M2 inspired balanced output preamp idea

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Pin 1-2 connection will draw > 50ma at 20Hz for 20Vpp output. Needs a buffer with enough grunt.

Correct. Good prediction.

For 10v p-p out, the -3dB frequency is 16.5 Hz. This is not bad for a $6 transformer. At 10 v p-p, there is no visible distortion.

As you predicted, for 20v p-p out at 20 Hz, the sine wave is visibly distorted.

So, if max p-p voltage out is needed at low frequencies, this is not a good solution.

But, you have to consider that acoustic music played at realistic levels does not contain vast amounts of 20 Hz. The cheap buffer boards driving the Edcor as a preamp may sound just fine for many applications. It is probably not a good idea for Dub Step music at a night club at 120 dB.
 
Good, let's worry about 20Vpp at 20Hz later. Edcor still has the PCW or the WSM 600/15k that will cut it. Time to start finding out what the PC 600/15k needs to sing well in circuit.

I agree. The cheap-o buffer boards are probably good enough to drive the Edcor in a balanced preamp configuration for real-world music.

I like to measure things so I went further and measured the Edcor for voltage swing and for bottom octave behavior.

No problem getting the Edcor to swing 100V p-p undistorted as low as 50 Hz. Below 50 Hz, and above 10 V p-p, there is visible distortion regardless of output impedance of the driver. So, I conclude that the distortion is core saturation.

This preamp idea would be to drive my Aleph 30 which would be driving an array of midrange speakers that will have a 100 Hz crossover making behavior below 50Hz not so important.

Thank you all for the inputs and suggestions. They were all helpful.
 
50 Hz 100V p-p output
 

Attachments

  • Tek001.png
    Tek001.png
    136.7 KB · Views: 156
It is a simple balanced ccs loaded source follower. Have fun. :)

Thank you for the well wishes. I have sorted through a pile of qty 50 MTP10N15 and have found enough matched pairs to go forward. I measured Vgs at 51.9 mA.

The matched pairs span Vgs from 3.85 to 4.11 with two quads at 3.93 and 3.94.

A pic of the test rig for measuring Vgs is attached for amusement.
 

Attachments

  • TestRig.jpg
    TestRig.jpg
    349.9 KB · Views: 185
Member
Joined 2011
Paid Member
How much DC current are you willing to push through the transformer primary? To avoid premature core saturation, to avoid permanent magnetization, and so on.

Thanks to Ohm's Law we can calculate it easily: Iprimary = Voffset / Rprimary

where Voffset is the offset voltage of the buffer amplifier, and Rprimary is the DC resistance of the primary. As I recall, Rprimary was about 250 ohms but you can measure your particular Edcor transformer's Rprimary with a DVM and find the exact number.

If you use a famous design procedure called r***** extraction to decide that you require Iprimary < 0.25 mA, then Ohm's Law says

Voffset < (0.25 mA) * (250 ohms)

Voffset < 62 millivolts

(Below 62 mV offset) is easy to do with ICs and no offset trimming (adjustment potentiometers).

(Below 62 mV offset) is possible to do with all-discrete circuits plus offset trimming.

(Below 62 mV offset) is possible to do with all-discrete circuits plus careful hand-matching of components, and no offset trimming.

It all depends on how much DC current you are willing to push through the transformer primary.
 
How much DC current are you willing to push through the transformer primary? To avoid premature core saturation, to avoid permanent magnetization, and so on.

Thanks to Ohm's Law we can calculate it easily: Iprimary = Voffset / Rprimary

where Voffset is the offset voltage of the buffer amplifier, and Rprimary is the DC resistance of the primary. As I recall, Rprimary was about 250 ohms but you can measure your particular Edcor transformer's Rprimary with a DVM and find the exact number.

If you use a famous design procedure called r***** extraction to decide that you require Iprimary < 0.25 mA, then Ohm's Law says

Voffset < (0.25 mA) * (250 ohms)

Voffset < 62 millivolts

(Below 62 mV offset) is easy to do with ICs and no offset trimming (adjustment potentiometers).

(Below 62 mV offset) is possible to do with all-discrete circuits plus offset trimming.

(Below 62 mV offset) is possible to do with all-discrete circuits plus careful hand-matching of components, and no offset trimming.

It all depends on how much DC current you are willing to push through the transformer primary.

Good point. Measure the primary resistance and make sure that DC current is low enough to not damage the transformer core.
 
Indra1's driver Asymmetric rails

Hello Indra1

I have interrogated your driver circuit. Clipping with + / - 9V rails is quite asymmetric. I adjusted the negative rail until clipping was mostly symmetric. It looks like the negative rail wants to be approximately two times Vgs more than the positive rail for symmetric clipping.
 

Attachments

  • AsymmetricRails.jpg
    AsymmetricRails.jpg
    166.9 KB · Views: 192
Yes Brian, clipping will be assymetric on symmetric rails. But it does not really matter for use up to 10Vpp you need to drive the Aleph 30 right?.
And use a 47 ohm dummy resistor instead of the transformer to first make sure DC offset is manageable as pointed out by Mark Johnson on your build.:)
 
Yes Brian, clipping will be assymetric on symmetric rails. But it does not really matter for use up to 10Vpp you need to drive the Aleph 30 right?.
And use a 47 ohm dummy resistor instead of the transformer to first make sure DC offset is manageable as pointed out by Mark Johnson on your build.:)

You are quite correct. Not necessary. I am just messing around having fun!

For sure I will use a dummy resistor initially.

Thanks and cheers. Have a great weekend.
 
answer to woofertester

Hello woofertester,


I have used those bufferboards mentioned in the beginning of this thread
to buffer the outputs of one of my active crossover projects (based on OPAmps). It behaved very well. But I wasn't measuring it. Only listening. Was a few months
ago.

But at the moment this project is sleeping. I made an PASS AXO and a M2X
inbetween......

Greets
Dirk
 

Attachments

  • ActXover and buffer boards.jpg
    ActXover and buffer boards.jpg
    127.4 KB · Views: 169
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.