Hello to all.
I've decided to recap my old ML-9 amplifier, which, I think, deserve it
At the moment, I'm looking for 1) finding as much and well the same references than the original ones (which is a fancy yet uneasy job 2) finding the right way (procedure) to replace the capacitors on the amplifier board (which seems uneasy too, considering the way it is built, as I'll explain).
I’m not a professional but I’ve got plenty of time, I already recapped several small amplifiers (like some Cyrus 1 and 2 and other electronic devices), and I’m equipped with all the necessary tools. But if it appears that it is impossible for someone like me to achieve the task, I’ll try to find a qualified professional to have the job done…
Then.
1. finding the same references than the original ones. Easy for certain items. Difficult or impossible for some others. Possible with a few references if you buy 10 or 300 of them (depending on the item). For instance, I’ve chosen to replace the original Sprague 36DX 36000uF/100V by new United Chemi E36D101HPN363UEM9N bought from M…r (same voltage/capacity/size) which, I believe, are widely used by Madrigal to recap old ML amplifiers. Those capacitors are not on the signal path and should do the job. Selecting Cornell or Vishay (Sprague) of the same type would have demanded that I take at least 9 of them, which is rather expensive.
I couldn’t find the 10/100V same polycarbonate caps (and as film caps I suspect they are not too bad even after 35 years of use), but I bought East German RFT NOS 10/100 polycarbonate caps, in case (they seem OK, cap 15uF and Vloss 2,5% - acceptable?). I could find NOS Sprague 39D series 520/100V caps, and I’ll check the Vloss and ESR when I got them. Hoping they are OK or that I could reform them if necessary. I could find NOS 10000/10 Frako (same question than previous). I could find new TE1105 250/6V at M…r.
Then I’m wondering what to do with the 672D Sprague 680/6.3V/105° which are used in the DCO suppressor circuit. If I’m not mistaking, they should be on the signal path. Impossible to buy less than 291 units… at 3,5/unit, I won’t do it. So I’m looking to find a substitute, but which? I found this: UPM0J681MPD6TD (UCC). Rippled current is at 825 mA which could match the particular position of these caps. But I’m fairly ignorant about this point (and many others…). A helpful advice will be happily considered
2. Assuming that I found all capacitors needed. The physical operation of replacing them has yet to be done. (!) Rather simple when access to the whole circuit board is open and clear. The power regulation board and the DCO board are in this case. BUT not the amplifier board, which is almost stuck to the large radiator, the transistors being on the outside of the radiator, their pins soldered on the inside of the board, where all the components are. A kind of sandwich, where to take out the board in order to have a full access to the soldering points of the components (the board seems to be a double sided board with different tracks on one side and on the other), you have to unsolder the 10 power transistors and un screw them. Which is a delicate job to do, especially when you see that some transistors pins are VERY close to some components. The pictures show this, I hope.
Then.
What to do? For the first part, advice is required for a suitable match for the two 672D Sprague 680/6.3V/105° in their particular location (DCO circuit).
For the second one, I haven’t taken everything apart yet, hesitating to do that at this point. Do I have to unsolder all transistors so I can have full access to both sides of the circuit board? Would it be necessary, would it be possible (even if not a proper job, I admit) to cut the leads of the existing caps and solder the new ones on them? (please don’t throw too many stones at me. Or just unsoldering the defective (read old) parts and solder the new ones on place with enough welding to go through the hole (but not too much)?
I’m trying to understand how this was done by the manufacturer by observing carefully the board and it seems to me that at least the two 520uF Sprague Caps couldn’t be soldered before soldering the transistors… if this had been done, how could they have soldered the last transistor with so little space available? (but this is Mark Levinson and their techs are probably above the average humans about soldering).
Thanks for reading I'll be more than happy to read all of your comments and considerations!
please excuse my poor English and the quality of the pdf docs.
I've decided to recap my old ML-9 amplifier, which, I think, deserve it
At the moment, I'm looking for 1) finding as much and well the same references than the original ones (which is a fancy yet uneasy job
2) finding the right way (procedure) to replace the capacitors on the amplifier board (which seems uneasy too, considering the way it is built, as I'll explain).I’m not a professional but I’ve got plenty of time, I already recapped several small amplifiers (like some Cyrus 1 and 2 and other electronic devices), and I’m equipped with all the necessary tools. But if it appears that it is impossible for someone like me to achieve the task, I’ll try to find a qualified professional to have the job done…
Then.
1. finding the same references than the original ones. Easy for certain items. Difficult or impossible for some others. Possible with a few references if you buy 10 or 300 of them (depending on the item). For instance, I’ve chosen to replace the original Sprague 36DX 36000uF/100V by new United Chemi E36D101HPN363UEM9N bought from M…r (same voltage/capacity/size) which, I believe, are widely used by Madrigal to recap old ML amplifiers. Those capacitors are not on the signal path and should do the job. Selecting Cornell or Vishay (Sprague) of the same type would have demanded that I take at least 9 of them, which is rather expensive.
I couldn’t find the 10/100V same polycarbonate caps (and as film caps I suspect they are not too bad even after 35 years of use), but I bought East German RFT NOS 10/100 polycarbonate caps, in case (they seem OK, cap 15uF and Vloss 2,5% - acceptable?). I could find NOS Sprague 39D series 520/100V caps, and I’ll check the Vloss and ESR when I got them. Hoping they are OK or that I could reform them if necessary. I could find NOS 10000/10 Frako (same question than previous). I could find new TE1105 250/6V at M…r.
Then I’m wondering what to do with the 672D Sprague 680/6.3V/105° which are used in the DCO suppressor circuit. If I’m not mistaking, they should be on the signal path. Impossible to buy less than 291 units… at 3,5/unit, I won’t do it. So I’m looking to find a substitute, but which? I found this: UPM0J681MPD6TD (UCC). Rippled current is at 825 mA which could match the particular position of these caps. But I’m fairly ignorant about this point (and many others…). A helpful advice will be happily considered
2. Assuming that I found all capacitors needed. The physical operation of replacing them has yet to be done. (!) Rather simple when access to the whole circuit board is open and clear. The power regulation board and the DCO board are in this case. BUT not the amplifier board, which is almost stuck to the large radiator, the transistors being on the outside of the radiator, their pins soldered on the inside of the board, where all the components are. A kind of sandwich, where to take out the board in order to have a full access to the soldering points of the components (the board seems to be a double sided board with different tracks on one side and on the other), you have to unsolder the 10 power transistors and un screw them. Which is a delicate job to do, especially when you see that some transistors pins are VERY close to some components. The pictures show this, I hope.
Then.
What to do? For the first part, advice is required for a suitable match for the two 672D Sprague 680/6.3V/105° in their particular location (DCO circuit).
For the second one, I haven’t taken everything apart yet, hesitating to do that at this point. Do I have to unsolder all transistors so I can have full access to both sides of the circuit board? Would it be necessary, would it be possible (even if not a proper job, I admit) to cut the leads of the existing caps and solder the new ones on them? (please don’t throw too many stones at me
. Or just unsoldering the defective (read old) parts and solder the new ones on place with enough welding to go through the hole (but not too much)? I’m trying to understand how this was done by the manufacturer by observing carefully the board and it seems to me that at least the two 520uF Sprague Caps couldn’t be soldered before soldering the transistors… if this had been done, how could they have soldered the last transistor with so little space available? (but this is Mark Levinson and their techs are probably above the average humans about soldering).
Thanks for reading
I'll be more than happy to read all of your comments and considerations!please excuse my poor English and the quality of the pdf docs.
Attachments
What about the big blue cola bottle caps.
These seem to me the first candidates to recap.
The large radial electrolytic caps on your board can be desoldered on one side, measured and if ok, be resoldered after having cleaned the mounting hole.
If way below the expected value, desolder the other side, clean the second mounting hole and fit the new component in reverse order.
Replacing the Rifa caps seems a waste of energy.
Hans
These seem to me the first candidates to recap.
The large radial electrolytic caps on your board can be desoldered on one side, measured and if ok, be resoldered after having cleaned the mounting hole.
If way below the expected value, desolder the other side, clean the second mounting hole and fit the new component in reverse order.
Replacing the Rifa caps seems a waste of energy.
Hans
Hi Hans and thanks for writing and your advice.
You're right, the 2 big caps - 36000uF/100V 76x220mm Sprague - will be replaced by new United ChemiCon E36D101HPN363UEM9N I ordered from M..r at a fair price ($35 each) and this reference can be bought by units (so better than being obliged to order 9 Cornell Dubilier at $85 each). They will be ordered directly from the US plant and available before June.
Thanks for the tip of desoldering just one side and see what's happening (first, the condition), I'll try this soon. Don't forget that this amp was built in 1985 and I doubt they're still OK even if of the best quality when they were made. But I'll see that when measuring. I was wondering if there was some contact on the "invisible" side of the board where the cap MUST be soldered to. If welding above is enough, pouring lightly on the other side when heating, I'm OK for that and feel confident in succeeding that kind of task
The RIFA needs replacement I think, because the plastic envelope is not in a perfect condition and they are known for that and I wouldn't like them to disfunction. They are inexpensive, not difficult to replace, and Kemet new caps will be safer I think. I may be wrong but it's not too demanding.
Thanks again. Take care
You're right, the 2 big caps - 36000uF/100V 76x220mm Sprague - will be replaced by new United ChemiCon E36D101HPN363UEM9N I ordered from M..r at a fair price ($35 each) and this reference can be bought by units (so better than being obliged to order 9 Cornell Dubilier at $85 each). They will be ordered directly from the US plant and available before June.
Thanks for the tip of desoldering just one side and see what's happening (first, the condition), I'll try this soon. Don't forget that this amp was built in 1985 and I doubt they're still OK even if of the best quality when they were made. But I'll see that when measuring. I was wondering if there was some contact on the "invisible" side of the board where the cap MUST be soldered to. If welding above is enough, pouring lightly on the other side when heating, I'm OK for that and feel confident in succeeding that kind of task
The RIFA needs replacement I think, because the plastic envelope is not in a perfect condition and they are known for that and I wouldn't like them to disfunction. They are inexpensive, not difficult to replace, and Kemet new caps will be safer I think. I may be wrong but it's not too demanding.
Thanks again. Take care
Thanks Rayma, I'll buy ALL absolutely new parts I can find. But sometimes it's uneasy OR you have to buy 300 of them from well-known distributors.
That's a part of my question...
Anyway I won't put any part that is not in specs even if I bought them, no worry. Your advices are always smart and welcome
That's a part of my question...
Anyway I won't put any part that is not in specs even if I bought them, no worry. Your advices are always smart and welcome
When distributors require that you buy a large quantity of a part, it means they don't normally stock it, or do
not want to break open a master pack. Instead, go to a firm that does stock the part, and sells them individually.
This is right Rayma but I couldn’t find any for some of the references I needed. Maybe it’s easier in the USA than here too. Especially for those caps made in USA. I’ll take any of the reliable shops you could know. Many of the ones I tried to order from sell NOS. Except in Asia but they’re mostly fake.
Yes Hans, if it's not broken don't fix it, I agree.
The ML-20.x were built after 1990 if I remember, and my ML-9 is from 1985 if I read the manufacturing dates on caps. That's 35 years from now.
Beside, several points about this particular ML-9 :
- it's been unused for a certain time in its life (bought it in 2009 from the US, but probably left unused for many years, difficult to know)
- the power is much less than it should be. If I want some sound volume, I have to put my JR Consonance preamp at volume 100 (on 200), as my Cyrus 2+PSX is only at 9 o'clock for the same volume.
- I can hear a light buzz in speakers (Apogee Stage or Centaur, depending) when the amp is on, which could be some sign of aging of the 2 big Coke cans you mentioned and I wouldn't like to stress the transformer (which would be difficult to replace). You are perfectly right about this and I think your view can extend to all electrolytics.
- All caps seem OK (view) but with some of them, the ESR is high as well as Vloss when measuring. All caps except the RIFA I talked about before, but that's another story and not related with the sound.
- All the parts seem to be original and untouched (which is good news), but no electrolytic capacitor is eternal, even from these manufacturers and from this period... (I agree with Rayma about their replacement).
- Still, the sound is excellent, for sure. But it should be much more powerful, not for power itself but to prove that the amp is in "original" specs and will not suffer from demanding speakers like the Apogee.
- I've got a Cello Encore from the 90s (50W double mono) which works stronger, but even this baby which has been mostly switched on ever since 1991 is loosing power (much less than the ML-9, but loosing), and it will benefit too from a supply caps replacement some day. This could be an interesting topic to start later
Glad your 20.x goes well anyway and thanks for your interest and concern.
Take care.
The ML-20.x were built after 1990 if I remember, and my ML-9 is from 1985 if I read the manufacturing dates on caps. That's 35 years from now.
Beside, several points about this particular ML-9 :
- it's been unused for a certain time in its life (bought it in 2009 from the US, but probably left unused for many years, difficult to know)
- the power is much less than it should be. If I want some sound volume, I have to put my JR Consonance preamp at volume 100 (on 200), as my Cyrus 2+PSX is only at 9 o'clock for the same volume.
- I can hear a light buzz in speakers (Apogee Stage or Centaur, depending) when the amp is on, which could be some sign of aging of the 2 big Coke cans you mentioned and I wouldn't like to stress the transformer (which would be difficult to replace). You are perfectly right about this and I think your view can extend to all electrolytics.
- All caps seem OK (view) but with some of them, the ESR is high as well as Vloss when measuring. All caps except the RIFA I talked about before, but that's another story and not related with the sound.
- All the parts seem to be original and untouched (which is good news), but no electrolytic capacitor is eternal, even from these manufacturers and from this period... (I agree with Rayma about their replacement).
- Still, the sound is excellent, for sure. But it should be much more powerful, not for power itself but to prove that the amp is in "original" specs and will not suffer from demanding speakers like the Apogee.
- I've got a Cello Encore from the 90s (50W double mono) which works stronger, but even this baby which has been mostly switched on ever since 1991 is loosing power (much less than the ML-9, but loosing), and it will benefit too from a supply caps replacement some day. This could be an interesting topic to start later
Glad your 20.x goes well anyway
and thanks for your interest and concern.Take care.
Farnell is one of the suppliers I've been visiting but their catalog is not as complete as Mouser's or Digikey's ones (for the usual known and reliable suppliers). Then I've tried Tedss from whom I received today the 39D 520uF/100V caps which seems good on my meter (ESR near 0 and Vloss<0.4%, cap in specs). The written date code seems to be 2014.
Just received the availability date for the 2 UCC cans, end of July... As I said, plenty of time to fix the rest before ending with the blue cans !
Thanks for your concern and while we are stuck at home, let's take our time for the important things (in France at least).
Just received the availability date for the 2 UCC cans, end of July... As I said, plenty of time to fix the rest before ending with the blue cans !
Thanks for your concern and while we are stuck at home, let's take our time for the important things (in France at least).
Thanks for your advice ticknpop. I bought 2 that's no big deal to let them unused. Any good electrolytic in the same range (like MAL213814103E3) would be OK?
The polycarbonate won't probably need to be replaced but I bought 4 just in case it would be necessary... (they're rarer year after year).
Beside, I like capacitors, I often find them "nice"
And my initial purpose is to keep original as much as I can when necessary.
The polycarbonate won't probably need to be replaced but I bought 4 just in case it would be necessary... (they're rarer year after year).
Beside, I like capacitors, I often find them "nice"
And my initial purpose is to keep original as much as I can when necessary.
Did you measure those caps ? I have inherited first version of Krell KSA100 and tested the caps for capacitance and ESR . They all test good & measure (ESR) better than new parts I had in the drawer. Those 40.000uF/70V jar sized caps will never dry have very low ESR so why bother ? I may just "bypass " them with something like 4700uF of small modern caps and have 470 uF at power transistor legs and that's it .
I can tell you that the 2 Sprague 36000uF/100V cans seem to be out of range... the meter (a Chauvin Arnoux multimeter which can't be suspected) has problems to measure them properly (more in the nF range than the uF range). But to be sure, I've got to take them out both, so take the transformer out and I'll do that not today
Once again, it's no problem if the 2 UCC that I ordered fit well. in terms of built quality, I think they're at least as good as the original Sprague.
Once again, it's no problem if the 2 UCC that I ordered fit well. in terms of built quality, I think they're at least as good as the original Sprague.
I guess it's up to you. I would definitely remove the transistors and apply new thermal pads: either Keratherm or mica + goo. Not sure how good the thermal interface is after 35 years of sweating it.
The Frako caps are imho pretty good sounding and hard to replace with modern alternatives if preserving the original sound signature is important.
The NOS caps should be carefully reformed and left under normal operating conditions (with ripple) for a week or so and then tested for leakage.
You don't mention anything about the relays. I would think it prudent to replace those as well and this may even bring some audible improvement.