Rod Elliot P.68 Phase?

[samutacaratti]

Hi all,

Having read numerous posts on the forum I have come to learn the importance of linear phase response and would like to take the audiophile route through my system.

I recently realised upon testing with a scope my cheapish subwoofer amplifier has very poor phase response deviating to almost 90 degrees at 120hz even if dc coupled. I found the capacitors which cause the problems but they are required for the amps stability and as the amp has had numerous faults and is running on virtually 0 bias current, I was looking into Rod's design substituting the MJ transistors with 2sc's.

I have emailed Rod but received no reply, could anyone advise from the design if it has a reasonably linear phase response with minimal deviation from 0 degrees (if anyone is able to deduce such information)?

Thanks

300/500W Subwoofer Power Amplifier

 

[Rafael L]

Hi all,

Having read numerous posts on the forum I have come to learn the importance of linear phase response and would like to take the audiophile route through my system.

I recently realised upon testing with a scope my cheapish subwoofer amplifier has very poor phase response deviating to almost 90 degrees at 120hz even if dc coupled.

Your subwoofer amplifier must have a crossover, the phase shift is normal.

 

[samutacaratti]

There is no crossover as I have stripped down the sub to the bare amp to use it with an PC based XO with a routed soundcard.

 

[richie00boy]

There must be a fault with the amp then. Any reasonable amp will have no phase shift to worry about in the audio band. The P68 will be fine and you can work out the phase shift points from the feedback network and input stage if it bothers you.

 

[samutacaratti]

Attached is a schematic of the stripped amp in use with the phase shift. The caps causing the big shift are circled in red. When removed the amp distorts heavily and is unstable.

Could anyone shine any light on this?

Thanks alot

 

[AndrewT]

Hi,
the input high pass filter 4u7F & {2k2+22k} is sufficiently low to allow all your bass and sub-bass signals to pass and yet still block DC and ultra low frequency.

The high pass built into the Negative FeedBack route cuts off a little higher. This should be lower than the input filter, change C4 to 180uF or 220uF.
Similarly C5 could usefully be a bit bigger, maybe 150uF.

 

[AndrewT]

what happened to ESP's P68?

Your new post shows the opamp with three caps directly connected to the -IN pin.
Many opamps become unstable with these in place. Maybe this amp has used these as compensation caps.

 

[samutacaratti]

Thanks for the advice with Rods design, from what people have said I gather Rods design will definitely have a linear phase response for atlest the lower frequency band (all "bass")?

My latest post was just to show the current manufactured design as Richie said it may be faulty - if the phase shift problem could be fixed I may put off building rods design for a while, ie if anyone can spot a flaw causing the phase shift in the original design that would be great I was resigned to having to totally make a new amp.

Thanks alot

 

[Rafael L]

How you measured the phase shift in your amp?

 

[wahab]

Increase C1 up to 68uF.
Increase C4 to 220 to 330uF.
C5 has a good value, increasing it bring no
improvement, quite the contrary.

Other than that, since it s a basses amp,
use a beefy power supply.

 

[samutacaratti]

I measured the phase shift just by connecting different scope channels to the IN and OUT. The shift was easy to see.

Thanks wahab, I assume those modifications are to help keep phase deviation down if I were to build the new amp?

 

[samutacaratti]

Just before I ditch the old design is there definitely no remidy to the problem in the existing design caused by c36 and 37?

(sorry wanted to edit my other post and add this but it wouldnt let me)

 

[samutacaratti]

Anyone?

 

[wahab]

I measured the phase shift just by connecting different scope channels to the IN and OUT. The shift was easy to see.

Thanks wahab, I assume those modifications are to help keep phase deviation down if I were to build the new amp?

Yes, it will reduce the phase shift to almost zero in the
low side of the audio band and settle the 1/f frequency
corner.

Increasing even more these caps value would be counter
productive in respect of the latter caracteristic mentionned.

 

[samutacaratti]

Increase C1 up to 68uF.

Did you mean 6.8uF?

Thanks

 

[wahab]

Did you mean 6.8uF?

Thanks

No, there s no mistake , it s 68uF.
100uF is the maximum for this cap.

 

[samutacaratti]

Thanks alot for the help wahab I'll be building it with those values over the next few weeks.

 

[Madaudio]

Attached is a schematic of the stripped amp in use with the phase shift. The caps causing the big shift are circled in red. When removed the amp distorts heavily and is unstable.

Could anyone shine any light on this?

Thanks alot

picture shown is a KEF PSW3000. I think such problems should not be there on a commercial amp like KEF

 

[pallab]

sir,i don't find mjl4281a/4302a ,anywhere near my location...for Elliot esp p-68 project..i have 6 pair 2sc3858/2sa1494 sanken transistor..if I use this instead of mlj pair is there any problem???please reply me..

 

[AndrewT]

compare the datasheet specifications.
if they are very similar they should be a straight swap.
If any one parameter is very different, then you need to look closer at what effect this may have on performance.
That's when you ask your question.