White Oak Audio Design PL14_20 Quick Test Steps - Quasi-Comp and Full-Comp

[Gepetto]

Control board debug instructions

The instructions below apply to the White Oak Audio Design PL14_20 Control PCBs. The Pad numbering cited is specific to the White Oak Audio Design PCBs with the Pad numbers silkscreened on the board. These same tips can apply to the original QUASI-COMP Phase Linear control boards but the user will have to do some schematic work to understand how to apply it to the particular vintage of Phase Linear board that is involved.

These are quick checks you can perform on your control board to ensure it is regulating properly.

Quasi-Comp Configuration, all PCB artwork revisions

• Pad references that follow will include L or R and will depend on the channel you are testing. The L an R are left off of Pad and component references for simplicity.
• Detach and tape off wires going to Pads 9, 8, 6, 4, and 3.
• Tack a jumper wire between Pad 3 and 12.
• Tack a jumper wire between Pad 8 and 6.
• Tack a 47 ohm 1/2W resistor between Pads 9 and 6.
• Tack a second 47 ohm 1/2W resistor between Pads 4 and 6.
• Bring up on DBT and Variac and ensure that the output on Pad 6 is zero or no more than 15 mV (much lower for RevE and G1).
• Check that you have B+ on Pad 11.
• Check that you have B- on Pad 12
• Ensure that the voltage across each of the 47 ohm 1/2W resistors is less than 2V and the voltage on the 47 ohm 1/2W resistor tacked between Pads 9 and 6 varies as you adjust the bias potentiometer.
• If all good on the control board search for your problem on the backplane or backwall.
• If not all good on the control board, use this configuration to debug what is wrong with the control board.

Full-Comp Configuration, PCB artwork Revision D, E and G1

• This set of instructions assumes you have the White Oak Audio Backplane Kit installed.
• Pad references that follow will include L or R and will depend on the channel you are testing. The L an R are left off of Pad and component references for simplicity.
• Detach and tape off wires going to Pads 9, 8, 6, 4, and PNP
• Tack a jumper wire between Pad 8 and 6 and 4.
• Tack a 47 ohm 1/2W resistor between Pads 9 and 6.
• Tack a second 47 ohm 1/2W resistor between Pads PNP and 6.
• Bring up on DBT and Variac and ensure that the output on Pad 6 is zero or no more than 15 mV (much lower for RevE and G1).
• Check that you have B+ on Pad 11.
• Check that you have B- on Pad 12
• Ensure that the voltage across each of the 47 ohm 1/2W resistors is less than 2V and the voltage on each 47 ohm 1/2W resistor varies as you adjust the bias potentiometer.
• If all good on the control board search for your problem on the backplane or backwall.
• If not all good on the control board, use this configuration to debug what is wrong with the control board.

 

[mlucitt]

Joe,
Have you or anybody else run distortion or other output measurement test as a function of varied bias settings? I would be curious to know how the WOPL 400/700 amplifier behaves with bias settings of .300, .350, and .400 volts. These are the span of the specs, I don't know what would happen at the extremes of the range of the bias pots; fully CCW and fully CW.

Thanks,
Mark

 

[oldphaser]

Joe,
Have you or anybody else run distortion or other output measurement test as a function of varied bias settings? I would be curious to know how the WOPL 400/700 amplifier behaves with bias settings of .300, .350, and .400 volts. These are the span of the specs, I don't know what would happen at the extremes of the range of the bias pots; fully CCW and fully CW.

Thanks,
Mark

Mark,

There is some verbiage about setting the bias levels in certain versions of the 6/76 revision of the Phase Linear 400 and 700B series 1 amplifier service manual:

NOTE: "Avoid setting the bias voltage level too high in order to obtain very low distortion readings--this may lead to excessive idling temperatures."

The 6/81 revision of the 400 and 700B series 1 amplifier service manuals state:
NOTE: "For amplifiers used in sound reinforcement or commercial use it is recommended that the bias be adjusted closer to 300mV across R38(R47). This will increase the distortion only slightly, but will provide for much cooler operating temperatures. The cooling fan is still recommended, however."

As for the bias settings that Phase Linear used for 400 and 700 series 1 and II amps, it will vary depending upon which version of the service manual you are looking at. In general, the range can be anywhere from 300 mV up to 450mV. As the original carbon composition 10 ohm resistors go up in value this bias value will also change. Also when the 10 ohm resistors are not closely matched you will get different results between both the positive and negative rails on a given channel. The method in the service manuals is a "static" method (after allowing the amp to idle for 5 minutes or so). What Phase Linear actually did on the production line and what they did in the service department was different. In the service department, they examined the crossover notch distortion at 1KHz or 2KHz at 1 watt with the amplifier at idle (cold) and also after heating the amp up and adjusted the bias as a compromise between both states. This method is what I call a "dynamic" method. Typically after Dean (the last factory service manager) adjusted the bias the distortion would be around .12% at 1KHz or 2KHz at 1 watt. In which case you would often see the bias greater than 400mV or higher depending upon the value of the 10 ohm resistor.


Ed

 

[Gepetto]

The bias must be high enough to produce reasonably low distortion levels but not so high as to cause output stage cross conduction which leads to catastrophic events such as cross conduction, shoot through, fuse blowing, transistor failure, etc.

0.350 is a reasonable compromise value.

 

[Gepetto]

Joe,

Does this information also apply to any stock Phase Linear PL14_14 pc boards? If not, perhaps the title should be changed to reflect that this thread applies to only White Oak Audio pc boards.

Ed

Hi Ed
The general principles apply to the quasi comp versions of the original PL boards as well. The issues in applying this will be:

That the Pad references tie to White Oak Audio boards where the Pads have been labeled with silkscreened legends on each. The original boards did not have labeling so some research would have to be done against the original schematic in order to apply this advice.

That the original PL boards had varying lower edge pinouts depending on the vintage of the board.

That the original PL boards did not have a concept of the PNP signal so again some schematic work would be necessary.

The PL14_20 nomenclature is specific to White Oak Audio, AFAIK, PL never used such naming for their boards. This is posted in the WOPL Archive section so as to apply to the White Oak board versions primarily.

 

[oldphaser]

Hi Ed
The general principles apply to the quasi comp versions of the original PL boards as well. The issues in applying this will be:

That the Pad references tie to White Oak Audio boards where the Pads have been labeled with silkscreened legends on each. The original boards did not have labeling so some research would have to be done against the original schematic in order to apply this advice.

That the original PL boards had varying lower edge pinouts depending on the vintage of the board.

That the original PL boards did not have a concept of the PNP signal so again some schematic work would be necessary.

The PL14_20 nomenclature is specific to White Oak Audio, AFAIK, PL never used such naming for their boards. This is posted in the WOPL Archive section so as to apply to the White Oak board versions primarily.

Joe,

I deleted my post within a few moments of posting it after I realized that the title of your post falls under >Home>Forums>Phoenix DIY Central!> WOPL Tips And Archive.

I would like to know how you were able to see a copy of it and then made a copy and then replied to it?
Does it have anything to do with the fact that you are also an "Administrator" on this forum?

Ed

 

[mlucitt]

Joe,
Thanks, this follows my general belief and understanding that there is a trade-off between low distortion and cooler operating temperatures with a compromise at .350 Volts. I like the idea of the dynamic test. I wonder if we could duplicate that with a crossover notch test procedure for the WOPL Control Boards? I'm sure the distortion would be much lower than .12% at 1KHz/2KHz at 1 Watt.

 

[Gepetto]

Joe,

I deleted my post within a few moments of posting it after I realized that the title of your post falls under >Home>Forums>Phoenix DIY Central!> WOPL Tips And Archive.

I would like to know how you were able to see a copy of it and then made a copy and then replied to it?
Does it have anything to do with the fact that you are also an "Administrator" on this forum?

Ed

I just replied to it Ed, I thought it was a fair question that an answer would be helpful on. No superpowers involved.

I edited the thread title and introduction as well to reinforce the points you made.

Joe

 

[oldphaser]

I just replied to it Ed, I thought it was a fair question that an answer would be helpful on. No superpowers involved.

I edited the thread title and introduction as well to reinforce the points you made.

Joe

Joe,

You didn't quite answer my question.
My post was soon after your initial post.
How does one respond to something that is no longer there?
Since I almost immediately deleted it and no further responses in this thread occurred until mlucitt responded at 2:56 p.m.
Do administrators have the ability to see deleted posts? The answer is Yes and you are an administrator.

I wish I could respond to deleted posts (that others have provided me with knowledge of). However, this might expose some to embarrassment.

Ed

 

[oldphaser]

Joe,
Thanks, this follows my general belief and understanding that there is a trade-off between low distortion and cooler operating temperatures with a compromise at .350 Volts. I like the idea of the dynamic test. I wonder if we could duplicate that with a crossover notch test procedure for the WOPL Control Boards? I'm sure the distortion would be much lower than .12% at 1KHz/2KHz at 1 Watt.

I have measured a number of Phase Linear amps and several WOPL Rev E amps.
To say that you are "sure the distortion would be much lower than .12% at 1KHz/2KHz at 1 Watt" has no basis or foundation (in my opinion) unless you or someone you know has actually measured it and I have.

As I stated earlier, the bias was adjusted on the Phase Linear amps (in the Phase Linear service department) to measure approximately .12% THD at 1 to 2 KHz at 1 watt after being in a cold state and a hot state. If you go back and measure the static bias after this dynamic method, it could be anywhere from 350-450mV depending upon the actual value of the 10 ohm resistor. NOTE: You may also for example, see a bias reading in a static state of 400mV and after warming up the value may go down to 375mV.

NOTE: The original Phase Linear 700 series 1 amps had the "bias" transistor (MPS5172) mounted on the pc board. In the fall of 1971 with the introduction of the 400 series 1 amps, the MPS5172 bias transistor was shrink wrapped and placed under a clip and mounted to the chassis in the area just behind the heat-sinks. (NOTE: Eventually Phase Linear began using 2N3402 transistors.) The mounting of the bias transistor allowed for thermal tracking. The 700's did not have a bias transistor mounted on the heat-sink until the 700B was introduced.

I have not installed any WOPL back-planes as of yet. However, I just recently noticed that the bias transistor is not mounted on the chassis directly. The bias transistor appears to be installed on the back-plane board and a aluminum stand-off spacer is mounted between it and the chassis. I have to wonder how much surface area is making contact and what the temperature difference is and its affects on thermal tracking and where the bias setting should be set at?

Old carbon 10 ohm composition bias resistors that Phase used had a tolerance of 10%. The upper limit of a 10 ohm +/- 10% resistor would be 11 ohms and I have seen my share of old carbon comp resistors measuring 11 ohms or greater. The value of the bias setting therefore plays a critical role in what % of THD will result.

The later fully comp Phase Linear 700 series II's that had a low distortion factory modification on the PL36 pc board and will all measure significantly lower. (NOTE: The Phase Linear 300 series II's had in them from the factory.) As an example, at 2KHz at 1 watt single channel driven it will be around .02% THD, and both channels driven .03% THD. In this case, the crossover notch is so small that it is hard to discern on an oscilloscope. Therefore, Phase Linear essentially had no choice but to use the "static" method of setting the bias and they set the bias to 385mV.

There are also a fair number of amps that have been "upgraded" that originally had this particular factory modified pc board in them. In these cases, the amplifiers may be in my opinion considered to have been downgraded (from the perspective of measuring the specs) after the installation of a WOPL pc board. I have shared on a number of people on this forum what the mod was.

I also do testing at 1/4 watt because the FTC standard was revised at one time to require that the distortion rating be given from 1/4 watt to full output both channels driven (not just one channel). In the case of the fully comp amp I noted above this same amp will do from 1/4 watt to rated output approx. .25 to .29% THD from 20Hz to 20KHz both channels driven. What you will find on many other Phase Liner 400 and 700 series amps (including WOPL rev E's and I suspect other revisions as well) from 1/4 watt to full output is approx. .8% to 1% THD from 20Hz to 20KHz both channels driven.

In general, with most amplifiers the THD level is worst at 20KHz and 20Hz, both channels driven. I have asked numerous times for people to publish their results and some have refused to do so. THD levels will be at their best at 1 KHz which some people on this forum now only quote.

NOTE: The bandwidth (which you can select on most distortion analyzers) you are using to measure THD will also affect the results.

NOTE: The duration of test is also important to note when testing an amplifier. The longer the duration of test the more likely the THD is also to go up. A variety of industry standards have been published over the years. At one time it used to be at 30 seconds. More recent standards were at "5 minutes or greater". I am willing to bet that a fair number of people are not testing their amplifiers in accordance with this requirement.

NOTE: The FTC tried to implement the 1/4 watt to rated output in their standard decades ago but the manufacturers pushed back on it. I can understand why considering that they would have to quote something on the order of 1% THD.



Ed

 

[mlucitt]

Ed, I think there was some misunderstanding here on my part. When I said, "I'm sure the distortion would be much lower than .12% at 1KHz/2KHz at 1 Watt." I was referring to the potential result of a test, that I assumed, no one had previously accomplished. That is - the dynamic or 'crossover notch test' on a White Oak Control Board retro-mod PL400/700 Series 1 in quasi-comp format with the bias adjusted for a minimum crossover notch in class AB operation. This is not a reference to the performance of the amplifier or a historical testing procedure, it is a reference to a request for a test to determine the lowest obtainable distortion by the dynamic method instead of a given bias voltage setting and a 8-Ohm loudspeaker load for both channels. Ideally, a graph showing distortion and temperature as a function of various bias voltages in several amplifiers (modded and not) would be helpful. Or, we could just set the bias at .350 volts and go on with life.

I agree that discussing distortion in the Series II amplifiers or other factory-modded amplifiers is a different topic altogether.

 

[grapplesaw]

I think Ed you need to write a procedure in order to get reliable comparative data. I for one do not publish much data because what I use for equipment may differ with others. Also types of Leeds and lengths, voltage of main supply, supply distortion Etc. etc. differ for all of us. When I test I compare to my last units built. this is the best can do without exact steps.

I have done extensive research on Bais. There is much information on amperage required to turn on the devices but a true explanation on correct voltage setting is very allusive.

I am very interested in this subject. With dual rail voltage amps your bias varies quite a bit from high to low rail. I am now making an intendant fixed voltage supply to power the control board regardless of rail voltage which will supply constant bias voltage regardless of rail voltages.


Any further information would help all of us.

 

[Gepetto]

Joe,
Have you or anybody else run distortion or other output measurement test as a function of varied bias settings? I would be curious to know how the WOPL 400/700 amplifier behaves with bias settings of .300, .350, and .400 volts. These are the span of the specs, I don't know what would happen at the extremes of the range of the bias pots; fully CCW and fully CW.

Thanks,
Mark

 

[laatsch55]

Looks like close to what I get..

 

[Gepetto]

Looks like close to what I get..

Basically it matters little...

 

[oldphaser]

I made a typo in my previous post. It should have been 2N3403 not 2N3402.

Joe,

In your measurement results posted in your pdf....

After adjusting the bias for both channels to your nominal values, did you perform these measurements single channel driven? Or both channels driven at the same time? I only see (1) set per each bias setting. Not a left and a right channel.

By the way, Phase Linear never used a 30KHz low pass filter in the service manual. Only 80Khz.

Did you perform (2) sets of measurements? (1) set after initial turn on and idling for (5) minutes and then again after driving the amp say at 1/3rd power into a 4 ohm load until the amp was hot to the touch? This was the process Dean used all the while observing the crossover notch.

Does the use of the aluminum stand-off spacers on the bias transistors located on the back-plane boards reduce the effects of heating the bias transistors to some degree as opposed to having them mounted directly to the chassis?

Were the results with a 2N5088 installed in the p-strap or with original 2N3403's installed?

Thanks!
Ed

 

[WOPL Sniffer]

OMG

 

[oldphaser]

NOTE: 30KHz low pass filters are used to "provide band limiting required by FCC for proof-of-performance broadcast testing".
80Khz are used "to reduce the effect of higher frequency noise present in the measured signal".

 

[mlucitt]

Speaking of distortion, can I get a few opinions on this unit for a quick (not perfect) distortion test. An AP Portable One Plus or a HP 8903B would be so nice, but this is all I can afford right now. Like I said, this would be for more of a quantitative test (pass/fail) than a qualitative test of the exact amplifier output distortion level at 1 KHz and 1 Watt. It has the specs listed in the description:

https://tinyurl.com/rxal3eb

 

[WOPL Sniffer]

The TDM-1911 can be used to measure the distortion and level parameters of audio products. The two-point signal distortion measured at 400 hz and 1 KHZ frequency is 0.01%~30% of 6 full scales.

I'd skip buying this..... Save your money, an 8903 can be had for another $100-$200 more