the same configuration as the 300-792 amp I believe. 18k and 180k give a 4.4dB boost at just over 31Hz. I'm not sure why PE claims 6dB boost for this config. The other numbers in their filter mod tables look right. This filter is unity gain since there is no resistor divider so you could disconnect the output of the U3B op amp and connect the output of the U3A op-amp here instead if you really wanted to get rid of the hi-pass filter completely. But I don't see the point when its so easy to select a no-boost gain and a center frequency wherever you want. Want 5Hz? Go with 450k and 225k. I'd stick to 12Hz or higher.

The next stage (with op-amp U4A) is the low-pass filter that's controlled by the "crossover" knob. Like the previous stage, its a unity gain 2nd order Sallen-Key. But this time the R's and C's have traded places to implement a high-pass instead of a low-pass. And pots are used to change the filter frequency. Again, you could route around this stage by disconnecting its output and connecting the output of the previous stage to the input of the following stage. This might make sense if you're wanting to use this amplifier to power a full range speaker instead of a sub. I haven't run the numbers to see what it would take to push the low-pass filter up to above 20kHz. (Edit: There's no guarantee that the amp will really perform up to 20kHz. The designers may have compromised high-frequency performance since this is meant to be a subwoofer amp - and a low cost one at that.)

I'm not sure yet about that last op-amp stage (with amp U4B). I suspect its some sort of all-pass filter that implements the phase control.

At the end we see a simple first order RC hi-pass filter that's used to get rid of any DC components. 3dB freq of this one is less than 1Hz so I wouldn't worry about it.

Last is a 2.2kohm series resistor on the output. It could be just for ESD/latchup protection.

Dave, if you're looking for no boost and F3=15Hz then I'd go with the 150k, 75k I listed above.

Whew! I sure hope somebody finds this interesting.

 

[Chris Carswell]

Is this how it goes?
0db of Bass Boost:
Fc=17.7Hz R26=62k R27=130k
Fc=15.0Hz R26=75k R27=150k
Fc=12.5Hz R26=91k R27=180k
1dB of Bass Boost:
Boost Freq. R26 R27 Filter Fc Filter Q
(Hz) (k Ohms) (k Ohms) (Hz)
20-24Hz 56 220 14 1
25-30Hz 47 180 17 1
31-35 Hz 39 150 21 1
36-40 Hz 27 120 28 1.1
2dB of Bass Boost
Boost Freq. R26 R27 Filter Fc Filter Q
(Hz) (k Ohms) (k Ohms) (Hz)
20-24 Hz 47 220 16 1.1
25-30 Hz 33 180 21 1.2
31-35 Hz 27 150 25 1.2
36-40 Hz 22 120 31 1.2
3dB of Bass Boost
Boost Freq. R26 R27 Filter Fc Filter Q
(Hz) (k Ohms) (k Ohms) (Hz)
20-23 Hz 33 220 19 1.3
24-30 Hz 27 180 23 1.3
31-35 Hz 22 150 28 1.3
36-40 Hz 18 120 34 1.3
4dB of Bass Boost
Boost Freq. R26 R27 Filter Fc Filter Q
(Hz) (k Ohms) (k Ohms) (Hz)
20-23 Hz 27 270 19 1.6
24-30 Hz 22 220 23 1.6
31-35 Hz 18 160 29 1.5
36-40 Hz 15 150 34 1.6
5dB of Bass Boost
Boost Freq. R26 R27 Filter Fc Filter Q
(Hz) (k Ohms) (k Ohms) (Hz)
20-23 Hz 22 330 19 1.9
24-30 Hz 18 220 23 1.7
31-35 Hz 15 180 31 1.7
36-40 Hz 12 150 35 1.8
I do have a ??? for you guys. I don't have the PE 250
(#300-794) but the older PE 150 (#300-800).
It has the same two resistors, r26 & r27 so I was thinking
I could apply this to my #300-800 to custom taylor my boost?
Thanks Guys

 

[Ryan Schnacke]

Alright, Chris, I threw every resistor combo you listed into my spreadsheet and my results are below. You'll notice that sometimes the boost or frequency is slightly off the range PE has listed but they were probably limiting themselves from using non-standard resistor values so they couldn't always choose the ideal values.

I do believe that the hi-pass section of the pre-amp circuit is the same in both the 300-794 amp and the new 300-792 150watt amp. I have no experience with the old 300-800 amp but it's likely using the same hi-pass. If you could determine the values of R26 and R27 on your amp then we could determine if those values even make sense for this filter. If they are 18k and 180k like the 300-792 then that would be proof enough for me.

Okay here's my spreadsheet results:

"1dB" of Boost
Boost, Boost Freq, R26, R27, Fc, Q
1.2dB 20.5Hz 56k 220k 14.3Hz 0.99
1.1dB 25.0Hz 47k 180k 17.3Hz 0.98
1.1dB 30.0Hz 39k 150k 20.8Hz 0.98
1.6dB 37.7Hz 27k 120k 28.0Hz 1.05

"2dB" of Boost
Boost, Boost Freq, R26, R27, Fc, Q
1.7dB 20.7Hz 47k 220k 15.7Hz 1.08
2.2dB 26.0Hz 33k 180k 20.7Hz 1.17
2.3dB 31.3Hz 27k 150k 25.0Hz 1.18
2.2dB 38.9Hz 22k 120k 31.0Hz 1.17

"3dB" of Boost
Boost, Boost Freq, R26, R27, Fc, Q
2.9dB 22.3Hz 33k 220k 18.7Hz 1.29
2.9dB 27.3Hz 27k 180k 22.8Hz 1.29
3.0dB 33.0Hz 22k 150k 27.7Hz 1.31
2.9dB 40.9Hz 18k 120k 34.2Hz 1.29

"4dB" of Boost
Boost, Boost Freq, R26, R27, Fc, Q
4.4dB 20.8Hz 27k 270k 18.6Hz 1.58
4.4dB 25.6Hz 22k 220k 22.9Hz 1.58
4.0dB 33.7Hz 18k 160k 29.7Hz 1.49
4.4dB 37.5Hz 15k 150k 33.6Hz 1.58

"5dB" of Boost
Boost, Boost Freq, R26, R27, Fc, Q
6.0dB 20.1Hz 22k 330k 18.7Hz 1.94
5.2dB 27.7Hz 18k 220k 25.3Hz 1.75
5.2dB 33.6Hz 15k 180k 30.6Hz 1.73
5.3dB 40.9Hz 12k 150k 37.5Hz 1.77

Okay, after working and typing that many numbers there's probably a typo or miscalcuation in there somewhere. If some value seems really off or really doesn't agree with PE's table then ask me and I'll double-check all the info on that line.

My spreadsheet is pretty useful for this kind of stuff since it works the equations both ways:
Enter the Q and frequency goals and read the resistor values
or
Enter resistor values and read the Q, boost and frequency values