ADA Depot - A Forum To Support Users of ADA Amplification Gear

Please login or register.

Login with username, password and session length
Advanced search  

News:

Temp "Time Format" issue Fix in Discussions

Pages: [1] 2   Go Down

Author Topic: Getting rid of noise in my MP-1  (Read 8231 times)

0 Members and 1 Guest are viewing this topic.

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Getting rid of noise in my MP-1
« on: Time Format »

Roughly ten years ago i acquired an MP-1 sporting a rev.C PCB and a v1.37 eeprom, which turned out to be very noisy.
Back then i followed the paint by numbers noise mod procedure, which sadly didn't cure the issues.

I've decided to give it another spin on the workbench, and see what happens :) Having acquired a degree in electronics engineering since then has hopefully raised the odds to my favor...

For reference - this is ground zero:
« Last Edit: Time Format by hansgraudal »
Logged

MarshallJMP

  • Superhero Member
  • ******
  • Offline Offline
  • Posts: 4702
    • marshalljmpmodshop.net
Re: Getting rid of noise in my MP-1
« Reply #1 on: Time Format »

How do you "measure" the noise level?
Logged

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Re: Getting rid of noise in my MP-1
« Reply #2 on: Time Format »

Step 1: Make sure the supply rails are clean and healthy!

Unfortunately i didn't save scope dumps for the first couple of steps, so you'll have to take my word for this:
Examining the supply rails i had almost 200mV ripple over the filter caps, but oddly there was still around 100mV 50Hz ripple coming out of all four DC regulators.
The +/-15V rails were closer to +/-14.5V and the 5V rails were both down to 4.7V
Also the 47n bypass caps around the regulators are way to small compared to the 330n caps specified in recent datasheets for the 78xx/79xx series.

As they must have had a hard life, I decided to replace the old 78M05, 78M15 and 79M15 regulators with new 7805, 7815 and 7915 regulators as they can handle currents in excess of 1.5A where the 7xMxx regulators are only rated for 0.5A.
I replaced the large filter caps with some 3300µF 50V low ESR caps (I'll have to do something about those. It's not too pretty as is...  ;D), and the 33µF 25V electrolytics (C121, C124, C126, C128) were upped to 220µF 50V also low ESR types.
The 47nF ceramics were replaced with 470nF MKP types.

The result of this was 100mV 50Hz ripple over the filter caps and just 6mV ripple on all four regulated supply rails. The frequency of the ripple on the regulated rails has shifted to something in the MHz range. I guess I'll have to look at the decoupling caps in the digital part...

All in all a great improvement. The rails are spot on their intended DC values and ripple is greatly reduced. All in a good hours work:
« Last Edit: Time Format by hansgraudal »
Logged

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Re: Getting rid of noise in my MP-1
« Reply #3 on: Time Format »

How do you "measure" the noise level?

I have an abundance of oscilloscopes as well as a R&S UPV audio analyzer at my disposal.  :thumb-up:
Logged

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Re: Getting rid of noise in my MP-1
« Reply #4 on: Time Format »

I have installed sockets for all op-amps in the circuit, and put in recent production TL07x for a start.
With shorted input terminals, OD1@10, OD2@0, master@10 and output volume at max, noise measured at the output jacks was 40mVpp (-35dBu).
With the opamps replaced throughout the circuit the measurement is down to 2mVpp (-61dBu).
Raising OD2 also raises the noise floor somewhat. The next I'll be doing is a recap of the tube board, and I'll try replacing the carbon film resistors on the tube board with metal film resistors as well.
Let's see what happens...
Logged

MarshallJMP

  • Superhero Member
  • ******
  • Offline Offline
  • Posts: 4702
    • marshalljmpmodshop.net
Re: Getting rid of noise in my MP-1
« Reply #5 on: Time Format »

I'm curious, how can you measure 50Hz ripple with a full bridge and full wave rectifier?
Logged

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Re: Getting rid of noise in my MP-1
« Reply #6 on: Time Format »

I'm curious, how can you measure 50Hz ripple with a full bridge and full wave rectifier?
Gah... That's a brainfart on my behalf. 50Hz rectified of course leaves 100Hz residuals. The frequency counter on my scope must have been consistently measuring over two waveform periods, as it definitely put out a reading of 50Hz :facepalm:
To be honest I was more stumped over the fact that the regulators didn't sort out the remaining ripple...
Logged

MarshallJMP

  • Superhero Member
  • ******
  • Offline Offline
  • Posts: 4702
    • marshalljmpmodshop.net
Re: Getting rid of noise in my MP-1
« Reply #7 on: Time Format »

Maybe you measured some common mode noise on those regulators?
Logged

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Re: Getting rid of noise in my MP-1
« Reply #8 on: Time Format »

I don't think I'd see much Common mode noise when measuring directly across the pins on the regulators.
Also: Wouldn't the noise issue have persisted, if the regulators weren't faulty?
Logged

MarshallJMP

  • Superhero Member
  • ******
  • Offline Offline
  • Posts: 4702
    • marshalljmpmodshop.net
Re: Getting rid of noise in my MP-1
« Reply #9 on: Time Format »

Usually those regulators work, or they don't, never saw ones that "kinda" worked. Now it's always a good idea to change those M type regulators because they fail sometimes.
About the common noise, why did your scope display 50Hz?

Also be carefull with putting 220µF caps on the output of those regulators, read the datasheet about using large caps.
Logged

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Re: Getting rid of noise in my MP-1
« Reply #10 on: Time Format »

My guess about the 50Hz reading is that the ripple probably isn't symmetrical for the top and bottom half of the rectified signal, and as such the scope sees a periodic signal at 50Hz even though it has a 100Hz sawtooth-ish signal "riding on top". I'll make some measurements tomorrow to verify if this is the case.
Logged

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Re: Getting rid of noise in my MP-1
« Reply #11 on: Time Format »

Right on the money. The brand new 500MHz Keysight scopes at work measure 50Hz over two ripple cycles, while my old 60MHz HP scope at home gets it right at 100Hz over one cycle ;) Attached is a dump of the ripple over the -15V cap, as the 100Hz ripple is most prominent there.

I brougt the unit home today, to play it for a bit. It's getting better, but there are still some issues to be solved.

When i got it, there was quite a bit of PSU hum bleeding through to the audio path. This is now completely cured.

With OD1 gains above 2-ish the unit starts oscillating, when a guitar is plugged in the front. When the guitar is played it's like the oscillation stops, and then returns when the guitar signal drops below a certain threshold. With no jack in the front input there is no oscillation.
I'll take a look at the alignment procedure next, and then the discrete input stage.

At high gain settings there is audible bleedthrough from the digital control circuitry to the audio path. This gets worse whenever buttons are pressed on the front panel. I'm thinking fresh decoupling caps throughout the board will help here.
Let's see what happens... :)
« Last Edit: Time Format by hansgraudal »
Logged

rnolan

  • Administrator
  • Superhero Member
  • ******
  • Offline Offline
  • Posts: 5998
Re: Getting rid of noise in my MP-1
« Reply #12 on: Time Format »

Cool  :thumb-up: , your getting there, and I'm enjoying the journey  :wave:
Logged
Studio Rig: Stuff; Live Rig: More Stuff; Guitars: A few

MarshallJMP

  • Superhero Member
  • ******
  • Offline Offline
  • Posts: 4702
    • marshalljmpmodshop.net
Re: Getting rid of noise in my MP-1
« Reply #13 on: Time Format »

Strange it gives you 50Hz, you can see clearly in the pic that it is 100Hz (10ms).
Logged

hansgraudal

  • Newbie
  • *
  • Offline Offline
  • Posts: 18
Re: Getting rid of noise in my MP-1
« Reply #14 on: Time Format »

That's what I'm saying too. The pic is from my scope at home, and it's obviously 100Hz. The 50Hz reading was from the frequency calculator in the scopes at work. I didn't even think about checking the measurement, as they're usually very accurate. Not this time though...

Anyway, the 100Hz ripple is far gone after the regulators, So that's not an issue anymore. Today I replaced all of the 100n decoupling caps as well as the 15uF tantalum, but didn't have time to measure the outcome. There's a job for tomorrow :)
Logged
Pages: [1] 2   Go Up