Phono "motorboating" solved(?)

[Steven Tate]

One more data point for motorboating. I recapped the P700 board in my 2230 and decided to replace the 2SC1000s and 2SC458s with KSC1815s based on multiple threads. I also upgraded the input and output caps to 1uF film and added the 47 ohm resistors to the outputs . . . . I got motorboating! put the 2SC1000s back in and it's fine. I'm leaning towards Leestereo's theory that the Cob of the 2SC1000s, listed as 6 - 10pF, is the primary reason they don't oscillate. All the KSC transistors have a very low capacitance. I suppose i could have tried a couple of 10pF caps across B-C but i'd already spent too much time on this one. I'll leave that experiment for the next time.

Since going down this road a few years ago, I have had one board that oscillated after changing all six transistors to KSC1815. In that case, I put the 2SC1000's back in and it worked fine. Although those transistors have a reputation for going noisy, I simply haven't experienced it. So, I now leave the 2SC1000's unless I'm finding significant noise in the phono mode, and only change the 2SC458's as automatic changes.

 

[goldswimmerb]

Since going down this road a few years ago, I have had one board that oscillated after changing all six transistors to KSC1815. In that case, I put the 2SC1000's back in and it worked fine. Although those transistors have a reputation for going noisy, I simply haven't experienced it. So, I now leave the 2SC1000's unless I'm finding significant noise in the phono mode, and only change the 2SC458's as automatic changes.

I just pop on some 4-10pf miller caps and call it done.

 

[Steven Tate]

I just pop on some 4-10pf miller caps and call it done.

Is that across B and C?

 

[Richard P]

@Steven Tate - I agree with you and I typically don't change transistors either unless they present problems, but on this receiver, the 2SC1000s (and/or the 2SA493s) on the preamp were causing hiss, so I shotgunned them on the phono board as well guessing they'd be noisy. Lesson learned . . . .

@goldswimmerb - have you had success adding the small caps across KSC1815s or KSC1845s on the phono board? All 4? I saw some photos of boards where they had been added to the Hx01 and Hx02 transistors, but i saw another reference to the Hx03 and Hx04 transistors being the source of the oscillation.

Rich

 

[goldswimmerb]

@Steven Tate - I agree with you and I typically don't change transistors either unless they present problems, but on this receiver, the 2SC1000s (and/or the 2S493s) on the preamp were causing hiss, so I shotgunned them on the phono board as well guessing they'd be noisy. Lesson learned . . . .

@goldswimmerb - have you had success adding the small caps across KSC1815s or KSC1845s on the phono board? All 4? I saw some photos of boards where they had been added to the Hx01 and Hx02 transistors, but i saw another reference to the Hx03 and Hx04 transistors being the source of the oscillation.

Rich

I added them to the same location they were on my 2325 I did a while back, 3-10pf at Hx01 and Hx02. Since the circuit relies on feedback I am not sure how much different it would be adding them to x03 and x04.

 

[willyrover]

This thread was helpful with my 2275 phono doing evil things after replacing all transistors with 1845s. Motorboating when input shorted (i.e., during automatic turntable functions), popping noises when adjusting balance, general noise, static, etc. Sine wave looked a bit goofy when adjusting the board for clipping.

For the record: I returned SC1344 to their original positions (H401-H404). I removed SC1845 and installed SC1815 (H405-H406) and added 10pF across B and C. Adjusted for balanced clipping. Sine wave looks good.

 

[Dingman]

This thread was helpful with my 2275 phono doing evil things after replacing all transistors with 1845s. Motorboating when input shorted (i.e., during automatic turntable functions), popping noises when adjusting balance, general noise, static, etc. Sine wave looked a bit goofy when adjusting the board for clipping.

For the record: I returned SC1344 to their original positions (H401-H404). I removed SC1845 and installed SC1815 (H405-H406) and added 10pF across B and C. Adjusted for balanced clipping. Sine wave looks good.

Thanks for that detail.

 

[rBuckner]

This thread was helpful with my 2275 phono doing evil things after replacing all transistors with 1845s. Motorboating when input shorted (i.e., during automatic turntable functions), popping noises when adjusting balance, general noise, static, etc. Sine wave looked a bit goofy when adjusting the board for clipping.

For the record: I returned SC1344 to their original positions (H401-H404). I removed SC1845 and installed SC1815 (H405-H406) and added 10pF across B and C. Adjusted for balanced clipping. Sine wave looks good.

Even more stable is to adjust for a small flat spot on the waveform bottoms.

 

[GlasgowGrip]

Another data point: Have a 2230 that I recapped and everything is sounding great, the power amp boards, the preamp, and the overall performance all test out better than original specs. Tuner is working well. The P700 phono board in this one is special though - it had a very bizarre mixture of old-school replacement transistors (ECG-199's, Motorola HEP S0015's) - really strange, since nothing else in this unit appeared to have ever been touched! I can only begin to speculate what may have happened... maybe the 2SC458's were misbehaving at some point, and someone decided to shotgun all the transistor spots on the board? Weird.

Anyway, I recapped P700 and replaced some of the signal path resistors with Dale CMF55's, and replaced all the silicon - 1N4448's and KSC1845FTA's, as per the common advice. At this point I found the performance when playing Phono through the Pre Out's was pretty good, but then tried to test the phono stage by connecting the analyzer to the Tape Mon Out jacks. The thing would immediately oscillate/motorboat.

First replaced KSC1845's at H705-H706 with KSC1815's - this slightly improved the results through Pre Out's, but did not solve the oscillation when connected to Tape Mon Out's. Next changed H701-H702 to KSC1815, still oscillated when connected to Tape Mon Out's. Finally, changed H703-H704 to KSC1815 (so all six positions are now KSC1815) and that finally solved the oscillation when connecting to Tape Mon Out's. THD and SNR test pretty well at 5mV input, but then looked at Frequency Response, and there is still a stability issue at higher signal levels. The FR stays flat up to about 1.77mV input, then begins to get funky as the signal level increases - by the time it gets up to 5mV it is whacked out at the lower frequencies.

I'm going to try putting 2SC1000's back in one channel - will start with H701 and test again, then will also change H703 and test again. At that point I may try adding 10pF Miller caps on the other channel just to see which works, and if either is better. Will report back tomorrow...

 

[GlasgowGrip]

Will report back tomorrow...

Update: I wanted to attack this with a semi-methodical approach, trying to rule things in or rule things out, hopefully without trying every permutation of 2SC1000's, KSC1815's, and Miller caps.

First up was to try 2SC1000's in the first two positions. My guinea pig was the Right channel...

Step 1: replaced KSC1815 at H701 with 2SC1000; Result was slightly reduced oscillation in Right channel, but no bueno.​

Step 2: replaced KSC1815 at H703 with 2SC1000; Result was additional reduction in oscillation in Right channel, better, but nowhere near a solution.​

So next up was to try Miller Caps with the KSC1815's, so that meant using the Left channel (H702, H704, H706). Based on the last post by @willyrover, I decided to cut to the chase and start by adding 10pF at H706...

Step 1: added a 10pF C0G MLCC across B-C of H706; BINGO!!! Result was zero oscillation in Left channel, flat Frequency Response at 5mV input even while the Right channel was still doing its funky dance.​

Okay, need to confirm by going back to the Right channel...

Step 1: added a 10pF C0G MLCC across B-C of H705; Totally Awesome result was zero oscillation in both channels, flat Frequency Response from 20Hz-20kHz at 5mV input. Cranked it up... 7, 8, 9, 10mV input... no oscillation. Sweet!​

To button this one up, I went back and removed the 2SC1000's from the Left channel, so the final complement on this board is KSC1815's at all 6 positions (H701-H706) and 10pF Miller caps added to H705 and H706 (I used C0G MLCC's).


Confirmed that Phono frequency response, both channels driven at 5mV and measured at the Tape Monitor Outputs is flat (-0.6/+0.5 dB) from 20Hz - 20kHz, using inverse RIAA weighting.



I do have a couple more 2230's here, so I will test the phono stages and report back on the repeatability of the problem, and repeatability of the solution.

 

[GlasgowGrip]

I do have a couple more 2230's here, so I will test the phono stages and report back on the repeatability of the problem, and repeatability of the solution.

Okay, another 2230, another update:

So in this second 2230, when I recapped/restored the P700 I had originally replaced H701-H706 with KSC1845's, but after reading several threads about the motorboating issues reported by others (@Steven Tate et al) I had gone back and changed out H705-H706 with KSC1815's as a precaution.

Begin test...

Step 1: plugged the analyzer into PhonoIn and PreOut, and the performance was reasonable; So far so good.​

Step 2: moved the cables from PreOut over to TapeMonOut and immediately the P700 starts motorboating/oscillating. Okay, problem replicated.​

Step 3: added a 10pF C0G MLCC across B-C of H705 and H706; Result was zero oscillation in both channels, flat Frequency Response from 20Hz-20kHz at 5mV input. Then turned it up to 10mV input... still no oscillation. Nice.​

So... the Miller cap fix to the motorboating problem is confirmed to work, at least in this unit. Final complement on this board is KSC1845F at four positions (H701-H704), KSC1815Y at two positions (H705-H706) and 10pF Miller caps added across B-C at H705 and H706 (again I used C0G MLCC's).

As an aside, my final observation is that this P700 tests much quieter than the one in my last post, about 8 or 9dB better Signal To Noise Ratio. Best guess is that this is likely unrelated to the transistor complement, since the culprit appears to be a much higher 60Hz spike on that first unit. This one is a later unit, S/N 174335 versus S/N 25250 for the first unit, which also may be unrelated. It will bear further investigation of that first unit.

Anyway, I'll leave with some test results of this second unit's P700 with the new Miller caps installed:


Also, I have one more 2230 here that I've mostly restored - the power supply, P750's, P400 and P450 - but it still has the original unmolested P700. My plan for tomorrow is to test the P700 as found, then I will recap it and will leave the 2SC1000's in place, and will just replace the 2SC458's with KSC1815's. At that point I'll confirm whether it then motorboats through the Tape Mon Out's, and if so I will install the Miller caps, test again, and will report back here...

 

[rBuckner]

Thanks for all the work and sharing it!

 

[GlasgowGrip]

Thanks for all the work and sharing it!

You're welcome Randy - just hoping to contribute a little something back to the community once in a while.

So for today, finishing up, here is my third data point...

This one is a 2230 that is still a work in progress: the switches have been deoxited, pots have been faderlubed, P750's, P400 and P450 have been recapped, main filter cap and output coupling caps have been replaced, so overall performance is solid and the unit is pretty well known to me at this point. It still has a fully stock/unmolested P700 board, so seems to be a good testbed for continuing this post.

For context, the original transistor complement is 2SC1000 BL at H701-H702, 2SC1000 GR at H703-H704, and 2SC458 A at H705-H706.

Baseline testing...

Step 1: plugged the analyzer into PhonoIn and PreOut, and the performance was reasonable; So far so good.​

Step 2: moved the cables from PreOut over to TapeMonOut and things are quiet, as expected. Tested THD and SNR at 5mV input, and Frequency Response at levels from 2.5mV up to 10mV. FR is within spec, although it does show just a hint of a stability issue beginning at the 10mV signal level. No problem here.​

P700 recap...

• C701-C702 and C714-C715 are "ugly brown" films, so leave them alone.
• C705-C706 and C707-C708 are NPO ceramics, so leave them alone.
• C709-C712 are polystyrenes, so leave them alone.
• C703-C704 are tantalums, so upgrade to 47uF Nichicon UES "Muse" bipolars.
• C713 and C011 (mounted on the adjacent terminal strip) are 50-year-old electrolytics, so upgrade to low ESR/high temp/higher capacity Panasonic FC's.
• H707-H708 diodes are small epoxy gob-type, so upgrade to 1N4448.
• H709 varistor is small epoxy gob-type, so upgrade to (2) x 1N4448 in series.
• H705-H706 are 2SC458 A (known to go noisy), so upgrade to KSC1815Y.

Before/After pics:


After recap...

Step 1: plugged the analyzer into PhonoIn and PreOut, and the performance was reasonable; So far so good.​

Step 2: moved the cables from PreOut over to TapeMonOut and things are quiet, this is a pleasant surprise. Tested THD and SNR at 5mV input, so far so good. Frequency Response at levels from 2.5mV up to 10mV... FR is within spec at 5mV, but starts to show a stability issue as level rises and is clearly becoming unstable at 10mV. So this one does not have the egregious motorboat oscillation problem as soon as the cable is plugged into TapeMonOut, but it's not as stable as it should be. I'm not taking any chances, so Miller caps it shall be.​

Step 3: added a 10pF C0G MLCC across B-C of H705 and H706; Result was zero oscillation in both channels, flat Frequency Response (+0.8/-0.6 dB) from 20Hz-20kHz at 5mV input. Then turned it up to 10mV input... does show just a hint of a stability issue beginning at the 10mV signal level, but no oscillation. This looks good, and is no worse than the FR measured with the original 2SC458's.​

So...
Leaving H701-H704 original and using KSC1815Y's as substitutes at H705-H706 was certainly "acceptable" for this particular unit - it likely would never exhibit the actual motorboating problem. However, it was measurably less stable than original response at higher levels, and placing the Miller cap fix was confirmed to restore the frequency response. So final complement on this board is the original 2SC1000's at four positions (H701-H704), KSC1815Y at two positions (H705-H706) and 10pF Miller caps added across B-C at H705 and H706 (again I used 10pF C0G MLCC's).

"AFTER" test results, for kicks:



My conclusions - your mileage may vary: I will be leaving H701-H704 as original in any future P700 boards I work on, unless there is some indication of a problem. If needed, either KSC1815Y's or KSC1845F's are fine substitutes at H701-H704, but will require the Miller cap fix at H705-H706. At H705-H706 I will always replace 2SC458's with KSC1815Y's as a matter of preventive maintenance, and will include Miller caps placed across B-C (using 10pF C0G MLCC's). Test for motorboating problem by connecting an analyzer or tape deck (or possibly just a set of cables?) to the Tape Monitor Output jacks and selecting Phono on the input selector.

I hope this was useful, at least I had fun and learned a lot. Cheers! -Steve

 

[Dingman]

Steve, this is fantastic work IMO - I am so pleased to be able to follow your diag and the report on each step. Credit to Steven for getting all of this started, and now we have what seems to be a final solution.
Now, I need to review all the marantz phono boards that I've done? Seems like ALL THE 1845/1815 replacements have this issue?

 

[GlasgowGrip]

Steve, this is fantastic work IMO - I am so pleased to be able to follow your diag and the report on each step. Credit to Steven for getting all of this started, and now we have what seems to be a final solution.
Now, I need to review all the marantz phono boards that I've done? Seems like ALL THE 1845/1815 replacements have this issue?

Hey, you're too kind!

So I would generalize two cases (again YMMV):

Case 1: If you put 1845's and/or 1815's at all six positions, I'm ready to assume you will have this problem, at least when connecting through the TapeMon Out jacks. Even if you don't use TapeMon Out, I would still assume a very good chance you will have motorboating with an automatic turntable that shorts its outputs when cycling (I believe that was the original case seen by @Steven Tate and others). More data points would always help, but anecdotally from other threads, plus my testing, that's my feeling.​

​

Case 2: If you leave H701-H704 original and only replace H705/H706 with 1815's, it would seem you are fine, although it quite likely may be case-specific when you go to the TapeMon Out scenario. I'd definitely want more data points to feel better about that conclusion, so I will just be belt-and-suspenders and put the Miller caps on anyway.​

 

[GlasgowGrip]

Now, I need to review all the marantz phono boards that I've done? Seems like ALL THE 1845/1815 replacements have this issue?

​

I hear you about revisiting past Marantz phono boards... The first two 2230's I posted about above were revisits of units I did for myself a couple of years ago. Thinking about it, I did a full restore of a 2270 for a friend last year, but I think at that point I had read enough about motorboating phono stages that I was feeling cautious and all I did was replace H705/H706 with 1815's - whew! - so that one "should" be safe, at least as long as nobody uses the Tape Monitor output.

You know what? I also have a couple of 1060's here - the P900 phono stage is similar, but slightly different so maybe they are okay. But I went full monty on those with 1845's so I will revisit them to see how they test... yikes!

 

[GlasgowGrip]

You know what? I also have a couple of 1060's here - the P900 phono stage is similar, but slightly different so maybe they are okay. But I went full monty on those with 1845's so I will revisit them to see how they test... yikes!​

@Dingman:
I just did some quick testing with the phono stage in a Marantz 1060 that was fully recapped and re-transistored with all KSC1845's. This is the P900 board, very similar to the P700 but a slightly different topology.

The good news is that it's more stable and has a slightly flatter frequency response, adjusting for RIAA, than the P700 in the 2230.

Testing Pre Out:

Step1: Analyzer plugged into PreOut jacks, I was not able to induce any motorboating. I simulated a shorting turntable by shorting each PhonoIn with 600R, and then with dead shorts. Not a hint of oscillation. Great!​

Testing Tape Monitor Out:

Step 1: Moved analyzer to the TapeMonOuts, I was not able to induce any motorboating by simulating a shorting turntable (shorted PhonoIn's with 600R, and then with dead shorts).​

Step 2: Then plugged in analyzer signal generator and things stayed quiet at idle.​

Step 3: Tested THD and SNR at 5mV, tested out fine. (Not quite as quiet as the 2230 but that appears to be 60Hz noise, will need to investigate that later.)​

Step 4: Tested frequency response, from 2.5mV up to 10 mV. Response within specs (-0.8/+0.2 dB) from 20Hz - 20kHz. Two observations: 1. FR all the way up at 10mV, there is the hint of instability starting, the FR is just beginning to get wavy. 2. Setting the input attenution on the analyzer too low will kick the FR into oscillation/motorboating. I need to investigate what is happening here - this may just not be a real world scenario. I will report back on this.​

So... Based on this one unit, my initial impression is that changing out the transistors on the 1060's P900 board using KSC1845F's is not a problem. I will follow up regarding the strange anaylzer interaction - hopefully that is not an important case. That, and I do have a second 1060 to test with, so hopefully I can confirm the good news tomorrow.

 

[rBuckner]

On the transistors I only change the C458s to the 1815’s. Now I will be also adding the 10pF caps! One vote for sticky status!!

 

[Ray Gianelli]

You're welcome Randy - just hoping to contribute a little something back to the community once in a while.

So for today, finishing up, here is my third data point...

This one is a 2230 that is still a work in progress: the switches have been deoxited, pots have been faderlubed, P750's, P400 and P450 have been recapped, main filter cap and output coupling caps have been replaced, so overall performance is solid and the unit is pretty well known to me at this point. It still has a fully stock/unmolested P700 board, so seems to be a good testbed for continuing this post.

For context, the original transistor complement is 2SC1000 BL at H701-H702, 2SC1000 GR at H703-H704, and 2SC458 A at H705-H706.

Baseline testing...

Step 1: plugged the analyzer into PhonoIn and PreOut, and the performance was reasonable; So far so good.​

Step 2: moved the cables from PreOut over to TapeMonOut and things are quiet, as expected. Tested THD and SNR at 5mV input, and Frequency Response at levels from 2.5mV up to 10mV. FR is within spec, although it does show just a hint of a stability issue beginning at the 10mV signal level. No problem here.​

P700 recap...

• C701-C702 and C714-C715 are "ugly brown" films, so leave them alone.
• C705-C706 and C707-C708 are NPO ceramics, so leave them alone.
• C709-C712 are polystyrenes, so leave them alone.
• C703-C704 are tantalums, so upgrade to 47uF Nichicon UES "Muse" bipolars.
• C713 and C011 (mounted on the adjacent terminal strip) are 50-year-old electrolytics, so upgrade to low ESR/high temp/higher capacity Panasonic FC's.
• H707-H708 diodes are small epoxy gob-type, so upgrade to 1N4448.
• H709 varistor is small epoxy gob-type, so upgrade to (2) x 1N4448 in series.
• H705-H706 are 2SC458 A (known to go noisy), so upgrade to KSC1815Y.

Before/After pics:
View attachment 3148187View attachment 3148186

After recap...

Step 1: plugged the analyzer into PhonoIn and PreOut, and the performance was reasonable; So far so good.​

Step 2: moved the cables from PreOut over to TapeMonOut and things are quiet, this is a pleasant surprise. Tested THD and SNR at 5mV input, so far so good. Frequency Response at levels from 2.5mV up to 10mV... FR is within spec at 5mV, but starts to show a stability issue as level rises and is clearly becoming unstable at 10mV. So this one does not have the egregious motorboat oscillation problem as soon as the cable is plugged into TapeMonOut, but it's not as stable as it should be. I'm not taking any chances, so Miller caps it shall be.​

Step 3: added a 10pF C0G MLCC across B-C of H705 and H706; Result was zero oscillation in both channels, flat Frequency Response (+0.8/-0.6 dB) from 20Hz-20kHz at 5mV input. Then turned it up to 10mV input... does show just a hint of a stability issue beginning at the 10mV signal level, but no oscillation. This looks good, and is no worse than the FR measured with the original 2SC458's.​

So...
Leaving H701-H704 original and using KSC1815Y's as substitutes at H705-H706 was certainly "acceptable" for this particular unit - it likely would never exhibit the actual motorboating problem. However, it was measurably less stable than original response at higher levels, and placing the Miller cap fix was confirmed to restore the frequency response. So final complement on this board is the original 2SC1000's at four positions (H701-H704), KSC1815Y at two positions (H705-H706) and 10pF Miller caps added across B-C at H705 and H706 (again I used 10pF C0G MLCC's).

"AFTER" test results, for kicks:
View attachment 3148142View attachment 3148143


My conclusions - your mileage may vary: I will be leaving H701-H704 as original in any future P700 boards I work on, unless there is some indication of a problem. If needed, either KSC1815Y's or KSC1845F's are fine substitutes at H701-H704, but will require the Miller cap fix at H705-H706. At H705-H706 I will always replace 2SC458's with KSC1815Y's as a matter of preventive maintenance, and will include Miller caps placed across B-C (using 10pF C0G MLCC's). Test for motorboating problem by connecting an analyzer or tape deck (or possibly just a set of cables?) to the Tape Monitor Output jacks and selecting Phono on the input selector.

I hope this was useful, at least I had fun and learned a lot. Cheers! -Steve

I was having motorboating issues on a 2230 after restoration as well. Used KSC1815s. Was pulling my hair out until I found a post from Catrafter. Tom had said to look at the 2325 phono board. It's almost the same as the 2230, but they added 47Ω resistors to isolate the outputs. That's what worked for me.



Tom, I know you no longer contribute here, but you are missed!

 

[rBuckner]

I was having motorboating issues on a 2230 after restoration as well. Used KSC1815s. Was pulling my hair out until I found a post from Catrafter. Tom had said to look at the 2325 phono board. It's almost the same as the 2230, but they added 47Ω resistors to isolate the outputs. That's what worked for me.

View attachment 3151407

Tom, I know you no longer contribute here, but you are missed!

Glad you thought of this Ray! It’s another automatic mod I do if they aren’t there.