# Specification sheet vis-a-vis measurement discrepancy?

#### mefistofelez

Greetings all,

I had built an amplifier wherein a diamond buffer (BUF634) is enclosed in an operational amplifier feedback loop (OPA1656), cf., https://www.diyaudio.com/community/threads/opa1656-buf634a-headphone-amplifier.396571/#post-7316599.

I was wondering, whether the amplifier will drive a compression driver on a wave-guide with measured SPL of 112 dB/W/1m, as the specification sheet claims 250 mA continuous output current, which into 8 Ohms would give P=8*0.25^2=0.5 W.

Before connecting the amplifier to the compression driver, decided to measure its performance into a 10 Ohm precision resistor. Using a 1,000 Hz sine-wave I was increasing the output voltage and the amplifier, powered from +-12 V power supplies, worked up to 7.2 V peak-to-peak (p-p) when the sine-wave started to be limited. That seems questionable as the current is I=(7.2*0.707)/10/2= 254 mA, which I interpret as “continuous”. Given the temperature of the case, see next paragraph, current limiting should not occur, cf., Fig.10.

What I further do not understand is that decreasing the voltage to 7 V p-p, causes the amplifier to work for hours, with the case temperature of the BUF634 being stable at about 51 deg C. It makes no sense from power dissipation perspective, because Junction to top of the case difference is 51-25=26 deg C. Since the Junction-to-case (top) thermal resistance per the specification sheet is 44.2 deg C/W, it would mean that the dissipated power is only P=26/44.2=0.5W, which is clearly incorrect. Cf. also Fig. 15 in the specification sheet.

Any help with whether I am misinterpreting the specification sheet, making error in my calculation, or any other reason for the above would be appreciated.

Kindest regards,

M

#### Attachments

• buf634.pdf
925.4 KB · Views: 52
Last edited:

#### PChi

Hello,
I think that the BUF634 current limit is for the peak, not rms current.
The internal junction temperature can't easily be measured amd will be a fair bit higher than the case temperature so I am not sure that the dissipated power calculation is correct.
Regards