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 > General Vintage Technology > Components and Circuits

Notices

Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets.

Closed Thread
 
Thread Tools
Old 15th May 2017, 3:22 pm   #1
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Question Cct to improve modulation here, please?

Here is the basis for my homebrew AM internet - I'm calling it that rather than a pantry transmitter as it has an unusual antenna arrangement, shown.

However, that is OT here., for the time being at least.

I'm getting a great carrier wave but modulation with the single transistor audio stage shown is very poor and unworkably so, probably around 10%, I'd guess., and this from an iPhone.

IPhones are meant to have an output of 1V at the appropriate impedance load but I'm measuring it at closer to 400mV and this circuit definitely needs probably 5 times greater voltage gain (we don't need any current amplification whatever) for 100% modulation.

Unless I've made some other error, I am at a loss as to how to increase the voltage gain in the simplest way possible, preserving linearity.

I'd like to use a single additional transistor if possible, op amp at a pinch.
I am aware of generic ideas, but need guidance down to the level of voltage gain and matching impedances, so actual designs rather than conceptual ones, please.

Any ideas very much appreciated!



Modifications.

I did decrease the value of the series resistor shown connected to the emitter bypass capacitor (220uF)in the modulator to just 120R in order to increase sensitivity but this didn't do enough. The emitter bypass resistor is 150R instead of 180.

Also, I have used a 220uH inductor for the output collector load, instead of 22uH. This was self-resonant at around 3MHz using a crystal oscillator with a fundamental of around 1MHz.

Very grateful for any tips, RF experts.

The resonant loop shown has a D-Qing resistor of 10K shown. I know this will have a beneficial effect on audio bandwidth once this is connected, but for the time being, all tests are being done with no load connected, just a 'scope. Modulation percentage has to be right before I get involved with tuning. Thanks!
Attached Thumbnails
Click image for larger version

Name:	AM internet (1).jpg
Views:	113
Size:	47.0 KB
ID:	142601  
__________________
Al

Last edited by Al (astral highway); 15th May 2017 at 3:37 pm.
Al (astral highway) is offline  
Old 15th May 2017, 3:50 pm   #2
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Default Re: Cct to improve modulation here, please?

Photo showing construction on QRP-me pads and inside an old sweet tin. The
BNC connector at picture top is actually to connect to the 23V power supply , not shown! The bottom one is RF out.
Attached Thumbnails
Click image for larger version

Name:	IMG_1499.jpg
Views:	92
Size:	70.1 KB
ID:	142602  
__________________
Al
Al (astral highway) is offline  
Old 15th May 2017, 4:13 pm   #3
peter_scott
Dekatron
 
peter_scott's Avatar
 
Join Date: May 2003
Location: Edinburgh, UK.
Posts: 3,273
Default Re: Cct to improve modulation here, please?

Hi Al,

You say you changed the 220uF series resistor from the 22 ohms shown to 120 ohms. You could try a smaller value than 22, say 10 ohms. This should increase the AC gain although it will also reduce the input impedance that your audio signal sees. It will affect the linearity too.

If connecting your modulator to the audio source causes significant loading you could interpose an emitter follower in the audio input circuit.

HTH

Peter
peter_scott is offline  
Old 15th May 2017, 4:28 pm   #4
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Default Re: Cct to improve modulation here, please?

Quote:
Originally Posted by peter_scott View Post
Hi Al,

You say you changed the 220uF series resistor from the 22 ohms shown to 120 ohms. You could try a smaller value than 22, say 10 ohms.

Hi Peter,

You spotted a mistake that I hadn't realised. I'd intended to decrease that resistor to 12 R, but lost the plot and selected a ten times higher value. No wonder the sensitivity didn't increase!

Well spotted, I'll try it out... Thanks.

Quote:
Originally Posted by peter_scott View Post
Hi Al,


If connecting your modulator to the audio source causes significant loading you could interpose an emitter follower in the audio input circuit.
Interesting. I hadn't thought of it as a loading problem, but that would make a lot of sense. I think IPhones are comfortable around 600R? What's the input impedance shown, approx? Is there a big old mismatch?

I'd thought of an emitter follower but for space reasons, it might be easier on second thoughts to pop a suitable SOIC op-amp before the stage shown, rather than after. I just need that little bit of voltage gain and don't really want to lose any linearity. I'm not sure what's best here, there seem to be loads of video amps that run on similar power rails. I think I have LM2904DG somewhere too.
Thanks for sharing your ideas
__________________
Al

Last edited by Al (astral highway); 15th May 2017 at 4:41 pm. Reason: Formatting of quotes for forum rules.
Al (astral highway) is offline  
Old 15th May 2017, 4:40 pm   #5
cmjones01
Nonode
 
Join Date: Oct 2008
Location: Warsaw, Poland and Cambridge, UK
Posts: 2,669
Default Re: Cct to improve modulation here, please?

Your iPhone will be happy down to about 32R, because that's what most headphones are. There's nothing to worry about with impedance matching in the audio circuit: as long as the output impedance of your iPhone is low enough to drive the input impedance of the modulator, you're OK. The iPhone's headphone output certainly be OK to drive any reasonable transistor circuit, even one with a low emitter resistance.

Chris
__________________
What's going on in the workshop? http://martin-jones.com/
cmjones01 is offline  
Old 15th May 2017, 5:09 pm   #6
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Default Re: Cct to improve modulation here, please?

Quote:
Originally Posted by cmjones01 View Post
There's nothing to worry about with impedance matching in the audio circuit: as long as the output impedance of your iPhone is low enough to drive the input impedance of the modulator, you're OK.
Cheers, Chris. But isn't the input impedance of the audio stage shown a few Kilohms, so there's a mismatch that way?
__________________
Al
Al (astral highway) is offline  
Old 15th May 2017, 5:36 pm   #7
mhennessy
Dekatron
 
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
Default Re: Cct to improve modulation here, please?

Unlike other areas, it's the norm in audio to have that mis-match. Output impedances are typically 50 ohms or less; input impedances are typically 10k or more.


(Obviously I'm talking about line-level stuff in general. There are occasions when matched impedances are encountered, and sometimes a particular input or output impedance might be required. But these are rare, so there's no need for pedantry )

Last edited by mhennessy; 15th May 2017 at 5:41 pm.
mhennessy is online now  
Old 15th May 2017, 5:53 pm   #8
mhennessy
Dekatron
 
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
Default Re: Cct to improve modulation here, please?

...anyway, the transistor is acting as a voltage controlled current source. So if you haven't got enough modulation, then a pre-amp is the best way to increase it as I'd expect that reducing the emitter resistance would reduce the linearity. A simple common-emitter job would be fine, given the context here.

But before doing that, have you been able to test for modulation depth/linearity using a signal source? We need to be absolutely sure that the multiplier is working correctly first.

As to a workable design, given that you need a gain of about 5, I'd probably start with the classic 4-resistor transistor circuit. A potential divider feeding the base, an emitter resistor and a collector resistor. Based on the 24V rail, I'd pick 150k and 10k for the potential divider, to give ~1.6V at the base, and 1k for the emitter resistor, and 6k8 for the collector. Gain will be approximately 6 times, depending on the input impedance of your current source. Reverse the polarity of your input cap, and add another cap between input and base. Simples!

HTH,

Mark
mhennessy is online now  
Old 15th May 2017, 6:30 pm   #9
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Default Re: Cct to improve modulation here, please?

Quote:
Originally Posted by mhennessy View Post

But before doing that, have you been able to test for modulation depth/linearity using a signal source? We need to be absolutely sure that the multiplier is working correctly first.
Well, something isn't right! Just can't work it out.

Even though there is a tiny impedance to AF (way less than 1R) via the electrolytic there in the input, the signal is attenuated to just 44mV RMS at the base of the audio amplifier transistor. Things are working just fine beyond that as there's a paltry 1.12V RMS out at the collector where it joins the two emitters.

That's a voltage gain of 25. Thing is, this indeed closer to the design voltage gain, since it would be 25V, approx 50% modulation, if things were working ok and there was around 1V at the input from the iPhone. 44mV in is ridiculous!!

But... how on earth do I get this spurious and infuriating attenuation at the very beginning?

Quote:
Originally Posted by mhennessy View Post

As to a workable design, given that you need a gain of about 5, I'd probably start with the classic 4-resistor transistor circuit. A potential divider feeding the base, an emitter resistor and a collector resistor.
Sounds good, only if we pursue the mystery attenuation problem, we may need to just double voltage gain instead. But I'd certainly like to see it with 50% modulation before we go there...

Thanks Mark... hope you can help suggest what's wrong.
__________________
Al

Last edited by Al (astral highway); 15th May 2017 at 6:36 pm.
Al (astral highway) is offline  
Old 15th May 2017, 6:37 pm   #10
Guest
Guest
 
Posts: n/a
Default Re: Cct to improve modulation here, please?

My iPod has a selectable "maximum volume" has the iPhone? Something to do with ear damage and law suites I think.
 
Old 15th May 2017, 7:12 pm   #11
mhennessy
Dekatron
 
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
Default Re: Cct to improve modulation here, please?

I've never tested an iPhone, but that's obviously a separate issue. Simplify. Use an audio oscillator with known characteristics first of all. Then, test the iPhone separately. As it is, there are too many variables. Is the iPhone being dragged down by your circuit, or is it being limited in the way MM suggests? Easy enough to figure out - although easy for us to say at this distance
mhennessy is online now  
Old 15th May 2017, 7:13 pm   #12
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Default Re: Cct to improve modulation here, please?

Quote:
Originally Posted by merlinmaxwell View Post
My iPod has a selectable "maximum volume" has the iPhone? Something to do with ear damage and law suites I think.
Hey MM, good point, but I checked. For those who're interested, go to Settings, Music, and you can delimit the volume in there. But the default setting is max. Applies to all devices by the same manufacturer.

I found two problems already, this with brand new components!!

1) A bandolier of brand new resistors had 10X the correct value stamped on the 1.8K resistors' paper backing strip (colour codes correct, but I wasn't looking). Yes! So I believed what I saw and inadvertently selected 1.8K instead of 18K! I think this is as hilarious as it is annoying. It could have taken a lot longer to discover this! It's definitely a printing error, so just down to poor quality standards in the factory of origin (without going too OT).

2) A 10uf decoupling capacitor, rated at ample voltage, had gone leaky. This is ridiculous. It took less than an hour of service for the component to fail.

These two things messed up a voltage divider across the base of the current source transistor, and also placed a high load on the power supply and attenuated the input. This is now remedied.

With I think 50% modulation, I still need more volts in at the base of the current source transistor. But I think I will check how things sounds with the resonant loop attached and tuned up, before going back to tweaking this...it could be a while before I've got it right as I need to hand wind some tiny inductors.!
__________________
Al

Last edited by Al (astral highway); 15th May 2017 at 7:27 pm.
Al (astral highway) is offline  
Old 16th May 2017, 10:19 am   #13
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Default Re: Cct to improve modulation here, please?

Quote:
Originally Posted by mhennessy View Post
Based on the 24V rail, I'd pick 150k and 10k for the potential divider, to give ~1.6V at the base, and 1k for the emitter resistor, and 6k8 for the collector.
Hey Mark, for my learning, just wondering why setting V at the voltage devider at approx just 1.6V?

Also how is voltage gain derived in this configuration, supposing I wished to alter it?

I still have much to learn on DC operating conditions of smal signal transistors!

Thank you
__________________
Al
Al (astral highway) is offline  
Old 16th May 2017, 11:44 am   #14
mhennessy
Dekatron
 
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
Default Re: Cct to improve modulation here, please?

It's a long story. I'll try to cut it short...

As you realise, in any amplifier, there are 2 sides of the coin:

1. DC conditions
2. The AC signal

You have to choose good DC conditions before you have a good amplifier, but it helps to think about them as separate entities as much as you can. The ability to mentally separate any signal into its AC component and DC offset is the secret to understanding amplifiers - not everyone find that easy.

For a common-emitter amplifier, we aim to put the collector at the mid-point of the supply rails - 12V in this case - to allow maximum undistorted swing. I didn't aim for that precisely in this case, as your modulator doesn't need the best part of 24V peak-to-peak(!), and also I was making the numbers easier for me, given the limited time I had available...

The first step is to choose the collector current; 1mA is often a good starting point. That means I needed to put 1V across the 1k emitter resistor to establish that 1mA, and knowing that there is ~0.6V DC between the base and emitter, 1.6V becomes the target for the base. The incoming signal level is modest, so 1.6V is not too low - the modulator should overload before the preamp. But there's nothing stopping you choosing a different value.

Having established 1mA of emitter (so hence collector) current, we're now able to work out what the DC collector voltage is. Simply, we have 1mA flowing in the 6k8 collector resistor, meaning there must be 6.8V across it. So with respect to earth, there is ~17V (24-6.8V) at the collector. Maximum voltage swing before clipping will be around 13V peak-to-peak. That is comfortably more than you needed, so that's all good


That's DC conditions, now what about AC signal?

In this circuit, the emitter resistor is un-decoupled. This means that a signal appears across the emitter resistor, and this signal subtracts itself from the incoming signal. In other words, there is negative feedback. As a result, the voltage gain is pretty close to RL/RE, or x6.8.

If you had put a capacitor across RE, then the gain is much, much higher. It's approximately gm times RL, where gm is 40 times the collector current, and RL is 6k8. That's 272. In practice, it'll be lower. Happy to do a fuller explanation, but this post is long enough already

But, the input impedance of the following stage is in parallel with RL (as far as AC signal is concerned), so that'll bring down the gain. RL can be increased in value to bring it back up, subject to the DC conditions remaining reasonable. E.g., if you changed RL to 10k, you'd have a theoretical gain of x10 in isolation, and the collector voltage would fall to 14V. Still perfectly workable. But if you increased RL to 22k, the collector voltage would fall to 2V, and that's no good for the size of signal you want. So, it's a juggling act. In practice, the input impedance of the modulator is quite low, so it'll have more effect on gain than RL. Decreasing RE might be better/easier...

While a single transistor stage like this is capable of tremendous amounts of gain, it's not very "clean". So you're much better asking for modest amounts - as we are - from a single transistor because the negative feedback does a really good job of cleaning up the distortion. It won't be as good as an op-amp, but that's why I said "given the context" earlier - it's an AM transmitter

It's possible to get a lot more precise about all this, but hopefully that's a explanation that's easy to understand given a couple of reads.

Mark
mhennessy is online now  
Old 16th May 2017, 6:16 pm   #15
Al (astral highway)
Dekatron
 
Al (astral highway)'s Avatar
 
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
Default Re: Cct to improve modulation here, please?

Hello Mark,

That is a wonderfully lucid explanation, thank you. I appreciate the time and thought you it into it and it is very helpful indeed in my decision making process.

I'll likely try two routes, both op-amp and single transistor , and compare them. I found a op amp in my spares that runs up to this kind of voltage and it will also come down to passive component count as there isn't much space on the board now and those QRP me pads need some breathing space ! (Although I do have smaller versions somewhere)

I am getting more analytical in my thinking, down to node level sometimes , but it is a relatively recent practice amd takes effort, and I find DC conditions to be more intricate, sensitive and important than I realised , the more I slowly discover ! I'm not an electronics engineer or even close and so this isn't hard-wired in me by any means. Loving the journey though .
__________________
Al
Al (astral highway) is offline  
Old 16th May 2017, 6:53 pm   #16
mhennessy
Dekatron
 
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
Default Re: Cct to improve modulation here, please?

Hi Al,

Glad it helped

Using an op-amp from a single supply will require a bit of extra effort to bias it. In practice, you might end up using more parts than the simple transistor. However, don't let that discourage you from experimenting!

To give you some tips about biasing the op-amp, here's an old thread which will be useful: http://www.vintage-radio.net/forum/s...ad.php?t=65007

Good luck,

Mark!
mhennessy is online now  
Old 16th May 2017, 7:29 pm   #17
Guest
Guest
 
Posts: n/a
Default Re: Cct to improve modulation here, please?

Quote:
where gm is 40 times the collector current
I simplify this to an equivalent emitter resistance of 25 ohms per (inverse of) mA, 1mA=25 ohms 2ma 12.5 ohms and so on. This pops out of the Ebbers Moll (spelling?) transistor equations. When you (Al) get it working for the DC conditions and haven't go enough gain and too much when merrily bypassing the emitter resistor put a resistor in series with the bypass capacitor, DC not affected and the AC the gain is (near enough) collector resistance divided by the total parallel emitter resistance.

Much as said by Mark, get the DC right first.
 
Old 16th May 2017, 8:34 pm   #18
mhennessy
Dekatron
 
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
Default Re: Cct to improve modulation here, please?

Yes, that's another outcome of Ebers-Moll...

In other words, re=1/gm
mhennessy is online now  
Old 16th May 2017, 11:16 pm   #19
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: Cct to improve modulation here, please?

The nice bit about the little re model approach is that it mkes it easy to incorporate with any resistance added in the emitter path.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 16th May 2017, 11:43 pm   #20
mhennessy
Dekatron
 
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
Default Re: Cct to improve modulation here, please?

Yes, there's usually more than one way to calculate things.

Take the example of a 100 ohm emitter resistor and a 4k7 collector resistor, plus a collector current of 1mA:

1. You could say the gain was 4700/(25+100), which is x37.6
2. Or you could say that gm is 1/(25+100), which is 8, hence gm.Rl also gives x37.6
3. Or you could calculate the "raw" open-loop gain (40 times 4.7), then apply NFB theory (gain = Ao/(1+AoB), which also gives x37.6

All three methods are mathematically equivalent, but give 3 different paths to understanding
mhennessy is online now  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 7:33 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.