| General > General Technical Chat |
| BJTs being phased out? |
| << < (9/10) > >> |
| tooki:
--- Quote from: coppice on October 16, 2020, 02:46:05 am --- --- Quote from: tooki on October 15, 2020, 11:04:39 pm ---It seems my hunch was right: while reading up on electrets to try and answer NANDblog’s question, I came across a statement from Shure (one of the big names in pro audio mikes) claiming that “probably 95% of the condenser microphones on the market are the electret type”. --- End quote --- Interesting. The fairly modern mics I've seen schematics for, or seen inside, had pumps to push the 48V phantom power up to a high voltage. Electrets have real benefits. They can create a really big bias voltage. I had a pair of Pioneer electret headphones in the early 70s. Condenser headphones, like the Stax models, could produce wonderful sound, but struggled to reach high volumes, because they didn't want an extreme bias voltage running down the cable. No problem with the Pioneer electret headphones. They would reach pain levels while maintaining high fidelity. --- End quote --- Given how mature both electret and non-electret microphone technology is today, I think that what ultimately matters is how carefully the capsule (and the device it’s in) has been designed, and that top quality can be achieved with either technology, when low cost isn’t the primary decision driver. |
| tom66:
Up until a few years ago discrete BJTs (e.g. MMBT3904) used to be half the price of equivalent NMOS transistors (e.g. 2N7002), useful for small loads. I see now the price has reached parity so it seems that for most applications any cost advantage there is now gone. |
| coppice:
--- Quote from: tom66 on October 16, 2020, 09:41:59 am ---Up until a few years ago discrete BJTs (e.g. MMBT3904) used to be half the price of equivalent NMOS transistors (e.g. 2N7002), useful for small loads. I see now the price has reached parity so it seems that for most applications any cost advantage there is now gone. --- End quote --- From my experience with small ICs, I assume the cost of packaging a small discrete, like a 2N7002, now greatly exceeds the cost of the silicon itself. So, the cost of the device is only marginally affected by which transistor is inside. |
| tom66:
--- Quote from: coppice on October 16, 2020, 09:49:03 am --- --- Quote from: tom66 on October 16, 2020, 09:41:59 am ---Up until a few years ago discrete BJTs (e.g. MMBT3904) used to be half the price of equivalent NMOS transistors (e.g. 2N7002), useful for small loads. I see now the price has reached parity so it seems that for most applications any cost advantage there is now gone. --- End quote --- From my experience with small ICs, I assume the cost of packaging a small discrete, like a 2N7002, now greatly exceeds the cost of the silicon itself. So, the cost of the device is only marginally affected by which transistor is inside. --- End quote --- Still, there was a price difference, 2N7002 being about twice the price of MMBT3904. We are talking about parts with a sub 1-2p BOM cost though. |
| NiHaoMike:
--- Quote from: coppice on October 16, 2020, 09:49:03 am ---From my experience with small ICs, I assume the cost of packaging a small discrete, like a 2N7002, now greatly exceeds the cost of the silicon itself. So, the cost of the device is only marginally affected by which transistor is inside. --- End quote --- Perhaps the next step is some sort of simple IC designed to take the place of a small transistor, the value add being built in protection against ESD, overcurrent, and overheating? Such a device, if cheap enough, would work nicely for interfacing sensitive logic to the external world. |
| Navigation |
| Message Index |
| Next page |
| Previous page |