I'm trying to select component values. I'm stuck on two items. The input DC blocking caps/filter caps, and Ri to set gain and filter.
Evaluation board uses 10Kohm Ri, and 10uF Ci
https://www.diodes.com/assets/Evaluation-Boards/PAM8406-User-Guide.pdf
Data Sheet
https://www.mouser.com/datasheet/2/115/PAM8406-247303.pdf
Has gain set by;
AVD = 20*log [2*(RF/RI)]
If i understand correctly RF is fixed at 142kohm. RI must be 18k minimum (confusing as the dev board above uses 10k RI)
Looking for a smaller gain as this is a 5v design, i'm planning to use to drive 3w speakers, i'm looking at using something like 62kohm Ri.
Ci is calculated by the following formula, based on the frequency cut-off(Fc) your looking for.
Fc=1/(2*pi*Ri*Ci)
Using Ri 62000, Ci would be 0.09uF, at 30hz Fc (lower than necessary for such a small driver)
Using Ri 10000 as on dev board, Ci would be 0.5uF at 30hz, despite the fact the dev board Ci is 10uF.
Are they just using a much larger cap than necessary? Or am i calculating something incorrectly?
Evaluation board uses 10Kohm Ri, and 10uF Ci
https://www.diodes.com/assets/Evaluation-Boards/PAM8406-User-Guide.pdf
Data Sheet
https://www.mouser.com/datasheet/2/115/PAM8406-247303.pdf
Has gain set by;
AVD = 20*log [2*(RF/RI)]
If i understand correctly RF is fixed at 142kohm. RI must be 18k minimum (confusing as the dev board above uses 10k RI)
Looking for a smaller gain as this is a 5v design, i'm planning to use to drive 3w speakers, i'm looking at using something like 62kohm Ri.
Ci is calculated by the following formula, based on the frequency cut-off(Fc) your looking for.
Fc=1/(2*pi*Ri*Ci)
Using Ri 62000, Ci would be 0.09uF, at 30hz Fc (lower than necessary for such a small driver)
Using Ri 10000 as on dev board, Ci would be 0.5uF at 30hz, despite the fact the dev board Ci is 10uF.
Are they just using a much larger cap than necessary? Or am i calculating something incorrectly?
I understand your confusion. The datasheet text is poorly phrased if my interpretation is right
Rf is not to be set to 142K maximum, it is 142K and we cannot change it (internal component we have no access to). The datasheet says that R1=Rin must be at least 18K. You have calculated that 62K suits you. That the evaluation board uses 10K against the recommendation of 18K in the datasheet is indeed confusion but it does not mean that the evaluation board will not work. Stick to your 62K. 100nF (0.09uF) for C1=Cin seems right for 30Hz roll-off. That the evaluation board uses 1uF (not 10uF; see page 3) means that it is configured for a lower roll-off frequency (16Hz). I do not have a PAM8406 but only a PAM8403 and a PAM8610. They can actually supply impressive low bass if your speakers can handle it.
You seem to be right all along. A happy 2020 to you!
Rf is not to be set to 142K maximum, it is 142K and we cannot change it (internal component we have no access to). The datasheet says that R1=Rin must be at least 18K. You have calculated that 62K suits you. That the evaluation board uses 10K against the recommendation of 18K in the datasheet is indeed confusion but it does not mean that the evaluation board will not work. Stick to your 62K. 100nF (0.09uF) for C1=Cin seems right for 30Hz roll-off. That the evaluation board uses 1uF (not 10uF; see page 3) means that it is configured for a lower roll-off frequency (16Hz). I do not have a PAM8406 but only a PAM8403 and a PAM8610. They can actually supply impressive low bass if your speakers can handle it.
You seem to be right all along. A happy 2020 to you!
Last edited:
No, I pretty sure that's not it.
The block diagram shows an internal input resistor on the first, inverting, gain stage. That must be 18k, so that constrains the maximum gain.
The second gain stage is x2 (its bridging), the first is x -(142k / (18k + Rin))
Thus the overall input gain is 20 log10 (2 x (142k / (18k + Rin)))
The datasheet calls the external resistance Rin, and the sum of this and the internal 18k as RI
The block diagram shows an internal input resistor on the first, inverting, gain stage. That must be 18k, so that constrains the maximum gain.
The second gain stage is x2 (its bridging), the first is x -(142k / (18k + Rin))
Thus the overall input gain is 20 log10 (2 x (142k / (18k + Rin)))
The datasheet calls the external resistance Rin, and the sum of this and the internal 18k as RI
I just reviewed my schematic (as this is a hobby, i spread my time around a bunch of different projects, and struggle to keep all the little details straight unless i write them down), and I was planning on putting trimpots in to allow adjustment of the gain.
Thanks for the help guys, I didn't fully understand what they meant by Rf and Ri. Rf is the feedback resistor, shown in the block diagram, and Ri is the input resistor, also shown in the block diagram. I can adjust the gain by adding external resistance (Rin) which adds to Ri to adjust the ratio of Rf/Ri.
I might order boards today!
Thanks for the help guys, I didn't fully understand what they meant by Rf and Ri. Rf is the feedback resistor, shown in the block diagram, and Ri is the input resistor, also shown in the block diagram. I can adjust the gain by adding external resistance (Rin) which adds to Ri to adjust the ratio of Rf/Ri.
I might order boards today!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.