Phase Linear 400 output transistors
 

On Sunday, June 1, 1997 3:00:00 AM UTC-4, 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.

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



If interested I have some XPL909 for sale

Phase Linear 400 output transistors
 

On 2/14/2014 5:23 AM, wrote:
On Sunday, June 1, 1997 3:00:00 AM UTC-4, 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.

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


If interested I have some XPL909 for sale



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:

http://www.vintageshifi.com/repertoi...ase-Linear.php

Phase Linear 400 output transistors
 

If interested I have some XPL909 for sale


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:

But, as the original posting and inquiry has a 1997 date on it, the
issue is probably moot (from the OP's point of view)/

Phase Linear 400 output transistors
 

Dammit ! Why are people regurgitating old posts now ? This is not hte first time this has happened. It has happened a few times lately.

Oh well, if anyone need advice on those thins I guess there it is.

Phase Linear 400 output transistors
 

Phase Linear 400 output transistors
 

Følgende er skrevet af Jon Elson:
wrote:

Dammit ! Why are people regurgitating old posts now ? This is not hte
first time this has happened. It has happened a few times lately.

Oh well, if anyone need advice on those thins I guess there it is.
It seems to be a problem with some news server that does this, then the
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!

These ghosts of christmas pasts almost always comes from somebody
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.

Phase Linear 400 output transistors
 

If interested I have some XPL909 for sale


** They the ones with inbuilt B-E resistor or is that the XPL910 ?

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

Phase Linear 400 output transistors
 

The B-E resistor really makes no difference since the amp has a 10 ohm across it anyway.

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.

Phase Linear 400 output transistors
 

The B-E resistor really makes no difference since the amp has a 10 ohm
across it anyway.

** It does if you mix up devices.


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.

** The relay contacts will arc and burn soon as they open under a DC fault.





..... Phil

Phase Linear 400 output transistors
 

"** 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. However you are right that it will matter because if you tray to push these things and the current isn't shared it will fry fast and the outputs will probably read zero all the way arond.

"** 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.

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 ****ing $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 ****ing 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 ****ing 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 **** figger dat out.

I see no downside to the caps. they are probably even cheaper than magnets. they are certainly cheaper to install.

Phase Linear 400 output transistors
 

"** 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.

** If you have 3 of one kind on one side, then 10 ohms drops to 7 and the
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.

** Far cheaper to use a triac, DC crowbar to protect speakers.

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