UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Other Discussions > Homebrew Equipment

Notices

Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members.

Closed Thread
 
Thread Tools
Old 21st Nov 2017, 2:39 am   #1
mictester
Triode
 
mictester's Avatar
 
Join Date: Dec 2007
Location: Sometimes Suffolk and other times Limburg, NL
Posts: 37
Default Basic limiter for pantry transmitters

I found this languishing on a hard drive last night and gave it a try earlier this evening. Granted it adds some distortion to the modulation when driven hard, but at reasonable levels, it's capable of shaving off the peaks and providing some basic tone control.

Click image for larger version

Name:	Compressor2.gif
Views:	558
Size:	8.3 KB
ID:	152622

I was pleasantly surprised by just how effective this simple circuit can be. I used trimmers for the two pots, as I see this as a "set and forget" circuit. I put it on the end of a simple FET-based compressor, mostly to handle overshoots. It's not going to be as effective as an expensive audio processor, but for the pennies it costs, it's worth building!

Last edited by mictester; 21st Nov 2017 at 2:43 am. Reason: Additional info
mictester is offline  
Old 21st Nov 2017, 3:18 am   #2
mictester
Triode
 
mictester's Avatar
 
Join Date: Dec 2007
Location: Sometimes Suffolk and other times Limburg, NL
Posts: 37
Default Re: Basic limiter for pantry transmitters

I also found this circuit. The op-amps were sections of a TL074, and the original was thrown together on strip board with excellent results. In my case, I used the fourth op-amp in the package as part of the modulator, driving into a medium power transistor to modulate the supply to a small Class C output stage. However, it occurred to me that the last op-amp could be used to stiffen the half-rail supply (at present just the two 10k resistors in a potential divider).

The time constants in the compressor are such that it attacks quickly, with a medium decay time. This seems to suit both speech and music quite well. Obviously, you could change the time constants to suit yourself.

I added a bit of "top lift" on the way in to the filter (the CR parallel network) to brighten the audio a bit. The filter is reasonably effective and the values are calculated for about 6kHz. If you're particularly concerned about bandwidth, you could cascade two of this filter for a tightened passband (another possible use for the fourth op-amp!).

There's no magic in the choice of 680n on the way in. 470n reduced the bass a bit, and 680n was the largest value polyester capacitor I had to hand at the time. A 1µ or even bigger might be appropriate, but the bass end sounds fine with the values shown here.

Click image for larger version

Name:	MWaudio.jpg
Views:	593
Size:	32.0 KB
ID:	152623

The 10µ capacitor in the gate circuit of the FET is the right way 'round - the positive terminal goes to ground, because the voltage derived by the diodes to feed the FET gate is below 0V.

Last edited by mictester; 21st Nov 2017 at 3:22 am. Reason: Further Info
mictester is offline  
Old 21st Nov 2017, 3:20 am   #3
Argus25
No Longer a Member
 
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
Default Re: Basic limiter for pantry transmitters

I found that two things were required to prevent issues with the AM modulator for an excellent pantry transmitter.

One is the soft peak clipper to limit the absolute modulation voltage on peaks (regardless of the input audio's average level) to a value that will prevent the carrier being modulated out at any time. Setting it at 90% for peaks with a peak indicator achieves that, so that the average modulation level is set so that only occasional peaks get clipped. The clipping has to to rounded or "soft" to prevent higher frequency harmonics.

The other thing is to level control the audio source, especially from iPods, as its very variable (despite the supposed volume levelers on some mp3 players that don't work very well).

There have been quite a few reasonable circuits over the years using fets, as you mentioned, set up in an AGC like configuration. After testing a number of these circuits with varying attack and decay time constants, I couldn't find a configuration I was happy with.

Then I tried the NE571 compander IC. Despite the time average volume compression of one stage of that relating to the time averaged square root of the absolute value, which one would think might be enough compression, for the AM modulator application it really required "double compression" that is to get the music transmitted to the radio to be as uniform in perceived average volume as that transmitted from a real radio station as well. That is an indication of the large average level variability from the mp3 player between songs/music acquired from different sources.

I cannot distinguish the transmission from my pantry transmitter to any other received commercial station, it is just as uniform and well modulated, they also strive for high modulation levels within the limits.

The NE571 is not an expensive IC and one IC does it for mono as the IC was designed with two sub-circuits for stereo use.

Using this system, a high average carrier modulation level and better transmission range can be attained without any frequent adjustments of the signal level driving the modulator too.

But I do agree with your sentiment of trying to keep it all as simple as possible as well.

Luckily in this case all the work is done in the NE571 and it just needs the external R's and C's.
Argus25 is offline  
Old 21st Nov 2017, 3:34 am   #4
mictester
Triode
 
mictester's Avatar
 
Join Date: Dec 2007
Location: Sometimes Suffolk and other times Limburg, NL
Posts: 37
Default Re: Basic limiter for pantry transmitters

I've used the NE571 in many products that I've designed over the years, and found it to be a very useful chip. These days it's getting a bit scarce, though it's still "second-sourced" by ON-semiconductor.

I'd certainly be interested in the way you configured your 571 - there is a neat "fast limiter" circuit provided in the datasheet that uses a couple of comparators for level sensing, which I've used in both mono and stereo configurations. There's also the "ALC" configuration that I've used successfully too. In the case of medium wave, I used the ALC into the fast limiter with pretty good results.

I have designed commercial medium wave processors, and have used BBD delay-line techniques to prevent overshoots and "pumping" effects. One of my designs is in widespread use and is considered to be almost an "industry standard" for low power medium wave stations in many parts of the world. It provides multi-band compression and limiting, using delay lines and also provides additional very tight bass end fast compression because that was demanded by the music radio broadcasters in Latin America in particular. The whole mess is on a single PCB and takes up the front half of a standard 1U 19" rack box.

Last edited by mictester; 21st Nov 2017 at 3:35 am. Reason: spilling error
mictester is offline  
Old 21st Nov 2017, 4:30 am   #5
Argus25
No Longer a Member
 
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
Default Re: Basic limiter for pantry transmitters

Quote:
Originally Posted by mictester View Post

I'd certainly be interested in the way you configured your 571
The circuit (sorry the IC is drawn as a box, not a good schematic and I didn't write the component values in but I have them if you need them) showing it is on post 15 of this thread:

https://www.vintage-radio.net/forum/...d.php?t=141419

but if you get out the data sheet you will see how its configured. That picture shows the IC as an SA571, supposedly identical to the NE, but I found some differences in the NE and SA versions on testing.

It can, with its R's and C's, be put on a fairly small pcb as shown in post 16.

Hugo.

Last edited by Argus25; 21st Nov 2017 at 4:40 am. Reason: add info
Argus25 is offline  
Old 21st Nov 2017, 9:07 am   #6
Argus25
No Longer a Member
 
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
Default Re: Basic limiter for pantry transmitters

Hi again, I've attached the NE571 circuit drawn as the circuit, with values and some of the equations when I designed it, its driven directly from the iPod, the output passes to the soft peak limiter and then to the amplitude modulator. I hope there is enough image contrast to see it.
Attached Thumbnails
Click image for larger version

Name:	amxmtr2.jpg
Views:	1037
Size:	49.5 KB
ID:	152626  
Argus25 is offline  
Old 21st Nov 2017, 11:29 am   #7
mictester
Triode
 
mictester's Avatar
 
Join Date: Dec 2007
Location: Sometimes Suffolk and other times Limburg, NL
Posts: 37
Default Re: Basic limiter for pantry transmitters

It occurred to me this morning that it might be fun to develop a feed-forward leveller, using op-amp all-pass filter stages for the delay. This is something I'll look at later. I did a telephone auto-patch / coupler for phone-ins using this kind of circuit.

Another idea was to use a switched capacitor filter IC for the bandwidth limiting. These can be configured for "brick wall" filtering (I've used them for 15kHz LPFs in Band II processors) and introduce few artefacts into the passband. With some smart design, it might be possible to use the LPF clock signal to drive a BBD for a short delay to overcome the compressor / limiter attack time....

It's funny - I've been designing and experimenting with this stuff for almost 50 years, and I still try to design for minimal cost / fewest components! It's as if I still have the commercial cost-saving imperative driving my work!
mictester is offline  
Old 21st Nov 2017, 5:12 pm   #8
Argus25
No Longer a Member
 
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
Default Re: Basic limiter for pantry transmitters

Quote:
Originally Posted by mictester View Post
It occurred to me this morning that it might be fun to develop a feed-forward leveller, using op-amp all-pass filter stages for the delay.
I think this is a good idea. The feed forward leveler for the application of getting more uniform modulation depth from varying source material levels I think is better than a feedback/AGC system.

I attempted one using a gain controlled Motorola IC. The tricky part was getting the right time averaging so the gain didn't obviously swing about during quiet and loud patches of a slow paced song.

One good piece of of music to test a leveler on is the Hotel California because it has a very quiet intro with basically just the guitar and a few notes, later the levels come right up as all the other instruments and vocals come in, so it is great for testing the performance of a leveler to check it doesn't cause issues within a recording which has a very wide range of levels.

In the absence of signal, the overall gain from the NE571 system comes up and the background noise increases a tad, but in practice, its not really audible listening to the radio.
Argus25 is offline  
Old 21st Nov 2017, 6:39 pm   #9
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,951
Default Re: Basic limiter for pantry transmitters

Could you use a delay-line?

[There was at least one VHF two-way radio that put the audio through a delay-line in order to prevent the nasty burst of 'sharsh' at the end of a message before the squelch activated]

Alternatively - given the low frequencies involved, you could probably do the whole lot in software for a few quid and a Raspberry Pi.

A/D converter [soundcard input], clipping/levelling/audio-AGC in software, then D/A converter [soundcard again]. I'm sure you could also digitally mix-in some fake co-channel interference [Radio Tirana?] to replicate that night-time-medium-wave-listening experience.
G6Tanuki is offline  
Old 21st Nov 2017, 7:18 pm   #10
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,059
Default Re: Basic limiter for pantry transmitters

Quote:
Originally Posted by mictester View Post
The 10µ capacitor in the gate circuit of the FET is the right way 'round - the positive terminal goes to ground, because the voltage derived by the diodes to feed the FET gate is below 0V.
Not to me it isn't!

The gate is driven negative to ground, so the positive terminal of the 10uF capacitor should be at ground. You have the negative to ground!

It looks a simple circuit, and thanks for putting up! 3-pole Sallen-Key filter; variable-gain amplifier; and rectifier to drive the gain-altering stage.

One possible refinement would be to increase the gate resistor to 1M (it passes zero current anyway, so this should not matter) and also add a second 1M resistor, drain-gate. This will put a bit of AC on the gate, so that mid-way along the channel there'll be zero AC. So the maximum amplitude of AC between the gate and any part of the channel will consequently be reduced - channel resistance modulation by the AC signal will be reduced, and you'll have less distortion.

The reason for using high-value resistances is of course when the gate is driven hard negative, the external resistors will shunt the FET and for maximum range you want this to be negligible.

You could even play games with additional FET gate drive from the output of the first stage (feed-forward AGC rather than feed-back) as well as the second - get this right and you'll have a dead-level output. With more feed-forward, you can even have an output that DECREASES with increasing input, though that's really of academic interest only!
kalee20 is offline  
Old 21st Nov 2017, 8:18 pm   #11
mictester
Triode
 
mictester's Avatar
 
Join Date: Dec 2007
Location: Sometimes Suffolk and other times Limburg, NL
Posts: 37
Default Re: Basic limiter for pantry transmitters

You're absolutely right - the capacitor IS reversed - that's what comes of posting when I should be asleep!

I tried your changes around the FET - you're right. I reduced the stage distortion when compressing by almost ¾ so that it's now under 1% with 10dB of gain reduction. I should have remembered this wrinkle from all those years ago!

There is a tendency for the compressor to "pump" with loud rhythmic passages of music, but this is much less objectionable than the effects of over modulation. The delay line approach is looking ever more appealing!

Here's the corrected and improved filter and limiter. I added a resistor from the input to ground to eliminate the annoying "pop" when I plugged in a source, and the additional resistor around the FET has improved things significantly.
Attached Thumbnails
Click image for larger version

Name:	MWaudio.jpg
Views:	953
Size:	35.2 KB
ID:	152667  
mictester is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 3:36 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.