They all have this shape in old Yamahas. But some were saying here that only those with LGB print are to be replaced. But there's plenty of old threads about it.Little " outhouse" TO-92 case style are the noisy ones.
So you replaced those c458s pressed against radiators on main amps? It had never been done. I asked some time ago here and every one said they kept them. What did you use and have you tested bias/idle currents if all is ok? Has it worked for some time now?Hey sssboa!
Thanks for your input. New tantalums, eh? I’ll have to think about that. Sounds deluxe!
From the state of the power supply, and how long it sat idle, I really expected it to sound tired and in need of more work. Actually, now that I have it all clean and working and hooked up to the NS-670s in the bedroom, I am pretty impressed by just how fresh and quiet-sounding it is and I may just leave well enough alone for now on the tuner recap and just listen to it for a bit.
I have read a lot of posts pro and con on the notorious c458s—but my (more limited I am sure) experience has usually made me lean towards replacement, LG version or no. It made a huge difference to a noisy Sansui 2000a I just did, and also my Kenwood KA-7002 that must have had like 30 of ‘em. In a couple of other amps, too, it has seemed to lower the noise floor, though of course they are not the only troublemakers out there as far as noisy vintage transistors. Seems to be most critical in tone and EQ stages; I haven’t dealt with them on tuner boards before. That said, they seem to be doing their job in this case, this CR-1000 doesn't have the noise issues of those others so far!
I’ve only seen one c458 outright fail; that one was biasing the outputs on a CR-820. The CR-1000 had those too, I already replaced those when I did the main amps.
I’ll be revisiting this, I guess, whenever I decide to get back in there on finish recapping...but right now there’s an HK 930 calling me with some similar power supply-to-tuner issues...this hobby (addiction) is neverending, isn’t it?
I'm doing the same for tants, replace with new or film where appropriate.Most of tantalums there are small 1uF/2.2uF/3.3uF one is 0.68uF two are 10uF. Replace them with aluminum lytics, I wouldn't do, as Master Yoda would say. Actually the most expensive option is to replace them with new tantalum which I was crazy enough to do on my 2 units (I used Kemet tantalum caps), just for 0.68uF I used film.
Out of 5 units of CR-1000/CR-800 I ever had only one original tantalum cap went bad causing crackling in one channel, it was 1uF on tone. I measured them all while replacing and only 0.68uF/1uF/2.2uF measured badly, 3.3uF and higher seemed ok.
On tone, filter and phono in one CR-1000 I replaced all tantalum caps with Kemet tantalum caps while in my other CR-1000 I replaced all 4.7uF and smaller with WIMA and EPOCS polyester, bigger than 4.7uF with Kemet tantalum. Difference in sound is enormou..., no actually there's no difference in sound, sorry all audiophiles.
There are like 12 of 2SC458 on tuner, then a few 2SC460 which are kind of the same, they may be both replaced with KSC1815 but it would be like 20 transitors to replace, I never did that.
Personally I don't understand that talk about how bad c458 are, I never saw any go bad and having them 12 on one board at least one should go. Actually on my CR-800 (same FM tuner as of CR-1000) I had 2SC460 gone bad. On schematics some c458s are marked as LG which some say are the worst kind of c458, they should have LG printed on them too.
If someone wants to use new tantalum caps I have a warning about non-branded or so called Multicomp caps which is a synonym for not branded parts from Taiwan. They are cheap like dirt. I used 2 of them on tone of CR-800 and they caused crackle in volume pot straight away.I'm doing the same for tants, replace with new or film where appropriate.
As for the 2sc458s, they have lasted 40 years and longer, just how bad can they be. I do replace with 2240s (while supplies last) but I'm replacing everything that can be replaced these days.
Very interesting you'd say that, I tried to find some lifespan-before-whiskers data on ROHS transistors recently and couldn't find anything concrete. Surely lab approximations must exist at least? I'd appreciate link/s if you have any, particularly case-size specific data.Unfortunately, since they stopped using lead, new transistors won't last as long (definitely not 40 years). Even less if they start forming tin whiskers on the pnp/npn wafers (maybe not plastic encapsulated but to-3 that are in a vacuum) where the wafer connects to the pins.
Most data/research I've read comes from hobbyist and technicians. I thought that encapsulation would stop the wiskers but it looks like it won't (looked into it after my post). I'm guessing the military will start looking into this when one of their f35s goes down because of a short circuit.Very interesting you'd say that, I tried to find some lifespan-before-whiskers data on ROHS transistors recently and couldn't find anything concrete. Surely lab approximations must exist at least? I'd appreciate link/s if you have any, particularly case-size specific data.
Do you think the encapsulation stops the whiskers forming internally? It wouldn't be hard to tin external leads with leaded solder, but I assumed encapsulating wouldn't actually stop the problem internally. Hopefully you have something concrete, it was only a month ago I was looking.
Damn guys. I changed so many vintage transistors for new. I didn't know I would have to shave their whiskers after 5 years. Government must do something about it! Redirect money from global warming research or something.Most data/research I've read comes from hobbyist and technicians. I thought that encapsulation would stop the wiskers but it looks like it won't (looked into it after my post). I'm guessing the military will start looking into this when one of their f35s goes down because of a short circuit.
From onsemi:
Tin Whiskers
Mitigation Strategies
Changing the lead plating material to 100% matte Sn (tin) from SnPb raises industry concerns about tin whisker growth. ON Semiconductor has implemented the following mitigation strategies to minimize the occurrence of tin whiskers.
Whisker Testing
- Increased the plating thickness from 5 µ m to 7.5 µ m minimum; 10 um nominal.
- Implemented a post - plate anneal of 150 ° C for 1 hour within 24 hours of plating
- Implemented strict plating process controls
ON Semiconductor has conducted tin whisker testing following the guidelines of JEDEC standard JESD22A121. Testing has been conducted on packages utilizing Alloy 42 and Cu leadframe materials with matte tin lead finish. Three test conditions have been used to evaluate whisker growth: Temperature Cycling (-55/+85°C) Ambient Storage at 30°C/60%RH, and High Temperature/High Humidity Storage at 60°C/87%RH.
ON Semiconductor's internal whisker acceptance spec is 50µ maximum length. All whisker test results to date have passed this spec.
That's some interesting info from OnSemi there fernarias, thanks for that. It is good to know they have a mitigation strategy though how does one evaluate its effectiveness vis-à-vis longevity? I am guessing we will just have to wait and see...
I have to say, I am a fan of improbable word combinations, and "internal whisker acceptance spec" is not a phrase I'd expected to encounter today. Personally speaking, my external whisker acceptance spec has gotten a bit lax as of late... definitely over the 50µ maximum length!
Silly EU, always looking after peoples' health and wellbeing... Don't know where you all are, but here in the US I think we just freed up a lot of that global warming research money, we could look into it...after all the Cabinet members' wives all get new diamond shoes of course.
Update on the ol' CR-1000: After an extended listening session it became apparent that ksc1845 was not a good replacement in the bias circuit. Went back in today to revert back to the c458s, and found to my bewilderment that when I had made the change some 12-18 months ago, I had only done one side. Can't really remember with any certainty now why I had done it in the first place. Oh well, back to normal now, running clean and cool. Discovered while setting idling current that one of the trimmers has gotten very touchy and probably needs replacing. I was able to set the current in the specified range but not near as exact as I would like. Thinking of replacing all 6, they are 4.7k ohms, any suggestion for replacements? is 5k alright? Maybe something like these?
https://www.mouser.com/datasheet/2/54/352-776447.pdf
The only other thing I am currently unhappy about an unbalanced volume control--I get almost nothing out of the left channel until about 1 on the volume control, it doesn't catch up with the right channel until about 1.5 but tracks with it after that. I do a fair amount of late-night listening at lower volumes (love the variable loudness slider for this!) and the weak left channel bugs me. I might have missed the volume pot when I cleaned the controls, going to clean the heck out of it tomorrow and see if that makes a difference; any other suggestions?
That's some interesting info from OnSemi there fernarias, thanks for that. It is good to know they have a mitigation strategy though how does one evaluate its effectiveness vis-à-vis longevity? I am guessing we will just have to wait and see...
I have to say, I am a fan of improbable word combinations, and "internal whisker acceptance spec" is not a phrase I'd expected to encounter today. Personally speaking, my external whisker acceptance spec has gotten a bit lax as of late... definitely over the 50µ maximum length!
Silly EU, always looking after peoples' health and wellbeing... Don't know where you all are, but here in the US I think we just freed up a lot of that global warming research money, we could look into it...after all the Cabinet members' wives all get new diamond shoes of course.
Update on the ol' CR-1000: After an extended listening session it became apparent that ksc1845 was not a good replacement in the bias circuit. Went back in today to revert back to the c458s, and found to my bewilderment that when I had made the change some 12-18 months ago, I had only done one side. Can't really remember with any certainty now why I had done it in the first place. Oh well, back to normal now, running clean and cool. Discovered while setting idling current that one of the trimmers has gotten very touchy and probably needs replacing. I was able to set the current in the specified range but not near as exact as I would like. Thinking of replacing all 6, they are 4.7k ohms, any suggestion for replacements? is 5k alright? Maybe something like these?
https://www.mouser.com/datasheet/2/54/352-776447.pdf
Not familiar with the CR1000, done a few CA-1000's. Will only use Bourn 3296P series, I used 5k, often 2 of the legs are tied together, in anycase the 5k isThinking of replacing all 6, they are 4.7k ohms, any suggestion for replacements? is 5k alright?
Not familiar with the CR1000, done a few CA-1000's. Will only use Bourn 3296P series, I used 5k, often 2 of the legs are tied together, in anycase the 5k is
within 4.7k tolerances.