Sir pharaon . . . . . . .
All of our prior attention was directed to the 4 FWB diodes, now in moving to the D5, D7 and D9 units, I have already explained the D5 circuit function.
I reworked the foil path portion of the board just below, in denoting the positions of the D7 and D9 diodes, the terminals of the power switching transistor .
If things have not changed, I am seeing the solder joints of D7 as having never been touched.
On D9, its visually being a different story.
I think where a mistake was made, is if the D9 diode was tested and found with its lower Vf junction reading being in the 400 millivolt range.
That tended to made you condemn it, because you are probably used to finding higher Vf's than that.
This can be attributed to to ones not fully knowing that Vf specifications can have variances between different types and families of diodes.
In the case of Schottly diodes , they don't come in too high of voltage ranges, but exhibit VERY fast switching rates and different current capabilities, of which, I have seen their Vf (low voltage drop) go down to as low as 200 millivolts on higher power rectifier function units,
In the case of D9 here , note that its function here is to provide a one way path from the emitter of the power transistor to ground.
And I am asuming that you were able to read the original D7 and D9 diodes markings and confirm that both were being 1N5819's.
Those two diodes fully meet the needed requirements of:
High Current Capability, Low Forward Voltage Drop and High Frequency / Fast Switching Rate capabilities.
Possible ERROR 1:
If you sub in a 1N4007, you still squeak by on the = < 1 amp current need and are making the circuit 50% worse on the voltage drop, but the largest error is in its switching speed capabilities.
If you had 50 ~ AC line power being involved, that s l o w switching time is taking f o r e v e r. . . .and EASILY handled .
BUT, in this power transistors case, you have switching times in the order 50,0000 . .or . .100,000 . . or. . .250,000 changes PER second.
That is where the shortcoming in the use of the 4007 is, as its specification for its STOPPING conduction or turnoff is so long that some of the then reversed polarity voltages, rush backwards through the still conducting junction.
The diode is conducting in ONE direction, while its the SLOW cutoff time characteristic, when being challenged by too high of a frequency
input, is the equivalency of a shunting resistor, which is letting both directions / polarities of + - voltages then pass through.
The diode revolts by running hot and the power transistor revolts by not operating normally / optimally due to that impending circuit deficiency.
Possible ERROR 2:
Your last possibility would be the use of a fast switching diode in that application . . . .which would be that newly found PR1004 diode.
If you sub in a PR1004, you still squeak by on the = < 1 amp current need, this unit will meet the switching speed capabilities, but its SHORTCOMING is its greater Vf specification, in its creating a GREATER voltage drop.
Soooo the proper candidates will be those 1N5819's which were in the circuits initially.
If you will test them again and find the Vf being in the 400mv range and swap the leads and get your "1" reading, by all means re install them, as I believe that they are telling you that they are good.
Pilot to ground control . . . . .over and out .....
Foil Side Drawing Rework:
( Mainly Pink and Red add ons)
73's de Edd