Phase Linear 400 Series II Driver and Output Transistors

spymac

New Member
I picked up a Phase Linear 400 Series II in unknown state of condition. I have some experience with electronics repair but not much experience with amplifiers in particular. I took it apart and found one of the main filter caps open. I have ordered replacements for both of these caps and also all of the other electrolytics in the amp. So far I have not found any other problems in the amp.

I noticed that this is the older PL400 Series II, without the complimentary output transistor banks. Everything is NPN. I also noticed that the left channel uses 2x RCA 66546 driver and 6x NEC 2SD555 output, and the right channel uses 2x RCA 66546 driver and 6x Motorola MJ15011. Everything tests good, no shorts or opens. But I am thinking more long-term about replacing so that both channels are the same.

After googling a lot, I could find no clear answer as to what modern transistors work the best. I saw plenty of people that don't like the 2SD555. It seems that people tend to like to use MJ15024 for the drivers and MJ21194 or MJ21196 for the outputs. I am a little bit confused about this though. When I look at the datasheets for these parts, they all seem pretty much identical. Why wouldn't I want to use the same transistor for all 16 of them, drivers, outputs, everything? For example, a quote from d3imlay about the PL400:

When I do a repair for someone I replace all outputs and drivers with 21196's as a matter of convenience and long term reliability.

What problems or disadvantage might I face if I used all MJ15024 for example? Since the driver and output transistors would be the same, wouldn't this make oscillation unlikely?

Also, another question more out of curiosity. Would there be any advantage to rewire this for complimentary outputs? I remember one person somewhere saying that they had rewired for complimentary, but they did not explain the reason why or mention any improvements from doing it.
 
Damn, I can't find my data update on this amplifier but if I remember, the 555 transistor can be used and also the Motorola MJ's but if I recall, you should have one or the other in both channels. Mine finally blew up so if you need any parts................I still have plenty of power and amps and systems so no big loss to me. :D
 
Not sure the PL400 is a good starting point. Suggest the following
reading and sites, (and maybe starting with a NAD)
http://www.whiteoakaudio.com/
http://synthetizer-sche.chez-alice.fr/power ampli/phase linear/repair.htm
http://forums.phxaudiotape.com/

My take is that in it's original form, the PL400 is somewhat fragile. There
is a race condition which may put rails voltage onto the speaker lines. The
amp does not have spkr protection circuitry (massive oversight). The boards
don't lend themselves to rework/tracks, eyes damage easily.

Consider the white oak upgrade, this may be within your skill level, though
it will cost $$$$'s.

I have 3x PL400's waiting, still undecided what's best, will white oak one.

No affiliation to any above.
 
Thanks century tek! I have a good feeling this thing will work after I replace the main filter caps. If so, I might try to bribe you for the volume knobs if you have a series II. :)

mbz: The PL400 series II has a special charm for me, my dad had one when I was growing up. The red LED bargraph displays and pops it makes after you remove power would hold value for me, etc. I will give it a try, what the heck. Thanks especially for your second link. I will refer to this extensively. I do have a variac, scope, and function generator that I was planning to use for this.
 
2sd555 are very good transistors (originals fetch a pretty penny, and are used in a lot of TOTL amps), as long as they are not counterfeits, but mj21194 or 21196 are better (IMO) Pioneer switched to them (2SD555 and 2SB600)when they took over Phase Linear. The later versions of the 400 series 2 were full complimentary. My opinion is to stick with the version you have, unless it's really messed up, and upgrade to the white oak boards. It's not a huge price (around $50) to change all the outputs to one type.YMMV
 
Sounds like someone rebuilt one channel at some point in the past. Not sure which would have been original on this one, PL changed what they used over the years.

You might want to check with the guys over at Phoenix Audio Tape about these. Thats the hub of Phase Linear and White Oak conversion information.
 
Okay, thanks for pointing me to the White Oak conversion. For now, I think I would like to get it working as-is and think about that as a future upgrade. I found several ceramic disc caps that were pretty far off! So I replaced all of the ceramic discs. I replaced the 10n and 50n ceramic caps with film caps. I guess I don't really know the implications of making this change, hopefully it is okay.

I have a question that is more of a general question. I see that there is a snubber directly on each output. In this case, it is 6 ohms in series with 100nF. What is the purpose of this? Is it preventing something from getting out of the amp, something getting into the amp, or is it some sort of stability thing? Mine were off by a lot, not sure what might have happened if I left them alone.

I measured 8 ohms and 9 ohms instead of 6 ohms (as indicated by the color bands; these are not shown in the schematic). The 100nF mylar caps were spot-on. I replaced with new resistors and went ahead and replaced the mylar caps with some nice 250VAC X-caps.

The amp came with all correct amp rating fuses but all 32V... I got 250V ones coming and also some 2A fuses for initial bring-up. Still waiting on the big electrolytics and some 47uF 100V.
 
Its a Zoebel network. Basically its to make sure the amp doesn't oscillate or output any high frequency garbage that might damage a speaker. Its original, or at least the two I've been into and the pics I've seen of others all had such a network.

The components used on that board were not the best of quality. When I rebuilt my friend's 700-II it had a pile of resistors that were way off value. Mine had several as well.
 
I think I would like to do the White Oak upgrades in the future. For now, I would like to get it running original, especially so that I can get used to it and then appreciate the difference when I am ready to spend the money.

I have seen mentioned at a few places, including here and Phoenix Audio tape, that the latching issue can be prevented on the PL400 series I by putting back-to-back diodes on C6 on each channel. For example:

If you don't go the new White Oak board route there are some tricks we can do with a few component changes to do away with a couple of their bad habits. Changing C6 to 470uf at 50 volts will flatten the lower frequency response considerably. Soldering two 1N4148's back to back on the back of the board to the leads of C6 will prevent random latch up of the positive rail voltage to the speaker outs. So if you change C6 leave about 3/16's of lead sticking out.

On a PL400 series II (with PL36 board), can anyone confirm that the same changes would apply, except with C111 / C211 instead of C6?
 
I am looking at getting some UPC1237's and building something like that for my amp, though I am aware that will take away my nostalgic turn off pops. I definitely need to do that in case I have an output transistor short. For the latch condition though, I would prefer to try to prevent the latching in addition to reacting to it. Though the 1237 approach would save the speakers, I wonder if the amp would still blow up, sitting there latched as I scratch my head wondering why I have no sound.

In fact I have access to 2 more of these PL400 series II (that aren't mine) and I would feel good to throw in a few cheap caps and diodes to help out there as well.

It looks like C111 / C211 have the same function as C6, just wondering if anyone could confirm the diode fix could apply here.
 
Update: I have received all of the caps that I ordered and replaced everything. I got a little creative with the main filter caps, it ended up costing me $10 to replace them both. I actually used common 35mm snap in caps to replace each of them. I bought 2x 6800uF 100V (thus the $10) and used 2x scrap-recovered 820uF 450V. I cut a PVC pipe the correct length and then one side of it lengthwise with a bandsaw. Then I just spread it open and slid in a cap from each end with the snap in prongs sticking out each end. Then I used some solid 14 awg wire to connect the caps together from end to end. I soldered ring tongues onto the 820uF end to align with original locations in the amp. Between the 2 ring tongues, I threw in a 1uF 450V film cap for good measure. I am very happy with the result, it worked out nicely.

I followed the instructions at http://synthetizer-sche.chez-alice.fr/power ampli/phase linear/repair.htm to bring it up slowly, and I am happy to report it is working great! I had symmetrical clipping at the couple of spots where I checked. I was lucky enough to use a Chroma AC source instead of a variac. I set a 0.5A current limit the first time bringing it up with 1kHz sine wave input and no load. Everything looked great. I used 8-ohm loads with a 2A input current limit the second time bringing it up. After this testing, I tried turning on and off directly into 110VAC. It passed that test but I had to turn the current limit way up on the AC source for the gigantic inrush. :)

Also I used a thermal camera to monitor anything getting hot. The R1/R2 input op-amp supply resistors got up to about 100 deg C, equally hot as each other, which is probably normal I guess. It doesn't make me too happy to see that, it seems kind of silly. I also noticed that the op-amp on the left side was warmer than the one on the right, by maybe about 10 deg C. Not sure why they would not be basically identical with outputs looking identical.

I have some TL071CP's, would it be worth trying them to see if the temperatures are more equal? Or should I not be concerned about that? I would not have been aware of it in any other way than the thermal camera.

I noticed my right display only goes down to the second dot with no input, occasionally flickering to the first dot but not often. The output has about 10mV of noise on each channel in this condition. Not affected when I add 8-ohm load or turn the pots on the front panel. I did replace all caps in the displays in the past weeks, including the main filter cap. The display does work normally other than this. I ordered some 4558 op-amps, will replace and see what happens. I found some knockoff UAA170's on aliexpress, 5pcs for $7. I am doubting that is the problem but definitely my dad has a PL400 with 2 bad UAA170s (left display is all glitched up on his amp), so I will report how that goes eventually.
 
On a PL400 series II (with PL36 board), can anyone confirm that the same changes would apply,

I don't *think* they do, but Lee (Laatsch55) could probably advise better than I. I seem to remember asking the same question when I was working on the one my friend blew up and I believe it was only a problem on the earlier board.

Is R1/R2 the large resistor at the left and the center of the board? That one gets toasty. Standing it off the board for better circulation would be a good idea if its not already that way. I seem to remember the board on mine having dark spots from that getting hot.
 
Thanks gadget73. I made a post about the caps and diodes at Phoenix Audio Tape:

http://forums.phxaudiotape.com/show...event-latch-up-how-about-on-Series-II-(PL-36)

Based on the reply from Gepetto, I did go ahead and make these changes. I used some bipolar 470uF 25V caps from Digikey and some 1N914A diodes that I had in my parts collection. Though I didn't test the amp at low frequencies, I did not have any problems at 1kHz or 2 kHz. It is a little bit unclear exactly what is going on to a software guy like me but it looks like the PL36 uses an input op-amp instead of discrete components used on the original PL400, and that is why the circuit looks a little different around C111 / C211.

I think R1/R2 are just designed to get hot on the PL36. All 3 of the PL36 I have seen are dark around these resistors and also the zener diodes connected to them. I replaced the zeners and at least stood them out from the board a little bit. The old zeners still tested just fine though.
 
Adding some photos of my work:

Replaced main filter caps using PVC pipe and 35mm caps (only $10!):
IMG_1486.JPG IMG_1487.JPG

PL-36 board with all new caps:
IMG_1488.JPG IMG_1489.JPG

Back of PL-36 board showing diodes and extra bypass caps I added to the op-amp rails:
IMG_1491[1].JPG
 
Last edited:
I don't *think* they do, but Lee (Laatsch55) could probably advise better than I. I seem to remember asking the same question when I was working on the one my friend blew up and I believe it was only a problem on the earlier board.
Okay, interesting. The responses I got at phxaudiotape were not particularly clear I guess. I wonder if it hurts anything for me to leave these on there?? Maybe I should take them back off, now I am not sure. :idea:
 
Read the 400II tech manual cover to cover. I posted it on hifi engine and at watts abundant. There are a number of updates on 400 II's in the manual. Also, the caps on display boards frequently fail.

Although putting the main DC caps in PVC pipe appears to be an elegant solution, it doesn't allow for any cooling which could lead to early/catastrophic failure.

I learned the hard way that some amps oscillate when the negative driver and outputs were the same. Keep the RCA drivers. The 2SD555's were great, but are long obsolete. MJ21196's are the preferred NPN devices.
 
Last edited:
Thanks d3imlay. I have actually read that entire service manual already, and also the service manual for the 700II. I used the manual somewhat when bringing up the amp for the fist time as well. AND, I also did replace all of the electrolytics in both displays before bringing it up. Some of them were a little crusty looking. I tested all of the old ones and I didn't find any that tested especially too far off from the marked capacitance. I will definitely let you guys know what the solution is on the second-dot thing. First I will take the easy approach and replace the op-amp. If it still does it, I will start getting the scope involved and tracing through.

I honestly never saw anything that said that the 400 series II could latch. Maybe it can't actually do it because of its input op-amp. It is not in the service manual and not particularly anywhere online. The only modifications I have made to this amp so far are replacing C111 / C211 with 470uF (which might still be a good thing) and then putting the back-to-back diodes (which might not be necessary and/or could cause a problem for all I know). I have not yet touched either of my dad's 2 400IIs. Mine is the test subject, then I ill apply to his once I have everything good. Plus he is getting a white oak upgrade in one of his.

I also am building a speaker protection board similar to yours d3imlay. I have all of the parts assembled on a board except I am still waiting for the UPC1237s in the mail. I will test it and post about it when it is complete, and hopefully I can get more of your advice if you see anything strange with the way I did it. I have built the Zobel networks into this board. When it opens the relays, it leaves the Zobel networks connected to the amp and not to the speakers.
 
Although putting the main DC caps in PVC pipe appears to be an elegant solution, it doesn't allow for any cooling which could lead to early/catastrophic failure.
Sorry I missed this d3imlay. I was a little worried about this, I completely agree that it should be considered.

Though this isn't ideal, PVC isn't that great of a thermal insulator. k = 0.2 W/(m*K), which is about the same as Nitrogen / air. However, as the PVC heats up, it doesn't float away like warm air would. So that is probably not a very fair comparison. Mica is about 0.7.

I could also argue that the PVC pipe has a greater surface area than the bare caps, so it may actually increase the heat dissipation. I think that in cases where caps are potted, they may have a similar effect where the thermal mass and surface area actually go up even though the cap seems sort of "insulated" inside of the plastic potting.

So I guess I am not really sold on either idea that it is okay to do this or bad to do this. I may have to do an extended run with a thermal camera and watch the temperature of the ends of the caps to see if it is making a big difference. Has anyone ever measured the temperature of their caps over time while working the amp hard before? Do the caps get hot on their own?
 
Last edited:
Back
Top Bottom