ms2125 showing clip after cap replacement!

marko · Post by **marko** » Fri May 25, 2007 12:37 pm

just swapped the 12 input caps on my ms2125 and went to fire it up in the chassis, no speakers connected and right channel is showing clip, what have i done wrong? i've done a few of these now and all has been well, i did notice a little pop too but thought that might have just been the caps charging up, nothing is visually blown and nothing gets hot either...

1moreamp · Post by **1moreamp** » Fri May 25, 2007 1:06 pm

marko wrote:just swapped the 12 input caps on my ms2125 and went to fire it up in the chassis, no speakers connected and right channel is showing clip, what have i done wrong? i've done a few of these now and all has been well, i did notice a little pop too but thought that might have just been the caps charging up, nothing is visually blown and nothing gets hot either...

Hi Marko,
Your pop was either a transistor developing a small hole in its plastic case, or one of the many current limits taking a dump on you due to a assembly short probably thru one of the insulators of the outputs or a power fet or a diodes case connecting to the chassis via a bad insulator between the device and the case. Or even thru the protective plastic mat on the bottom of the case, any of the above can and does happen

I sincerely doubt that any of your work with the caps is the fault, especially since your use to doing this work on a regular basis. So that leaves a short circuit between the power devices and the case somewhere.

Shite happens man ! but I think this is what is plaguing your reassembly. It happens to everyone

So don't feel bad

marko · Post by **marko** » Fri May 25, 2007 1:15 pm

embarisingly i did notice some very fine wire strands knocking around the fets, never noticed these till afterwards as they were so fine, maybe these shorted out? need to start checking stuff, what a nightmare being without music and i sold my spare ms2125 last year

with so many fets where do i start, they all look good with no signs of damage! maybe i should just take the board out of the sink and wire it up with no load and see what happens?

thanks mark.

1moreamp · Post by **1moreamp** » Fri May 25, 2007 1:30 pm

marko wrote:embarisingly i did notice some very fine wire strands knocking around the fets, never noticed these till afterwards as they were so fine, maybe these shorted out? need to start checking stuff, what a nightmare being without music and i sold my spare ms2125 last year

with so many fets where do i start, they all look good with no signs of damage! maybe i should just take the board out of the sink and wire it up with no load and see what happens?

thanks mark.

Check this link out, I hope yours was not as bad

http://phoenixphorum.com/how-not-to-ins ... t1919.html

Oh careful with the out of sink ops, They have heat issues even unloaded, and try to use a 5 or 10 amp power supply and turn your current limit down to about 3 amps just for safety, as with no sink the outputs surely will head to the nether regions if keep idling too long without a sink

marko · Post by **marko** » Fri May 25, 2007 11:59 pm

ok, made few discoveries, f4 (rail fuse) is showing 6v instaead of 35v like the rest, this is the nearest to the right hand devices. took me a while to spot something but i found a very small scorch mark on the back of one of the toshiba outputs, not the 1302 ones the other ones, 281 rings a bell, anyway my batteries are flat in the camera but i'm sure this isn't right and there's the remains of about 3mm worth of copper strand there

now i have had a rumage around and found one spare parts m50 and it looks to be in good condition too so will transplant that over then check rail voltage again!

1moreamp · Post by **1moreamp** » Sat May 26, 2007 12:24 am

most likely there will be blown fuse resistors , 22 ohm comes to mind, they are meant to pop like fuses, please replace with same safety parts.
If the fuse resistors did there job then the damage was limited, but check all transistors in that section driving the outputs, this is a DC coupled amp and except for those fuse resistors there is no other protection to prevent a fold back of the rails to the drivers. I have seen lessor amps take out the entire channel all the way back to the itty bitty input transistors for the channel.

This why I love my PG they designed safety into the amp to protect from such major failures, But anything can happen.

Good Luck C

PS for God's sake use a air wand and a soft brush to clean all the loose copper wires out of the amp before you do any more testing at all. I would bet there are more copper sprigs floating around in there. Oh and I have seen this before all the way back in 1991. Things happen.....

marko · Post by **marko** » Sat May 26, 2007 12:30 am

just been and swapped the driver only to realise it's on the left channel which wasn't showing clip! still the same anyway, i checked all the resisters on the right chanel including those little white ones and all seams well....

how do i check the devices? all red leds's are lit up on the board still...

marko · Post by **marko** » Sat May 26, 2007 1:39 am

how do i check them, it seams very strange that paul (kub3) had the exact same problems when he changed the caps on his ms2125 that was new, right clip light showing...

1moreamp · Post by **1moreamp** » Sat May 26, 2007 1:50 am

marko wrote:how do i check them, it seams very strange that paul (kub3) had the exact same problems when he changed the caps on his ms2125 that was new, right clip light showing...

Lets digress for a moment... You just swapped the 12 volt caps correct ?
Not the main internal rail caps < the big blue ones> ??

There is no direct connection to the clip circuit and the 12 volt side input caps, so unless you have a bad rail cap I see no connection. I have a MS-2125 someone gave me, It had a clip light to, and it had semi burnt resistors near the big rectifiers in the power supply.

Well come to find out one of the big blue caps legs had broken off and someone had hot glued the cap back to the board. Heedless to say the glue did not work, and once I replaced that cap the amp worked fine, I.E. no more clip light.

What do your rails read like in that channel ??

marko · Post by **marko** » Sat May 26, 2007 1:56 am

one reads 6v (f4)and the other 35v, only changed the 12 input ones as the rail caps were fine and amp worked great before for 3 years ok!!

1moreamp · Post by **1moreamp** » Sat May 26, 2007 2:04 am

marko wrote:one reads 6v (f4)and the other 35v, only changed the 12 input ones as the rail caps were fine and amp worked great before for 3 years ok!!

Well the 6 volt rail is the issue. You need to find out where that is getting pulled down 29 volts. Thats why you clip is turned on, I believe.

How much current is the amp drawing at idle 12 volts no load or signal ???It should not be more than 2 amps and depending on bias sets it really should be around 1.6 amps at idle... LMK please

marko · Post by **marko** » Sat May 26, 2007 2:19 am

yep, 1.5amp at idle.... f4 still showing 6volts as well, f3 f2 and f1 all 35volts which i believe is correct.

1moreamp · Post by **1moreamp** » Sat May 26, 2007 2:52 am

marko wrote:yep, 1.5amp at idle.... f4 still showing 6volts as well, f3 f2 and f1 all 35volts which i believe is correct.

Its early here I need some sleep, Please try to find where it drops to 6 volts from 35, by removing the fuse, Also check the fuse if you could, Ill be back with you in a few hours.... C

marko · Post by **marko** » Sat May 26, 2007 2:53 am

ok, cheers, will have a look around

marko · Post by **marko** » Sat May 26, 2007 5:33 am

well my curiosity got the better of me so i very quickly tried the amp on a low volume and both channels work! still got the 6volts on f4 though and the clip light, maybe it was like this all the time? still needs looking at though to place my mind at rest...

1moreamp · Post by **1moreamp** » Sat May 26, 2007 9:45 pm

Hi Marko,
Have you found anymore info on the 6 volt issue ?
That fuse should read 35 like the rest, actually two of the fuses read - 35 volts and the other two read + 35, and there should be a + and - 35 volts to each amp section.

I looked at my notes and such and unless your not using the correct ground point you might get 6 volts as thats the difference between the rail and the raw voltage lines in this amp. And thats the only way I can see you getting 6 volts at that fuse.
Have you checked the case to see if it has voltage on the case as referenced to the RCA shield < for simplicity sake RCA shield ground should be the same as center tap of the toroid, LMK if your case has voltage on it referenced to RCA ground and ground of your 12 volt source please ?

marko · Post by **marko** » Sat May 26, 2007 10:59 pm

i'm getting 77volts accross both left and right rectifiers.

ok, black probe on rca shield i'm getting +39 and -39volts on both left and right rail fuses so this seams consistant.

where would i check the case if doing it that way?

i can't get the rca shield to measure any ground when put the red probe on the +12 input toroid leg. (at this stage my board is still out of the heatsink)

1moreamp · Post by **1moreamp** » Sun May 27, 2007 12:00 am

marko wrote:i'm getting 77volts accross both left and right rectifiers.

ok, black probe on rca shield i'm getting +39 and -39volts on both left and right rail fuses so this seams consistant.

where would i check the case if doing it that way?

i can't get the rca shield to measure any ground when put the red probe on the +12 input toroid leg. (at this stage my board is still out of the heatsink)

Hi Marko,
Always use the RCA shield as a ground for all tests on the audio side of the amp, Always .

You now see that all the rail voltages are steady and good, something i could could tell by the amplifiers idle current draw < you will also after some time with these amps.

I really think you should trust me on this and replace those two transistors i sent you in the PM pic. believe me I am trying to save you time and effort. They are small signal 2SC2389 and 2SA1038. They are the clip LED drivers, and cost very little to replace.
I know your amp works OK as it has all normal operating voltages inside, and that clip light is turned on constantly even without a signal. Now think about this test, as it will clue you on whats on my mind here.

Test the DC output of each channel with you voltmeter by reading across the output terminals with the amp idling and NO RCA's or signals. You should get 10 millivolts DC or less on the output of each channel. Thats 0.010 volts DC or less , it varies from amp to amp based mostly on condition of the amplifier circuitry.
If your channel was blown it might have 39 volts on that output, or all the outputs would be flashed and fused and blown and them the amp would be shorted and drawing a ton of current.

These are the two common failure modes of any amp. Either shorted on itself internally and outputs scorched and smoked and board blackened. OR High DC voltage on the speaker output terminals. these are two most common failure modes, and a simple voltmeter test checks for both very quickly.

If your not getting HIGH DC voltage out and the outputs are not exploded then neither failure mode exists. BUT the clip driver transistors could be popped and turned on all the time giving you that false RED LED signal your getting. Please give this a thought and maybe a try, those two little transistors cost very little to replace, and it might just save you some time, and work.
And if the two failure modes i spoke about before Do not exist them your sorta cornered into trying those two little transistors, cause thats all thats left Man...

"i can't get the rca shield to measure any ground when put the red probe on the +12 input toroid leg. (at this stage my board is still out of the heatsink)"

After you re-assemble the amp do the following step:

Now as this point for your ground has been established you can now test the case with the red probe, and see if your getting voltage on the case. This tests to see if the insulators are all ok and that the case has not become charged with rail voltage <yes this happens and will blow up most decks and eq's without a second thought. so it is a important QC check of your reassembly workmanship.

Take those readings I requested above and get back with me my friend... C

marko · Post by **marko** » Sun May 27, 2007 12:26 am

i'm getting 0.01amps on both outputs

actually the left is ever so slightly lees than right but i wouldn't of expected them to be 100% accurate as i remember the zpa being a real pain to get anywhere matched perfeclty!

will raid my blown m50 again today to see if it has the same clip trasistors i can rob! is it possible these transistors have been iffy for a while or will it have been brought on by me tinkering?

1moreamp · Post by **1moreamp** » Sun May 27, 2007 12:58 am

marko wrote:i'm getting 0.01amps on both outputs actually the left is ever so slightly lees than right but i wouldn't of expected them to be 100% accurate as i remember the zpa being a real pain to get anywhere matched perfeclty!

will raid my blown m50 again today to see if it has the same clip trasistors i can rob! is it possible these transistors have been iffy for a while or will it have been brought on by me tinkering?

Well you can't adjust this amplifier like the ZPA, there is no DC offset adjustment. So you must live with the offsets unless you want to start replacing leaky transistors from input to output, to lower the offset. There is a circuit mod to add offset adjustment, but at 10 milli-volts or less I would not bother as this amp will probably be used for bass, and you will never hear the offset.

The main failed transistor is the 1038, but I replace both for general principals. They cost me less than 22 cents US, and that won't buy a pack of gum around here.

Do I think you did this tinkering around, NO. Your work could not have caused this, BUT constantly over-driving the amp with bass could cause this.
And most folks don't pay attention to the peak/clip indicators when they are booming around town, or at shows. But PG was kind enough to try and tell you when you were clipping the amplifier.

Time to rest, take care Marko, LMK if you have anymore issues... C

marko · Post by **marko** » Sun May 27, 2007 1:05 am

thanks again cecil!

this ms2125 drives mids only, 200hz up and the other ms2125 drives tweeters so they have a very easy life, it did run a pair of subs 2ohms mono for about a year though

it's bank holiday weekend over here so will take time to swap those clip transistors out, if i don't have one it looks lke i'll be scouring ebay!

noticed the same configuration on the left channel for clip transisitors, is this identical?

marko · Post by **marko** » Sun May 27, 2007 3:40 am

changed the 1038 and was the same so i thought i may as well change the 2389 and guess what, no clip light!! it's great having a junker m50 i can raid spares from, glad pg used same parts in all their old school amps

all is left to do now is build the amp up and check for voltage on the case, this time i'll be paying attention to cleaning and make sure the pads aren't holed

is there an alternative to the pads i can use?

marko · Post by **marko** » Sun May 27, 2007 6:51 am

well it's all back together now and has been playing away on my kef's in the garage for a good half hour at "realistic levels

), there wasn't any dc on the case.

thought it had all gone pear shapped though when i didn't get any sound and realised i forgot to re-solder the remote lead on back on (insert "bangs head against wall smiley" here).

1moreamp · Post by **1moreamp** » Sun May 27, 2007 10:25 am

congrats, that makes two for us now doesn't it

You will find that on these older cases the screw holes for the transistors tend to distort right at the hole. I suggest you chamfer these with a drill bit, so the power devices still have a flat mounting surface. I just did that to Nico Boom's also as PG used a older sink to build his 2250 on.

And yes the clip circuit is duplicated once in each channel. Good Work !

Happy weekend and Cheers C

marko · Post by **marko** » Sun May 27, 2007 12:19 pm

not quite sure what which bit you mean to chamfer! now i got the other ms2125 to re-cap but i'll take extra care with installation on that one!!

ms2125 showing clip after cap replacement!

ms2125 showing clip after cap replacement!

Re: ms2125 showing clip after cap replacement!