Attention: 5 Ohms is the output impedance, Rds_on is certainly much less than 1 Ohms. If you like, I can measure some of mine for comparison.
2SK77 are somehow still available in Japan, I found my last quad 2SK77 set about two years ago on Yahoo.
In meantime the hype, initiated by Nelson Pass on this forum probably ate up all spares.
The B-I doesn't need to have fried outputs to give you an overprotection, there could also be a defect in -200V biasing (which may destroy the power V-FETs as well) and the driver/small signal V-FETs (2SK75, 2SK78 and some other high voltage FETs) could also be an issue. The electronic fuse circuits may have a defect also, but again this would most likely take the power V-FETs with it as well.
Unfortunately the availablility for 2SK78 and 2SK75 is probably 100x less than for a 2SK77 ... I never saw any available anywhere.
2SK77 is a very robust transistor, it's only weak point is, that it needs a voltage accross its gate to switch off.
2SK77 are somehow still available in Japan, I found my last quad 2SK77 set about two years ago on Yahoo.
In meantime the hype, initiated by Nelson Pass on this forum probably ate up all spares.
The B-I doesn't need to have fried outputs to give you an overprotection, there could also be a defect in -200V biasing (which may destroy the power V-FETs as well) and the driver/small signal V-FETs (2SK75, 2SK78 and some other high voltage FETs) could also be an issue. The electronic fuse circuits may have a defect also, but again this would most likely take the power V-FETs with it as well.
Unfortunately the availablility for 2SK78 and 2SK75 is probably 100x less than for a 2SK77 ... I never saw any available anywhere.
2SK77 is a very robust transistor, it's only weak point is, that it needs a voltage accross its gate to switch off.
THIS has been a world of info on the notorius 2SK77's. I have promised this dead saint to a friend who pleads me not to touch it, but .......... It is alot easier to get to now (after some 25 years) and that I have the service manual and all of you DIY friends. I'll post what I find when time allows.
THANXS DIY AUDIO, JON LORD, tiefbassuebertr et al
THANXS DIY AUDIO, JON LORD, tiefbassuebertr et al
one set was available 40 days ago that i checked with the local Yamaha service ... hmmm not any more ... 2 pairs ( sets ) of 2SK76 paired with 2SJ26 is available at the prize of 116 euro per set .
In between i will make a few calls to mama in Japan and see what they can tell me about it
kind regards sakis
In between i will make a few calls to mama in Japan and see what they can tell me about it
kind regards sakis
Yes indeed ... but it's great to see these wonderful devices being rediscovered for what they are.
Back when these first made their audio debut Bascom H. King wrote a great review of them as used in the B-1, the then new 2SK77 V-FET power FET.
Anybody remember these reviews?
Attachments
Last edited:
Hi Sakis,
Always an honour and a treat to "meet" you.
Hi downandy,
If you can get a set of 2SK77, I recommend that you do. In the event that you don't need them, you can easily re-sell them.
The only other way to replace these devices would be to build up a small circuit board that will allow you to properly mount and use something else. That direction is always possible. You could very probably lead the circuit with a depletion mode device that controls the gate drive for the replacement enhancement mode parts. That way you preserve the original circuitry in the amp. When sparking this thing up, make sure the gate bias is appearing quickly! Any open series gate resistors will sink this ship. Same for normal mosfet outputs, and this was the most common mistakes I have seen other shops make. You could be bringing it up on a variac with open gate resistors and not see any current draw until the bang. You know, the bang that ruins a tech's day.
I no longer have a manual for this amp. Does anyone have an electronic version they could send my way?
Hi Yamahear,
Those were the days, back when audio service could support a living and the properly equipped shop. I actually had a B-1 in for service a billion years ago. The outputs were already NLA and Yamaha couldn't (wouldn't?) get them. The customer didn't realize how much the outputs were worth for one, and refused to pay for development of an alternate output section. Many calls for service on these over the years, but we knew better by that time. This was before the internet existed, the days of the BBS. It was expensive to dial into a US support site for tech info for the various semiconductors. The phone "fax-back" service was the other alternative short of speaking with a human and having a data book shipped. Schematics (if available) cost a fortune too. Those were not good days for service people.
-Chris
Always an honour and a treat to "meet" you.
Hi downandy,
If you can get a set of 2SK77, I recommend that you do. In the event that you don't need them, you can easily re-sell them.
The only other way to replace these devices would be to build up a small circuit board that will allow you to properly mount and use something else. That direction is always possible. You could very probably lead the circuit with a depletion mode device that controls the gate drive for the replacement enhancement mode parts. That way you preserve the original circuitry in the amp. When sparking this thing up, make sure the gate bias is appearing quickly! Any open series gate resistors will sink this ship. Same for normal mosfet outputs, and this was the most common mistakes I have seen other shops make. You could be bringing it up on a variac with open gate resistors and not see any current draw until the bang. You know, the bang that ruins a tech's day.
I no longer have a manual for this amp. Does anyone have an electronic version they could send my way?
Hi Yamahear,
Those were the days, back when audio service could support a living and the properly equipped shop. I actually had a B-1 in for service a billion years ago. The outputs were already NLA and Yamaha couldn't (wouldn't?) get them. The customer didn't realize how much the outputs were worth for one, and refused to pay for development of an alternate output section. Many calls for service on these over the years, but we knew better by that time. This was before the internet existed, the days of the BBS. It was expensive to dial into a US support site for tech info for the various semiconductors. The phone "fax-back" service was the other alternative short of speaking with a human and having a data book shipped. Schematics (if available) cost a fortune too. Those were not good days for service people.
-Chris
Yamahear:
thanks a lot for this hint. Do you also have the complete article available ?
All:
Regarding the hype, I can understand, that even a very well known professional power amp designer, who showed and supported a very stubborn behaviour (ala "not invented here" and "Jap is crap" or "those copycats") over all the years, surprisingly now has understood somehow, what they are and that Japan and not the US finally made it.
Unfortunately he probably didn't understand for decades, that these fine devices have triode characteristic and thus always wrongly called them Power J-FETs or claimed Power J-FET = V-FET which would mean something like Triode = Pentode). This forum is full of these (his) wrong assumptions of the past and forum members, who adopted these without any criticism, because their Guru said it.
On the other hand side, this highly respected designer had his chance to understand and to finally implement V-FETs in his own designs decades ago (when they were freshly available) and now he reintroduces these devices not only for what they are, but with a high emphasis on his own commercial interrests, to create a market for his ideas around depletion mode SiC power semiconductors. For me this smells a little bit fishy ...
Unfortunately this hype doesn't have any respect for those open minded (and -eared) people, who are long time fans of the original Japanese V-FET amplifiers, and who really already have a very difficult spare part situation. Just take a look at this thread, how difficult this situation is !
I didn't want to start my "V-FET Blues" here again, but I'm bitterly disappointed by the recent developments on the V-FET front (regarding DIY), especially on this forum.
Yamahear:
thanks a lot for this hint. Do you also have the complete article available ?
All:
Regarding the hype, I can understand, that even a very well known professional power amp designer, who showed and supported a very stubborn behaviour (ala "not invented here" and "Jap is crap" or "those copycats") over all the years, surprisingly now has understood somehow, what they are and that Japan and not the US finally made it.
Unfortunately he probably didn't understand for decades, that these fine devices have triode characteristic and thus always wrongly called them Power J-FETs or claimed Power J-FET = V-FET which would mean something like Triode = Pentode). This forum is full of these (his) wrong assumptions of the past and forum members, who adopted these without any criticism, because their Guru said it.
On the other hand side, this highly respected designer had his chance to understand and to finally implement V-FETs in his own designs decades ago (when they were freshly available) and now he reintroduces these devices not only for what they are, but with a high emphasis on his own commercial interrests, to create a market for his ideas around depletion mode SiC power semiconductors. For me this smells a little bit fishy ...
Unfortunately this hype doesn't have any respect for those open minded (and -eared) people, who are long time fans of the original Japanese V-FET amplifiers, and who really already have a very difficult spare part situation. Just take a look at this thread, how difficult this situation is !
I didn't want to start my "V-FET Blues" here again, but I'm bitterly disappointed by the recent developments on the V-FET front (regarding DIY), especially on this forum.
A long time ago on this forum I have participated or opened discussions on the v-fet or sit (view my profile).
Time ago for the first time, I also have announced on this forum a chance to buy original Sony or Nec V-fet on Ebay.
Then I presented a complete single-ended amplifier with ALL V-FET devices: the result was very little interest
.
Then I submitted another Vfet scheme, this time in Push-pull complementary configuration: the result was very very little interest.
However then I was able to create interest someone else , and these people have more followers than me and have written articles on Vfet on this forum. Maybe I am not able to do the guru, but I'm glad about this movement of revaluation of Vfet and SIT
Best regards
Francesco Corrado
service manual B1
If I could only figure out this computer in the time between my wife's interuptions I would make an electronic copy of the manual for you all. It cost me ten bucks but to the few of us -strike that- few of you that could use this it would be invaluable.
..SPARE TIME: the amount of time the wife has set away for honey do's
now where is the spell checker?
If I could only figure out this computer in the time between my wife's interuptions I would make an electronic copy of the manual for you all. It cost me ten bucks but to the few of us -strike that- few of you that could use this it would be invaluable.
..SPARE TIME: the amount of time the wife has set away for honey do's
now where is the spell checker?
OK here's a quicky. a B1 has just finished bidding on EECK-BAY and in the description it says clearly that it is for parts and repair. it also states "The amplifier is intact with all circuit boards and output V-FET 2SK77 transistors. These transistors had a static diode test done without power applied and one tested shorted. The other three did not test shorted. " they implied that these three were ok.
Doesn't that go against what I've been learning about these ARNOLD type TO-3'S?
Doesn't that go against what I've been learning about these ARNOLD type TO-3'S?
I was watching that auction too just for fun and I don't know this for sure but from the seller's test results it could be that perhaps they have things backwards and there's really only 1 FET that's possibly still good and the other 3 are toast maybe ...
I wonder how hard it would be to use the IXYS IXTH20N50D depletion mode part in a B1?
http://ixdev.ixys.com/DataSheet/DS99192B%28IXTH-T20N50D%29.pdf
http://ixdev.ixys.com/DataSheet/DS99192B%28IXTH-T20N50D%29.pdf
YAMAHEAR I thinks you are right but DJK!! @**%^@#! dON'T you think that a part that small might fall thru the bottom or get lost somewhere inside? Great Specs I need to find out more! There goes my Sunday afternoon----when I should be setting all of my clocks ahead
now let me see
http://ixdev.ixys.com/DataSheet/DS99...T20N50D%29.pdf
now let me see
http://ixdev.ixys.com/DataSheet/DS99...T20N50D%29.pdf
Simply put - no, you cannot use one of these unless you extensively modify the unit, to the point it would not be a Yamaha B1.
The B1 has a rather unusual topology for a solid state amplifier, quite close to that of a Futterman tube OTL amplifier, but with triode outpurs.
The data for the IXYS device clearly indicates that it's mostly enhancement mode, and expects positive gate voltages to get some real current flowing.
The K77 in common with other SITs electrically tests like a JFET so there is a diode from the gate to the channel structure (basically S+D), hence the maximum positive voltage you can apply is one Si diode voltage drop - and it has to be current limited to about 1A peak. A negative 20V or so (varies quite a lot between instances of the same type VFET) will turn the current down to a few mA at 200V across the device. Hence, the driver stages are designed to provide such a signal.
It should also be mentioned that the IXYS device has pentode characteristics hence the final gain of the stage it's used in depends heavily on the load. A SIT has triode characteristics and has a low internal resistance and a defined (and rather low) gain, which makes it's application somewhat different even in follower mode. Even if the driving voltages and currents were compatible, frequency compensation wouldn't (and it's quite complex in the B1).
Finally, the K77 has lower transconductance and capacitances than the IXYS part, however although the capacitances are about 1/2 to 1/3rd of that of the IXYS device, it's the IXYS that has MUCH more non-linear behaviour in this respect.
Contrary to popular belief, VFETs (SIT) do NOT always have a negative tempco. It actually turns positive at higher currents and temperatures but since the transconductance is low and the gain has a fixed 'ceiling' it is not enough to cause device thermal runaway. The B1 has no means of temperature compensation, similar to all VFET designs I have seen.
The B1 however, has a rather extensive protection circuit which does not act in the amp itself but rather through the power supplies. The B1 has fully regulated power supplies, which is one consequence of the way SIT work, not only due to the need to apply bias first (or they act basically as a short), but also because just like ia triode, ALL pins are inputs - current depends on Vgs AND Vds. To anyone not familiar with triodes, VFET/SITs are not exactlye asy to work with properly.
Finally, in a pinch (and with a mechanical adaptor) you could replace one 2SK77 with two 2SK27 in parallel, PROVIDED you could find enough to build good pairs. In fact two K76 and two K26 work dead parallel (no means of current balancing) in a Yamaha B2, which is why Yamaha only ever supplied matched sets. You could take the two K76 used in a B2 and use them in parallel in a B1 - assuming you can buy them somewhere. A better bet would be using two Sony K82, which are available on ebay regulairly, through Acronman Electronics. A long time ago a bought a number of these and their P-ch counterpart, J28, and they are all genuine.
It should be said that there are two general philosophies in building the VFETs of old - Sony and the rest (Yamaha, NEC, Hitachi). Sony parts have lower transconductance but are more linear and have signifficantly lower and more linear capacitances. They are in general geared toward complementary operation and their characteristic curves are tailored to cancel non-linearities in appropriately biassed complementary amps - AS ARE THE CAPACITANCES! This approach is somewhat unique historically and offered great potential for some truly outstanding designs, which has sadly not been properly utilized to this day. The down-side is higher Rdson, i.e. rail loss. Yamaha et al. went for higher transconductance and higher capacitance, less lossy devices, which are slightly more non-lienar and not as well matched when it comes to complementarity, but offer more power per pair. Sit technology that followed (including all but a few out of dozens of later offerings from Tokin) went the same way, so does the Nelson Pass SIT1. It is possible that some of the smaller Tokin SITs could be used instead of the K77, they even use the same casing. Still, if one had to ressurect a Yamaha B1 with dead K77s, usinf a pair of Sony K82 instead of each K77, there would be a small loss in maximum power butt the smaller capacitance may have positive sonic effects. A member of this site that is no longer active, verified this approach replacing K70/J20 in a Sansui BA-1000 VFET amp (one of the rarest and the only VFET amp made by Sansui) with Sony parts.
It's nice to see that there's still an interest and understanding out there of these wonderful devices!
Doesn't matter what's enough for me, but for the B1 as it was suggested as a replacement for the 2SK77. Most SITs have Idss approaching the maximum design current as it's the whole idea of a depletion only device. That being said, they also exhibit a small enhancement region when Vgs becomes positive (within limits of a Si diode voltage drop of course), sometimes it actually signifficantly increases Id from the Vgs=0 condition. (*)
Please understand I'm not plucking this data out of thin air. Having a good number of different types of VFET/SIT parts and a Tek 576 curve tracer at hand helps greatly. It's also really the only way to verify proper operation of a part, various far east 'suppliers' have gotten wind of the value of these parts in the last 5-6 years or so and there are many fakes around, as evidenced in other threads.
(*) In fact, there is a ex-USSR development of the SIT which exclusively uses the enhancement region of a SIT, has pentode characteristics and since it has (seizable) gate current and positive Vgs, it actually has Beta like BJTs. It was used as a replacement for high voltage switching transistors. It is, however, a rather slow part owing to the classical problem of base or in this case, gate charge removal, but it is immune to secondary breakdown. At first glance it measures tlike a BJT but the breakdown region (where C-B begins to leak adding to base current) is different, looking like a zener breakdown, which shows that it's really Vgs governing the current, not Ig. This has been verified on real devices, again, I'm not just telling tales.