Technics RS-B965

[A.N.T.]

I've got that top of the range deck from a friend, for a check-up and a required service. After I've done that and the deck was fully operational, I was very impressed with the build quality and the mechanism, but extremely disappoined with the sound quality, which I've found tiring, flat and "digital" in a worst sense. After I've done some measurements and looked into the schematics it became quite clear why the sound is so horrible. Two things - high order distortion in the recording channel that spoiled the sound even before it was recorded on the tape, and 5 (five!) 4066 chips in the signal path, some in the PB chain and some in the recording chain. Recently I've modified the RS-B965, removing 6 muting transistors from the recording input circuitry (that took care of the high distortion) and replaced all 4066 chips with my hybrid ANT4066. As a result the deck now is sounding better than my Nak BX-300 and Cassette Deck 1 and quite close in quality to my reference Aiwa XK-S7000.

On the pictures are:

1) Distortion spectra - before and after the mod, at 3 kHz Dolby level (0dB) and -10 dB, from the CD Direct input to the Line output in the monitor "Source" mode.

2) Part of the schematics showing the transistors to remove in red. I've replaced all 6 transistors with resistors between collector and emitter points - Q9 and Q10 with 150K resistors, Q11 and Q12 with 3.9K, Q13 and Q14 - with links.

Alex

[Dimitar Georgiev]

On the pictures are:

1) Distortion spectra - before and after the mod, at 3 kHz Dolby level (0dB) and -10 dB, from the CD Direct input to the Line output in the monitor "Source" mode.

2) Part of the schematics showing the transistors to remove in red. I've replaced all 6 transistors with resistors between collector and emitter points - Q9 and Q10 with 150K resistors, Q11 and Q12 with 3.9K, Q13 and Q14 - with links.

Alex

Very interesting Alex! I have seen those high order peaks with my primitive Sound card and Virtins Audio analyzer software when I send some sinusoid in the KHz range. I thought that the higher order harmoncis were either due to inapropriate input impedance or problems due to the cheap sound card. I did not imagine they were real...

D.

[Scorpion8]

Would it have been better to replace Q9 - Q12 with a better transistor? There should be a more suitable one out there. The original circuit designer put them there for a reason, no?

[A.N.T.]

Would it have been better to replace Q9 - Q12 with a better transistor? There should be a more suitable one out there. The original circuit designer put them there for a reason, no?

The original purpose of these transistors is to mute one (unused) input out of two, and to change the sensitivity of the preamp section. However both are unnecessary and, as you may see, seriously damage the sound quality. A replacement by different transistors types would not make things better, probably only worse. The only proper way to keep the muting feature would be to use relays, however it is very complicated. Most of other decks with two switchable inputs do not have this muting circuitry and perform just fine.

Alex

[A.N.T.]

Very interesting Alex! I have seen those high order peaks with my primitive Sound card and Virtins Audio analyzer software when I send some sinusoid in the KHz range. I thought that the higher order harmoncis were either due to inapropriate input impedance or problems due to the cheap sound card. I did not imagine they were real...

D.

It really depends on the card - you should try to feed the card output directly to the card input ("loop-back") to see what the card is capable of. I use a very good quality RME Digi 96/8 PAD soundcard and after some modifications ;) it has distortion levels for the loop-back close to 0.0003%

Alex

[A.N.T.]

I've uploaded 3 files, recorded from RS-B965 playing back the same tape but using different chips for switching:

Original MN4066B (http://rapidshare.com/files/294016050/RSB965_MN4066B.zip),

Fairchild MM74HC4066N (http://rapidshare.com/files/294016910/RSB965_MM74HC4066N.zip)

ANT4066A (http://rapidshare.com/files/294017828/RSB965_ANT4066A.zip)

All files are zipped WAV 96 kHz 24 bit, about 42 Mb each.

Alex

[Marc Hugo]

I've got that top of the range deck from a friend, for a check-up and a required service. After I've done that and the deck was fully operational, I was very impressed with the build quality and the mechanism, but extremely disappoined with the sound quality, which I've found tiring, flat and "digital" in a worst sense. After I've done some measurements and looked into the schematics it became quite clear why the sound is so horrible. Two things - high order distortion in the recording channel that spoiled the sound even before it was recorded on the tape, and 5 (five!) 4066 chips in the signal path, some in the PB chain and some in the recording chain. Recently I've modified the RS-B965, removing 6 muting transistors from the recording input circuitry (that took care of the high distortion) and replaced all 4066 chips with my hybrid ANT4066. As a result the deck now is sounding better than my Nak BX-300 and Cassette Deck 1 and quite close in quality to my reference Aiwa XK-S7000.

On the pictures are:

1) Distortion spectra - before and after the mod, at 3 kHz Dolby level (0dB) and -10 dB, from the CD Direct input to the Line output in the monitor "Source" mode.

2) Part of the schematics showing the transistors to remove in red. I've replaced all 6 transistors with resistors between collector and emitter points - Q9 and Q10 with 150K resistors, Q11 and Q12 with 3.9K, Q13 and Q14 - with links.

Alex

A thoroughly enjoyable read - well done on the tweaks. I heard an earlier version of that deck (the RS-B905) and was not very impressed either. That was in about 1990 or so. I always liked the 808/828 a lot more. It does have an excellent transport though. And HX-Pro from what I recall.

Thanks for that!!

MH

[SaSi_Sidi]

Well, I have an RS965 and did no modifications. It seems to perform great and considering the inherent distortion of the tape itself, it is perhaps not much of an issue.

What I really like on this deck is the dual capstan transport. Rock solid and steady, IMHO rivals the Revox.

[A.N.T.]

Well, I have an RS965 and did no modifications. It seems to perform great and considering the inherent distortion of the tape itself, it is perhaps not much of an issue.

What I really like on this deck is the dual capstan transport. Rock solid and steady, IMHO rivals the Revox.

The transport is very good indeed on this deck, however in a non-modified form it is (IMHO) useless for recording from a good quality analogue source and even for playback duties it is crippled and will not do justice to a good quality analogue recording. If you use this deck to tape from CDs than the sound is already dead enough and you will not hear the difference, so it does not matter.

Alex

P.S. before the mod RS-B965 is an average deck sound-wise. After the mod it is up there with the very best.

[Marc Hugo]

Hi ANT - the recording was indeed from CD - it was the Naxos "Slavonic Dances" to a 1991 Maxell XLII-S. I thought to myself at the time that my 1983 Alpage AL-50 (which was all I had at the time) would have sounded much more animated and detailed. But one must be polite. He just paid R3200 for it (then about +-UKS400).

I think this guy still has this deck so I should make him an offer. Naturally, you're in London which makes things tricky!!

Cheers - Marc

[SaSi_Sidi]

The transport is very good indeed on this deck, however in a non-modified form it is (IMHO) useless for recording from a good quality analogue source and even for playback duties it is crippled and will not do justice to a good quality analogue recording. If you use this deck to tape from CDs than the sound is already dead enough and you will not hear the difference, so it does not matter.

Alex

P.S. before the mod RS-B965 is an average deck sound-wise. After the mod it is up there with the very best.
Now that I think about it, I've not (yet) used it for recording. Just a couple of casual tests that blew me off with dbx and noise disappearing. I also recall that with dbx enabled, the dynamic range seems to expand hugely.

I will connect it up (it currently serves at the little used living room system) to the main system and try some recordings from the TT with an Ortofon MC25FL.

[scan80269]

The transport is very good indeed on this deck, however in a non-modified form it is (IMHO) useless for recording from a good quality analogue source and even for playback duties it is crippled and will not do justice to a good quality analogue recording. If you use this deck to tape from CDs than the sound is already dead enough and you will not hear the difference, so it does not matter.

Alex

P.S. before the mod RS-B965 is an average deck sound-wise. After the mod it is up there with the very best.

Hi Alex, I just picked up an RS-B965 deck in good working order, and considering to perform your mods for improved sound quality.

Which components do you believe are most detrimental to the sound quality of this deck, the MN4066B switches or the muting transistors?

I do plan on putting IC sockets at the 4066 switch locations, but also trying to decide if the muting transistors are worth the effort.

BTW, the three ANT4066 devices are working great in my Sony TC-K707ES, which also recently had belts and capstan motor replacements.

[A.N.T.]

Hi Alex, I just picked up an RS-B965 deck in good working order, and considering to perform your mods for improved sound quality.

Which components do you believe are most detrimental to the sound quality of this deck, the MN4066B switches or the muting transistors?

I do plan on putting IC sockets at the 4066 switch locations, but also trying to decide if the muting transistors are worth the effort.

BTW, the three ANT4066 devices are working great in my Sony TC-K707ES, which also recently had belts and capstan motor replacements.

Hi Sam,

It is potentially a very nice deck, well done! The muting transistors on the inputs are (obviously) do screw up the recording side, however original 4066 switches are bad both for recording and playback sound quality. If you plan to record on your RS-B965, then it would be wise to remove muting and gain switching transistors and replace these as described in the first post - it is an easy mod and makes a lot of difference in the recording quality.

I am glad that you like ANT4066 in your Sony, thank you for the feedback.

Regards

Alex

[scan80269]

Hi Alex, I have completed the mods to my RS-B965: sockets at the five MN4066B switch locations, and "3-pin sockets" at the locations of the four muting transistors and the two gain adjusting transistors, to allow these to be swapped in and out for comparison purposes.

The MN4066B ICs have been replaced with Fairchild MM74HC4066N, but there is a strange problem. With IC9 being MM74HC4066N, I get loud pops on my headphones when toggling between SOURCE and TAPE. There is also loud hum on the right channel with the SOURCE setting. If IC9 is changed back to MN4066B the pop noise is reduced to a faint click, the hum disappears and everything is back to normal. Two different MM74HC4066N ICs have yielded the same results.

At the moment, I have MM74HC4066N at IC7, IC8, IC11 & IC12, but I'm staying with MN4066B for IC9 until I get a better understanding of what is happening. The six transistors have been removed as described in your first post, but I used 200Kohm resistors for Q9 and Q10 since I don't have any 150Kohm resistors on hand. I don't have the service manual, which is the next item to get.

I may ultimately decide to switch (no pun intended) to ANT4066, but with the RS-B965 needing five chips this will be quite a hefty sum.

[A.N.T.]

Hi Alex, I have completed the mods to my RS-B965: sockets at the five MN4066B switch locations, and "3-pin sockets" at the locations of the four muting transistors and the two gain adjusting transistors, to allow these to be swapped in and out for comparison purposes.

The MN4066B ICs have been replaced with Fairchild MM74HC4066N, but there is a strange problem. With IC9 being MM74HC4066N, I get loud pops on my headphones when toggling between SOURCE and TAPE. There is also loud hum on the right channel with the SOURCE setting. If IC9 is changed back to MN4066B the pop noise is reduced to a faint click, the hum disappears and everything is back to normal. Two different MM74HC4066N ICs have yielded the same results.

At the moment, I have MM74HC4066N at IC7, IC8, IC11 & IC12, but I'm staying with MN4066B for IC9 until I get a better understanding of what is happening. The six transistors have been removed as described in your first post, but I used 200Kohm resistors for Q9 and Q10 since I don't have any 150Kohm resistors on hand. I don't have the service manual, which is the next item to get.

I may ultimately decide to switch (no pun intended) to ANT4066, but with the RS-B965 needing five chips this will be quite a hefty sum.

Hi,

First of all - here is the service manual for RS-B965 (http://www.ant-audio.co.uk/Tape_Recording/Service/TECHNICS_RSB965.pdf). Use "save as" option from your browser. As you may see from the circuit, the voltages on 4066 in RS-B965 are a bit high even for MM74HC4066N, and IC9 has some non-standard driving voltages - that will explain the problem you have with it. Drop me a PM re. ANT4066 - I can offer you a deal on 5 of these.

Alex

[A.N.T.]

Now I have two of these in my own collection - one in a pretty good condition overall, the second is severely scratched on top but otherwise looks OK. Will modify them in due course and use as recording decks - after the mod the recording quality on these is very good. More to follow.

Alex

[scan80269]

Alex, thanks for all your help with my RS-B965. It has benefited tremendously with the ANT4066 switches and the muting transistor removal.

My latest acquisition is not a cassette deck (!), but rather an RME Fireface 400, which I plan to use as a high-quality headphone amp, as well as ADC & DAC. Since I don't have any decent analog source (other than perhaps my Sony TC-KA3ES decks), I am going to explore the possibility of using the DAC in the Fireface 400, along with high-resolution digital content such as 24/96 recordings (HDtracks.com, etc.) Listening to good CD recordings with a pair of headphones plugged into the front of the FF400 has yielded a level of sonic clarity I have never experienced. I used an old Sony DVP-NS999ES DVD/SACD player to play CDs over Toslink S/PDIF, and the FF400 cleans up the clock embedded in the S/PDIF stream, and does a clean DAC to drive the headphones output. I also have an OPPO BDP-83 Blu-Ray player that can play DVD-A & SACD discs to use as a high-res digital audio source. I intend to feed the FF400 analog outputs to my decks, as soon as I find some TS to RCA phono adapters.

The FF400 is also useful with AudioTester. I think I'll keep busy using the FF400 as a high-quality external sound card to do measurements on my various decks.

BTW, here's a discovery that an audiophile colleague and I made recently: A good PC can serve as an audiophile quality digital sound source! Yes, I know PCs have never come close to being considered hi-fi in sound quality, but here are a few tricks to get the best audio out of a modern PC:

* Use Toslink optical S/PDIF output (can be from sound card or motherboard)
* Use Windows 7 OS
Win7 supports the latest implementation of Microsoft's WASAPI (Windows Audio Session API), with ability to bypass the kernel mixer to deliver low-latency audio streaming. Without WASAPI, the PC sound is arguably NOT audiophile quality
* Use an audio player with WASAPI support, e.g. Foobar2000 with WASAPI plug-in
* Uninstall non-essential HW and their device drivers to help cut down on DPC latency spikes, which can cause drop-outs in real-time audio/video streaming
* Optional: use clock-jitter reducing devices such as Apogee Big Ben to clean up clock jitter embedded in S/PDIF
* Optional: use high-quality standalone DAC (Benchmark 1, etc.)
* Use high-resolution audio content, e.g. 24/96 recordings

The FF400 is really versatile, and can be easily moved from PC to PC. I'm probably going to build a small PC dedicated to being an audio source to use with the FF400 and my decks.

Cheers!

[A.N.T.]

Alex, thanks for all your help with my RS-B965. It has benefited tremendously with the ANT4066 switches and the muting transistor removal.

My latest acquisition is not a cassette deck (!), but rather an RME Fireface 400

Sam, thank you for a good feedback. I think that modified RS-B965 is a great deck. Re. RME - I like their products, actually I am using 3 RME Digi 96/8 PAD/PST cards in my computers. After some mods these work very well in hi-res audio (24/96), the sound quality is comparable to a good tape deck. However it is a complete off-topic here!

Cheers

Alex

[scan80269]

My RS-B965 started giving me fits the other day. It chewed up a TDK D and a Sony UX-Pro. Since I was watching the deck closely while it played the UX-Pro, I was able to stop the chewing before it got too bad.

I had to use my Sony mirror cassette to help with adjusting the lateral position of the supply-side roller/tape guide assembly. After fiddling with this a bit, the deck has managed to play through a few different tapes without chewing them, and even the previously chewed tapes can now play through, after a bit of help to flatten the wrinkled areas. This deck apprarently suffered from the same ailment as my Sony dual-capstan decks: shifted supply side roller/guide assembly causing the tape to slip off the guide.

This RS-B965 also sometimes fails to open the cassette compartment on eject. When I first got the deck, there was a thin spring loose at the left side of the cassette compartment and interfering with compartment motion. I couldn't figure out how to latch it back in place so I removed it. Now, the left-side hub of some cassettes can get a little stuck during eject and the compartment would not swing open. If the spring is supposed to provide additional eject force to help the compartment swing open, then the next time I work on the RS-B965 I'll look into putting the spring back.

This unit is also in need of major azimuth alignment work.

[A.N.T.]

My RS-B965 started giving me fits the other day. It chewed up a TDK D and a Sony UX-Pro. Since I was watching the deck closely while it played the UX-Pro, I was able to stop the chewing before it got too bad.

I had to use my Sony mirror cassette to help with adjusting the lateral position of the supply-side roller/tape guide assembly. After fiddling with this a bit, the deck has managed to play through a few different tapes without chewing them, and even the previously chewed tapes can now play through, after a bit of help to flatten the wrinkled areas. This deck apprarently suffered from the same ailment as my Sony dual-capstan decks: shifted supply side roller/guide assembly causing the tape to slip off the guide.

This RS-B965 also sometimes fails to open the cassette compartment on eject. When I first got the deck, there was a thin spring loose at the left side of the cassette compartment and interfering with compartment motion. I couldn't figure out how to latch it back in place so I removed it. Now, the left-side hub of some cassettes can get a little stuck during eject and the compartment would not swing open. If the spring is supposed to provide additional eject force to help the compartment swing open, then the next time I work on the RS-B965 I'll look into putting the spring back.

This unit is also in need of major azimuth alignment work.

Yes, that spring is a nuisance - I had to remove it on my RS-B965 as well and now cassettes (mostly Metal) sometimes stuck. I think about adding a spring somewhere to fix it.

Re. chewing tapes - if you have a tension measuring cassette - use it and see if the back tension is correct (about 8-10g on my deck) - sometimes left reel becomes loose and tension goes down - that in itself is enough in a dual capstan configuration to mishandle the tape. The position of the tape guide on the left may affect few things, including REC/PB phase response (which looks like an azimuth misalignment between REC and PB heads), but it should not affect the tape travel unless it is way off. Something else is wrong and I would not consider fiddling with the tape guide as a fix.

Alex

[A.N.T.]

I've spent all day today servicing, modifying and aligning my own RS-B965. It came from Germany in a pretty good condition some while ago however I didn't have time to do anything with it. Now I did all the mods and also found that the last bit that spoils the sound is the output "Class AA" amplifier stage. When I took it out the impression was like a last veil was removed from the soundstage, everything snapped into focus. Now I have to do something about boosting the output (the amp had a gain about 3 and I took it out (IC5 and IC6 near output sockets) and replaced each with a wire link between pins 5 and 7. Which means that I have 1/3 of the output voltage and level meters do not work properly *sigh*. But the sound is just great *thumbsup*.

Alex

[scan80269]

RS-B965 appears to make liberal use of "Class AA" amps. In addition to IC5 & IC6 (M5218L) driving the line outputs, the headphone outputs are driven by IC701 & IC702 (another pair of M5218L) also in "Class AA" configuration connected to line-out. Does this mean the headphone outputs should sound even worse than the line-outs?

There is also an IC10 (M5218L) in the voltage-controlled amp section of the recording-equalizer amp, though the current drive stage is built with discrete transistors.

The RS-B965 operating manual has a section dedicated to "Linear Magne-Field Class AA", talking about the design of the recording-equalizer amp. I suppose this is largely marketing hype, and such designs do not necessarily translate to better sound.

Regarding the tape chewing, I have a mirror cassette but not a torque cassette. I need to be on the lookout for one of the latter. My RS-B965 has stopped chewing tapes, but I'm keeping a watchful eye on it.

[Dave Cawley]

I am the friend Alex originally talked about. It's a great deck and I'm waiting for the final mod from Alex!

Regards

Dave

[A.N.T.]

RS-B965 appears to make liberal use of "Class AA" amps. In addition to IC5 & IC6 (M5218L) driving the line outputs, the headphone outputs are driven by IC701 & IC702 (another pair of M5218L) also in "Class AA" configuration connected to line-out. Does this mean the headphone outputs should sound even worse than the line-outs?

Yes, the headphone output does sound worse than the line output. What I would like to try is just to cut resistors to the second amp (in case of the line output I'll cut R31, R32, R37 and R38, leaving the second amp as just a follower with the output disconnected). Perhaps even for the headphone amp just cutting R721/R722 and R725/R726 may do the same trick.

There is also an IC10 (M5218L) in the voltage-controlled amp section of the recording-equalizer amp, though the current drive stage is built with discrete transistors.

The RS-B965 operating manual has a section dedicated to "Linear Magne-Field Class AA", talking about the design of the recording-equalizer amp. I suppose this is largely marketing hype, and such designs do not necessarily translate to better sound.

At the moment I am considering changing all remaining opamps in the signal path - IC1, 2, 3 and 10. For a start I will just install sockets for all of them.

"Linear Magne-Field" is a separate issue from "Class AA" and is a good thing for a change - it is a current output stage for the recording head. It does reduce the distortion, especially at high levels with CrO2 and Metal tapes. On the SONY UX-S 3% 1kHz THD reached at +6.5 dB from DIN level (i.e. +10.5 dB on VU scale ! ) - as good as most Metal tapes. And, obviously, Technics did not invent this - same technique was used as early as in Nakamichi 1000 and Sony TC-229SD.

Alex

[A.N.T.]

An update - the input opamp (IC3) affects the sound noticeably after the output opamps are replaced. I've changed it for LM4562 and it is way better. I now will try to change the rest of opamps as well.

Also I've found at last a suitable new belt for it and that improved the W&F from 0.12% to 0.06% DIN wtd on my unit.

Alex

[Scorpion8]

Alex, you should consider creating and posting a list off all the mods that you made to this deck so they are captured in a single document. That'd be great.

[A.N.T.]

Alex, you should consider creating and posting a list off all the mods that you made to this deck so they are captured in a single document. That'd be great.

Yes, I will do that eventually. RS-B965 is an amazing deck - it actually has a complete bypass of the Dolby chips both on PB and on REC. Now, after I've replaced the opamps it is just an awesome deck - even with the stock heads. Japanese version of RS-B965 was equipped with amorphous heads and a superbias. All the component positions are there and that will be the next thing to try - to double the bias frequency. It could be the best deck in my collection!

Alex

[scan80269]

Yes, I will do that eventually. RS-B965 is an amazing deck - it actually has a complete bypass of the Dolby chips both on PB and on REC. Now, after I've replaced the opamps it is just an awesome deck - even with the stock heads. Japanese version of RS-B965 was equipped with amorphous heads and a superbias. All the component positions are there and that will be the next thing to try - to double the bias frequency. It could be the best deck in my collection!

Alex

Alex, I was going to comment to Scorp that you may not be quite done with your RS-B965 mods...

It looks like you may end up replacing every analog switch and op-amp in the record and playback signal paths of your deck, in pursuit of improved sound! Having sockets at the op-amp locations really make for easy A/B comparisons. I can only imagine how good sounding your RS-B965 is becoming and would be thrilled to hear your final recommendations on replacement op-amps.

I'm also trying to understand why Technics chose to implement "class AA" amps for line and headphone outputs. The second half of IC5 & IC6 (also IC701 & IC702 for headphone outputs) are puzzling to me. Are they used to alter the output impedance of the outputs? Whatever their purpose, it is unfortunate that this circuit topology degrades the sound as you have indicated.

One feature that I sometimes wish the RS-B965 had is a Dolby HX-Pro enable/disable switch, similar to the one on my Sony TC-KA3ES. I did some quick measurements on the KA3ES with HX-Pro on and off, and found that recording with the On setting yielded a bit higher distortion (3rd harmonic), though perhaps not in a very audible way. I don't know how easy (or hard) it is to add an HX-Pro on/off switch to the RS-B965. Perhaps a slider switch can be added somewhere at the deck's rear panel, similar to how some decks have an MPX filter on/off switch at the back. IC11 & IC12 each appear to have one of four analog switches unused within their packages, but I'm not sure if it is feasible to exploit these analog switches to implement HX-Pro on/off control via a user-accessible slider. Anyway, this is just a wild idea.

With two or more decks on their way to me, more mods queuing up for the RS-B965 and TC-K707ES, and the author of AudioTester asking me to help evaluate the upcoming wow & flutter measuring capabilities, I guess I'll keep busy during my spare time...

I, too, have to say that you never cease to amaze me, Alex!

[A.N.T.]

Just an update - after listening to few different opamps on the output of RS-B965, I've chosen NE5534. They are marginally better than LM7171 and noticeably better than LME49710. In any case the original opamps produce quite a poor sound IMHO. I've uploaded a 40 Mb ZIP file (http://rapidshare.com/files/364823570/RSB965_output_opamps_compared.zip) to RapidShare - if you can listen to 96 kHz 24 bit audio you can download the file and compare 2 40-second samples of the same fragment from an old pre-recorded cassette - with the original opamps in place and with NE5534.

Alex

[scan80269]

Alex, have you considered the Burr Brown OPA627 "Difet" op-amp? I found a place that carries an "OPA2627 SuperChip" which is a small PCB containing a pair of OPA627 chips, and terminates in a DIP8 pinout compatible with standard dual-channel op-amps:

http://www.audiophileproducts.com/opamps

They also have an op-amp adapter "OPA104" to map single-channel op-amp pinout back to dual-channel op-amp pinout.

Would some Sony decks benefit from op-amp changes like with the RS-B965?

[A.N.T.]

Alex, have you considered the Burr Brown OPA627 "Difet" op-amp? I found a place that carries an "OPA2627 SuperChip" which is a small PCB containing a pair of OPA627 chips, and terminates in a DIP8 pinout compatible with standard dual-channel op-amps:

http://www.audiophileproducts.com/opamps

They also have an op-amp adapter "OPA104" to map single-channel op-amp pinout back to dual-channel op-amp pinout.

Would some Sony decks benefit from op-amp changes like with the RS-B965?

I am not very fond of the OPA627 , especially of their price and especially when NE5534 already does a very good job *thumbsup*. All opamps in RS-B965 are in SIL8 packages and I had to make my own adaptors to use DIL8 there. I'll post some photos later.

On Sony decks - some top decks are using good opamps already (TC-K970ES and TC-K870ES use NE5532 in the PB amp, for example). In other cases the same NE5532 can be a very good replacement for usual Japanese 4570, 4558 etc etc. There are also meter rectifier opamps in many Sony decks that could benefit from replacement but that matter is discussed in another thread (http://www.tapeheads.net/showthread.php?t=7585).

Alex

[A.N.T.]

Today I had a good listening session with my newly modified and aligned RS-B965. I have to say that after these last few modifications it is easily the best playback deck I've ever heard, period. It has much better sound than my modified Aiwa XK-S7000 (my usual reference - not many decks have the same level of playback performance as that Aiwa) . I am listening again to many of my pre-recorded cassettes. I have a new "gold standard" by which all other decks will be evaluated from now on *thumbsup*

*reelspin**reelspin**reelspin*

Alex

[Dave Cawley]

And I have to say, it was me, the owner of a TC-K81 from new, that found this deck and persuaded Alex to modify it! Technics made the seminal SP-10 turntable and now the best cassette recorder?

Cheers Alex!!

Dave

[A.N.T.]

And I have to say, it was me, the owner of a TC-K81 from new, that found this deck and persuaded Alex to modify it! Technics made the seminal SP-10 turntable and now the best cassette recorder?

Cheers Alex!!

Dave

Hi Dave,

Yes, I have to thank you for bringing this deck to my attention and for your persuasion - as I was really appalled by the horrible sound that deck produces in its unmodified form and was not sure at all that it was worth the effort.

I have to add that it is the best playback deck I've heard so far. I am not sure about the "recorder" bit. It can be very good at it too but I did not evaluate it properly for the recording quality yet.

Cheers!

Alex

[perry]

What about using those NE5532s in place of the 4559s in the B215? They are everywhere.

[A.N.T.]

What about using those NE5532s in place of the 4559s in the B215? They are everywhere.

Not all of them in the signal path. IC1, IC3, IC9, IC11 on the Input/Output board, IC4, IC9, IC10, IC12, IC15, IC16, IC19 on the Record Ctrl board can certainly be replaced with NE5532 from TI. It is a much better opamp than 4559.

Cheers

Alex

[perry]

CRAP! 11 op-amps and sockets! That's pretty much everywhere to me! What the hell, "in for pint, in for a pound" or something like that; I've replaced everything else in the B215. If it makes as much a difference as you said (and I have no reason to doubt you, as your other recommendations have been excellent), the B215 is probably the easiest to do this on, eh?

[A.N.T.]

Did some work on my second RS-B965 yesterday evening. This one was in a much more battered condition and required a full casework disassembly for a good cleaning operation. I did that and also serviced the mechanics, including a new capstan belt. The performance is in the same ballpark as the first unit - 0.06% DIN wtd W&F. The main culprit, as before, is the capstan belt. Old belt thought not weakened and still working, has lost its flexibility. As a result the W&F before the belt replacement is twice as high, with the unweighed DIN figure reaching 0.25% (only 0.12% with a new belt).

Meanwhile I am enjoying the performance of the first RS-B965M where "M" stands for "modified". I've recorded a couple of Maxell XL-IIS tapes on it - from a good old vinyl record of Maria Callas, and it produced an excellent sound, very close to the original, with just a bit of added background noise (as I am not using any NR).

Alex

[A.N.T.]

I've uploaded a sample from an old Maria Callas vinyl (http://rapidshare.com/files/368549183/Callas_Fragment1.zip)- recorded and then played back on the Technics RS-B965M. 38 Mb, 96 kHz 24 bit wav file, zipped.

Cheers

Alex

[scan80269]

Yes, I will do that eventually. RS-B965 is an amazing deck - it actually has a complete bypass of the Dolby chips both on PB and on REC. Now, after I've replaced the opamps it is just an awesome deck - even with the stock heads. Japanese version of RS-B965 was equipped with amorphous heads and a superbias. All the component positions are there and that will be the next thing to try - to double the bias frequency. It could be the best deck in my collection!

Alex

Alex, it sounds like your RS-B965 decks have permalloy heads and use 80KHz bias frequency? My RS-B965 has fixed 120V 60Hz AC input, and the serial number sticker in the back says "RS-B965PP-K" above the serial number. The Rec/Pb head enclosure has a label that says "rec/play AX". Do you think my RS-B965 already features amorphous Rec/Pb heads and 160KHz bias frequency?

By the way, my RS-B965 shows a slight speed difference between playing a tape at the beginning vs. the end. With a 3KHz W&F test tape, AudioTester gives a reading of 3009Hz near the beginning of side A, and 3001Hz near the end of side B. This is a bit surprising since RS-B965 has a quartz-locked DD capstan motor. I suspect the deck may be sensitive to take-up tension differences between beginning and end of tape.

In contrast, my Sony TC-KA3ES plays the W&F test tape rock steady, with a 3001Hz reading from beginning to end on both sides. My recently acquired Pioneer CT-93 is also steady, but with a reading of 3027Hz, the pitch is high by almost 1%. This deck's azimuth is also in major disagreement with my two TC-KA3ES, and the FL meters read about 4dB low, so it needs a bit of work.

You probably know this already, but I find AudioTester extremely helpful for adjusting deck play speed, in addition to the other useful functions like frequency spectrum plots, oscilloscope, frequency generator, etc. I hope Urich can get the W&F function working soon, then I'll be ecstatic...

Cheers!
Sam

[A.N.T.]

Hi Sam,

Alex, it sounds like your RS-B965 decks have permalloy heads and use 80KHz bias frequency? My RS-B965 has fixed 120V 60Hz AC input, and the serial number sticker in the back says "RS-B965PP-K" above the serial number. The Rec/Pb head enclosure has a label that says "rec/play AX". Do you think my RS-B965 already features amorphous Rec/Pb heads and 160KHz bias frequency?

Yes, PP version has AX (amorphous) head and 160 kHz bias. Lucky you! I only have MX heads and 80 kHz. Nevertheless it records very nicely at the moment.

By the way, my RS-B965 shows a slight speed difference between playing a tape at the beginning vs. the end. With a 3KHz W&F test tape, AudioTester gives a reading of 3009Hz near the beginning of side A, and 3001Hz near the end of side B. This is a bit surprising since RS-B965 has a quartz-locked DD capstan motor. I suspect the deck may be sensitive to take-up tension differences between beginning and end of tape.

You may need a new belt for your RS-B965. I also found that it is useful to lift the cassette by about 0.3mm up. I did it by putting a suitable thickness heat-shrink tubing on both cassette support bits. It reduces W&F at the and of the tape - on some tapes quite noticeably.

Cheers

Alex

[scan80269]

You may need a new belt for your RS-B965. I also found that it is useful to lift the cassette by about 0.3mm up. I did it by putting a suitable thickness heat-shrink tubing on both cassette support bits. It reduces W&F at the and of the tape - on some tapes quite noticeably.


Thanks for the recommendations, Alex! Save me a couple of those RS-B965 belts you have. I'm preparing to order these, along with the A.N.T. alignment tapes and a few more ANT4066A.

Cheers!
Sam

[scan80269]

Yes, PP version has AX (amorphous) head and 160 kHz bias. Lucky you! I only have MX heads and 80 kHz. Nevertheless it records very nicely at the moment.


Alex, the AX head for RS-B965 (RBR4CY004-C) may still be available. I found it at two places in the US (PartStore.com, AndrewsElectronics.com) with $150-170US pricing, though I can't tell if either store has it in stock.

If you change your RS-B965 to AX head and 160KHz bias, comparing it with the MX head & 80KHz bias would be an interesting exercise!

Cheers!
Sam

p.s. I like your new avatar!

[A.N.T.]

Alex, the AX head for RS-B965 (RBR4CY004-C) may still be available. I found it at two places in the US (PartStore.com, AndrewsElectronics.com) with $150-170US pricing, though I can't tell if either store has it in stock.

If you change your RS-B965 to AX head and 160KHz bias, comparing it with the MX head & 80KHz bias would be an interesting exercise!

Cheers!
Sam

p.s. I like your new avatar!

Thank you, Sam! The picture is taken from RS-B965M :) .

I have two RS-B965 at my disposal and I hope to get another one. I also can borrow an AX head from my RS-M253X, or put a new one from Tascam, which I have. In any case, I may experiment with one of two decks after I will make the second one up to the same standard as the first!

I attach some pictures, as promised earlier.

Cheers

Alex

[scan80269]

Thanks for the pictures, Alex!

May I ask about the type and value of the white caps soldered between V+ and V- pins under the IC sockets?

Cheers!
Sam

[A.N.T.]

Thanks for the pictures, Alex!

May I ask about the type and value of the white caps soldered between V+ and V- pins under the IC sockets?

Cheers!
Sam

Hi Sam,

Caps are 0.1uF 63V film (polyester).

Cheers

Alex

[Dave Cawley]

Mine is a RS-B965EG-K, what are the variants?

Thanks

Dave

[A.N.T.]

Mine is a RS-B965EG-K, what are the variants?

Thanks

Dave

Hi Dave,

That is from the service manual. It is not a full list, thought, as it does not cover other colours and some areas.

Cheers

Alex

[Dave Cawley]

So how do you tell which head?

Dave

[A.N.T.]

So how do you tell which head?

Dave

Hi Dave,

The only one RS-B965 I know of with an amorphous alloy head is "PP" version, which is not even mentioned in the list I've posted. However it is mentioned on the schematics and in the parts list. The rest have standard (and very good) hard permalloy "MX" heads.

Cheers

Alex

[scan80269]

Hi Alex,
Do you have a final list of RS-B965 mods?

I take it that you did at least the following:
- Replace MN4066B switches (IC7,8,9,11,12) with ANT4066A
- Replace line-in M5238L op-amp (IC3) with LM4562
- Replace line-out M5218L op-amps (IC5,6) with NE5534

Did you also modify the following:
- PB head M5219L op-amps (IC1,2)
- Rec head M5218L op-amp (IC10)
- headphone M5218L op-amps (IC701,702)

Your RS-B965 mods have made me curious in whether other high-end decks, e.g. Aiwa XK-S7000/9000, Sony TC-KA3ES, Nakamichi Dragon, Pioneer CT-93, etc. can also benefit from having their 20+ year-old Japanese op-amps replaced with modern western brands, such as Burr-Brown/TI, Analog Devices, National Semiconductor, etc.

I am probably going to put sockets into a number of op-amps in my better decks to facilitate op-amp swapping for experimentation and auditioning. Any recommendations for good op-amps to use for the above deck functions (PB head amp, line-out amp, headphone amp, line-in amp, Rec head amp, etc.) would be much appreciated.

[scan80269]

I finally managed to add IC sockets for a number of op-amps in my RS-B965. Along with some DIP8-to-SIL8 adapters I bought from Cimarron Technology, Inc. it is now possible to experiment with different op-amps to see how the sound is affected.

The only disappointment would be the two headphone op-amps on the headphone board, which is in a tight space against the left edge of the deck. It doesn't look like there is room to allow the SIL8 op-amps to be replaced with DIP8 packages. *fit*

Now, it's time to go on an op-amp shopping spree... *Hi5*

[scan80269]

Alex, I just completed some mods for my pair of RS-B965 decks, in preparation for op-amp changes and line-out/headphone out amp reconfigurations for better sound:

Headphone PCA:
* Relocated IC701 & IC702 ICs to solder side for better accessibility
* Installed mini-sockets for resistors R721, R722, R725, R726, R719 & R720

Main PCA:
* Installed mini-sockets for resistors R31, R32, R37 & R38

Attached are a few pics showing the work.

Last year I installed SIP8 sockets for op-amps IC1, IC2, IC3, IC5, IC6 & IC10 and the five 4066 switches for my first RS-B965, along with mini-sockets for the muting transistors. Today I added the above mods, and got the second unit modified to the same level.

I had wanted to experiment with replacing the headphone amp ICs, but on the headphone PCA the amp ICs faced inwards toward the mechanism, and there was no room to accommodate sockets, so I made some right angle 8-pin sockets by hand and soldered them to the PCA solder side, so now I can change the ICs without having to unscrew and separate the headphone PCA from the unit.

Now comes the fun part of checking how the sound changes with op-amp IC changes.

[A.N.T.]

Hi Sam,

That looks like a very neat work *thumbsup* . By the way, have a look if your deck has a mod described in the manual (see the attachment). If not, it is useful to do that as it noticeably improves the handling of very thin tapes.

Cheers

Alex

Alex, I just completed some mods for my pair of RS-B965 decks, in preparation for op-amp changes and line-out/headphone out amp reconfigurations for better sound:

Headphone PCA:
* Relocated IC701 & IC702 ICs to solder side for better accessibility
* Installed mini-sockets for resistors R721, R722, R725, R726, R719 & R720

Main PCA:
* Installed mini-sockets for resistors R31, R32, R37 & R38

Attached are a few pics showing the work.

Last year I installed SIP8 sockets for op-amps IC1, IC2, IC3, IC5, IC6 & IC10 and the five 4066 switches for my first RS-B965, along with mini-sockets for the muting transistors. Today I added the above mods, and got the second unit modified to the same level.

I had wanted to experiment with replacing the headphone amp ICs, but on the headphone PCA the amp ICs faced inwards toward the mechanism, and there was no room to accommodate sockets, so I made some right angle 8-pin sockets by hand and soldered them to the PCA solder side, so now I can change the ICs without having to unscrew and separate the headphone PCA from the unit.

Now comes the fun part of checking how the sound changes with op-amp IC changes.

[scan80269]

Hi Sam,

That looks like a very neat work *thumbsup* . By the way, have a look if your deck has a mod described in the manual (see the attachment). If not, it is useful to do that as it noticeably improves the handling of very thin tapes.

Cheers

Alex

Thanks for the pointer, Alex! Looking at that diode from the PCB component side, I cannot tell if it's been replaced. The supplement does not mention whether the zener diode voltage was changed. Do you know the "from" and "to" zener voltages?

BTW, the RS-B965 has one of the cleanest internal layouts I've ever seen in a cassette deck. Besides the head cables, there are only 4 ribbon cables that plug onto the main board, along with the two daugherboards. Contrast this with decks like the Nak Dragon, Aiwa AD-F990, JVC DD-9, etc. with their multiple PCBs and huge cable messes. *eyepop*

[A.N.T.]

Thanks for the pointer, Alex! Looking at that diode from the PCB component side, I cannot tell if it's been replaced. The supplement does not mention whether the zener diode voltage was changed. Do you know the "from" and "to" zener voltages?

BTW, the RS-B965 has one of the cleanest internal layouts I've ever seen in a cassette deck. Besides the head cables, there are only 4 ribbon cables that plug onto the main board, along with the two daugherboards. Contrast this with decks like the Nak Dragon, Aiwa AD-F990, JVC DD-9, etc. with their multiple PCBs and huge cable messes. *eyepop*

Sam , the zener voltage is changed from 5.1V to 4.3V, it reduces the take-up tension. And I would agree on the layout - the RS-B965 is one of the best decks in that respect.

Cheers

Alex

[scan80269]

Sam , the zener voltage is changed from 5.1V to 4.3V, it reduces the take-up tension. And I would agree on the layout - the RS-B965 is one of the best decks in that respect.

Cheers

Alex

Thanks again, Alex! I just measured 5.1V across D625 on both my RS-B965, so I'll need to replace these zeners.

Takeup torque readings are presently 35-45 g-cm for the first deck and 40-50 g-cm for the second, as measured with a Sony torque meter cassette.

[scan80269]

Alex, I noticed that both my RS-B965 decks did not come with guard washers installed on either of the capstan shafts inside the cassette compartment. Are your RS-B965 like that? I'm beginning to wonder if mine had been "handled" by previous owners/technicians who promptly lost the guard washers after pulling out the capstan/flywheel assemblies... *scratchchin*

[A.N.T.]

Alex, I noticed that both my RS-B965 decks did not come with guard washers installed on either of the capstan shafts inside the cassette compartment. Are your RS-B965 like that? I'm beginning to wonder if mine had been "handled" by previous owners/technicians who promptly lost the guard washers after pulling out the capstan/flywheel assemblies... *scratchchin*

I think these mechs never had washers on capstan shafts! None of my decks has them and they not shown in the service manual.

Cheers

Alex

[scan80269]

I think these mechs never had washers on capstan shafts! None of my decks has them and they not shown in the service manual.

Cheers

Alex

Thanks for the confirmation, Alex! Frankly I'm a bit surprised to see no capstan guard washers for a TOTL Technics deck. I'm tempted to make some washers, but will need the diameters of the two capstans.

Today I discovered some capacitors on the Rec Amp board of my first RS-B965 with AX head that don't agree with the service manual:

C53, C54: SM = 0.022uF, actual = 2200p (222J)
C319, C320: SM = 0.018uF, actual = 0.1uF (104)

My second RS-B965 is even more strange. It has several rework wires on the main board and the daughterboards, PCB layout differences (at least on the Rec Amp board), and completely different PCB color (cream). I counted some 15 differences in resistor & component values on the Rec Amp board alone, and some of the values don't match either the ones for AX head (PP) or MX head (EB, GN) in the SM! There is not even a serial number sticker at the back of the unit! I think I may have picked up a pre-production unit *eyepop*, though it does have an AX head!

[jsmiddleton4]

"posting a list off all the mods"

I agree.

I'm not experienced nor confident enough to take a soldering gun to this level of audio equipment. What I'd like to see is the list of mods and directions that a qualified service technician, if someone were to bring this deck to a tech, could follow and make the mods.

So I could get the parts, the instructions and take all of them with the deck to someone who is confident and experienced in such things.

I know I'm asking a lot, really I am and I understand, but it would be nice to have the mods listed according to impact on REC and Playback so if one didn't want to spend the money to modify the opamps for the headphone jack you could let that one go.

Seems to me by reading this thread the 4066 chips and the resistors in place of the transistors are the main and most imporant mods. Yes?

If that makes sense.

[Warped Bezel]

I agree but think there should be a subforum called called Modification Corner and at least some should become stickies by model or brand.

One for reel, sources and speakers.

This would guarantee easy access across the years.

If you want them in Specialites on the Literature section next to Njord's databases, that's up to you.

[jsmiddleton4]

A B965 just went for 450 dollars on eBay....

Wow.

[vince666]

just bought a second RS-B965 on italian ebay for only 78euros with "Buy It Now"! *devil*

it should arrive here in few days. *hope*