I'm working on a PL700b for a friend. As the title says, its completely original and currently working fine. My friend wants it gone through completely as well as having it converted from the original quasi complementary output stage to a full complementary output. Before getting started I've done a lot of research on this amp. I've found quite a bit of info on what to do in the way of upgrading it, but the conversion info is spotty.
My big concern is that I perform the quasi to full comp conversion properly. I've redrawn the original schematic for the PL20 control pcb to show the planned changes for the conversion. I'm hoping there is someone familiar with these amps that can tell me if I'm on the right track or if I've missed something.
I've attached the original and modified drawings. Any info would be appreciated.
My big concern is that I perform the quasi to full comp conversion properly. I've redrawn the original schematic for the PL20 control pcb to show the planned changes for the conversion. I'm hoping there is someone familiar with these amps that can tell me if I'm on the right track or if I've missed something.
I've attached the original and modified drawings. Any info would be appreciated.
Attachments
You should do the usual cap replacements etc. and have him use the amp for a while before even considering such a drastic circuit change.
You could get into big trouble going beyond this.
Last edited:
I started working on the control pcb already. C6 was toast in both channels and his low freq response was down about -9dB at 20 Hz in the L channel and about -6dB in the R. I planned to do most of the control pcb work, install the wattsabundant output protection board while and then use the amp as a quasi for a bit to see how it does.
I'm also installing a new set of main caps and a new bridge rectifier.
So the full comp conversion was going to be the last stage after I had completed basically all other work and verified there was no oscillation or any other problems present.
The thing is that this amp has the XPL910 outputs in one channel and Fairchild 909's in the other. My friend plans to use this to power his Acoustats so it will be working on a 4 ohm load. I don't expect it to be that reliable without some more modern output transistors in there.
Good approach. The comp conversion could change the stability enough that you might have real problems.
Just make sure that he understands this, and is willing to pay what it will cost for your time and the parts.
Replace C11 with a 100 volt low leakage type. Any leakage at at all and you'll get wierdness that can take days or weeks to diagnose because the symptoms won't make any sense.
Or return wire #13 on the board to the output (where it normally connects on most amps) instead of the base of the output transistor bank. This can increase crossover distortion in the normal quasi-comp under-biased class B mode. Converting to full comp mitigates this distortion and allows you to run pretty much any bias in the outputs without oscillations. The full comp conversion actually improves stability margin if you use the recommended 4-7 MHz drivers and outputs. Anything faster than the predrivers (15MHz) will cause oscillations, but they aren't as rugged anyway. Heat sinking is still a limiting factor on quiescent bias, but you can crack them open a bit (a few mA per device) in full comp mode and it improves things a lot over running the outputs in cutoff.
Or return wire #13 on the board to the output (where it normally connects on most amps) instead of the base of the output transistor bank. This can increase crossover distortion in the normal quasi-comp under-biased class B mode. Converting to full comp mitigates this distortion and allows you to run pretty much any bias in the outputs without oscillations. The full comp conversion actually improves stability margin if you use the recommended 4-7 MHz drivers and outputs. Anything faster than the predrivers (15MHz) will cause oscillations, but they aren't as rugged anyway. Heat sinking is still a limiting factor on quiescent bias, but you can crack them open a bit (a few mA per device) in full comp mode and it improves things a lot over running the outputs in cutoff.
wg_ski, that's good info thanks.
I am going to replace Q7/Q10 with some new 2n3440/2n5415 pairs that I have had in my parts drawer for about 15 years. I also plan to add base stopper resistors on them as well as the output transistors.
I am going to replace Q7/Q10 with some new 2n3440/2n5415 pairs that I have had in my parts drawer for about 15 years. I also plan to add base stopper resistors on them as well as the output transistors.
Hummmm .... Mixing old with new = trouble. These old classic amps were
designed and compensated with the old low ft devices as a major factor.
Like the old leach amp . I just started from scratch .... instead of trying to
"update" it.
PS - the old outputs are most likely MORE reliable (with this circuit).
OS
Take a look at the Phoenix Audio web site.A good deal of info is available there.Increasing C6 to 470uf and installing back to back diodes on the back of C6 as well as matching the input diff pair will improve stability.RCA 410 drivers with MJ21196 outputs seems to be a common improvement.I tried the full comp output stage and heard little or no improvement but it is your choice.Cheers.
The last official output transistors that the factory recommended in their service bulletins were MJ15024's. Those have the same 4mhz fT as the MJ21196, its just the MJ21196's have a much better SOA. Luckily I will not be breaking new ground in this regard as many others have used the MJ21195/96 in their PL amps with great success.
I welcome the comments though, thanks!
Re-wiring the output sockets and emitter resistors for full comp is a huge undertaking.
I would just go up to MJ21196, and be done with it.
I would just go up to MJ21196, and be done with it.
Yeah I know! Its going to be a lot of work. I've surveyed the interior of the amp quite a bit.
At least I don't have to build a chassis for this project. After my last few DIY amps where I built the chassis from the ground up, I finally decided that ONLY having to deal with the electronics/wiring for a project wasn't anything to complain about. I easily spent 30 - 40X as much time machining and putting the finish I wanted on my DIY amp cases as I did in getting the electronics and wiring completed.
Maybe I'm just a glutton for punishment.
I had seen where Steve Mantz (ZED audio) was rebuilding these amps from the ground up about 4 years ago. For about $400 the work included new output transistors, reworked control pcbs, new main caps, converted to full comp, etc. I was thinking that seemed like a bargain for the spaghetti of hard wiring that had to be dealt with in these old 70's amplifiers.
This brings up an interesting point. Its seems the designers, Mantz and White Oak push the full complementary conversion as being superior. While I've seen comments by actual users that they saw little sound quality benefit from the amps original Quasi output stage.
I'm wondering if the gains people have seen in the full comp conversion had more to do with the fact they also updated and upgraded most other parts as well, i.e. replacing 40+ yr old control pcb electrolytics, new larger main PS caps, fixing original wiring issues, etc.
I read rlr274's comments earlier, anybody have any opinions on this?
I'm wondering if the gains people have seen in the full comp conversion had more to do with the fact they also updated and upgraded most other parts as well, i.e. replacing 40+ yr old control pcb electrolytics, new larger main PS caps, fixing original wiring issues, etc.
I read rlr274's comments earlier, anybody have any opinions on this?
Chamberman,
I just finished converting a 400II quasi to full comp. I would not say it is "a huge undertaking", but it will require some patience and planning. The 400II conversion took me about 8 ~ 10 hours (Two nights after work). This was taking my time including a few interruptions.
I understand the 700 will require a bit more effort, but not too much more.
It's late now, but I'll post more info tomorrow to help you get started on the conversion. Converting the driver board for full comp is the easy part. It is simple and I will document it all for you.
I just finished converting a 400II quasi to full comp. I would not say it is "a huge undertaking", but it will require some patience and planning. The 400II conversion took me about 8 ~ 10 hours (Two nights after work). This was taking my time including a few interruptions.
I understand the 700 will require a bit more effort, but not too much more.
It's late now, but I'll post more info tomorrow to help you get started on the conversion. Converting the driver board for full comp is the easy part. It is simple and I will document it all for you.
djk,
Steve Mantz says the opposite. He says the IC front end needs to be replaced. Here's a link to the thread where he details the work ZED audio performs to the various versions of PL amps. Its in post #28 where he says the IC front end has to go.
Last edited:
Some people actually prefer the older discrete design over the later IC op-amp design of the Series II.
One thing is certain... you will get a lot of opinions and very little information. It's nice to see people like Mantz and White Oak Audio making the hobby fun.
The op amp front end is a bit 'edgy' because the compensation scheme is marginal for that much open loop gain. It needs some (more) degeneration in the VAS, local feedback from output to the 2nd stage, or both, to make it optimum. Do that and it will be better even with the original LF351. These mods are do-able, if you were so inclined.
Sometimes when you're working on these things it can be hard to separate out what's a local output stage oscillation and what's just a global loop problem - because you can get either one and the cures are different.
Sometimes when you're working on these things it can be hard to separate out what's a local output stage oscillation and what's just a global loop problem - because you can get either one and the cures are different.
Do you have more specific information (component values) regarding your proposed circuit modifications? Have you modified the 100pF and 120pF caps values and performed extensive testing to verify your claim of "better?