|
Sample-and-hold product detector
Dr. Ulrich Rohde described in one of his articles (QST ?) a novel
product detector based on CD4066 MOS-switch, which is supposed to have a superior performance compared to other types. I have tried the schematic and found the distortion quite unacceptable. I have not found any other reference to this type of product detector and may have misunderstood the cocept somehow. Therefore I would like to know if this idea is still alive and/or if somebody has any experience of it. Otherwise I will rely on either DMB or MOSFET in my homebrew receiver which is intended to be an advanced . project. I have already successfully constucted the front end with Fairchild FST3125 switching mixer followed by a strong postmix amp, 12-pole xtal filter and the following if-stages. It would a shame to destroy the whole thing with an inferior detector/af-portion when I am trying accomplish something better than my previous design which is heavily based on the W7ZOI Progressive receiver. 73 de Heikki |
Heikki,
I'm not sure exactly what this particular circuit looks like, but I strongly suspect it's just a balanced mixer made from 4066 analog switches. I have used this technique and it is quite good at low IF frequencies (a few MHz or less). Are you sure you have the switches biased properly? The thru-path should be biased at about 1/2 of the 4066 DC supply voltage in order to avoid severe signal clipping. Can you point to a link where the circuit is shown?? Joe W3JDR "Heikki Ahola" wrote in message ... Dr. Ulrich Rohde described in one of his articles (QST ?) a novel product detector based on CD4066 MOS-switch, which is supposed to have a superior performance compared to other types. I have tried the schematic and found the distortion quite unacceptable. I have not found any other reference to this type of product detector and may have misunderstood the cocept somehow. Therefore I would like to know if this idea is still alive and/or if somebody has any experience of it. Otherwise I will rely on either DMB or MOSFET in my homebrew receiver which is intended to be an advanced . project. I have already successfully constucted the front end with Fairchild FST3125 switching mixer followed by a strong postmix amp, 12-pole xtal filter and the following if-stages. It would a shame to destroy the whole thing with an inferior detector/af-portion when I am trying accomplish something better than my previous design which is heavily based on the W7ZOI Progressive receiver. 73 de Heikki |
Heikki,
I'm not sure exactly what this particular circuit looks like, but I strongly suspect it's just a balanced mixer made from 4066 analog switches. I have used this technique and it is quite good at low IF frequencies (a few MHz or less). Are you sure you have the switches biased properly? The thru-path should be biased at about 1/2 of the 4066 DC supply voltage in order to avoid severe signal clipping. Can you point to a link where the circuit is shown?? Joe W3JDR "Heikki Ahola" wrote in message ... Dr. Ulrich Rohde described in one of his articles (QST ?) a novel product detector based on CD4066 MOS-switch, which is supposed to have a superior performance compared to other types. I have tried the schematic and found the distortion quite unacceptable. I have not found any other reference to this type of product detector and may have misunderstood the cocept somehow. Therefore I would like to know if this idea is still alive and/or if somebody has any experience of it. Otherwise I will rely on either DMB or MOSFET in my homebrew receiver which is intended to be an advanced . project. I have already successfully constucted the front end with Fairchild FST3125 switching mixer followed by a strong postmix amp, 12-pole xtal filter and the following if-stages. It would a shame to destroy the whole thing with an inferior detector/af-portion when I am trying accomplish something better than my previous design which is heavily based on the W7ZOI Progressive receiver. 73 de Heikki |
W3JDR wrote: Heikki, I'm not sure exactly what this particular circuit looks like, but I strongly suspect it's just a balanced mixer made from 4066 analog switches. I have used this technique and it is quite good at low IF frequencies (a few MHz or less). Are you sure you have the switches biased properly? The thru-path should be biased at about 1/2 of the 4066 DC supply voltage in order to avoid severe signal clipping. Can you point to a link where the circuit is shown?? The circuit can be found in QST Nov 1992 (Recent Advances in Shortwave Receiver Design)! I t is unclear whether only two switches are used or if each switch is actually two in parallel. The biasing should be OK! 73 de Heikki |
W3JDR wrote: Heikki, I'm not sure exactly what this particular circuit looks like, but I strongly suspect it's just a balanced mixer made from 4066 analog switches. I have used this technique and it is quite good at low IF frequencies (a few MHz or less). Are you sure you have the switches biased properly? The thru-path should be biased at about 1/2 of the 4066 DC supply voltage in order to avoid severe signal clipping. Can you point to a link where the circuit is shown?? The circuit can be found in QST Nov 1992 (Recent Advances in Shortwave Receiver Design)! I t is unclear whether only two switches are used or if each switch is actually two in parallel. The biasing should be OK! 73 de Heikki |
Heikki, I dug this issue out of the archives and looked at the circuit. I'm not surprised that the performance is not real good. This circuit has no balance or switching symmetry. It is a simple half-wave switch which samples the IF at the BFO frequency. The output will have significant energy at the IF frequency as well as the BFO frequency, and these components might be overloading the audio amplifier. I'd suggest something that has full-balance. You could modify the '4066 circuit to have this property if you could generate 2 opposite phases of the IF and drive them separately into the inputs of the two '4066's. If you set it up right, you will get very large dynamic range. Alternatively, a traditional double balanced IC circuit such as an NE602 or even a diode DBM will work just fine as long as there is adequate selectivity in front of it to ensure that it is normally working with one IF input signal. Joe W3JDR "Heikki Ahola" wrote in message ... W3JDR wrote: Heikki, I'm not sure exactly what this particular circuit looks like, but I strongly suspect it's just a balanced mixer made from 4066 analog switches. I have used this technique and it is quite good at low IF frequencies (a few MHz or less). Are you sure you have the switches biased properly? The thru-path should be biased at about 1/2 of the 4066 DC supply voltage in order to avoid severe signal clipping. Can you point to a link where the circuit is shown?? The circuit can be found in QST Nov 1992 (Recent Advances in Shortwave Receiver Design)! I t is unclear whether only two switches are used or if each switch is actually two in parallel. The biasing should be OK! 73 de Heikki |
Heikki, I dug this issue out of the archives and looked at the circuit. I'm not surprised that the performance is not real good. This circuit has no balance or switching symmetry. It is a simple half-wave switch which samples the IF at the BFO frequency. The output will have significant energy at the IF frequency as well as the BFO frequency, and these components might be overloading the audio amplifier. I'd suggest something that has full-balance. You could modify the '4066 circuit to have this property if you could generate 2 opposite phases of the IF and drive them separately into the inputs of the two '4066's. If you set it up right, you will get very large dynamic range. Alternatively, a traditional double balanced IC circuit such as an NE602 or even a diode DBM will work just fine as long as there is adequate selectivity in front of it to ensure that it is normally working with one IF input signal. Joe W3JDR "Heikki Ahola" wrote in message ... W3JDR wrote: Heikki, I'm not sure exactly what this particular circuit looks like, but I strongly suspect it's just a balanced mixer made from 4066 analog switches. I have used this technique and it is quite good at low IF frequencies (a few MHz or less). Are you sure you have the switches biased properly? The thru-path should be biased at about 1/2 of the 4066 DC supply voltage in order to avoid severe signal clipping. Can you point to a link where the circuit is shown?? The circuit can be found in QST Nov 1992 (Recent Advances in Shortwave Receiver Design)! I t is unclear whether only two switches are used or if each switch is actually two in parallel. The biasing should be OK! 73 de Heikki |
| All times are GMT +1. The time now is 10:46 AM. |
Powered by vBulletin® Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
RadioBanter.com