On the diode body, the cathode is the end with the stripe and/or the point of the arrow, like the schematic shows. Determination of anode vs. cathode is not a function of the polarity of the circuit. It is determined solely by the construction of the diode. The red/black lead (in your case "blue/white," even though it isn't really blue/white) goes to the bias diode's cathode, which, again, is the end with the stripe and/or point of the arrow. Look at the schematic again. It indicates the red/black wire as "R-BLK" and clearly shows it connected to the end of the bias diode having the stripe and arrow-point.
What makes this confusing is that the bias is set up to provide a negative bias control voltage. Dynaco accomplished this by flipping the polarity of the two electrolytic capacitors in the bias circuit, which results in a negative DC voltage on the diode's anode while there is an AC voltage (the "+" sign in the older manual's schematic) on the diode's cathode. The diode is functioning as a half wave rectifier. When the AC voltage from the red/black lead swings positive, the diode does not conduct. When it swings negative, the diode conducts because the current is flowing (strange as this might sound) from ground "through" the two bias supply electrolytics and through the bias diode, because ground is actually at a greater potential than the negative swing of the transformer on the diode’s cathode. This charges the bias supply electrolytics with a negative DC voltage for the bias supply.
While the diode must be wired properly, whether it solves your amplifier's problems or not might be another matter entirely.
I have looked at your photos and though I can see your "blue wire" (that I call "blue/white" and believe to actually be the faded red/black wire), I cannot see the diode and how it is currently oriented. You might actually have the diode properly oriented already, I don't know. Even with the correct transformer lead connected to the correct end of the diode, you may still have other issues.
My suggestion would be to disconnect your "blue" wire, secure it somehow so that it is physically stable and does not touch anything, and then measure the voltage between the wire and the chassis. This will tell you if it is indeed the red/black wire and if it is still functioning, provided the appropriate ground references are connected. As Bob indicated previously, you should get something in the range of around +55 volts AC provided your meter is set to an appropriate AC voltage (not current) scale, your meter leads are connected to the right places on the meter and lead polarity is consistent (red meter lead to "blue" wire, black meter lead to chassis). With meters, remember always to switch between appropriate AC and DC scales, voltage or current, and, if necessary, accordingly change the lead connections to the meter depending on what you intend to measure. Easy enough to miss this kind of stuff.
Last edited by PeterCapo on Fri Nov 06, 2015 11:30 am; edited 4 times in total (Reason for editing : Add comments. Spelling. Clarify.)