Talk:Direct-conversion receiver

This article is somewhat misleading.

The Direct Conversion Receiver does NOT normally have the Local Oscillator synchronised.

The L.O. can certainly be phase locked (see Costas Loop), but more often is free running.

?In most commercial applications, including, but not limited to cellular, it is certainly phase locked. When considering modern communication technology I find it a strange statement that it is normally free running.

Direct Conversion designs are frequently used by Radio Amateurs to build minimalist receivers for SSB or CW.

117.120.16.131 (talk) 05:58, 30 September 2008 (UTC) Agree with the above comments. There article appears to use an unreasonably broad understanding of what constitutes a direct conversion receiver. I personally would prefer to see Zero IF, and homodyne removed, as these are not limited to direct conversion applications. 117.120.16.131 (talk) 05:58, 30 September 2008 (UTC)

Agree, this article is a mix of the misleading and the erroneous, and badly presented to boot. 75.140.251.185 (talk) 01:58, 3 July 2009 (UTC)

Superhet comparison removed

On March 15, 2011 User:132.170.55.192 removed from the intro the lines: "A direct-conversion receiver requires only a single stage of detection and filtering, compared to a superheterodyne receiver design. Superhet designs use two stages of filtering to convert signal carrier frequency to an intermediate frequency first before extracting the audio modulation" with the comment: "The lines removed are misleading and in error; perhaps a reference should be made to the single-conversion superheterodyne for a comparison." It seems to me the deleted lines (I didn't write them) are fairly accurate, and express a notable advantage of the direct-conversion circuit over the dominant superhet. --Chetvorno^TALK 02:35, 16 March 2011 (UTC)

I agree that the earlier wording was useful. I am rewriting it and will try to be clearer than the previous wording. Interferometrist (talk) 13:38, 16 March 2011 (UTC)

OK, I changed it, and it's much longer than before! That means that the rest of the article has to be expanded so that the new lede is of proportionate size ;-) Interferometrist (talk) 14:29, 16 March 2011 (UTC)

Oh, after all that, I think I understand the confusion:

The lines removed are misleading and in error; perhaps a reference should be made to the single-conversion superheterodyne for a comparison.

Well, the original reference was indeed in relation to a normal single-conversion (single IF frequency) superhet, but that had been written in the previous version as "two stages of filtering" and indeed in my rewrite it mentions "two (or more) frequency conversions" since the demodulation stage itself is a frequency conversion (from IF to baseband). It might not seem like a frequency conversion using incoherent detection (envelope detector etc.) but it is, especially when using a product detector and BFO in a superhet which is what the DCR must be compared to. So it amounts to only ONE frequency conversion compared to 2 (or more) and only ONE filtering stage (after demodulation) rather than also needing IF filtering and indeed RF filtering to remove the image frequency (though there are other reasons you'd like filtering in the RF stage, such as preventing intermodulation from strong out of band signals). I hope that helps. Interferometrist (talk) 15:23, 16 March 2011 (UTC)

I'm glad that a knowlegeable editor is working on this, it needs work. But I really feel the new introduction is way too technical. It would be more appropriate for an engineering manual. The intro has to be understandable by general readers. What they are looking for is the simplest possible explanation (say, high school science level) of what it is and how it works, what products it's used in, some notability (a summary of advantages and disadvantages), and maybe some historical data (who invented it). This introduction has only about one sentence on how it works; that should be expanded. Most of it is taken up with technical advantages and disadvantages; great material, but the few important ones should be summarized (it's simpler than the superhet, but its got DC offset problems and requires a PLL to receive AM and FM so it's used in less demanding digital receivers) and the rest moved down into "Advantages" and "Disadvantages". Also it's uses should be mentioned: cell phone and WiFi receivers. Anyway, that's my view. --Chetvorno^TALK 16:09, 16 March 2011 (UTC)

I see. Perhaps I could move a lot of that to the body. But I'm not sure what stays in the lede then. This is solely a technical article anyway since this isn't an item you can just go to the store and ask for, or read consumer reviews of. What content SHOULD stay in the lede? Do you also think the previous version (before this issue came up) was too technical? Or by "too" technical do you just mean too wordy. The problem is that if you try to make it short, it becomes inexact, and I think that led to the complaint and removal of text by 132.170.55.192 Interferometrist (talk) 17:37, 16 March 2011 (UTC)

I guess I'll try to rewrite the lede to include some of what you suggest, and move the other stuff to the body. I hadn't really meant to get into this article, but was just trying to deal with a shortcoming (that you pointed out) when some technical language was removed. But so it goes..... Interferometrist (talk) 17:43, 16 March 2011 (UTC)

Well, I didn't change the text too much, but did break it up. Now the lede is less technical (because it says less). Maybe I'll try to integrate the two new sections better into the existing content. Interferometrist (talk) 18:36, 16 March 2011 (UTC)

That looks a lot better to me. You're right, there's a limit to how much this subject can be simplified, but in the lede it's important to try. I think more of the advantages and disadvantages of the circuit could be mentioned in the lede; it's just that they should be summarized, and the detailed technical explanations saved for the following sections. Also its uses should be included in the lead; something like: "In the form of the regenerative receiver it was used in the early days of radio in the 1920s and 1930s, but due to DC offset problems was replaced by the superheterodyne. In recent years, because if its simplicity it has seen a resurgence of popularity in digital radio applications, and is used as the receiving circuit in some cell phones and WiFi and Bluetooth enabled devices." --Chetvorno^TALK 22:26, 16 March 2011 (UTC)

Well thanks for the feedback, good and bad! I'll try to do a little more work (in fact I had some better text written, but then my browser crashed and I didn't feel like looking through a 2GB "core dump" to find it :-(

Also its uses should be included in the lead; something like:

Actually YOU certainly seem to know more about it historically and about its current application. Why don't YOU write a small section on applications. In fact, when I work on it I'll include a stub section where you can paste in a few things like you just pulled off the top of your head. But perhaps let me first finish what I was doing so it fits in better (I was going to further mention quadrature detection and processing, for instance). Also, I slightly disagree with you here:

In the form of the regenerative receiver it was used in the early days of radio..

Of course this is a matter of symantics, but I think it doesn't do justice to either technology to say that. Of course I understand what you're saying, that when the regenerative receiver is oscillating that dominates the signal so all the stations beat against it and the demodulated signal is just like the DCR. But 1) the regenerative receiver could be used without oscillating doing straight envelope detection but still taking advantage of the increased selectivity around the tuning frequency due to the Q approaching infinity (due to the positive feedback), a property the DRC doesn't have; 2) The DCR is very refined compared to the regenerative, and can for instance use a frequency synthesiser, do quadrature outputs, carrier level and absolute phase detection and PLL the whole receiver; 3) The regenerative was largely designed to get much more gain from 1 tube than you could possibly expect; hardly a concern with the present price of transistors!

There were also some shades of opinion (but not very nicely expressed!) above on the talk page about definitions -- basically I don't know of any glossary of definitions we're up against and this article should be inclusive, I'd say. In fact I think with some software radio it gets a little ambiguous. If you have a sample/hold and A/D at 100MHz, then is the sample/hold a part of the A/D, or is it a 25MHz carrier frequency DCR (implemented with the sample/hold) with 4 quadrature outputs (output serially) which get sampled by the A/D? Definitions. Interferometrist (talk) 01:22, 17 March 2011 (UTC)