Thanks! I'll try that too!
Right now, I've got to replace two transistors. I blew up two.. don't ask. It was stupid.
Transistors replaced, still shows about 60 mv noise @ 1MHz, (maybe 50mv rms) ungrounded probe.
Is that a reasonable amount of 'noise'? I know that the F5 sports 60 uV noise, or 1000 time less, which would be just awesome to achieve.
Darn! I missed my chance at measuring the noise on an F5 that was at my place tonight!
Is that a reasonable amount of 'noise'? I know that the F5 sports 60 uV noise, or 1000 time less, which would be just awesome to achieve.
Darn! I missed my chance at measuring the noise on an F5 that was at my place tonight!
Hey Big E
Just don't expect too much from such an old amp. And btw how much power can you get from f5? Don't compare it with good old legendary speaker destroyer ,,Blaze Linear,,
I read somewhere that someone has nightmares about connecting Phase Linear on his Lowthers......I can imagine that......
Just don't expect too much from such an old amp. And btw how much power can you get from f5? Don't compare it with good old legendary speaker destroyer ,,Blaze Linear,,
I read somewhere that someone has nightmares about connecting Phase Linear on his Lowthers......I can imagine that......
Phase Linear oscillations
Four ways to get rid of oscillations completely.
First is insert 56-100 ohm base stoppers in the base leads of the first driver transistors (between VAS outputs either side of the Vbe multiplier) and the the first driver stage. This will remove almost all the oscillations. The input differential pair MUST have emitter degeneration resistors added (whether it is an early NPN diff design or the later PNP diff designs). A value of about 330 ohm works well. The open loop gain of this design is way to high (Carver was of course going for low THD numbers and then screwed it up by bad grounding). Phase Linears unmodified seldom show lower than 0.1% THD as the bad grounding techniques simply prevent the THD from being low.
Observe carefully where the take off point of the main series feedback resistor is taken, it comes from an point on the centre node where current is flowing. Simply taking this feedback point off from the speaker binding post for each channel will lower THD by an order of magnitude.
Phase Linears are at best poorly designed amplifiers and it is not very difficult (but is time consuming) to make them into very nice sounding amplifiers. Of course those wimpy supply caps must also go, we replace the PL400s with 18,000mfd per rail and the PL700s with 22,000 mfd per rail.
Second is rewire the grounding of the amplifier. As you see there is a strap between the RCA ground and the ground of ONE channel's speaker output.
The RCA grounds should only be connected to the PCB ground eylets and no other place. The main ground wire must go to the main ground buss between the main supply capacitors.
Third is replace those ugly PL909 output devices and if you have a quasi comp version replace these with MJ15024. Even better is convert the amplifier to a full complementary output nu this requires some extensive rewiring and a good knowledge of this amplifier.
Last is add 1-3 ohm base stoppers in EACH base leg of every output device.
These mods apply to both the PL400 and 700's.
We completely rebuild these for our customers and even offer a new PCB which has been designed with a full complementary design from input to output.
Those later PL400/700s with the LF351 opamp, I always remove this and replace with a differential amplifier pair with emitter degeneration.
Steve Mantz
Zed Audio Corp
Four ways to get rid of oscillations completely.
First is insert 56-100 ohm base stoppers in the base leads of the first driver transistors (between VAS outputs either side of the Vbe multiplier) and the the first driver stage. This will remove almost all the oscillations. The input differential pair MUST have emitter degeneration resistors added (whether it is an early NPN diff design or the later PNP diff designs). A value of about 330 ohm works well. The open loop gain of this design is way to high (Carver was of course going for low THD numbers and then screwed it up by bad grounding). Phase Linears unmodified seldom show lower than 0.1% THD as the bad grounding techniques simply prevent the THD from being low.
Observe carefully where the take off point of the main series feedback resistor is taken, it comes from an point on the centre node where current is flowing. Simply taking this feedback point off from the speaker binding post for each channel will lower THD by an order of magnitude.
Phase Linears are at best poorly designed amplifiers and it is not very difficult (but is time consuming) to make them into very nice sounding amplifiers. Of course those wimpy supply caps must also go, we replace the PL400s with 18,000mfd per rail and the PL700s with 22,000 mfd per rail.
Second is rewire the grounding of the amplifier. As you see there is a strap between the RCA ground and the ground of ONE channel's speaker output.
The RCA grounds should only be connected to the PCB ground eylets and no other place. The main ground wire must go to the main ground buss between the main supply capacitors.
Third is replace those ugly PL909 output devices and if you have a quasi comp version replace these with MJ15024. Even better is convert the amplifier to a full complementary output nu this requires some extensive rewiring and a good knowledge of this amplifier.
Last is add 1-3 ohm base stoppers in EACH base leg of every output device.
These mods apply to both the PL400 and 700's.
We completely rebuild these for our customers and even offer a new PCB which has been designed with a full complementary design from input to output.
Those later PL400/700s with the LF351 opamp, I always remove this and replace with a differential amplifier pair with emitter degeneration.
Steve Mantz
Zed Audio Corp
thanks for the tips!
Can you clarify please?
I see a 150 ohm base stopper on Q11 already. Do you replace this with 56-100 ohms?
Also, I see the first series resistor in the feedback loop, R15 (12K) is connected between the two protection circuit transistors (Q8 and Q9). Should this connection be moved directly to the speaker positive?
As currently designed there are two ground traces on the PCB.... one is the shield ground/input signal reference, the other the protection circuit reference. They are split by R2.
For grounding, I've disconnected the shield ground from the speaker negative and ran a new shield ground from between the RCA's directly to power common, between the caps.
Are you suggesting instead to create a ground bus, from RCA in, to PCB to speaker negative (through R2) to power common?
I tried this. There was a massive change to the sonics....much tamer highs.
But, there is additional resistance added to the input impedance of the first stage (Q1/Q2) via R2, which unbalances the differential pair. Approx 1 mv dc offset is added per 3 ohms at R2. So, to keep the current flow in the ground bus away from the inputs, I replaced R2 with 15 ohm resistors. The DC offset then would not stabilize -- it would always oscillate, +/- 3 mv. (I only had a DMM at that time, I now have a scope.)
Do you remove R2 instead and replace with jumpers?
Your posts always have great information! Thank you!
PHASE LINEAR 400 AND 700 MODIFICATIONS
There are several versions of the PL400 drive board.
a) Early versions with PNP differential amplifiers, VAS driver and triplet output stage
b) The next version used an NPN differential amplifier, an intermediate PNP pair, VAS driver and triplet output stage.
c) The last version used an LF351 op-amp, voltage translator, VAS and COMPLEMENTARY triplet output stage.
* All three versions suffered from oscillation and turn on/off thump issues as well as reliability due to those nasty PL909 output devices.
* Grounding of the various elements in the amplifiers is extremely poor as was the routing of the main plus and minus rail supplies to the output stages.
* When I re do a PL400 I always replace the following mechanical parts: BINDING POSTS, RCA SOCKETS.
* I always replace the wimpy 5900mfd main supply capacitors with 18,000mfd 100v parts.
* I remove the feedback take off points for each channel (Grey wire) and take them directly to the binding posts. This together with the new grounding scheme reduces THD by a huge order of magnitude.
OK now to the individual modifications for the various PCB types
PNP differential version PL400C:
a) Add emitter degeneration resistors to the PNP differential transistors Q1 and Q2. I use 340 ohm 1% 1/8w parts and I replace these with high gain low noise parts which we use in production.
b) Rewire the bias potentiometer so that it is simply in series with R15 (1K ohm) between the base and emitter of Q3.
c) Replace C8 with a low ESR 470mfd 16 or 25v part and parallel with a 0.1mfd film capacitor.
d) Add a 22pF 250v disc ceramic between collector and base of Q4.
e) Replace C6 with a low ESR part.
f) Replace R52 with a 6K2 1w resistor
g) Replace R18 with a 330 ohm 1/4w part.
h) Replace those nasty germanium VI transistors with high gain medium current silicon parts. These are Q5 and Q6. I use KTC1027 and KTA1023 again we use these in production.
i) Insert base stopper resistor in the base leads of Q8 and Q7. value not critical at 56 to 82 ohms. The PCB traces must be cut to insert these. I use 1/8w parts as they are small and easy to solder to the skinny PCB traces.
j) Rewire the feedback take off points as I describe above.
k) Rewire the amplifier’s grounding. Remove the small wire link that is between the RCA ground and the speaker ground of the right channel. Make sure that there is only a single ground wire from the PCB ground to the main capacitor ground strap. This wire MUST be soldered to the star point on this copper strap. The speaker return grounds are fine. The RCA grounds run to each channel on the PCB and are isolated with the 56R and 2R7 resistors. I put back to back 1N4004 diodes across one of these resistors.
l) Bypass the two 33v zener diodes with 4.7mfd 50v parts.
m) Please replace those PL909s with MJ15024 or MJ21194 and add in base stopper resistors in each base lead. These can be 1.5 ohm to 2.7 ohm 1/2w parts. You have to remove the base buss wire and solder on the base stoppers and then make a new base buss from the driver transistors.
n) Remove the + and – 80v high current buss wires from the right channel and solder in new longer wires and take them directly to the fuse holders. The amplifiers are wired with the right channel sharing the high current feed to the left channel. I actually prefer and do remove all the +/-80v wiring, bypass the fuse holders and take each channel directly to the main buss capacitors. This removes ANY shared wiring between channels.
o) Remove capacitor C9 (180pF) from the PCB in two places.
For those a little more experienced the addition of pure constant current sources for the differential and VAS stages will improve the performance dramatically.
NPN Differential amplifier PL14A and PL14B
The same modifications should be performed as described above except that the differential amplifier is NPN. Again I change those ugly MPS5172 to low noise high gain types.
The bias pot does not need to be rewired as it is now correct on these versions.
The input infrasonic filter needs to be changed on all versions.
The first type has a 2K7 resistor at R1. Change to 1K.
Change R13 the down feedback resistor to 453 ohm 1%
The second type has R1 now at 10K, change this to 1K.
Change R13 to 620 ohm
The last PCB type on PL400 uses the LF351 in the front end and this in my opinion is just a disaster. I always change this to a PNP differential amplifier with a constant current source driving the tails of the diff pair. There is simply NO good cure for this bad op-amp. Replacing with one of the high end “OPA” types will not improve matters at all.
Phase Linear 700
All versions of this used the NPN diff pair with PNP diff pair, then VAS and finally drivers. The last iteration used the same LF351 opamp.
These I change as described above and they also used the same dumb infrasonic filter as the later 400’s. If you see R1 as a 10K, change to 1K and change the down feedback resistor to 620R.
PL700s are wired the same as PL400s except of course that each channel has its own DC fuses, BUT these fuses for each rail are fed by a common wire from the main capacitors.
Run separate wires to each fuse holder from the main capacitors.
I replace these 10,000mfd with 27,000mfd 100v as well.
Changing PL400/700 from a quasi to a full complementary output stage.
This is of course always my preference but it requires again knowledge of circuit design and requires almost a full rewiring of the output stages. Do not forget that doing this requires the replacement of all the output and TO-3 driver transistors.
The bottom PNP driver on the PCB now becomes and emitter follower instead of a phase inverter. The Baxandall diode and associated parts are removed as the output drive is now taken from the emitter of this device.
These Phase Linears can be made into very nice amplifiers with some work.
We charge $350 to completely refurbish, rewire, convert to full complementary, new main 18,000mfd caps, new binding posts and new Tiffany RCAs for PL400 and $400 for PL700.
Hope that the above helps.
Steve Mantz
Zed Audio
There are several versions of the PL400 drive board.
a) Early versions with PNP differential amplifiers, VAS driver and triplet output stage
b) The next version used an NPN differential amplifier, an intermediate PNP pair, VAS driver and triplet output stage.
c) The last version used an LF351 op-amp, voltage translator, VAS and COMPLEMENTARY triplet output stage.
* All three versions suffered from oscillation and turn on/off thump issues as well as reliability due to those nasty PL909 output devices.
* Grounding of the various elements in the amplifiers is extremely poor as was the routing of the main plus and minus rail supplies to the output stages.
* When I re do a PL400 I always replace the following mechanical parts: BINDING POSTS, RCA SOCKETS.
* I always replace the wimpy 5900mfd main supply capacitors with 18,000mfd 100v parts.
* I remove the feedback take off points for each channel (Grey wire) and take them directly to the binding posts. This together with the new grounding scheme reduces THD by a huge order of magnitude.
OK now to the individual modifications for the various PCB types
PNP differential version PL400C:
a) Add emitter degeneration resistors to the PNP differential transistors Q1 and Q2. I use 340 ohm 1% 1/8w parts and I replace these with high gain low noise parts which we use in production.
b) Rewire the bias potentiometer so that it is simply in series with R15 (1K ohm) between the base and emitter of Q3.
c) Replace C8 with a low ESR 470mfd 16 or 25v part and parallel with a 0.1mfd film capacitor.
d) Add a 22pF 250v disc ceramic between collector and base of Q4.
e) Replace C6 with a low ESR part.
f) Replace R52 with a 6K2 1w resistor
g) Replace R18 with a 330 ohm 1/4w part.
h) Replace those nasty germanium VI transistors with high gain medium current silicon parts. These are Q5 and Q6. I use KTC1027 and KTA1023 again we use these in production.
i) Insert base stopper resistor in the base leads of Q8 and Q7. value not critical at 56 to 82 ohms. The PCB traces must be cut to insert these. I use 1/8w parts as they are small and easy to solder to the skinny PCB traces.
j) Rewire the feedback take off points as I describe above.
k) Rewire the amplifier’s grounding. Remove the small wire link that is between the RCA ground and the speaker ground of the right channel. Make sure that there is only a single ground wire from the PCB ground to the main capacitor ground strap. This wire MUST be soldered to the star point on this copper strap. The speaker return grounds are fine. The RCA grounds run to each channel on the PCB and are isolated with the 56R and 2R7 resistors. I put back to back 1N4004 diodes across one of these resistors.
l) Bypass the two 33v zener diodes with 4.7mfd 50v parts.
m) Please replace those PL909s with MJ15024 or MJ21194 and add in base stopper resistors in each base lead. These can be 1.5 ohm to 2.7 ohm 1/2w parts. You have to remove the base buss wire and solder on the base stoppers and then make a new base buss from the driver transistors.
n) Remove the + and – 80v high current buss wires from the right channel and solder in new longer wires and take them directly to the fuse holders. The amplifiers are wired with the right channel sharing the high current feed to the left channel. I actually prefer and do remove all the +/-80v wiring, bypass the fuse holders and take each channel directly to the main buss capacitors. This removes ANY shared wiring between channels.
o) Remove capacitor C9 (180pF) from the PCB in two places.
For those a little more experienced the addition of pure constant current sources for the differential and VAS stages will improve the performance dramatically.
NPN Differential amplifier PL14A and PL14B
The same modifications should be performed as described above except that the differential amplifier is NPN. Again I change those ugly MPS5172 to low noise high gain types.
The bias pot does not need to be rewired as it is now correct on these versions.
The input infrasonic filter needs to be changed on all versions.
The first type has a 2K7 resistor at R1. Change to 1K.
Change R13 the down feedback resistor to 453 ohm 1%
The second type has R1 now at 10K, change this to 1K.
Change R13 to 620 ohm
The last PCB type on PL400 uses the LF351 in the front end and this in my opinion is just a disaster. I always change this to a PNP differential amplifier with a constant current source driving the tails of the diff pair. There is simply NO good cure for this bad op-amp. Replacing with one of the high end “OPA” types will not improve matters at all.
Phase Linear 700
All versions of this used the NPN diff pair with PNP diff pair, then VAS and finally drivers. The last iteration used the same LF351 opamp.
These I change as described above and they also used the same dumb infrasonic filter as the later 400’s. If you see R1 as a 10K, change to 1K and change the down feedback resistor to 620R.
PL700s are wired the same as PL400s except of course that each channel has its own DC fuses, BUT these fuses for each rail are fed by a common wire from the main capacitors.
Run separate wires to each fuse holder from the main capacitors.
I replace these 10,000mfd with 27,000mfd 100v as well.
Changing PL400/700 from a quasi to a full complementary output stage.
This is of course always my preference but it requires again knowledge of circuit design and requires almost a full rewiring of the output stages. Do not forget that doing this requires the replacement of all the output and TO-3 driver transistors.
The bottom PNP driver on the PCB now becomes and emitter follower instead of a phase inverter. The Baxandall diode and associated parts are removed as the output drive is now taken from the emitter of this device.
These Phase Linears can be made into very nice amplifiers with some work.
We charge $350 to completely refurbish, rewire, convert to full complementary, new main 18,000mfd caps, new binding posts and new Tiffany RCAs for PL400 and $400 for PL700.
Hope that the above helps.
Steve Mantz
Zed Audio
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.