Thank you all for your valuable input!! I have just competed the electrical adjustment process and have to state that all values were spot on! The popping noise is the last remaining issue. I noticed this before the caps went bad so this is an independent issue. It is not that loud that it would make me worry about the speakers - just noticeable. I will try and see how easy the 22uF caps are accessible, Hans. This is indeed a possibility. On the other hand, the popping noise is on both channels...
Popping noise is in both channels because relays are driven simultaneously on both channels and not independent.
So when one 22uF is not OK you will notice this in both channels.
Hans
So when one 22uF is not OK you will notice this in both channels.
Hans
When you have an oscilloscope, you can find out which relay coincides with the pop.
One channel on the LS output and the other channel successivily connected to the relay steering signals to find the one that causes the pop.
That could be the one with a broken 22uF cap.
Hans
One channel on the LS output and the other channel successivily connected to the relay steering signals to find the one that causes the pop.
That could be the one with a broken 22uF cap.
Hans
You can do this with a $60 analog VOM. DVM are too slow, average the signal over 2-4 seconds. A used one with 2 vdc scale will show small voltage differences.
I wouldn't hesitate to replace any electrolytic cap in a 25 year old product. I sound check the product after I replace two, in case I made a mistake. If the product gets worse instead of better, I then know right where the problem is. What I just replaced. A 22 uf cap will be $.15, the freight to buy just one from a reputable distributor is $9. I install e-caps with no less than 3000 hour service life. There is no reason to buy 1000 hour trash from a local store, as I used to do when I started repairing over-aged amps. I've replaced the e-caps in one amp 4 times in 41 years, using store shelf caps 3 times. Now newark sells to people with debit cards. You had to have a business credit rating in 1970. Newark & digikey will provide the service life in the selector table. Mouser requires you to download and read the datasheet on each cap. Alliedelec stocks nothing but 1000 hour caps.
Last edited:
pop / click on power up.
My No. 336 always made a slight click (could be considered a popping sound) upon powering it up. It was consistent from day one and over the 10 years of ownership. I miss that amp, what a beast!
My No. 336 always made a slight click (could be considered a popping sound) upon powering it up. It was consistent from day one and over the 10 years of ownership. I miss that amp, what a beast!
working in stages
I always do the same, like flagging trees and focal points, along a hiking trail.
Plus it is easy to pattern what component, what work or change makes, meaningful differences.
Keep forging ahead, happy to see the progress Agosto.
I always do the same, like flagging trees and focal points, along a hiking trail.
Plus it is easy to pattern what component, what work or change makes, meaningful differences.
Keep forging ahead, happy to see the progress Agosto.
On the protection board, there is a relay K1 that enables the whole chain with 22uF caps.
What I don’t see is a diode in opposite direction to prevent large voltage spikes when turned off.
This diode maybe somewhere else but I can’t find it.
So you could attach a diode over the coil and see wether this helps.
The point is that you hear the pop in both channels, so the cause must be in this protection board.
Hans
What I don’t see is a diode in opposite direction to prevent large voltage spikes when turned off.
This diode maybe somewhere else but I can’t find it.
So you could attach a diode over the coil and see wether this helps.
The point is that you hear the pop in both channels, so the cause must be in this protection board.
Hans
On the upper side of the chain you have the signal LED_DRIVE.
This signal comes from the Xilinx FPLA, U105.
But since your amp seems to switch between standby and operational, it’s a bit unlikelily that the cause of the pops is here.
But to be sure, you could measure on both sides of the chain to see whether top or the bottom of the chain switches when going from standby to operational and back.
Hans
This signal comes from the Xilinx FPLA, U105.
But since your amp seems to switch between standby and operational, it’s a bit unlikelily that the cause of the pops is here.
But to be sure, you could measure on both sides of the chain to see whether top or the bottom of the chain switches when going from standby to operational and back.
Hans
So I thought I connect the amp to my "real" audio system to check for sound quality and the "popping" noise effect: The "pop" is significantly weaker than on my bench and it is in line with what "ozark" described in his experience with the amp. So I decided to leave it alone (for now) and hope that the amp is now good for another few years...
I am still without clue what triggered the capacitor discharge in the first place. Perhaps indianajo and Hans are correct that those caps were initially connected backwards!?
In any case: Many thanks to all of you in assisting me - patiently - through my exploration of the 333 circuitry!! As a byproduct I am now familiar with the disassembly process and the principal operation of this amp.
I am still without clue what triggered the capacitor discharge in the first place. Perhaps indianajo and Hans are correct that those caps were initially connected backwards!?
In any case: Many thanks to all of you in assisting me - patiently - through my exploration of the 333 circuitry!! As a byproduct I am now familiar with the disassembly process and the principal operation of this amp.
I have a 336 (original owner). It’s at Pyramid in Texas getting rebuilt. I sourced Cornell’s for the caps. My amp was sent back to Madrigal 2x for the same issue - bad batch of caps (unk brand). This time right channel went and I observed black soot on cover and throughout amp yikes (I removed top). Not a tech, so I contacted pyramid (33x experts). On/standby switch failed also. Cornell’s sound very good (dated 2017) others mentioned here - not too sure. Do the caps have a shelf life if they haven’t been powered or used? Amazing amp when connected to dedicated 20 amp line.
Yes indeed sounds exactly like the problem I had. It was quite some work to take the unit apart - and the weight does not make things easier. The right channel board was covered with electrolyte so I had to take the entire board apart for cleaning purpose. I don't know about the shelf life for unused caps.
It has now been more than four weeks since I repaired the amp. No more issues so far....
It has now been more than four weeks since I repaired the amp. No more issues so far....
Omg it’s heavy. I sent my amp to TX partially disassembled. Good that you can repair your own gear. This is a marvelous piece of engineering - want to keep it forever. * Initial two separate cap failures (both crappy Philips) were covered under 5 year wnty repaired by madrigal. This failure is more ominous. Thanks for your comment! 👍
Definitely over $1k. But I’m ok with it since the beast is a keeper. Can’t wait to hear it recapped & connected to 20A dedicated line.
I had an older Proceed HPA3 recapped. It’s a nice sounding amp (250w-8ch) and I use it exclusively until my 336 comes back. New caps make a big difference - so does a 20a line. Opens up the soundstage.
THE ORGINAL ML-333 POWER SUPPLY CAPS WAS MADE BY PHILIPS MADE IN CHINA 3186GJ503M125DPA2 (50000uF 125V. ML SERVICE CENTER USES UNITED CHEMICON U36D612 (E36D1C1CPN503TEE9U) 50000uF 125V. ALL OF THESE WILL FAIL BECAUSE OF THE RAIL VOLTAGE. I HAD CORNELL DUBILIER MAKE DCMC513M160DP2D ABOUT 11 YEARS AGO 51000uF 160V WITH 200V SURGE SAME DIAMETER AND LENGTH. NOW THESE CAPS ARE RELEASED TO MOUSER MADE IN USA VERY EXPENSIVE $159 . YOU ALSO NEED TO REPLACE THE REGULATOR CAPS 1900uF 150V WITH CDE 1900uF 250V DCMC192T250AJ2B. SINCE RAIL VOLTAGE IS 143V. ALSO THERE IS AN UPDATE MANY SMALL CAPS BETTER SURGE RESISTORS CUSTOM MADE. I HAVE THE SERVICE MANUAL FOR 333.
