Question about output transistors in preamp (Pm1000)

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

JW

Well-known member
Joined
Jun 8, 2005
Messages
1,100
Location
Portland USA
Do these transistors need to be matched for symmetrical clipping? ( Q8 and Q9) Schematic below. 

It was suggested to use MJE172/182. Not sure if they need to be matched though.
 

Attachments

  • InputStrip.png
    InputStrip.png
    69.8 KB · Views: 67
> need to be matched for symmetrical clipping?

No.

Clipping will never be symmetrical. The Q7 stage won't slam both sides the same. The likely intent was that the user would "never" clip.
 
Does anyone know what the trimmer labelled Fact. Adjust is actually doing? I can't find mention of it in the service manual. Seems like it would affect the gain?
 
jazzcrisis said:
Does anyone know what the trimmer labelled Fact. Adjust is actually doing? I can't find mention of it in the service manual. Seems like it would affect the gain?

Yes, it does trim the gain. Its range is from about 15dB to about 25dB.

Cheers

Ian
 
Thanks folks,

Interesting that somebody asked about that trimmer. What I've done to a couple channels is measure the trimmer and replace with a straight resistor. It's usually in the 2K to 2.5K range (with the intent to switch between that resistor and a higher one) By adding resistance here, you can quite a bit more gain. I got something around 23K to add another 12dB of gain, but I'm not sure about it yet. It's at the cost of more distortion, but I'm just using my ears.

Should I not be going as high as 12dB?
 
ruffrecords said:
Yes, it does trim the gain. Its range is from about 15dB to about 25dB.

Cheers

Ian

So it's functioning as a rheostat, a variable resistor, and not a pot, to control voltage...?  I can never figure out the difference on a schematic...  :-\
 
PRR said:
> need to be matched for symmetrical clipping?

No.

Clipping will never be symmetrical. The Q7 stage won't slam both sides the same.

Can you explain why there won't be symmetrical clipping?  Are you referring to the diode or a general principal which effects P-P output pairs?  Or something else...?

In general, do P-P output pairs benefit from matching?

thanks.
 
tommypiper said:
Can you explain why there won't be symmetrical clipping?  Are you referring to the diode or a general principal which effects P-P output pairs?  Or something else...?
I think you are investing too much magic into the output stage topology (P-P)... Solid state devices will saturate fairly consistently without matching and that saturation will be dominated by the drive circuitry and/or other external factors (like DC operating point for a single supply amp). If that drive is not symmetrical or identical for pull-up and pull-down, it should not be expected to clip symmetrically, and from observation that amp's drive circuits are not symmetrical.

The drive pull-up is bootstrapped , so it will likely saturate in that direction when the output device runs out of PS rail voltage. OTOH it has an active pull-down circuit that involves some internal voltage drop to work, so it will saturate in that direction before reaching ground.

If you really care about symmetrical clipping perhaps tweak the DC operating point (slightly above V/2), but I am REALLY reluctant to encourage you in that pursuit.  ::)
In general, do P-P output pairs benefit from matching?

thanks.
Not for this design... and not for solid state in general***...  In fact I would not expect P type and N type devices to ever be very well matched.  The only kind of matching (perhaps) would be for devices configured in parallel to increase output current, but even they generally use emitter degeneration resistors to force sharing making matching unnecessary.

Rather than kill brain cells over matching to clip similarly, perhaps don't clip that stage. If you plan to clip it routinely the recovery from clipping behavior is probably more audibly significant. Anti-saturation clamps can prevent device saturation and speed up device recovery after clipping, but for that crude design, I'd just avoid clipping it.

JR

*** MOSFET power devices when paralleled might benefit from Vgs matching (does anybody still make lateral MOSFETS? )
 

Latest posts

Back
Top