Phase Linear Model 700

[PB2]

Hi Pete,
I used to stock 2SD424 and 2SB554 as well as 2SD555 and 2SB600. Wonderful output devices. I can't say that I had any trouble with oscillation with these.

Of course I also stocked the jedec numbers, or ordered them as required.

I would suspect that the layout may have something to do with devices oscillating. Yes, the 2Sx series seemed to be faster with higher gain.

-Chris

Not sure I'm following, did you try them in a PL700, and not see oscillations? If Phase Linear was right about only using 2SD555s it points to the MJL4281 as the right part. On the other hand, it's well documented that the PL700 output deviced were selected for LOW beta, why would they suggest a high beta replacement part? On the other hand the MJL4281 SOA is only 1A, 100V, 1 sec, whereas the 21194 is 2A, 100V, 1 sec. 21194s win from an SOA perspective and I believe it's worth figuring out if a mod is needed to make them stable.

 

[PB2]

PL20:

R61: 18k
C1: 120pF
C10: 180pF
R16: 33r
R3:39k
R17: 18k/1W
R18: 3.9k
R10:39k

/Hugo

Thanks Hugo, this helps.

 

[anatech]

Hi Pete,
I went with manufacturer's recommendations where ever possible. I'll bet that I used MJ15024 / 5 in an emergency for a road amp without too many troubles. I would not generally use a Japanese part for a US one or vice versa.

I am so use to seeing amps oscillate that it would not be memorable if I had to modifiy the feedback cap or bandwidth limiting caps in some way. I would refer to a manufacturer's in print changes for other models if need be. Even C-B caps could be used.

-Chris

 

[d3imlay]

Last year I repaired a 700II with the original 2SD555's in it. It had burnt the output R/C network, specifically the cap in both channels. It was oscillating. I tried the usual fix of 150pf on the bc of the predrivers to no avail. Decoupling caps, and all of the tricks I know didn't work. I switched to MJ2119X and all was well. Question of the day: Did it leave the factory like that? Who knows.

 

[anatech]

Hi d3imlay,

Question of the day: Did it leave the factory like that? Who knows.

No. The existing decoupling and power supply caps have degraded with time.

-Chris

 

[Netlist]

I could be wrong but I think 2SD555 were never used in the factory.

/Hugo

 

[anatech]

Hi Hugo,
You are right, but they are supposed to be an approved substitute. See the Phase Linear service bulleton 1-80 concerning output transistors.

That amplifier may have been serviced at an autorized warranty center or other good shop. Bad shops generally would not use 2SD555's due to cost.

-Chris

 

[PB2]

Hi Hugo,
You are right, but they are supposed to be an approved substitute. See the Phase Linear service bulleton 1-80 concerning output transistors.

That amplifier may have been serviced at an autorized warranty center or other good shop. Bad shops generally would not use 2SD555's due to cost.

-Chris

Hi Chris,

As I read it the faster 2SD555 were only for the older PL700s, and the slower MJs were for the newer series II. So that would be the wrong part in that amp.

Pete B.

 

[anatech]

Hi Pete,
Good point. There might be some differences in compensation component values between them. I can't see board layout getting worse.

-Chris

 

[d3imlay]

2SD555's and 2SB600's were used on the production line in the series II amps. I used to buy them for $4.50 each in small quantities before they were obsoleted.

 

[anatech]

Hi d3imlay,

2SD555's and 2SB600's were used on the production line in the series II amps.

Interesting. Of course it's hard to tell since most amps I saw had been hacked many times first.

-Chris

 

[PB2]

It seems that there is no concensus on what is a stable configuration using replacement parts, even those suggested by the factory. Makes it difficult to help.

Pete B.

 

[anatech]

Hi Pete,

It seems that there is no concensus on what is a stable configuration using replacement parts, even those suggested by the factory.

Just make the most intelligent decision you can at the time and deal with any problems as they come up. The life of a technician.

-Chris

 

[Netlist]

I found some time to finish modeling the 700B.
Picture is the best I can produce, Transistors are what was reasonably available in Microcap except for the Ge's Q8 and Q9 which are standard models.
I can pass the model text file to anyone interested.

/Hugo

 

[sixtek]

Hi Netlist,

I'm interested in SIM results. Are there any data (THD, noise, square-wave responese, etc...) you could share with us?

Thx,

 

[Netlist]

Device conditions:

 

[sixtek]

PS:
Hmmm... Strange.. PL had not used any rail bypass capacitor? Something polyester in 47n-470n the range, optionally bypassed with 10u-100u elkos?

 

[Netlist]

Gain:
With a 1Khz sine wave of 6.44Vpp at the input the output slightly clips at 189.58Vpp into 8ohm.
If my calculation is correct: VRMS = Vpp * 0.3535, the ouput power would be 561 WRMS.

/Hugo

 

[Netlist]

sixtek,
In this simulation, I have omitted the PSU caps as this is supposed to be battery operated.
In real life, PL only use two big capacitors, not bypassed.

I can post any results you want, given that I know how to perform them in Spice.

/Hugo

 

[Netlist]

Square wave input 1Vpp, output 30Vpp.
I'm still searching how to change the frequentie of the square wave model to make it say 10kHz.
Edit: This should be 10kHz.

/Hugo