with a properly chosen pair of load capacitors, the built-in oscillator will function fairly well and with reasonable reliability at frequencies below 25 MHz, though they work MUCH more so at frequencies below 20 MHz. As has been stated by me and others, they aren't intended to work with 3rd harmonic crystals.

The O/P made it clear that he wanted to observe a speed increase as a consequence of installing a crystal that is marked with a frequency three times as high as the one that he previously used.

Anyone who's made the mistake that the O/P made, namely attempting to use a low-frequency oscillator circuit (as indicated by the overly large capacitors of 30 pf that he indicated he's using) with a 3rd overtone crystal knows that the nominal frequency at which that oscillator will operate, if at all, will be about 1/3 the labeled frequency. That's exactly what was observed and described in the O/P's original post.

I recommended that he try using a crystal oscillator, which it is likely he already had on hand, or, at least, could easily find. You, on the other hand, insisted that he didn't need a crystal oscillator. Well, nobody said he needs a crystal oscillator, but I did say it might make things work better, as this is just an experiment.

Just what did you think you were contributing by your inane remark that he should use the KEIL baudrate parameter calculator? What did you think you were contributing by your remark that he didn't need a crystal oscillator?

Had you suggested a method for modifying the Pierce oscillator built into his MCU so that it would operate at 3rd harmonic, that would have been helpful. Had you suggested a way in which he could obtain a first-overtone crystal of the frequency he needed, that would have been helpful.

Do you really believe that telling him he should use a SiLabs MCU would help him get his ATMEL-based circuit to function at the nominally 33 MHz rate?

You're quite apparently promoting KEIL and SiLabs products, Erik, but it's seldom indeed that you actually focus on the O/P's query. Your most common response to, "How can I do this?" is, "Why would you want to do that? You should do this instead!" Do you really think that helps an experimenter do what he wants?

Why would you help an old grandma across the street if she doesn't want to go there?

RE