Phase Linear 400 output transistors
John Whitmore
I've recently acquired a Phase Linear 400. This was
an historically important audio amp, as it was the first of the
wave of high-powered amplifiers that came out of the 1970's.
This unit, though, has been through some tough times.
I've found lots of shorted and fried components, but the
main difficulty is in the output transistors.
There are four different types plugged in to the
sixteen TO-3 sockets. The outputs are four composite-Darlington
transistors, each having a single driver feeding three
paralleled output transistors. Someone in the past
replaced some of the paralleled outputs with various types.
So, I don't trust the types of transistors I see there.
Also, I don't know what the original drivers were.
Driver types: GE-37, DTS-701, RCA410
The first two are 700V, 1A; the third is 300V, 5A rated.
Output types: GE-37, PL-909
The first is clearly inappropriate in this application,
and the second is (apparently) a Fairchild special part
number for Phase Linear. I assume it's intended to be
a high current type, and possibly similar to 2N5264
(a Fairchild power transistor of the right era).
From the emitter degeneration resistance (0.22 ohms)
I suspect the output transistors should be roughly 5A rated.
Does anyone know? Should I just jam in MJ15003's?
Or maybe MJ410's (very similar to the RCA410)?
And as for the drivers; should they all be the same
type as the outputs? What were the original drivers?
Thanks for any info; this can't interest many folk,
so e-mail is preferred.
John Whitmore
wh...@hipress.phys.washington.edu
Tom Ciaramitaro
The outputs were the PL909's; long since gone. Sometimes you can find a
used handful at a higher end repair shop; else change the whole bank of
outputs and use 2SD555s. We have done a significant number of these
with success. If you end up with a few spare PL909s save them for the
other channel if ever needed.
-To
Moe Kunkle
Tom Ciaramitaro wrote:
>
> John Whitmore wrote:
> >
> > I've recently acquired a Phase Linear 400. This was
> > an historically important audio amp, as it was the first of the
> > wave of high-powered amplifiers that came out of the 1970's.
> >
> > This unit, though, has been through some tough times.
> > I've found lots of shorted and fried components, but the
> > main difficulty is in the output transistors.
Now this amp is truelly wonderful device. A lot of rock-n-roll
garage bands just had to have them. After a while I heard them
refered to as "Flame Linear". Most of the failures were due
to improper load, shorted load line, DC on the input etc. I even had one
guy
who reasoned: Hey, I can get 400 watts if I tie the two red
posts together. Up in a puff of smoke it went. You don't wanna
know what that did to voice coils of that stack of JBL 15's. hehehehe
I put myself through school on these babies.
Back in the good'ol days you could call the factory after hours
and Bob Carver himself would answer the phone.
> >
> > There are four different types plugged in to the
> > sixteen TO-3 sockets. The outputs are four composite-Darlington
> > transistors, each having a single driver feeding three
> > paralleled output transistors. Someone in the past
> > replaced some of the paralleled outputs with various types.
> > So, I don't trust the types of transistors I see there.
> > Also, I don't know what the original drivers were.
> >
> > Driver types: GE-37, DTS-701, RCA410
> > The first two are 700V, 1A; the third is 300V, 5A rated.
I have used GE-35 for the outputs in a pinch before. Never
had one to come back.
> >
> > Output types: GE-37, PL-909
> > The first is clearly inappropriate in this application,
> > and the second is (apparently) a Fairchild special part
> > number for Phase Linear. I assume it's intended to be
> > a high current type, and possibly similar to 2N5264
> > (a Fairchild power transistor of the right era).
> >
> > From the emitter degeneration resistance (0.22 ohms)
> > I suspect the output transistors should be roughly 5A rated.
Sounds real low. Should be 15 or 20 amp.
> > Does anyone know? Should I just jam in MJ15003's?
MJ15024's might be better here. The 2SD555's that John
recommends would be better still.
> > Or maybe MJ410's (very similar to the RCA410)?
> >
> > And as for the drivers; should they all be the same
> > type as the outputs? What were the original drivers?
The 410's will serve well as your driver. It's been
better than 15 years since I've looked at the service manual.
> >
> > Thanks for any info; this can't interest many folk,
> > so e-mail is preferred.
> >
> > John Whitmore
> > wh...@hipress.phys.washington.edu
> The drivers are original, they don't fail often.
> The outputs were the PL909's; long since gone. Sometimes you can find a
> used handful at a higher end repair shop; else change the whole bank of
> outputs and use 2SD555s. We have done a significant number of these
> with success. If you end up with a few spare PL909s save them for the
> other channel if ever needed.
> -To
Good advice. The 2SD555's are very good in this application. As for the
PL909's, I remember that after a couple of failures, the leakage on the
one's that survived was all over the place. They were originally matched
sets if my memory serves me here. I never installed matched sets on a PL
repair but did at least make sure that all the outputs were the same
type.
You might want to bring it up slow with a variac. All those outputs will
set you back a few $$$.
Good luck with it.
Moe Kunkle
Mr Fidget
I luv these amps. I am the proud owner of 6 x PL700's. :)
Replace all the transistors on the heatsink with MJ15024s including the
driver. Some of the PL700's had complimentary outputs. Replace these
with the MJ15025's.
(2SD555's cost me 3 times as much so I can't comment)
Before you power up. Pull out the rail fuses and connect a 10W 150 ohm
to 270 ohm resistor across the fuse holders. When you power up, check
the voltage across each of these. They should be roughly equal in the
range of 1.5 - 5V. If you get 30 - 80V you have a major problem.
Doing this limits the current to the outputs and will save you from a
disaster. You can then check the output for 0V DC, then run a low level
signal with a speaker attached to check the output for obvious
distortion. Dont crank it too high or your protection resistors will
fry. (See thread on Infrared :) )
Its a good idea to replace the electrolytic caps on the PCB. This tends
to clean uo a lot of wierd noises that sometimes come after replacing
blown outputs.
Substitute transistors that I have used successfully are
2N3403 BC338
2N3569 BC547
2N5455 BC557
MPSA93A 2N5415
RCA40327 2N3440
MM4003 2N5416
RCA40412 MJE340
PL909 MJ15024
However would be interested in any suggestions for the above.
Especially a replacement for the RCA40412. Replacing it with the MJE340
works but there is a noticable degridation in the clarity and smoothness
in the top end. (Mind you as they are usually used for amplifying crappy
grunge bands to pissed punters who really cares 8^] )
Also any one got a suggestion for a better Opamp than the LF351's I have
been using for the B models would be much appreciated.
Have a good time...all the time
Chris
ps. I am from Oz so any suppliers in the US may not be all that
practical for me.
Dan Fraser
In article <5msdgl$1f...@nntp6.u.washington.edu>,
wh...@hipress.phys.washington.edu says...
>
>
> I've recently acquired a Phase Linear 400. This was
>an historically important audio amp, as it was the first of the
>wave of high-powered amplifiers that came out of the 1970's.
>
> This unit, though, has been through some tough times.
>I've found lots of shorted and fried components, but the
>main difficulty is in the output transistors.
>
>sixteen TO-3 sockets. The outputs are four composite-Darlington
>transistors, each having a single driver feeding three
>paralleled output transistors. Someone in the past
>replaced some of the paralleled outputs with various types.
>So, I don't trust the types of transistors I see there.
>Also, I don't know what the original drivers were.
>
> Driver types: GE-37, DTS-701, RCA410
>The first two are 700V, 1A; the third is 300V, 5A rated.
>
>The first is clearly inappropriate in this application,
>and the second is (apparently) a Fairchild special part
>number for Phase Linear. I assume it's intended to be
>a high current type, and possibly similar to 2N5264
>(a Fairchild power transistor of the right era).
>
> From the emitter degeneration resistance (0.22 ohms)
>I suspect the output transistors should be roughly 5A rated.
>
> And as for the drivers; should they all be the same
>type as the outputs? What were the original drivers?
>
I've used Motorola MJ413's to fix these amps. Please note the early Phase
Linears while sounding great were not all that stable and they tend to blow
outputs a lot. You may be embarking on an adventure that may be expensive and
have an unhappy ending. I've been there.
tu...@400.com
dfr...@direct.ca (Dan Fraser) wrote:
>> I've recently acquired a Phase Linear 400. This was
>>an historically important audio amp, as it was the first of the
>>wave of high-powered amplifiers that came out of the 1970's.
Ever work on a Carver C500?
george...@gmail.com
> I've recently acquired a Phase Linear 400. This was
> an historically important audio amp, as it was the first of the
> wave of high-powered amplifiers that came out of the 1970's.
>
> I've found lots of shorted and fried components, but the
> main difficulty is in the output transistors.
>
> There are four different types plugged in to the
> sixteen TO-3 sockets. The outputs are four composite-Darlington
> transistors, each having a single driver feeding three
> paralleled output transistors. Someone in the past
> replaced some of the paralleled outputs with various types.
> So, I don't trust the types of transistors I see there.
> Also, I don't know what the original drivers were.
>
> Driver types: GE-37, DTS-701, RCA410
> The first two are 700V, 1A; the third is 300V, 5A rated.
>
> Output types: GE-37, PL-909
> The first is clearly inappropriate in this application,
> and the second is (apparently) a Fairchild special part
> number for Phase Linear. I assume it's intended to be
> a high current type, and possibly similar to 2N5264
> (a Fairchild power transistor of the right era).
>
> From the emitter degeneration resistance (0.22 ohms)
> I suspect the output transistors should be roughly 5A rated.
> Does anyone know? Should I just jam in MJ15003's?
> Or maybe MJ410's (very similar to the RCA410)?
>
> And as for the drivers; should they all be the same
> type as the outputs? What were the original drivers?
>
> so e-mail is preferred.
>
> John Whitmore
> wh...@hipress.phys.washington.edu
george...@gmail.com
Smarty
layouts, instructions, are available at numerous places for free. A
Google search turned this one up, among others:
http://www.vintageshifi.com/repertoire-pdf/Phase-Linear.php
jurb...@gmail.com
http://thecarversite.com/
THAT is where you get information on these. I bought one with a fried right channel a couple years ago and due to improper servicing it had blown it all the way back to the current limiters. Surprisingly it did not blow the drivers, but it did blow one predriver. I forgot what I used for those but I DID look at spec sheets etc. because you don't just go to NTE and get transistors for something like this. If you need anything other than the outputs I will find you the correct transistors, don't trust anyone on this. A couple of them are germanium.
The current correct output transistor type is MJ15024 available from Digikey for reasonable, about five bucks each. Replace them ALL in one bank, no exceptions. Each in a bank of three has to equally share the current, and rated down to four ohms with 85 volts rails that pretty much stay put, you don't need any current hogs. That's what happened to mine, it blew the emitter resistor on the one in a bank used for current detection. This is why it blew the predrivrs but not the drivers.
Remove the transistors for checking, all the resistances are too low in the whole thig to do it in circuit. Check ALL emitter resistors. After all is said and done, you can run it on the bench without a load connected with all of the outputs removed and it should look right on a scope, with possible SLIGHT early clipping on one side of the waveform.
There are two types of these, most use all NPN outputs and drivers, the other uses complemmentary. As far as I know only some of the Series Two amps were complementary. If it has the LED meters you don't know, those were built around 1979 and 1980 right after Pioneer bought the company. These have an IC on the board rather than the usual diff pair at the input.
The older ones with the analog meters as fas as I know all used all NPN outputs and drivers.
After I got the right channel fixed I was jamming it and the damn left channel fried. The reason was that where the heavy wires act as a bus and connects all the collectors together in one of the banks of outputs had come loose, it broke the solder over the years. Check ALL connections to ALL outputs because even one of them off can fry the whole channel.
Really, if you are not used to working on high power amps like this, you could waste alot of silicon.
The price break on the MJ15024s comes at ten pieces, so just get twelve and replace them all. With any luck they will all be from the same batch so matching will not be a big problem. There is a matching procedure required after replacing the outputs to assure they are properly matched. If not they could fail prematurely when pushed hard, and you buy this amp to push it hard.
A short at the output terminals should not fry this amp, but connecting it to certain low impedance loads can. The book says never go below four ohms, but many four ohm systems do just that. Bob Carver knew this and set the current limiting to handle probably down to about three ohms because of those gnarly woofers and infinite baffle systems of the day, but any lower and it is vulnerable.
Aso note these amps have abslutely no speaker protection whatsoever, no relay no nothing. They recommend you use speaker fuses calculated to your particular speaker handling capabilities, the formula is in the book.
If you buy the manual you have been scammed, it is available for free not only from carversite, but a few other places. I think you can get it at bama as well, but the best one with the addendums etc. is at carversite. There is a note about certain Infinity speakers that use some ungodly high value cap in their crossover that affects the power down of these amps and can result in a transient if turned back on too quickly. If you got this, you got the right manual. Also note that you do have to sift through that manual as more than one version is covered.
When you get it working you probably won't want to sell it. Even at flat response it sounds better due to an extraordinarily high damping factor. This is one of the few amps that can really benefit from heavy guage (and short) speaker wires. The cones of the speakers act like servomotors actually. Mine is not for sale except under extreme duress, and I mean for alot more than it's worth. You can't buy an amp like this for less than twice what you could sell it for.
So fix it right.
There is a guy on eBay who claims to fix these right for about $350 no matter what is wrong with them. Of course two way shipping is probably extra. The point is, there is a reason people don't fix these things for a hundred bucks.
Get the ohmmeter out before even plugging the thing in and keep us posted.
David Platt
>> george...@gmail.com
>Service manuals for Phase Linear 400s, 700s, etc. with schematics, PCB
>layouts, instructions, are available at numerous places for free. A
>Google search turned this one up, among others:
issue is probably moot (from the OP's point of view)/
jurb...@gmail.com
Oh well, if anyone need advice on those thins I guess there it is.
Jon Elson
messages propagate to other servers which have long since forgotten
they ever had this message, so it shows as new. Yup, seen a LOT of these
relics from the VERY distant past!
Jon
Leif Neland
replying to a message on google groups.
I wish Google would add a test: "You are replying to an old message. Do
you really want this?"
Leif
--
Husk kørelys bagpå, hvis din bilfabrikant har taget den idiotiske
beslutning at undlade det.
Phil Allison
<george...@gmail.com>
> If interested I have some XPL909 for sale
A DMM will show about 68 ohms B-E with either polarity at the probes.
BTW:
Phase 400s had in *interesting* design flaw that would cause both channels
to go DC when one channel suffered an output device failure, due to using
common DC fusing for both channels. It worked like this:
1. A single output device fails short in channel A sending that output full
rail DC.
2. The same channel tries to pull the output back to zero by driving the 3
output devices on the other side hard - so one or more of them fails short
immediately due to SOA being grossly exceeded.
3. Channel A's output stage is now shorted rail to rail.
3. One of the two DC rail fuses instantly opens removing the short on the
PSU.
4. Both channels now have their DC supplies connected to a single rail of
either + or - 85V.
5. One of the "flyback" diodes wired from DC rail to speaker output become
forward biased in channel B.
6. The speakers connected to both channels now start smoking .....
FYI:
One can normally remove either DC rail fuse on a PL400 with no ill effect.
The PL700 has 4 DC rail fuses, which eliminates the problem.
Many stereo power amps have NO rail fuses, which also eliminates the
problem by forcing the AC supply fuse to blow.
.... Phil
jurb...@gmail.com
Yes, I am not really happy with the fusing arraingement in these, they should be separate for each channel. I can't seem to find 8 amp fuses easily, right now mine has 5s. It works you just can't crank it up too much. I'll p[robably put 19s in it eventualy. It would probably not be a good idea to eliminate them though because in the case of a short the thing can put out significant current. Hell, you can probably get over 10 out of it under normal conditions, 85 volts... and it WILL run into 4 ohms.
One little thing I was thinking of doing is to build a little protection circuit for such amps. At these power levels a I could use the NC contacts of a relay, use a couple resistors and a cap to filter out DC and rectify it to operate the relay and possibly save the woofers. If not at least prevent a fire.
A guy contaced me about a GAS, so for the hell of it I looked up the Ampzilla. It also has no relay, but it does at least have speaker fuses. I bet those things smoke pretty good because they have the outputs in series rather than parallel. I think the power out is close, and the Ampzilla uses only four outputs instead of six but is cooled by forced air.
Anyway, the diodes and the fuse situation is why I told the guy to take the outputs out to test.
Phil Allison
<jurb...@gmail.com>
The B-E resistor really makes no difference since the amp has a 10 ohm
across it anyway.
One little thing I was thinking of doing is to build a little protection
circuit for such amps. At these power levels a I could use the NC contacts
of a relay, use a couple resistors and a cap to filter out DC and rectify it
to operate the relay and possibly save the woofers.
.... Phil
jurb...@gmail.com
>"** The relay contacts will arc and burn soon as they open under a DC fault. "
take a nice 47 uF or so bipolar cap across the contacts. Period. think aabout it no before to go saying anything.
First of all the cap is not in the circuit during normal operation. The relay only kicks "out" by the application of high DC, which of course WOULD produce an arc due to the inductance, but a capacitor will take care of that. We are not concerned with AC here, just DC blowing your fucking $3,000 Tannoy. Plus when amps put out full DC, they are not puttin gout any AC except for maybe some power supply ripple.
You know, that was a damn goo danswer about why they put those magnets around the speaker realys in amps that make the Ampzilla look like Kermit the frog, but my question is why didn't they just put a fucking capacitor across the terminals ?
You tell me. I am not being sarcastic. I can't tnink of a single reason why they couldn't just put a 10/300 bipolar across there and be done with it anstead of fucking with magnets. that solution probably cost directly in manhours to implement, a cap could simply be on the board.
What's more, from a fidelity standpoint, if the realy had any inductance the cap would improve its performance there. Maybe even better the slew rate.
I wonder why these highly paid enjunears cunt figger dat out.
I see no downside to the caps. they are probably even cheaper than magnets. they are certainly cheaper to install.
Phil Allison
<jurb...@gmail.com>
>"** It does if you mix up devices. "
Not much. In each bank you got the bases all connnected directly together
and the emitters are only separated by 0.33 resistors.
pre bias is skewed.
>"** The relay contacts will arc and burn soon as they open under a DC
>fault. "
Well now I got a solution to that believe it or not and I am surprised that
noone does it. (or maybe they do)
take a nice 47 uF or so bipolar cap across the contacts. Period. think
aabout it no before to go saying anything.
OR a changeover relay wired to open the output link PLUS short the speakers
to ground under DC fault conditions.
Must have installed 50 of those in PL400s and PL700s years ago.
... Phil