Uncle Paul's B-2 Rehab

As you point out, there is no need for 630V, they are just that rating because more and more low-voltage series have been taking out of production (or not kept on stock)
For the snubber caps around the rectifier diodes, assume that they could face full Vpp of the transformers' AC, so 200V would be the minimum to take
Capacitor voltage ratings come in standard values like 250V/400V/630V
Although multiple voltage rankings exist on the datasheets..... retailers don't like to keep each uF/V item in stock, so ..... you simply end up with what is available, not what you theoretically would like to get. Same issue with the Nichicon FG series: 50V and 100V are normally available at Mouser, but the 63V version is often out of stock.

Yes, finding the parts with the right dimensions (leg space & overall) versus available capacitance and voltage rating can be 3x more work than actually putting them in.

As for rectifier snubber caps, keeping inductance (ESL) low is also key, that's why short legs and installing them as near as possible the rectifier legs is the best practice.
Which is also the reason why ceramic disc capacitors are often used for snubber functions: they have a much lower ESL than film capacitors.
You can see in many Yamaha amps (CA-1000/2000, CR-2020, etc) that they used ceramic disc caps for snubber functions
So if you can't find the right snubber caps in film format, don't hesitate to go for ceramic disc caps.

As for the bypass caps around the transformer secondaries (the two existing green mylar caps near the rectifier diodes) , try to stay with a polypropylene film cap.
 
The Electrolytic Capacitor Board is now complete. I switched back to the original plan of locating the large filter caps to the bottom of the board. I used the CAD drawing I made earlier as a template and drilled two 5/64" (aprox 2mm) diameter holes for each terminal.
upload_2017-3-27_17-14-12.jpeg
upload_2017-3-27_17-44-1.jpeg

I replaced the jumpers with some Teflon jacketed Mil-Spec wire. The wire itself is white, so I used some colored heat shrink tubing to keep to the original color code. The snubber caps and C251/C252 are all .01uF Panasonic ECW-F6103JL
upload_2017-3-27_17-11-52.jpeg


Here we see how the filter caps connect to the PCB. The two on the right came in right at the corner of the solder pads used by the originals. The two on the left were more problematic. I wound up using the legs of the 1 uF bypass caps to make the electrical connection and provide support. The filter caps are also attached on the other side via the mounting pads that came pre-attached.
upload_2017-3-27_17-12-37.jpeg

The ideal connection would have been to remove the epoxy coating to make new pads like the ones on the right, but it's made of tougher stuff than modern coatings and I was concerned about damaging the trace underneath. It would have been a bit more elegant, but I don't think it will make a meaningful difference.
 
Last edited:
Great work, but I also think you need to remove the coating around the pins of the filter caps themselves and make a solder connection there. I wonder if the bypass caps legs are fit to do this job.
 
Great work, but I also think you need to remove the coating around the pins of the filter caps themselves and make a solder connection there. I wonder if the bypass caps legs are fit to do this job.

The connection was bugging me as well, so I COMPLETELY redid the big electrolytic cap connections. A LOT of work, but it is better then the previous attempts both electrically and mechanically.
upload_2017-3-29_13-44-38.png

I started by unsoldering and removing the two left bypass caps, then created four new pads in line with the holes I drilled. The new pads were a lot of very delicate work, but well worth it in the end. The process was to very delicately scribe a line through the green epoxy with an exacto knife just barely touching the copper. Then I carefully scraped a fine hatch mark in that area. Next I used some MEK to clean up the epoxy dust and further soften the epoxy still stuck to the copper. I repeated the scraping and MEK treatment a couple of times to get down to the bare copper before polishing it smooth with some green 3M automotive final stripping pads cut into tiny pieces. I cleaned with alcohol and tinned the areas before continuing.

Each EC lead was then bent to to lay flat on the new pad at an angle to maximize contact and then soldered in place

The two caps on the right were simply cleaned up by removing the bypass cap "tails" that I wound around the capacitor leads, and then bending the EC leads like the two on the left and soldering.

upload_2017-3-29_13-51-46.pngupload_2017-3-29_14-10-52.png

Everything now looks great and lines up nicely
upload_2017-3-29_13-43-51.png
upload_2017-3-29_13-42-55.png

Now waiting on a few parts for the PS Board I forgot to order along with figuring out my parts list for the driver boards. You can be sure I'm paying rapt attention to your thread at this point!
 

Attachments

  • upload_2017-3-29_13-58-49.png
    upload_2017-3-29_13-58-49.png
    296.3 KB · Views: 5
The connection was bugging me as well, so I COMPLETELY redid the big electrolytic cap connections. A LOT of work, but it is better then the previous attempts both electrically and mechanically.
View attachment 898877

I started by unsoldering and removing the two left bypass caps, then created four new pads in line with the holes I drilled. The new pads were a lot of very delicate work, but well worth it in the end. The process was to very delicately scribe a line through the green epoxy with an exacto knife just barely touching the copper. Then I carefully scraped a fine hatch mark in that area. Next I used some MEK to clean up the epoxy dust and further soften the epoxy still stuck to the copper. I repeated the scraping and MEK treatment a couple of times to get down to the bare copper before polishing it smooth with some green 3M automotive final stripping pads cut into tiny pieces. I cleaned with alcohol and tinned the areas before continuing.

Each EC lead was then bent to to lay flat on the new pad at an angle to maximize contact and then soldered in place

The two caps on the right were simply cleaned up by removing the bypass cap "tails" that I wound around the capacitor leads, and then bending the EC leads like the two on the left and soldering.

View attachment 898878View attachment 898901

Everything now looks great and lines up nicely
View attachment 898876
View attachment 898875

Now waiting on a few parts for the PS Board I forgot to order along with figuring out my parts list for the driver boards. You can be sure I'm paying rapt attention to your thread at this point!


Looking good Uncle Paul:trebon::thumbsup:
 
I haven't been writing much lately as I have been busy with some non hobby related activities. When I've had has been been spent researching other threads regarding the driver boards and capacitor replacement. IMO the B-2 deserves better than mylar caps in the signal path, so like Mr. Yamaha I am replacing all with polypropylene.

The problem is that all the general purpose polypropylene caps with the right specs that I can purchase in singles come with plastic box covers and are pretty rigid where lead spacing is concerned. For example, many of the caps used on these boards have spacings of 4mm and 11 mm, so it's tough to get a 5mm or 10mm cap to fit. So, not only is it mechanically awkward, they also don't fit the visual aesthetics of these boards.

Now I can find RF Microwave polypropylene caps like the ones I used for the power supply bypass and snubber duty. I'm just not sure that these are also suited for the driver boards. I've done a LOT of research but can't find out just what the difference is between a general purpose and an RF microwave capacitor except for a few statement saying that the microwave varieties are more stable at higher frequencies.

Does anybody know what the actual difference between the types and if the microwave caps such as Panasonic ECWF(A) are good fits for this application? Or perhaps certain suppliers that have what I need in smaller quantities. For example, one part I'd like to use is the Xicon MPP 1429-4103, but I can only find it in quantities of 100 pieces. I found one supplier that had it listed for single units, but then subbed a mylar version since the was "the replacement from Xicon".
 
Last edited:
I haven't been writing much lately as I have been busy with some non hobby related activities. When I've had has been been spent researching other threads regarding the driver boards and capacitor replacement. IMO the B-2 deserves better than mylar caps in the signal path, so like Mr. Yamaha I am replacing all with polypropylene.

The problem is that all the general purpose polypropylene caps with the right specs that I can purchase in singles come with plastic box covers and are pretty rigid where lead spacing is concerned. For example, many of the caps used on these boards have spacings of 4mm and 11 mm, so it's tough to get a 5mm or 10mm cap to fit. So, not only is it mechanically awkward, they also don't fit the visual aesthetics of these boards.

Now I can find RF Microwave polypropylene caps like the ones I used for the power supply bypass and snubber duty. I'm just not sure that these are also suited for the driver boards. I've done a LOT of research but can't find out just what the difference is between a general purpose and an RF microwave capacitor except for a few statement saying that the microwave varieties are more stable at higher frequencies.

Does anybody know what the actual difference between the types and if the microwave caps such as Panasonic ECWF(A) are good fits for this application? Or perhaps certain suppliers that have what I need in smaller quantities. For example, one part I'd like to use is the Xicon MPP 1429-4103, but I can only find it in quantities of 100 pieces. I found one supplier that had it listed for single units, but then subbed a mylar version since the was "the replacement from Xicon".
Panni ECWF would work great if you can fit them in. The ones you can still buy are higher voltage and larger in size
 
I love the Xicons, but they indeed are not available in all needed values and there aren't much other compact PP films with the retro look available. I've searched a lot. I became pretty keen on the Nichicon (Q)XP series, available at Mouser and Digikey. Their sizes are slightly bigger (only length and height, not width) than the Xicons.

And I did not have much problems with lead spacing, Although not everybody would like this approach, but I don't see any drawbacks in bending leads like this:

IMG_3981.JPG
 
Besides options in brands and models, the values of most film caps on the B-2 driver board are not critical (meaning, you can take the next higher value)
Which opens in some cases the way for using Xicon MPP caps which are still available with Mouser.

Do not play with the values of the small pF ceramic/polystyrene caps though; those are critical (and should not be changed to film caps).

C110/111/112/113/115: no need for PP (MKP), you can keep those in MKT/polyester.
(still red-brown lacquered like in the photo in the previous post)
 
Besides options in brands and models, the values of most film caps on the B-2 driver board are not critical (meaning, you can take the next higher value)
Which opens in some cases the way for using Xicon MPP caps which are still available with Mouser.

Do not play with the values of the small pF ceramic/polystyrene caps though; those are critical (and should not be changed to film caps).

C110/111/112/113/115: no need for PP (MKP), you can keep those in MKT/polyester.
(still red-brown lacquered like in the photo in the previous post)
@Oilmaster
What about silver mica to replace the ceramic/polystyrene? I have used those before with very good results.
About the 2sc1452 2sa810. Do you replace them for reliability or only if they have failed? I also noticed on a couple of boards of the PSU that the 10DC4 bridged diodes have been replaced. Needed or just precaution! Thank you
 
The ceramic/polystyrene can always be replaced by silver/mica; it is just the most expensive option, notably when you select tight tolerance classes.
Keeping the rated voltage low (100V ~ 200V) will keep the size small (i.e Cornell-Dubilier CD5 series)
Unfortunately some dudes grab what they can (i.e. 1000V or higher rated silver/mica caps), ending up with 5x larger sized silver/mica caps than the original ceramic caps; looking really idiot....

As for semi-conductors, for me it's a combination of reliability/improvement/lower noise/matching, if it is my amp, or if the 'client' wants all replaced.
In other cases only selected trannies/diodes go...
Enough debates have been held on AK on that matter, no need to throw another one in...
 
the small pF ceramic/polystyrene caps though; those are critical (and should not be changed to film caps).
Regarding changing ceramic (not PS) to FKP, which particular applications is this bad for?
Could you explain why?

Thanks
 
Regarding changing ceramic (not PS) to FKP, which particular applications is this bad for?
Could you explain why?

Thanks

That is an advanced topic that requires a lot of nuance, even down to which manufacturer and exact series.
But to give you a generic answer: stacked film caps (i.e. non-metallized film caps) are rapidly disappearing, leaving more and more metalized film caps as the only available film cap.
The issue with metalized film caps is that they (thus) exhibit a certain amount of inductance (ESL), as where the length of the legs also adds up.
It would not be recommended to introduce unnecessary inductance to certain locations in the amp circuit (such as miller caps, high-pass filters, etc) in the small signal input stages.
I have had small resonance frequencies after some rebuilds which were only solved by swapping the new PP caps back to ceramic or silver/mica

Furthermore, film caps can have much poorer temperature stability characteristics, which are sometimes no problem for the application, but sometimes they might be....

Yes, I have seen very small non-metalized inductance-free-winded film caps down to 10 pF, but they are nearly impossible to get hold off (and cost $$$).
And why bother anyway when readily available inductance-free caps like NP0/C0G MLCC and silver-mica exist with nearly flat tempco....?
 
Back
Top Bottom