Any Phase Linear Experts out there?

[laatsch55]

I recently restored a 400 SerI and managed to mount 8 - 6800uf 80 volt caps in the original space of the 2- 5900uf. They were snap-ins that i butt connected, crimped and soldered top to top and fabricated new brackets.
No issues with inrush or rectifier. It continues to play for 15-20 seconds after removal of main ACpower.

 

[d3imlay]

Any NEC 2sd555's are almost certain to be fakes. When they were available I used 100's of them. They are faster than MJ5xxx and I think more prone to oscillate. They were great transistors.

Bharper: Is there a reason you use them as drivers? When the factory switched to MJ15024's they also used them as drivers. the 400 can be modified to full complimentary. I've done it. A lot of work. The last series II's were full comp.

If ythe bridge is changed out you may as well add the .01uf caps on each diode of the bridge to ward off RF interference.

 

[bharper]

one of my basket case 400's had about a half a dozen of the nec 2sd555's in it, and having read that they were a very robust transistor, went ahead and used them in the driver position. new mj15024s for the outputs.

the newer mj211xx series may be the best current replacement, pretty similar to the 15024, but supposidly designed for audio use.

 

[laatsch55]

The factory used 15024's as drivers while also using them as outputs? Seems i read they didn't recommend using the same trans as a driver and output. Is there any reason you couldnt?

 

[laatsch55]

BTW d3imlay,
You were right about the MJ15011 being borderline. The first 4 ohm load I ran them on and got loud they failed rather quickly. Have now swapped out the 011's with 15024's and 21196's on the 4 amps I had them in. Thanks, Lee

 

[d3imlay]

the newer mj211xx series may be the best current replacement, pretty similar to the 15024, but supposidly designed for audio use.

The 211xx have similar specs to the 15024 with a higher safe operating area (SOA).

 

[releone71]

Any NEC 2sd555's are almost certain to be fakes. When they were available I used 100's of them. They are faster than MJ5xxx and I think more prone to oscillate. They were great transistors.

I know from several discussions around the web and PL service manual that trend and the use of caps to stop the phenomenon. But why exactly they tend to oscillate? Will you be so kind as to explain in simple words? Thanks.

Bharper: Is there a reason you use them as drivers? When the factory switched to MJ15024's they also used them as drivers. the 400 can be modified to full complimentary. I've done it. A lot of work. The last series II's were full comp.

My 700 II is a full complementary too. Anyway, I've heard from people who own both early and late models the quasi complementary output stage sounds better. They talk about a more fashinating, magical sound. What do you think about it? As for me I can talk about my experience with my Quads 303. I have three of them, one 100% original, the other two upgraded by Net Audio UK. The one with the original quasi complementary output stage actually sounds warmer and more magical, where the others with full complementary output stage are rigorous, detailed and crisp performers capable of driving low impedance speakers. They both sound nice but absolutely different.

If the bridge is changed out you may as well add the .01uf caps on each diode of the bridge to ward off RF interference.

Good to know. Which type of caps do you suggest? How can I wire them together with the diodes in the bridge? Could it be good a new bridge like this one? http://it.mouser.com/Search/Product...51virtualkey61370000virtualkey625-GBPC3508-E4
I've heard also it would be good to add bypass caps across the main filter caps. Why? Which type and value in case? Thanks.

 

[bharper]

higher bandwidth, and higher gain transistors can create the oscillations, because the circuits were not orginally designed or tested with them in mind.

usually a small cap can be added base to emitter to help solve this, but ymmv.

problem here too, is some times the oscillations can be above your hearing, and would require the use of a spectrum analyzer to know that they were there.

 

[mikethebike]

My Phase Linear caught fire and if it wasn't for the protection circuit it would have damaged the speakers too.

We called them 'Flame Linear' after a bunch of them caught fire on stage at a WHO concert back in the early 70's.
I lost 2 tweeters to Phase Linear and Bob Carver amps in the early 80's. Stayed with Crown after that.

 

[d3imlay]

I know from several discussions around the web and PL service manual that trend and the use of caps to stop the phenomenon. But why exactly they tend to oscillate? Will you be so kind as to explain in simple words? Thanks.

Good to know. Which type of caps do you suggest? How can I wire them together with the diodes in the bridge? Could it be good a new bridge like this one? http://it.mouser.com/Search/Product...51virtualkey61370000virtualkey625-GBPC3508-E4
I've heard also it would be good to add bypass caps across the main filter caps. Why? Which type and value in case? Thanks.

Depending on who you talk to there are a number of opinions for the oscillations. It's been said before that some of the worlds best oscillators started out as audio amplifiers. I know of one individual who claims that adding a small value resistor on the base of the predrivers stops all oscillations as it takes advantage of the capacitance in the transistor. The factory put ferrite beads on the base bus in the drivers. I don't know what difference an output choke would make.

Attached is the page of the manual that refers to the rectifier. the rectifier shown is a typical device. Probably wouldn't hurt to have .1uf decoupling caps on the power supply. Electrolytic caps have poor high frequency characteristics. Ceramic and plastic types are much better at higher frequencies..

 

[releone71]

Attached is the page of the manual that refers to the rectifier. the rectifier shown is a typical device. Probably wouldn't hurt to have .1uf decoupling caps on the power supply. Electrolytic caps have poor high frequency characteristics. Ceramic and plastic types are much better at higher frequencies..

Interesting page. :thmbsp: Is that a PL service bulletin? I have some bulletin attached to a copy of the 400 service manual but I've never seen that page before. Is it part of the 700 II service manual?

Regarding the ceramic and plastic caps, are they good also to bypass the main filter caps? Which value in case? Is there a rule to choose the correct value? Do I have to calculate a percentage (1% of the main capacitance for example)? Thanks.

 

[d3imlay]

Interesting page. :thmbsp: Is that a PL service bulletin? I have some bulletin attached to a copy of the 400 service manual but I've never seen that page before. Is it part of the 700 II service manual?

Regarding the ceramic and plastic caps, are they good also to bypass the main filter caps? Which value in case? Is there a rule to choose the correct value? Do I have to calculate a percentage (1% of the main capacitance for example)? Thanks.

I use .10 to .50 mfd depending on the application. The cap needs to be as close to the outputs as possible to be effective.

 

[P.L.F.]

Beta matching - makes sense?

As always, many great advice on D3imlay moderated page :thmbsp:... I am renovating two PL700II gears, export versions - one with quasi and the other with fully comp output. My current focus are PL36 boards. However one can find this kind of message regarding the output transistors matching for PL700II:

"(...) Bob Carver discovered that other amplifiers of the time that he designed the Phase Linear series, used a dual darlington output stage. This configuration had a distortion curve that would rise quickly with an increase in the current in the output transistors. Beta or gain collapse was the cause for this. Bob's triple darlington significantly improved the performance, but you still could not get optimum performance by just replacing the transistors with off the shelf units. To achieve the highest performance requires matching the Beta of both sections of the darlington totem pole. (...) We use state of the art epitaxial power transistors (i.e. MJ15024) and test and match them in the same totem pole configuration as in Phase Linear amplifiers. The upper and lower totem poles are independently characterized for beta at 32 amperes collector current and then burned in for 72 hours. (...)"

Do you agree? I was planning to use off the shelf MJ21196/95 still available at e.g. Mouser. Would I need any specific selection to take into account? Thanks for your comments!

 

[releone71]

As always, many great advice on D3imlay moderated page :thmbsp:... I am renovating two PL700II gears, export versions - one with quasi and the other with fully comp output. My current focus are PL36 boards. However one can find this kind of message regarding the output transistors matching for PL700II:

"(...) Bob Carver discovered that other amplifiers of the time that he designed the Phase Linear series, used a dual darlington output stage. This configuration had a distortion curve that would rise quickly with an increase in the current in the output transistors. Beta or gain collapse was the cause for this. Bob's triple darlington significantly improved the performance, but you still could not get optimum performance by just replacing the transistors with off the shelf units. To achieve the highest performance requires matching the Beta of both sections of the darlington totem pole. (...) We use state of the art epitaxial power transistors (i.e. MJ15024) and test and match them in the same totem pole configuration as in Phase Linear amplifiers. The upper and lower totem poles are independently characterized for beta at 32 amperes collector current and then burned in for 72 hours. (...)"

Do you agree? I was planning to use off the shelf MJ21196/95 still available at e.g. Mouser. Would I need any specific selection to take into account? Thanks for your comments!

That seems to be the ad of the guy who usually sells on eBay those upgrade kits for the output stage.
MJ21195/21196 are the best choice at the moment. They are better than MJ15024/15025 in the safe operation area (3Amperes). I'm going to replace them in my 700 II together with the main caps and the electrolytics. I purchased the replacements from Mouser: really efficient people there!!:tresbon:
Congrats for the PL rack!!

 

[cademan]

OK, so while reading these posts, which Motorola transistor would be the best to use in place of my FPL 909's on my 400, series II, with the least chance of oscillations?

 

[d3imlay]

I buy 21196's for 400's and 700's. I'm not convinced that matching is necessary, but I do it as a matter of Q.C. My H.P. test gear doesn't have the resolution (<.05% THD) to measure the difference. I check each transistor on a curve tracer, I check for max voltage breakdown (Vce) and current gain. Rarely do I find one that doesn't meet spec. Each transistor in the amp has an emitter resistor to force them to share the load which accounts for differences from one to the other. I'd be very surprised if the P/L factory matched the transistors.

When I worked at Westinghouse R&D in the early 70's I did transistor testing on stuff like 2n2222's. Borrring. I could do anywhere form 200 to 800 in 4 hours.

In the early 70's, I could believe that there were wider variations from one batch to the next. I buy a hundred at a time and they always have the same date code in the lot. I match them to within about 2%. In the lot of 100 they're within 8-10 percent as is. The service manual calls out a load sharing test as part of the checkout process. The spec is +/-20%.

By the way, I occasionally check transistors that I pull from repaired units. They're all over the map. I use these for destructive testing.

 

[kcollins4]

An interesting bit of info there. One of these days I'll so something with my
PL-400 and 4000. Just too many projects...

 

[P.L.F.]

I buy 21196's for 400's and 700's. I'm not convinced that matching is necessary, but I do it as a matter of Q.C. My H.P. test gear doesn't have the resolution (<.05% THD) to measure the difference. I check each transistor on a curve tracer, I check for max voltage breakdown (Vce) and current gain. Rarely do I find one that doesn't meet spec. Each transistor in the amp has an emitter resistor to force them to share the load which accounts for differences from one to the other. I'd be very surprised if the P/L factory matched the transistors.

When I worked at Westinghouse R&D in the early 70's I did transistor testing on stuff like 2n2222's. Borrring. I could do anywhere form 200 to 800 in 4 hours.

In the early 70's, I could believe that there were wider variations from one batch to the next. I buy a hundred at a time and they always have the same date code in the lot. I match them to within about 2%. In the lot of 100 they're within 8-10 percent as is. The service manual calls out a load sharing test as part of the checkout process. The spec is +/-20%.

By the way, I occasionally check transistors that I pull from repaired units. They're all over the map. I use these for destructive testing.

Thank you d3imlay for this advice. Much clearer now to plan renovation works at reasonable cost! Don't have much time theses days for my hobby but when the job is done I'll share the results.

Releone71, very kind of you to notice. Was an effort to bring all this vintage to Poland. But it's worth it! My music listened with P/L still drives zillions of emotions ...

BTW, I'm looking for PL700II aluminum faceplate replacement. I'd very much appreciate any suggestion on how to acquire this element.

Cheers

 

[d3imlay]

BTW, I'm looking for PL700II aluminum faceplate replacement. I'd very much appreciate any suggestion on how to acquire this element.

Cheers

Do you know of a sheet metal shop? The panel is just stock aluminum that has been anodized. If you get one made, make a bunch and recover your cost. It seems like they pop up every now and then missing the panel.

 

[P.L.F.]

Do you know of a sheet metal shop? The panel is just stock aluminum that has been anodized. If you get one made, make a bunch and recover your cost. It seems like they pop up every now and then missing the panel.

Correct! Thank you D3imlay. I am looking around then to find the right place. My plate was devastated a bit by audio recording studio in Germany, where they had to fit the P/L amp into a rack... But the second one amp I have is in perfect condition - so feasible to make a 'copy'.

Best regards!