Nate, I've been designing 8870s and 8880s in things for over 15 years. Good chip and it has lasted.
Pin 3 is simply the output of the internal op-amp and using a feed back resistor to 2 and driving with another one can set the AC gain over a wide range. The gain is feedback/driving resistor unless you wish to drive the non-inv input. The 1/2 Vcc source, pin 4, was cleaver by Mitel to simply use with a single 5 V supply. As to the other gain pins it is best not to use these for some other manufactures of the chip do not provide them. California Micro Devices, probably one of the biggest manufacture, does not. Really with the op-amp and allowing of adjusting its gain one does not really need these. For the 8880 the analog input is the real easy part. The CPU interface is much more difficult, but really simple design. The only same pin out of the 8870 to 8880 is the input op-amp. After that all changes. I normally do away with the steering circuit for RC adjusting the attack/decay decode times. I do both in software and tie pins 16 & 17 together on 8870, 18 & 19 on 8880. This give instant decode/release time, but use software for the timing. As with any op-amp using a dif amp, as the 8870 & 8880 do, single ended or dif input can be done. Looks like you cut and pasted your posting, hi. Referred to figures not included. Oh well. 73, ron, n9ee/r >From: Nate Duehr <[EMAIL PROTECTED]> >Date: 2008/03/28 Fri PM 08:09:36 CDT >To: Repeater-Builder@yahoogroups.com >Subject: Re: [Repeater-Builder] False DTMF Muting On CAT-1000 Controller > >Ron Wright wrote: > >> The voltage levels on most DTMF decoders are set by the IC designer. On >> decoders like the 8880 and 8870 pin 3 is what the internal op-amp of the IC >> is producing and this parameter is determined by the IC designer. As with >> any IC one needs to supply the proper levels to the device to allow it to >> achieve what is needed. > >Pin 1 is the non-inverting op amp input, Pin 2 is the inverting op-amp >input, and Pin 3 is Gain Select, described below. > >From the MT8870 datasheet: > >"GS: Gain Select. Gives access to output of front end differential >amplifier for connection of feedback resistor." > >"The input arrangement of the MT8870D/MT8870D-1 >provides a differential-input operational amplifier as >well as a bias source (VRef) which is used to bias the >inputs at mid-rail. Provision is made for connection of >a feedback resistor to the op-amp output (GS) for >adjustment of gain. In a single-ended configuration, >the input pins are connected as shown in Figure 10 >with the op-amp connected for unity gain and VRef >biasing the input at 1/2VDD. Figure 6 shows the >differential configuration, which permits the >adjustment of gain with the feedback resistor R5." > >Same pin-out on the 8880. > >You can build it as a single-ended input or a differential one, and mess >with GS... depending on how you built your audio circuit... > >The input op-amp, has a decent amount of gain available in the op-amp >itself, too... because it' meets the specs to pull a tone up from -31 >dBm (dBm referenced to 1mV into a 600 ohm impedance input -- the typical >telco reference for dBm) and won't fully reject tones until -37 dBm. > >These chips can hear DTMF tones that are incredibly low amplitude... on >quiet circuits. The problem is... radios aren't quiet circuits. > >The published error rate at -31 dBm is 1 in 10,000 without background >noises. They're very good decoders. > >But depending on how they're hooked up, generic statements like 3 V >P-t-P aren't ALWAYS correct... unless the repeater controller is feeding >the op-amp in the unity-gain single-ended configuration. > >Nate WY0X > > Ron Wright, N9EE 727-376-6575 MICRO COMPUTER CONCEPTS Owner 146.64 repeater Tampa Bay, FL No tone, all are welcome.
