Phase Linear 2000 Pre-Amp Rebuild

[punchback]

I’ve recently acquired the subject PL2000 preamp. I’m going to go through it and rebuild it as follows:

1. Replace Electrolytics.
2. Replace CF resistors with 1% Metal Film.
3. Replace op-amps with modern op-amps.
4. Install regulated +/- 15 power supply.

For the Electrolytics it will be the standard stuff. The 2 in the power supply will be replaced by Nichicon UPW’s. There are a few that appear to low leakage and they will be replaced by Nichicon UKL’s. The rest will be replaced by UFG’s. Some of them are axial so I hope the radial Nichicon leads will reach.

I am replacing the carbon film resistors with metal films. One question I have is what is the wattage rating? In the service manual on two different pages it says ½ watt then ¼ watt. It seems that I read in another post @Gepetto said (I’m paraphrasing) that Phase Linear spec’d ½ watt but it wasn’t really necessary and ¼ watt would do. Hopefully that’s the case as I already bought the ¼ watt resistors. If it isn’t… oh well it’ll be yet another Mouser order.

I’ve read where @laatsch55 said that he’d attempted to replace the RC4739 with a few different modern op-amps but was met with oscillations. Despite this information I’m going to venture down that road also. I was wondering if both the RC4136 and RC4739 were replaced. The RC4136 has all signals routed through it and there’s one RC4739 is for the rear channel Ambience circuit and another for the phono preamp/RIAA circuit. Does the oscillation manifest itself with just power applied to the PL2000 or do you need to connect it to the amp? Is the oscillation measured at the outputs?? One of the op-amps (OPA2277) I selected has a very similar gain and slew rate as the RC4136 and RC4739. I purchased some NE5532D, TL072, and OPA2277 to try out.

I’m going to shoehorn an adjustable regulated +/- 15 power supply that will take the AC output from the transformer and create +/- 15 VDC. The unregulated power supply is +/-18 VDC, will the +/-15 be OK? Is it just powering the 3 opamps on the pcb? Will it be a problem to leave in the 470 uF capacitors from the stock unregulated supply? I’m planning on desoldering the transformer leads from pcb and connect them to the new power supply. I’m not sure exactly where I’m going to land the output from the new power supply. The output from the new power supply could just land on the +, -, and ground pads of C40/41 of the unregulated PS output.

Alan

 

[BlueCrab]

Very ambitious. My recommendation is to start simply, make one set of changes, such as power supply electrolytics, and then retest that it still works, but I'm sure you're doing that. With regard to resistor wattage, it's impossible to make a blanket statement that 1/4 watt are sufficient - it has to be determined on a case by case basis. I like to have things run cooler, not hotter, so I tend to go up in wattage if I have to. In pre-amps there is generally little power being dissipated, so 1/4 watt or even smaller are often used, but I'd go with what the service manual states and if there is a discrepancy in that, just go with what's on the board now or the larger of the two.

I assume you've checked that the op amp substitution that you want to do have the same pinouts as the current ones, right? I'd go with sockets, so you can make the changes easily. I suspect the oscillations were there whether the preamp was connected to an amp or not. I would recommend that you do most, if not all, testing without an amp at all until you are sure there are no spurious oscillations that could risk some speakers.

Why an adjustable power supply? I can understand regulated, but I'm not sure what adjustable buys you. You should be able to determine by looking at the schematic what the +/-18vdc is powering. I think you need to determine where the physical placement of you're new parts are going to be.

Hope this helps. Good luck. Keep us posted of your progress.

 

[62vauxhall]

It's not a tecnical reason but since current 1/4 watt resistors are physically much smaller then their predecessors, so are their leads. And that bothers me, the wispiness of the wires by comparisson. I've needed to replace 1/4 watt resistors and have always gone up a notch just because thickness of their leads is closer to that of the old 1/4 watt resistors I'm replacing.

 

[mlucitt]

I did the PL2000 upgrade several years ago. I replaced the electrolytics with Muse for the radials and regular Nichicon for the axials moving up in voltage and capacitance when the vintage values were no longer available. Make sure you add the polarization marks for the electrolytics. I checked the signal path resistors at random and all were within spec, so I left them and spent my rebuild budget on dual OPAMPs like the OPA2134 or the LM4562 and the 060601 Brown Dog adapter to replace the original RC4136, which is still made today. The RC4739 dual OPAMP chips in the preamp were obsolete, so I replaced them with two more LM4562 device and the Brown Dog 180801 adapter for the different pinout.
Then I re-flowed all the solder connections and scraped off all the old solder flux rosin. Cleaning the controls with DeoxIT also helped to make the preamp quieter. I never considered fooling with the power supply because it seemed to work just fine.

 

[wattsabundant]

!/4 watt resistors are fine. None of them dissipate any real heat. I would go with 12 volt regulators. The reason is the regulator needs 2-3 volts more on the input than output. If the line voltage drops and the input is only say 16 volts the regulator may be unstable. The opamps will work fine at +/- 12VDC. Make sure to use a .1 uf ceramic decoupling cap on the output as well as a small electrolytic. Also, the opamps should have decoupling caps on their power supply pins which will help with oscillations.

For the power supply I'd keep the transformer, rectifier and 470uf caps and tap off of the caps to go to the regulator. That will allow a very small perfboard. Use 7812/7912 regulators in TO92 packages to reduce the size of the board. You'll also want a 10uf or so cap on the input to the regulators. This assuming you have wires from the rectifier and the 470uf caps to the regulator board. Keep wires as short as reasonable possible.

I used these techniques on my 3300 preamp. The 3300 also isolates the power supplies for different opamps through a cascading R/C circuit. That would be desirable here but not likely practical as traces on the main board would need to be cut.

I've attached the 3300 manual

 

[punchback]

I will be doing one thing at a time. Probably start with the Electrolytics. I'm glad to hear someone say that 1/4 watt is fine. I didn't think there would be a need for 1/2 watt resistors. I had already ordered the 1/4 watters by the time I noticed the service manual also said that the resistors should be 1/2 watt.

Im hoping the AC output from the transformer will be enough to keep the new power supply in regulation. It's adjustable so I can dial it into a voltage that stays in regulation.

 

[wattsabundant]

There is no magic to determine the power dissipation of a resistor. It's the voltage across it, squared, divided by the resistance. Easier to say then put in text here. The only place that would be a concern in most solidstate preamps is the power supply.

The spec for AC input on the 2000 only says 117 Volts. Unlike the 3300 there is no range. The 3300 is 90-132 VAC. Power supply regulation is typically tested at +/-10 percent of nominal AC input. So if 117 VAC is the standard, then the power supply should regulate down to 105 VAC and shouold be able to tolerate 128 VAC. In residential use it would probably be acceptable to regulate +/- 5%. If a variac is available the power supply should regulate down at least 110 VAC.

 

[mlucitt]

In reply to both previous posts: Bob Carver figured out long ago that a regulated power supply can give no more and no less that the regulators will allow. However, an unregulated power supply will deliver more voltage (at the expense of some current) when necessary. This is much more important and apparent in an amplifier, but I think it holds for a pre-amplifier as well. During an especially dynamic, low frequency transient will the voltage demand from the OPAMPs not increase? If so, and this can be measured on the o'scope, a regulated power supply will not provide this voltage and distortion will be the result.

Maybe others can chime in.

 

[marcok]

I suggest to connect a 16 A relay between power switch and AC switched outlet to save the
power switch .
Ciao
Marco

 

[wattsabundant]

In reply to both previous posts: Bob Carver figured out long ago that a regulated power supply can give no more and no less that the regulators will allow. However, an unregulated power supply will deliver more voltage (at the expense of some current) when necessary. This is much more important and apparent in an amplifier, but I think it holds for a pre-amplifier as well. During an especially dynamic, low frequency transient will the voltage demand from the OPAMPs not increase? If so, and this can be measured on the o'scope, a regulated power supply will not provide this voltage and distortion will be the result.

Maybe others can chime in.

In a power amplifier, a large signal could put a high power (volts and amps) demand on the the amplifier power supply and approach clipping. The caps help regulate the DC voltage. A linear regulator would by nature lower the DC voltage to the amp in order to regulate, thus reducing the ouput power of the amp. So it's quite debatable that a regulated power supply serves any purpose in the output stage. The White Oak PL20 board uses a well regulated power supply and decoupling caps for the front end opamp and the following differential amplifier.

In a preamp the output voltage it sends to the power amp will never exceed a couple volts or so and the power demand on the power supply is milliwatts. A regulated power supply is useful in a preamp to maintain a low noise floor and keep the various stages from affecting each other. Assuming opamps are used, it is common to see .1uf dceoupling caps on the power supply pins of each opamp to further minimize noise and the possibility of oscillations.

 

[mlucitt]

You are going to have to show me how regulating the preamplifier main power supply by using a linear (or adjustable) voltage regulator which automatically adjusts the resistance via a feedback loop, accounting for changes in both load and input, while keeping the output voltage constant, "maintains a low noise floor" better than a straight piece of wire (no regulator).
I agree with the use of decoupling caps mounted close to the power pins of the OPAMPs, but the subject is more about: is it better to maintain a constant voltage (regulate), or allow the main power supply to react to the demands of the downstream circuitry (no regulator)?
This also assumes there is adequate filtering in the main power supply and any other devices/circuitry to minimize ripple voltage.

I understand the WOA Control Board uses precision DC regulators for the OPAMPs and other circuitry. My understanding is that was done to stabilize the DC voltage supply on the Control Board when there are heavy transients caused by dynamic power excursions in the amplifier, a preamplifier would not have those heavy transients.

This is my understanding, but I am open to other viewpoints.

 

[marcok]

The best solution is :
LM 7812 , LM 7912 and two 220 uF 35 V caps at the output of these ICs , considering that you can use at the input the 470uF caps already
existing .
Consider also that at that time this kind of ICs didn't exist and that now billion units were sold .
Ciao
Marco

 

[BlueCrab]

This may shed some light: https://www.analog.com/en/technical...oise-measurements-for-linear-regulators.html#

 

[punchback]

All of the amplification than happens in the 2000 is by either the RC4136 or the RC4739. I know just enough to be dangerous but I believe opamps are limited by the power supply rails. In my case the 2000 with its unregulated supply is making +/- 18.85 volts which is more than the RC4136 and RC4739 are rated for. Maybe that's because of current day AC voltage being a little higher than in the 70's. It seems that most modern opamps have a recommended power supply voltage +/- 15 VDC or lower. Most have a max power supply of 36 VDC which I believe is the span of the two supplies. In the case of my 2000 that would be 37.7 VDC. I wouldn't think that running the opamps at their max supply would be the best practice. The NE5532 has a max supply 44 VDC but also has a recommended supply of +/-15. Assuming I can use the modern opamp it seems to make sense to run them at +/-15.

 

[Gepetto]

All of the amplification than happens in the 2000 is by either the RC4136 or the RC4739. I know just enough to be dangerous but I believe opamps are limited by the power supply rails. In my case the 2000 is making +/- 18.85 volts which is more than the RC4136 and RC4739 are rated for. Maybe that's because of current day AC voltage being a little higher than in the 70's. It seems that most modern opamps have a recommended power supply voltage +/- 15 VDC or lower. Most have a max power supply of 36 VDC which I believe is the span of the two supplies. In the case of my 2000 that would be 37.7 VDC. I wouldn't think that running the opamps at their max supply would be the best practice. The NE5532 has a max supply 44 VDC but also has a recommended supply of +/-15. Assuming I can use the modern opamp it seems to make sense to run them at +/-15.

You should run them at +/-15VDC. Not the first time PL ran components beyond ratings.

 

[mlucitt]

I agree that when the PL2000 was made most households were at a nominal 110VAC at the wall plate (2 wires and no ground). Now, with almost everyone enjoying 120-125VAC (and a grounded outlet) the PL2000 power supply is a little on the warm side.

 

[wattsabundant]

The 7815 spec says minimum 17.5 VDC input. Punchback measured 18.85 VDC. That doesn't give much wiggle room. Will it cause a catastrophic failure? No. But if the line voltage sags during the summer or when the frig cycles on any advantage of using regulated supplies is lost. When I reworked my 3300 the rectified DC was 16.8 volts at 110 VAC. My house typically runs between 121 -123. I went with 7812's. By the way, I used the TO92 regulators which minimized the board size.

Was it really common to have 110 VAC 40 yeras ago? I've never seen anything specified for 110 in the U.S. 117 and 120 yes. 9 out of 10 electricians "hook it up to 110". Was that just something that was handed down?

 

[laatsch55]

I think so Don. 110 is not common...

 

[marcok]

The the transformer used in PL 2000 is 12 /0/ 12 ,so LM 7812 / LM 7912 are the right solution ,
because this kind of regulators needs a 3V voltage drop to work in the right way .
The difference from 12 to 15 V is negligible for an audio preamp .
See schematics of other manufacters . ( for instance Carver C 4000 )
Ciao
Marco

 

[punchback]

Anyone have any experience with LDO Voltage Regulators? They only need 1.75 volts headroom. I bought a power supply kit but it's going to be a pain to fit in. The more I think about it some perf board and a couple of components would convert the 18 volts into regulated 15 or 12. I guess I'll have to cut some traces to patch it in.