34.55? i think there has to be a typo.

How did I come up with that number?!
21.6 x 1.414 = 30.542.
21.6 x 1.4142 = 30.546. Rounded to two decimal places 30.55. Yeah, 34.55 must have been a typo.

My bad. GEEZ I hate that statement. It's like saying "I was wrong but not sorry." Sorry for the misinformation.
Still with all the diodes in play, you're still a half volt over spec. Will that cause problems immediately? Maybe not. But what about down the line, in a month or two. Or year. Or two.

I tried using diodes to limit the amount of voltage from a 5V supply down to 3.3. It worked for several months but then failed. Two diodes would have dropped the voltage down to 3.8VDC. An additional diode would have dropped it down to 3.2 (assuming 0.6Vf for 1N4001's).

Diodes are cheaper than bridge rectifiers. Stringing a series of diodes as voltage drops might be cheaper than one or two BR's.

 

I see the statement "but nothing has changed!!" But now the DC to DC converter gets hot. That seems like a change. so my first suggestion is measure the current thru that device. It may also be good to try disconnecting the various grounds, to see if it is an unintended ground someplace in the wiring.

 

Just as a remotely related comment here; I'm running a 13.8VDC power supply into a rocket launch control panel (another project on this website). Because I'm using binding posts for the battery hookup of what will be a 12V SLA battery and the possibility a child could connect power inadvertently backwards I've built into the project a BR. My PS is 13.8VDC but the gauge on the controller is 12.7VDC. So I'm only going through two diodes and dropping 1.1 volts total. For the controller that's not an issue. Since this post opens with "Just as a remotely related comment here;" I'm sharing this to demonstrate how diodes will drop some voltage. On average, my BR is dropping 0.55Vf per diode. Don't ask for the number, it's buried between a board and a heat sink.

Not off topic, just related to the voltage drop the TS needs.
24VDC minus (0.6 x 2) = 22.8VDC. Adding a filter cap will increase that to 32.24VDC. Minus one diode Vf = 31.64. Minus a second Vf = 30.44VDC. A third drops to 29.84VDC. So 3 diodes in series would drop the filtered voltage to 29.84 volts DC. Before or after the cap but certainly after the BR.

Does anybody have issue with the following diagram?
View attachment 358885
The blue lines connect to your existing circuit.